STUDIES IN ANCIENT HISTORY STUDIES IN ANCIENT HISTORY THE SECOND SEEIES COMPKISING AN INQUIEY INTO THE OKIGIN OF EXOGAMY BY THE LATE JOHN FERGUSON M'LENNAN EDITED BY HIS WIDOW AND ARTHUR PLATT ILonfcon MACMILLAN AND CO., LTD. NEW YORK: MACMILLAN & CO. 1896 PEEFACE THE second series of Studies in Ancient History is now published at so long an interval, not only from the publica- tion of the former series but also from the author's death, that a word of explanation is called for. The delay is owing to a succession of most lamentable misfortunes. Mr. M'Lennan had originally contemplated the produc- tion of a great work on early society, which should have collected and presented in a better form all his pre- viously published investigations, together with the results of his later study, and which should in particular have thrown some light on the origin of exogamy. Despairing, in his state of health, of ever executing this project, he decided to publish an intermediate volume to supply its place, but even this intention was frus- trated by his death. His brother, Mr. D. M'Lennan, then took up the task, for which he was eminently fitted. He completed from Mr. J. F. M c Lennan's posthumous papers, and published in 1885, The Patri- archal Theory, a preliminary and polemical inquiry meant to clear the way, while at the same time em- 239030 vi STUDIES IN ANCIENT HISTOR Y bodying positive results. He also brought out a new edition of the first series of Studies in Ancient History, with some notes added by himself (1886). But, while still employed in working on the most important part of his task, he was also carried off, and the work passed into the hands of Professor Eobertson Smith. Unfor- tunately he was already attacked by that long illness to which he finally succumbed, and nothing has been found among his papers which has any bearing upon the work. Under these circumstances, we determined to do what seemed to us possible in the way of arrang- ing the mass of material which Mr. M'Lennan had left for the book. This consisted of a great quantity of notes, several chapters partially or entirely finished, and a list of chapter headings, which show clearly the principal lines on which the work would have been developed. One of these papers, that on Agnation, was used by Mr. D. M'Lennan for the Patriarchal Theory. Besides this, there was the " Essay on the Wor- ship of Animals and Plants" (Totemism), published in the Fortnightly Review in the years 1869 and 1870 ; this would certainly have been embodied in the work by the author, indeed parts of it have been used, as it is, in the ninth chapter of this volume. The rest, with some trifling changes, is here reprinted in an Appendix. Of the chapters which were written for the new book, PREFACE vn none has hitherto appeared except that on the Origin of i Exogamy, with additions by D. M'Lennan (English His- torical Review, January 1888). They are here arranged in the order indicated by the list of chapter headings already referred to, deficiencies being supplied by short connecting passages, which give a skeleton of what should have been there. But the eighth and ninth do not correspond to any heading in the list, and we have thought it best to place them after the rest to stand independently. Our own additions are in smaller type and in brackets. The whole was to have been arranged in three books, the second of which would have included the evidence for the different races of man all over the globe. It would have come between our third and fourth chapters, according to the list left by the author, but it has swelled to such a size that we pre- ferred to depart in this instance from his arrangement, and to place it separately as Part II. of this volume. The vast collection of matter which had been made for this section of the work was left in a condition of most unequal development. Some parts were completely finished, others were mere heaps of notes either marked in books or copied out, others worked up into shape in various degrees. The most important chapters, which could be printed as they stood, appear here in large type. America should have come first, but the notes for America were so fragmentary for the most part that viii STUDIES IN ANCIENT HISTORY we prefer to give the precedence to Australia and the Pacific Islands, as being in the most satisfactory con- dition. This is largely due to Mr. D. M'Lennan's having worked at them after his brother's death. He rewrote a great part of, and added to, the chapters on Fiji, Samoa, and Australia. He also did the same for that on Ashantee, and wrote the essay on the Kamilaroi and Kurnai which is printed in the Appendix. Mr. Ralph Cator of Lincoln's Inn has arranged the rest of the African evidence, which was left in a nearly com- plete form, and has given us valuable assistance in other respects. The American evidence, as hinted above, was so fragmentary that little could be done with it, and it seemed better to leave the reader to draw his own conclusions from it than to attempt to give Mr. M'Lennan's. As for the rest of the world, there were notes in abundance, and in particular upon Arabia, but nothing in any approach to literary form. It proved quite impossible in this division of the work to carry out consistently any distinction between the author's text and the alterations which it was neces- sary or advisable to make ; and so the brackets had to be generally discarded. What additions have been made were taken from the author's own notes, with only one or two exceptions, the principal of these being the references to Theal in Chapters XXVIII. and XXIX. The nature of the present work sufficiently PREFACE ix explains by itself the absence of any reference to recent writers on the subjects concerned, in particular Messrs. Frazer, Letourneau, and Westermarck. It remains for us to thank Mr. W. H. D. Kouse for correcting that part of the work where a knowledge of Sanscrit was required, and for adding references ; two notes of his are signed te K." Professor Strong was kind enough to assist us in the Arabic. We have also to thank the proprietors of the Fortnightly and English Historical Reviews for permission to reprint from them the papers above mentioned. ELEONORA A. M'LENNAN. ARTHUR PLATT. With the deepest regret I have to add that while the following pages were in the press Mrs. M'Lennan passed away. During her husband's life she was his untiring assistant and amanuensis, and from his death every moment of her time was given up to the prepara- tion of this book. Despite continual ill -health, she persisted in working at it, and after Professor Robert- son Smith's death determined to complete it herself. It is due entirely to her unceasing devotion and resolu- tion that it now sees the light. A. P. CONTENTS PART I CHAPTER I PAGE CASES RAISING THE QUESTIONS FOE INVESTIGATION CHAPTER II ON THE METHOD OF INQUIRY IN EARLY HISTORY CHAPTER III THE MODE OF HANDLING EVIDENCE . .27 CHAPTER IV DEFINITION OF TERMS . CHAPTER V KINSHIP, TOTEMISM, AND MARRIAGE CHAPTER VI THE ORIGIN OF EXOGAMY . . .57 xii STUDIES IN ANCIENT HISTOR Y CHAPTER VII PAGE FEMALE INFANTICIDE ... .74 CHAPTER VIII CAN EXOGAMY AS MARRIAGE LAW BE INFERRED IN ANY CASE FROM THE LAW OF SUCCESSION? . . .112 CHAPTER IX SOME EXAMPLES OF FABRICATED GENEALOGIES ADDUCED TO SHOW THE KEADINESS OF MEN IN ALL TIMES TO FABRI- CATE GENEALOGIES . . . . .117 PART II SECTION I THE PACIFIC ISLANDS AND AUSTRALIA CHAPTER X HERVEY GROUP, OR COOK'S ISLANDS . .189 CHAPTER XI FIJIAN GROUP . . .206 CHAPTER XII THE SAMOA OR NAVIGATORS' ISLANDS . 232 CHAPTER XIII THE TONGAN OR FRIENDLY ISLANDS . 241 CHAPTER XIV NEW ZEALAND . . .255 CONTENTS xiii CHAPTER XV PAGE AUSTRALIA . . . . . . .278 SECTION II AMERICA CHAPTER XVI NORTH AMERICA, EAST AND SOUTH . . . .315 CHAPTER XVII NORTH AMERICA, WEST . . . . .357 CHAPTER XVIII NORTH AMERICA, NORTH ..... 367 CHAPTER XIX MEXICO AND CENTRAL AMERICA . . . .378 CHAPTER XX PERU . ..... 392 CHAPTER XXI GUIANA . tfr .400 SECTION IllJt AFRICA CHAPTER XXII GUINEA THE KINGDOM OF ASHANTEE . . .405 xiv STUDIES IN ANCIENT HISTORY CHAPTER XXIII PAGE GUINEA THE GOLD COAST . .417 CHAPTER XXIV GUINEA : SLAVE COAST FID A . . .427 CHAPTER XXV LOANGO, ETC. . . .433 CHAPTER XXVI EQUATORIAL AFRICA . . . .440 CHAPTER XXVII s SOUTH AND SOUTH-EAST AFRICA . 445 CHAPTER XXVIII KAFFIRS AND ZULUS . . . . . .477 CHAPTER XXIX BUSHMEN AND HOTTENTOTS 483 APPENDIX I. THE WORSHIP OF ANIMALS AND PLANTS . . .491 II. THE KAMILAROI AND KURNAI. BY D. M'LENNAN . 570 PAKT I CHAPTER I CASES RAISING THE QUESTIONS FOR INVESTIGATION [FOR the first chapter of the proposed work, Mr. M'Lennan has left nothing but a few hints. He has jotted down notes of certain marriage and succession laws in civilised countries, "France, Iberia (Basques), Wales"; and again in other uncivilised countries, " Australia (Grey, Eyre), America (Lafitau), Africa (Bowdich)." It is to be presumed that he would have begun by develop- ing the idea that the law of any country is apt to preserve , many relics of an order of things long passed away, relics unin- telligible in themselves and embedded in a stratum of later formation. As in Primitive Marriage he had started from a custom which has degenerated into a mere symbol, that of capture in wedding ceremonies, so in his later work he appears to have intended to begin by giving some cases of curious survivals from the_past in civilised communities, and to show how intelligible they become .in the light of the knowledge derived from the study of primitive society. For the first chapter, then, he had noted under the above heading some cases among uncivilised races of- the law that no man may marry a woman of his own clan, the law to which he has given the name of exogamy, existing along with the law that children, taking their mother's name or totem, are of her clan and succeed as her heirs. The cases are chosen from Australia, America, and Africa, doubtless to show the wide B 2 STUDIES IN ANCIENT HISTORY CHAP. prevalence of these customs. He had also made notes of several cases of succession laws in the customaries or codes of races in more highly advanced stages of civilisation, by which the children are their mother's heirs long after male kinship has been recognised in all other respects. These laws would be absolutely unintelligible if we had not the explanation of them as a persistence of the ancient succession law after the C. , V^ marriage law which mustjiave once accompanied it had fallen into disuse and been forgotten. The persistence with which such fragments of the laws of inheritance of our savage ancestors hold their place among nations of the highest civilisation is truly remarkable. Thus Chateaubriand, in his Analyse raisonne'e de I'histoire de France, tells us that " dans certaines provinces le ventre annoblissait, c'est a dire que la noblesse etait transmise par la mere." And Baugier says in his Mtmoire sur la Champagne : " Cette noblesse, que la mere transferait a ses descendants dans cette province, ne commenga d'etre attaquee qu'en 1566 ; le pro- cureur de la cour pretendit que cette coutume avait ete toleree par ne"cessite et pour remplir le pays de noblesse ; que, la cause e*tant cessee, Feffet devait aussi cesser." ] Again, in Ancient Wales " there are three women against whose issue there are to be no pleas respecting the inheritance of their mother," the third case being that of "a worm whose son shall avenge a man of his mother's kindred, anc lose his patrimony on account of that crime." 5 Thirdly, among the Basques, Le Play tells us at length that an ancient custom mentioned by Strabo (iii. 165) lasted down to the era of the Code Napole'on, and even beyond that was practised by evasion. " The husbands," says Strabo, " bring a dowry to their wives, and the daughters undertake 1 These passages were quoted also in the article on the Levirate and Polyandry (Fortnightly Review, May 1877, p. 706), where they illusti the extreme tenacity of laws of inheritance. It may be added that in tl author's opinion we may infer in such a case the former existence of law of exogamy ; see infra. 2 Ancient Laws of Wales, vol. i. p. 775. i CASES FOR INVESTIGATION 3 the business of giving their brothers in marriage." The eldest daughter inherits from her mother, and the husband of the eldest daughter takes her name. Customs and laws of this kind have for centuries perplexed all who have met with them. It may be judged by the instances quoted above from Baugier of what value are the attempts to account for these anomalies, when they are looked at by themselves and considered as deliberate expedients of civilised races to secure some rational end. . But recent research into the condition of life and law among uncivilised nations throws abundant light on their real origin. To Lafitau first among the moderns, however ridiculous many of his conclusions now appear, is due the credit of having compared civilised law with that of men yjjt .living in the savage state. A man of great acuteness, well acquainted with classical antiquity, he was struck with the resemblance of the customs of the North American Indians, among whom he laboured as a missionary, to those of the ancients. He found, to his surprise, the same state of things there as Herodotus and other classical authors declare to have pre- vailed in Lycia. " Quell e que soit I'origine des Iroquois et des Hurons, ils ont conserve cette forme de gouvernement . . . car outre cette ginecocratie, qui est absolument la meme que celle des Lyciens, et ou le soin des affaires n'est entre les mains des homines, que comme par voye de procuration, tous les villages se gouvernent de la meme maniere par eux- memes, et comme s'ils etoient independans les uns des autres." 1 We have similar reports from all quarters of the world. Though the form of the family, says Bowdich, would permit of kinship being traced through fathers, yet in point of fact, in Ashantee, it is traced through women only. 2 Grey tells the same of Australia : " Children of either sex always take the family name of their mother." 3 1 Mceurs des Sauvages Ameriquains, torn. i. p. 463. Paris, 1724. 2 Mission to Ashantee, pp. 234, 254. London, 1819. 3 Travels in North- Western and Western Australia, vol. ii. p. 226. 4 STUDIES IN ANCIENT HISTOR Y CHAP. The condition of the family, therefore, in which female kinship prevails is no isolated phenomenon ; if the French noblesse retained it so late, it was as a relic of a very early state of things; the gynsecocracy, as Strabo calls it, of the Basques was a natural result of it, and the Welsh law becomes intelligible at once. For if the son chose to avenge a man of his mother's kindred that is, to take up the blood-feud on the mother's side he thereby elected to remain of her kin, and accordingly had no right to succeed to his father's in- heritance. The statement of the law, that he lost his patri- mony " on account of that crime," is of course another example of a false reason given for an existing fact when the true reason had gone out of sight. The law dates from a time when male kinship was beginning to be understood, and a son might choose which side of the family he would belong to, a period intermediate between a system of female kinship only and the modern system of male kinship. But it has been tacitly assumed in all these statements that the child is not only of the mother's kin, but also is not of the father's. This involves another custom equally widely spread that is to say, that the father and mother cannot belong to the same kindred. " A man cannot marry a woman of his own family name," says Grey (loc. cit.) Thirdly, we ask how kindred is denoted, what is meant by the family name ? And this raises the question of totemism, which is everywhere found as the mode of marking the stock or blood the kin as first understood. "Each family," con- tinues Grey, " adopts some animal or vegetable as their crest or sign or kobong. ... A certain mysterious connection exists between a family and its kobong" Again : " A most remarkable law is that which obliges families connected by blood upon the female side to join for the purpose of defence and avenging crimes; and as the father marries several wives, and very often all of different families, his children are repeatedly all divided amongst themselves." * Thus the problems which particularly interested the author 1 Travels in North-Western and Western Australia, vol. ii. pp. 228, 230. i CASES FOR INVESTIGATION 5 in preparing his new work were these two. " How came there to be a law declaring it to be incest for a man to marry a woman of the same stock or blood as himself, however far removed from him she might be by degree of consanguinity, or how came marriage to be interdicted between persons of the same totem ? " And secondly, How did it come about that there are traces of a system of kinship on only the female side, or of the succession law proper to such a system, in the codes of more civilised nations ?] 1 Fortnightly Review, 1877, p. 885 CHAPTER II ON THE METHOD OF INQUIRY IN EARLY HISTORY THE common notion is that history begins with monu- ments and literary records that is, when civilisation is already far advanced. On this view some of the earliest stages of human development must, as belonging to the beginnings of civilisation, have been passed through in prehistoric times. Properly speaking, however, no time should be called prehistoric, if, by any means, we can ascertain the general character of the events that took place in it. In the earliest times of which we have record, several nations appear in different parts of the world, fairly compacted socially and politically, with arts and sciences well elaborated, and religious faiths and philo- sophies in luxuriance. Each of these nations must have had a long anterior history which is unrecorded ; for, to confine our attention to a single matter, it must have taken the men of every nation an exceedingly long time to acquire the arts of writing, architecture, and monumental sculpture, which made records pos- sible ; and we cannot assume in a scientific inquiry, nor CH. ii METHOD OF INQ UIR Y IN EARL Y HISTOR Y 7 is it by any one seriously suggested, that these arts were divinely communicated to men. Before the date of the first records, accordingly, the expejriences-o the? , "earlier generations would have been forgotten, and those records would be almost entirely occupied with! contemporaneous or very recent events. Any anti- ' quities they might contain would be only such as had become the subject of a steady tradition through being 1 set in some literary form or otherwise, and even thesej could not possibly be of very old date. The whole of the earlier life of these nations is thus left by their records nearly a complete blank, so that it would be useless to try to compile the early history of mankind from records. In these circumstances the question arises whether it is possible for us now to ascertain, in general outline at least, what the early history of mankind must have been. Analogous cases suggest that the obstacle due to the want of records, however formidable, is not in- surmountable. The direct testimony of witnesses and documents is not always the most trustworthy, as we learn in courts of law ; and in many obscure affairs we are accustomed to see the truth established to the satis- faction of all reasonable minds, independently of such testimony, by attention to the import of facts and cir- cumstances ; in short, by what is called circumstantial evidence. It is not chemistry, whose subjects can be pounded in a mortar, or reduced in the crucible, that is the most perfect of ^ the science's,' but astronomy, that owes its completeness to the use of the telescope and 8 STUDIES IN ANCIENT HISTORY CHAP. spectroscope, and to the application of sound methods of reasoning to facts ascertained by their means from great distances. ^Unrecorded time equally with un- visited space may be open to exploration.] Though it would have been impossible for the men of antiquity to have explored the history of the past, that is no conclusive proof of our inability to do so. We inherit the records of the ancients, and know much of all of them, whereas they knew little or nothing of one another. Moreover, the facts which they have handed down we see in the light of modern events and of a knowledge of the whole world and of all races of "men, and from this vantage ground we can contemplate those facts in bearings utterly unknown to contem- porary generations. It is mainly in this way that we who come after the ancients may see better than they could what took place before them, because we see all they saw and more. ) If for this reason, then, an inquiry into the earl history of man may be hopefully engaged in, the question arises, By what method is it to be conducted ? Does that wider information of ours afford us any_clue by which we may penetrate the unrecorded past ? ( Now one thing we see is, that (the forms of life among the most ancient historical nations, as shown in their earliest records, are not so " primitive " as many such forms which have been quite recently exhibited, or may be even yet observed.' Within the last four hundred years numerous sections of mankind have become known to us, not one of which had when discovered a te ii METHOD OF INQUIR Y IN EARL Y HISTOR Y 9 fragment of written history ; l in other words, they were all less advanced than the ancient nations were in the earliest historical times. The study of these barbarous peoples has disclosed a great diversity of phases of civilisation. > Some of them, like^tne TMlncopies of the Andaman Islands, led almost brutish lives ; while others, like the Mexicans and Peruvians, were com- paratively cultured, were already maturing the arts which give permanency to records, and were, in short, so far as those arts were concerned, but little behind the most distinguished of the ancients. Now it readily \ \ suggests itself, as a comparison of the varieties of life thus brought to view, that the more advanced of those recently discovered peoples had formerly been in the same condition with the others ; and on a comparison of certain of them with the ancient nations, that the latter had come through phases of development similar to those presented by the former. In short, it is sug- , gested to us, that the history of human society is that/ 1 of a development following very closely one general law^jand that the variety of forms of life of domestic and civil institution is ascribable mainly to the un- 7 equal development of the different sections of mankind. \ If this suggestion is not misleading ; if there has been a"" development of all human powers and habits according to a general law ; and if the phases of the progress arising from inequality of development which have been observed and recorded are sufficiently numerous ; 1 E.g. the Polynesians, Micronesians, and Indonesians ; Americans, Australians, Bushmen, Negritos, and Mincopies. io STUDIES IN ANCIENT HISTORY CHAP. if they are interconnected ; still more if they shade into one another by gentle gradations, then manifestly it V may be possible by an induction of the facts and by careful reasoning to draw a clear outline of the whole course of human progress, as well of that part of il which is more remote as of that which is more recent. The preface to general history may be compiled from the materials presented by ^ barbajism._ That we are entitled to use these materials for that purpose will be made more plain, if it is shown (1) how inequality of development 1 is capable of producing the existing diversity of social forms ; and (2) what kin< of positive evidence remains to us by which we ma] ? still be able so to connect form with form as to infer f the historical order in which they arose. But before ( proceeding further it may be well, with the view of ) guarding against misapprehension, to explain in a word that the expressions human progress and human de- velopment are here used in an entirely neutral sense, to denote the fact that mankind have undergone a succession of changes, without meaning to convey any judgment as to the quality of the successive changes themselves. Js[p . doubt it follows from the nature of the forces at work that, on the whole, improvement must have been the usual result of change, as it has been the invariable object of men in seeking it ; but still ^changes have not always been improvements?) This, however, is immaterial for the employment of the method of inquiry about to be explained, which, is equally applicable to ascertain any course of positive ii METHOD OF INQUIRY IN EARLY HISTORY 11 changes whether it be held to be one of growth or t^. one of decay if only a sufficient number of connected stages be disclosed to view. And there is a conveni- ence in using the word " progress " to indicate any suc- cession of changes and "progression" to express the / circumstance of experiencing successive changes, what- ever their character may be. (In this sense it is assumed in the present investigation that mankind is a progressive community, that is, a community which has been always undergoing change in its institutions, h_abits, and ideas, by virtue of the operation of its own internal forces^ This assumption will be admitted by all save those who hold that the destinies of men have not been dependent on their own powers, but have been influenced from without by revelation. But those who entertain this view will be the readiest to concede that to take account of this faith in a scientific inquiry would be to abandon the inquiry. With these ex- planations I shall proceed to consider the two points above specified. _^^ My first proposition is that inequalities of develop- ment result from the conditions under which most of the causes of development operate. They may be observed in every community which is not in a state of stagnation, and the greater the community the more remarkable the inequalities. In the human family, as a whole, they must therefore be more numerous and striking than in any separate tribe or nation, since the human family comprises all communities. The progress of a community is to some extent v, 12 STUDIES IN ANCIENT HISTORY CHAP. determined by causes which affect all its sections equally, ^o a greater extent, however, and in more important particulars, r it is promoted by causes which affect the sections unequally.^! I do not pretend here to prove this, but merely to illustrate it. The leading spirits in a community, it is apparent, act chiefly on only certain of its sections in the first instance. The men of genius, who by their discoveries add, from time to time, to human knowledge and power, and by their speculations and aspirations dignify our life ; the philo- sophers and critics, who are foremost to purify, amplify, and quicken ideas ; and the favourites of fortune, who are so circumstanced as to be immediately benefited by discoveries, and influenced by new standards of propriety, form a class by themselves in every com- munity. What is gained by the few leaders is first appreciated, taken over, and secured by those who are nearest to them in acquirement and capacity, or who are most immediately, by social connections, under their influence ; from the latter it passes similarly to a wider social circle ; and so on, through ever widening circles, till the whole society is more or less affected by it. Now its transmission from circle to circle, or from rank to rank, to the rear rank, and its adoption and preserva- tion there, are manifestly dependent on the arrange- ments for that end existing ; that is, on the general educational apparatus of the community. This appa- ratus is imperfect in 'every community ; and again, in every community it is less perfect for the lower ranks of society than for the higher; that is, the wider the ii METHOD OF INQUIRY IN EARLY HISTORY 13 social rank, the more imperfect the educational ap- paratus. And since, in the greater masses, the force of custom is more decided than in the less, while the means of diffusing new ideas are more inadequate, v the greater masses must tend to^^lmnge Jiessjs^idl^^ha^' jbh^jss/jb^ other wprds,\owing to inequality in gifts and opportunities, ana "the conditions which favour or hinder the discrimination of new ideas and methods, the different sections of every society must present inequalities of development, and the larger the society, the more numerous and remarkable must jthese be._ I We ^should not look For very different modes of life in a small group, and we should be surprised not to find them in a large group, for there, on the view I have been stating to take no other they are normal jjid-Jiecessary. ^ -i If inSb^alities~of~dev^iDpment necessarily arise in every single community from the causes I have now mentioned, it is a fortiori inevitable that they should appear among mankind as a whole. 1 For while the XX CJ ^^^^ ^^^^^^^^y ^ rationale ~of"their production is the same in the one case as in the other, it is manifest that the inequalities of gifts and opportunities to refer to nothing else must have been indefinitely more numerous for the totality of iumanv races- munity. Some of the races have had a much better geographical situation than others. They have been more favoured by climate and soil, and by those in- fluences of locality which have been described under the name of "aspects of nature." Then the cireum- STUDIES IN ANCIENT HISTORY CHAP. i stances, whatever they were, which determined the variety of human types the assortment of physical features, on the possession of which in common rests the conception of men as belonging to the same race must have exerted a similarly powerful influence on the rate, and perhaps in some respects the character, of the progress of the races exposed Jbojtti. -_____ It is involved in the conception of a community as progressive that all its sections are moving though not pari passu ; and it results from the nature of the influences that mainly determine the progress, am from the general character of the restraints on thei: action which determine the inequalities, notwithstand- ing that many of those restraints must in every com- munity have been what we call accidental, that al] the sections are moving ir^ the same general direction. If, then, we imagine a progressive community to have been isolated from its origin, and to have develope< entirely by virtue of its internal forces, we may con- ceive a time when the course of development and the >, . ******** inequalities commenced, and a condition that was till then, speaking broadly, common to all its members. The majority would remain in this condition, while a favourably situated minority, taking the first step of r progress, left them behind. r ln proceeding, however, the minority could not but retain in their new life marks and traces of the old., The same thing would take place throughout in every advance of a section, down to the lowest and least favourably situated, as ii METHOD OF INQ UIR Y IN EARL Y HISTOR Y 1 5 \ \ the progress went on through its various stages, and I \ included section after section in its scope. 1 It will be admitted to be conceivable that [long 1 after the commencement of the progress supposing the - community ^o be very-4afger-and composed of many sections^-the inequalities of various sorts observ- able in the sections should be so numerous^ and the stages of the development so linked together by the f " *' **'*- ** traces of preceding changes forming c a continuous chain from the highest to the lowest, as that the series should afford a disclosure ^ at once of almost the initial condi- tion of the community and of the whole subsequent progress of its most advanced section. r Now the method of investigation I adopt is founded on the assumption that mankind presents us with the case of a communit in the situation -xoneeivable._/ The lowest forms of life of which we have accounts enable us confidently to judge of the primitive state of man, and a study of the highest forms yields evidence that they have grown up from the lowest. fThe stages of the growth, moreover, can be made surprisingly clear. The species has been : unequally developed that almost every phase of pro- gress may be studied as a thing somewhere observed and recorded ; and the philosopher, fenced from mis- take by the interconnection of the stages and by their 1 I am contemplating evolutions that are in their nature exceedingly slow ; such as are accomplished only in great periods of time. Every phase, I suppose, would last long enough to stamp an enduring impress on the forms of life it affected, and of this influence at least traces would long remain in these forms as modified in subsequent stages. I/ 16 STUDIES IN ANCIENT HISTORY CHAP. shading into one another through gentle gradations, may draw a clear and decided general outline of the whole course ~of human history. It will be made apparent hereafter that it is not merely the unrecorded portion of history, but the recorded also that must have the light of primitive facts thrown upon it to make it - intelligible. Granting the inequalities of development to be as numerous as I allege, and the stages of development to toe interco^jiejctejd^jih_e__^[uestion arj^e^/How are we ceed in order to ascertain^thVseries of stages ? It follows from all I have said that |the first thing to be done is to inform ourselves of the facts relating to the least developed races 1 r To begin with them is begin with history at the farthest-back point of tim to which, x except by argument and inference, we can reach. JTheir condition, as it may to-day be observed, is truly the most ancient condition of man^J It is the lowest and simplest, and, as I have elsewhere observed, in the science of history old means not old in chrono- logy but in structure. That is most^ ancient which lies nearest the beginning of human progress considered as a development, and that is most modern which is farthest removed from it. 1 1 In Colonel Lane Fox's admirable collection, illustrative of the development of the boomerang, no one could hesitate as to where the series of stages began, nor, indeed, as to the succession of stages between a common stick and the boomerang. Again, a stone picked off the ground at random for the purpose of throwing, and a stone prepared in the best shape for slinging are at immediately recognisable extremes of progression. The sling and slinging stone, however, compared with the Henry rifle and Boxer-Henry cartridge are as manifestly at opposite ii METHOD OF INQUIRY IN EARLY HISTORY 17 -, \ Having acquainted ourselves with the more primi- tive forms of life, we must proceed to compare them \ with the more advanced, and to study their inter- connections with a view to 'their classification. We shall find that our power of arranging in their order the stages of human advancement depends as a rule on the interconnections, and thatl;the commonest way in which the lower prolongs itself into the higher is of the | nature of a symbolism^ The persistence of the lower in the higher, however, takes place also in a variety of other ways, as, for example, when some law^ institu- tion, or custom being displaced by a new one, there is yet left remaining in the latter some unmistakable 1 ' secondary feature of . the^J^rjner.,. Valuable assist- ance is, moreover, occasionally obtained in the work of classification from < general considerations which co-operate with other evidence to support a judgment that one form of life is essentially ruder than another ; and sometimes the most useful of all aids is available, r cases being found which illustrate the phases of transi- tion from one stage to another in such a way as to exclude doubt regarding their mutual relations. A power of distinguishing the comparative degrees of rudeness of forms of life, apart from any induction and minute analysis of structure, seems to be implied at the outset in selecting the races to be in the first instance studied. There may exist some tests of degrees extremes of a progression. We shall hereafter see many social states that are as readily capable of being discriminated as being the one higher and the other lower, as any phases of mere mechanical developments. C 1 8 STUDIES IN ANCIENT HISTORY CHAP. of rudeness which would command general assent, but it would be most hazardous to rely upon them if unsupported by a certain amount of other evidence. Should we find anywhere tribes of men herding together without clothing, almost without tools or weapons, using nature-provided shelters, without laws, government, or religion, improvident and trusting for subsistence to fruits, roots, and shell-fish, we may believe that no condition could well be at once human and lower. There are not a few tribes that have been said to answer pretty closely to this description, but, truth to tell, the accounts are not very trustworthy, nor, except as indications of the condition from which men may have emerged, could such ca'ses be instructive r in a scientific view, even were they established. 1 f When I speak of the least advanced races as falling to be studied in the first instance, I mean all the least advanced races indiscriminately ; and in selecting from these the type of group to start from in any investigation one may be making, any experience is that 'the inquirer must be guided by the circumstance that the group presents to him forms of life which, in casting about for origins, he has become trained to perceive to be 1 I know of no groups of men from evidence that can be counted trustworthy that are not in the strictest sense groups of men gifted with speech and speculative, in the sense of being metaphysical ; that is, in my opinion we have no accounts of man as a mere brute to be handed over as a proper subject of study to the naturalist. "On the other hand, it will be found that we have the means of historically tracing back the condition of man, by a series of irresistible logical inferences from well-established facts, to a stage where his earlier history belongs not to the historian, but to the natural historian. ii METHOD OF INQ UIR Y IN EARL Y HISTOR Y 19 germal. To test his impression that they are of this\ nature he must compare them with kindred forms, studying their differences and correspondences till he learns whether or no the differences are of the character of growths from, or modifications of, a germal form. Having lighted upon such a form, he must trace its subsequent modifications from point to point. Very frequently this is not a difficult process, as r one stage grows so obviously out of another as to suggest what preceded it a suggestion which, it is manifest, no mind could receive except as having already a know- ledge of the main features of the earlier stage. Indeed, not a few of the most instructive germal forms have been actually lighted upon under the guidance of such a suggestion ; and it is only what should be expected that in inquiring into a subject so perplexed, and the facts respecting which have only so recently begun to be systematically collected, the selection of the start- ing-point is, as a matter of necessity, determined by a judgment on the whole facts brought before the mind during years of research. The importance of a minute study of the whole body of facts gleaned from observation of the ruder races cannot be exaggerated. These facts are at once the materials from which the earlier chapters of general history must be compiled, and an essential requisite in rendering intelligible many events recorded in written histories; that is, they are of primary import- ance to general history throughout. It is not too much 20 STUDIES IN ANCIENT HISTORY CHAP. to say that v there is no phase even of our modern civilised life but has some feature which, when pointed out, will appear puzzling and strange to one ignorant of primitive facts, but which becomes suddenly significant .when these facts are explained. It is chiefly to a want of knowledge of these that we must ascribe the failure of what pretends to be "comparative jurisprudence" to explain ancient institutions, as well as the failure of philology to penetrate any depth into the past. We must also ascribe to the want of this knowledge the errors of philologists and others in handling the ancient mythologies, the worship by the ancients of animals and plants, and the gods compounded of bestial and human forms that appear in systems of faith in course of being worked over into a consistent anthropomor- phism; to mention no other failures. 1 To the same also must be attributed frequent misapprehensions of the order of succession, as well as of the meaning, of events occurring within the period of records a species of misapprehension so common as to show that written history may be not only meaningless but delusive, when read apart from a knowledge of primitive modes of life 1 I have dealt with these in another place, and hope to recur to them hereafter. Here I wish to point out how things may be recorded and the record mean nothing, be unintelligible, how, in short, to infuse a meaning into much of ancient history, it must be approached after forming the preface to it. With regard to the failure of philology in ethnology as well as history, it is interesting to recall its high preten- sions : " The evidence of language is irrefragable, and it is the only evidence worth listening to with regard to ante-historical periods." Professor M. Miiller, " Last Results of Sanskrit Researches," quoted in Muir's Sanskrit Texts, part ii. p. 309. ii METHOD OF INQUIRY IN EARLY HISTORY 21 and their classifications. Of this kind of misapprehen- sion I have elsewhere given examples. They might be given in any number. The student who would exercise his mind in seeking them for himself could not do better than read Ortolan's History of Roman Law, or the Ancient Law of Sir Henry Maine. Nor need I here show, what I have elsewhere shown, liow necessary a knowledge of primitive life is, as a preparation of the mind to profit by the suggestions which symbolic forms afford as to the connection of stages, and therefore as a preparation for the work of classification generally, since a symbol is the commonest trace of itself which the lower form of life leaves in the higher that supersedes it. The symbolism is probably due to reverence for the past ; it is due at least to r our incapacity whatever may be its cause to drop all at once any mode of proceeding which has long been customary. This incapacity the student of history may be allowed to think fortunate, since but for it there could be no high degree of certainty attained in penetrating the secrets of unrecorded time. The inequalities of development, into however perfect a series they might fall, could not per se impress the mind as they do when by means . of this symbolisinstage isjiinked to stage. As it is,__ we are aMe to trace everywhere, frequently under striking disguises, in. the higher layers of civilisation, the rude modes of life with which the examination of the lower makes us familiar, and are thus made sure of the one being a growth from the other. 22 STUDIES IN ANCIENT HISTORY CHAP. The causes which produce the symbolism are j active now as ever they were ; symbols are every da arising and attesting the continuity of human affairs. Feudalism is practically dead in Scotland, but the Scotch law still teems with feudal ideas and forms, and should the ideas die and be forgotten, many of the forms or traces of them will certainly long remain. The instinct or incapacity which secures their pre- servation knows no bounds in its application, and may be illustrated as easily from the sewage system- of a country as from its jurisprudence. By means i of these symbols which replace extinct practices we are able to trace out old customs in districts in which they have gone into disuse." I shall cite two illustrative instances. It is said to be a custom in Borneo in building a house that holes are made to receive the posts, and men are killed and placed in the holes, " so that the house being founded in blood may stand." According to the Eev. Mr. Taylor, a similar custom once prevailed in New Zealand. It is now extinct, but the ancient practice is symbolised on the building. The wall-plate of the verandah is carved to represent the prostrate figures of the victims on whose bodies stand the pillars that support the house. 1 We get a different case from Captain Cook in his account of human sacrifices, in his time, in Otaheite. "It were much to be wished," he says, "that this deluded people may learn to entertain the same horror of murdering their fellow -creatures in 1 Te-Ika A-Maui, p. 388. By the Kev. R. Taylor. 1855. * ii METHOD OF INQ UIR Y IN EARL Y HISTOR Y 23 order to furnish an invisible banquet to the god (the sacrificed were buried by the altar, and the notion was that the god fed on their souls), as they now have of feeding corporeally on human flesh themselves. And yet we have good reason to believe there was a time when they were cannibals. We are told (and, indeed, partly saw it) that it is a necessary ceremony when a poor wretch is sacrificed, for the priest to take out the left eye. This he presents to the king, holding it to his mouth, which he desires him to open ; but instead of putting it in, he immediately withdraws it. This they call ' eating the man 7 or ' food for the chief/ and perhaps we may observe here some traces of former times, when the dead body was really feasted on." 1 Knowing that cannibalism was a practice of some of the congeners of the Otaheitians, we cannot doubt the correctness of the inference that the practice of cannibalism was here symbolised. In cases like these the presence of the symbol, if supported by collateral evidence, the requisite amount of which must vary from case to case, would enable us to arrive at a knowledge of a precedent practice in a super- seded order of things, even had we no direct evidence of such an order having existed among the people possess- ing the symbol. For since numerous cases exist in which ^ there is direct evidence of symbols arising out of pre- C vious practices, we are fairly entitled to assume, on finding a symbol in use among a people, that there was in an /A earlier period a reality corresponding to it, of which i/o 1 A Voyage to the Pacific Ocean, vol. ii. p. 44. London, 1874. 24 STUDIES IN ANCIENT HISTORY CHAP. remains as the shadow, especially if we know of such a reality prevailing elsewhere. Usually it will suffice as collateral evidence that we know a practice to be com- mon of which the symbol is the natural relic. This condition I have shown, in Primitive Marriage? satisfied in regard to the Eoman marriage by coemptio. We seem to be justified in inferring that a section at least of the Koman people had had experience of the system of procuring wives by actual sale and purchase ; that is, had been, in regard to marriage, as rude as races we esteem barbaric. So also the Libripens in Eome, officiating with his scales at the making of a will or act of adoption, seems to illustrate the source whence all ideas of formal dispositions were derived ; while the formalities in the Legis Actio Sacramenti may be con- sidered to prove that the citizens of Eome were anciently ignorant of legal proceedings and dependent for a settle- ment of their disputes on the force of arms or the good offices of neutral persons interfering as arbiters. Enough, I hope, has now been said to convey a general idea of the mode of proceeding I propose to adopt. The method is not new, though perhaps the I rigid application of it to history has never been ' attempted on any great scale. It is as ..old at least as Thucydides, who has recourse to it in the introduc- tion to his history, in which occasion arises for consider- ing the ancient condition of Greece. The following passage shows that this historian appreciated the instructive significance of inequalities of development. 1 [Studies in Ancient History, first series, p. 6.] ii METHOD OF INQUIR Y IN EARL Y HISTOR Y 25 " In the old times/' he says, " on navigation becoming common, the Greeks as well as the barbarians on the sea-board and in the islands turned pirates under their leading men as captains, some of them seeking gain, others a maintenance ; they fell upon open and unfor- tified towns, on the plunder of which they mainly subsisted ; a degree of glory rather than disgrace attaching itself to such conduct. The correctness of this statement appears from some dwellers on the continent still regarding acts of piracy cleverly per- formed as honourable ; also from the ancient poets representing men as asking navigators who reached their shores whether they were pirates, in a way imply- ing that the employment was one the questioned were not ashamed of, or their interrogators disposed to reproach them with. A system of mutual plunder also prevailed on the continent, and the Greeks to this day in many districts continue the system ; as the Locri Ozolse, the JEtolians and Acarnanians, and others in that part of the continent, among whom also the fashion of wearing arms continues from their persisting in piracy. The whole of Greece, indeed, used to wear arms, owing to the insecurity of their dwellings and communica- tions ; the people, like the barbarians, even wore their weapons when engaged in their ordinary avocations ; and we have a proof of the universality of this custom in Greece at one time in its still surviving in certain parts." This interesting passage is a fair example of the application of the method, for the tradition of the former \ 1 Thucydides, book i. 2. 5. 26 STUDIES IN ANCIENT HISTOR Y CHAP, n order of things is sustained by a reference to persist- ing instances of it on the one hand, and by an inference from the tone of the ancient poets on the other; the causes of the persistence at points being specified, and their ceasing to operate at other points being assumed in explanation of the disappearance on the whole of the ancient barbarities. Were this the place for doing so, numerous instances might be cited of the survival to this day of the order of things Thucydides described as having before his time prevailed in Greece. 1 If within the scope of my inquiry I can connect the ancient civilisations with forms of civilisation existing among barbaric races of which we have trustworthy accounts ; and the latter with lower forms prevailing among neighbouring races more barbaric than them- selves ; and these again with still lower forms in use among peoples counted by common consent to be savages ; and if I succeed in showing the connections throughout the series of connected stages to be con- fections of growth and development the higher to \ have in every case passed through the lower in be- J^ coming itself then I shall be entitled to conclude (1) that the doctrine of development is established as applicable to human society, and (2) that in the series of stages investigated I have, so far as my subject is ^concerned, the Preface to General History. 1 One notable case appears among the tribes round Munniepore as described by M'Culloch. (Records of the Government of India. For. Dep. No. XXVII.) i CHAPTEK III THE MODE OF HANDLING EVIDENCE IN making an induction of facts, the first essential is to ascertain that the facts we use rest on good evidence. A false fact is much more mischievous than a false theory, for it eludes detection more easily, and it not only vitiates the argument it is employed in, but often stands in the way of reaching a better. A false argu- ment, says Mr. Darwin, is sure to be taken to pieces in the course of subsequent investigation, but a false fact often perplexes speculation for ages without there being any means of getting rid of it. Facts become of import- ance exactly in proportion as the system they are used in is a reasoned system. A single erroneous fact may then prove fatal to a whole series of logical inferences, and no ingenuity can save the author from error. But for the fact of the earth going round the sun Ptolemy's ingenuity had dispensed with Newton. jjtn~an investi- gation like the present, the accuracy of the individual i \) data is a matter of even greater consequence than usual, / on account of the cumulative nature of the argument by | 28 STUDIES IN ANCIENT HISTORY CHAP. which our conclusions are reached. The whole argument is built on circumstantial evidence, and in circumstantial evidence, if a single link is unsound, the whole chain ives way. "J For example, a full explanation of the origin of exogamy requires it to be made out that wherever exogamy prevailed, totemism prevailed ; that where totemism prevailed, blood -feuds prevailed ; that where blood-feuds prevailed, the religious obligation of vengeance prevailed ; that where the religious obliga- tion of vengeance prevailed, female infanticide pre- vailed ; that where female infanticide prevailed, female kinship prevailed. A failure to make good any one of tie&e--particulars would be fatal to the entire argument. tit belongs to the nature of such an^argument that it cannot be conclusively established while even the most inconsiderable of the parts that compose it is insuf- ficiently supported, and the issue may often turn on the truth or falsehood of a single fact. Philo, for instance, says that in Lacedsemon a man might marry his sister by the same mother but not by the same father, whereas we know that at Athens and in most other places all over the world at the same period, a man might marry his sister by the same father but not his sister by the same mother. Is Philo's statement true ? We have no means of knowing ; but it is manifest how the statement, if true, would affect specu- lation. Again, the singular phenomenon of the couvade used to be explained on the principles of agnation, but Mr. ; Brett, in his book on the Arawaks, has pointed out the couvade in coexistence with kinship through in THE MODE OF HANDLING EVIDENCE 29 women only. 1 In mentioning the circumstance, Mr. Brett had in his mind no thought of the origin of the couvade, but if his statement is true and there is no reason to doubt it it upsets all the speculations on the origin of the couvade which had preceded it. In this case, however, unlike that of Philo, we have some means of judging of its accuracy, for the result of all recent research has been to show that agnation rarely got itself established anywhere, and the couvade is very general in cases where we do not know the form of kinship. 2 Now, in an inquiry where single facts may carry so much weight, it is indispensable to make sure, before accepting them, that they are well authenticated. It is not enough to find a thing stated, and the statement ever so often repeated. Account must be taken of the authority of the first narrator, and inquiry made as to whether the subsequent statements are not of the nature of mere echoes. The original authority must be weighed by considering what opportunities he possessed for cor- rect observation, how far he had capacity and willing- ness to make good use of his opportunities, and to what extent, if at all, he was disposed to mix up with his statements of fact any element of speculation or 1 [The Indian Tribes of Guiana. London, 1868, pp. 98, 101.] 2 [It may be as well to observe that the statement in the text must not be taken to mean that Brett's statement offers any difficulty to the theory that the couvade was a sign of transition from female to male kin- ship ; agnation, of course, means a great deal more than the recognition of fatherhood. The couvade is now generally admitted to have been a method of establishing paternity and acquiring some rights over the child ; it follows that it must have begun in a state of things where female kinship was still the rule.] 30 STUDIES IN ANCIENT HISTORY opinion of his own. In short, as far as possible, th< witness must be cross-examined respecting the evidenc< he tenders, at least as closely as he would be in a coui of justice, so that the true value of his evidence may ascertained. For we must keep in view not merely thai all men are apt to be untruthful, but that very man] of the witnesses we shall have to deal with have what astronomers call personal errors, for which reckoning must be made, not to mention the rarity of the gifts of correct observation and of correct expression and literary form. f The data for our investigation are supplied to us from sources of various kinds.) For tl^e ancient world, our chief authorities will be its own literatures, i1 sacred books, like the Bible and the Vedas, its codes law, its epic and dramatic poems, its philosophic; treatises, which contain constant references to past an< existing conditions of society, its geographical and his torical works, written sometimes by men who wei themselves also travellers, such as Herodotus, Pau- sanias, and Strabo. ("This class of evidence is in no wise limited to the more prominent nations of antiquity, like the Chinese, the Hindoos, or the Greeks, but it is available even for such peoples as the Finns and the Peruvians, and in large quantity for the Scandinavians\ What the Homeric and Hesiodic poems are to us for the Greeks, what the Zendavesta and the Shah Nameh of Firdusi are to us for the Persians, what the Vedic literature and other ancient books are to us for the Hindoos, what the Chinese sacred books are to us for in THE MODE OF HANDLING EVIDENCE 31 the Chinese, that the Kalevala is to us for the Finns, the Popol-vuh for some of the Americans, the Nibelungen for some of the Germans, and the Sagas of the North for the Scandinavian races generally. They bring vividly before us ancient institutions and manners of thought, feeling, and action, and present them in such a way as to leave no doubt that we have evidence for them above all suspicion. Another source of evidence for the nations of antiquity lies in the monuments and inscriptions belonging to them which still remain to us, and whose record has at last, in the cases of Egypt and Assyria in particular, become intelligible in recent times. The evidence as to the past which is afforded by the collections of folklore and popular tales so frequently made in the present day such as Grimm's Tales, Grey's Polynesian Mythology, Schoolcraft's Algic Researches, and Campbell's Tales of the West High- lands is certainly valuable, but perhaps not so trustworthy as that derived from the sources already mentioned. As regards those countries which have become known to us only within the last few hundred years, and which taken together constitute more than three - fourths of the whole inhabited world, the evidence we possess is furnished mainly by soldiers of fortune, traders, and missionaries, who have lived or travelled in those parts, and have put on record the results of their observations. Owing to many different causes, the immense record which has been thus obtained is very imperfect. We find the earliest accounts exceedingly 32 STUDIES IN ANCIENT HISTORY CHAP. uninformed, because, on the one hand, the observers knew not what to look for, and, on the other, they were only too anxious to excuse their own rapacity or cruelty by depicting the tribes they conquered as mere brute beasts whom it were charity to sweep off the face of the earth. Nothing is more common in these old narratives than to find the peoples who were being sacrificed to European cupidity described as living in purely animal state, without government, laws, or r< ligion, and yet the student will sometimes be able spell out from these very narratives themselves that tl peoples so described were intensely religious, and thi ; they dwelt under the constant pressure of a rigid bod] of customary law, and what we would call a highl; developed system of constitutional government. This disposition of the first explorers to depreciate th< natives of the countries they invaded has often bee] made the subject of comment. Dr. James, for example in his interesting narrative of Tanner's captivity amon^ the North American Indians, says (p. 12): " In tl writings of the early historians, particularly of th< Puritanical divines of New England, we find the* people commonly described as a brutal and devil-drw race, wild beasts, bloodhounds, heathen demons. N< epithet was considered too opprobrious, no execration too dire, to be pronounced against them." Even Acosta, a man of much sense and learning, discovers endless proofs of the all - pervading influence of the devil in every place he visited, and even among the Peruvians, the most advanced of the American races, he finds no in THE MODE OF HANDLING EVIDENCE 33 religion, but much idolatry everywhere. (The narratives of the early explorers, therefore, are on the whole of little value to us,) since they are pervaded by a spirit of bitter hostility to the peoples they describe, and by a love of detraction entirely unrestrained by any percep- tion of philosophic truth. It is impossible to take as trustworthy authorities men like the Jesuits of West Africa, who, exercising the power of the sword over the natives, baptized them by the thousand with a hose, and destroying their idols wherever they found them, erected crosses and cheap madonnas on their vacant sites. They had a strong interest to vindicate the fury of their religious zeal by blackening the character and condition of the tribes on whom they poured it. Another circumstance which in some cases diminishes the value of the earlier evidence we possess, is that the manners and religious feelings of the peoples it relates to had been considerably corrupted by the influence of European traders before any writers undertook to describe them. An example of this is furnished by the North American Indians. Our first knowledge of these tribes is supplied by the Jesuit fathers, but English, Dutch, and French merchants had been already nearly a hundred years at work among them before the Jesuit fathers began to have any understanding of their social life and customs. Their trade was mainly a barter of European products for fur, and it is impossible but that any traditional religious regard for animal life which the natives may have possessed should have largely broken down under the prolonged influence of a trade D 34 STUDIES IN ANCIENT HISTORY CHAP. in which they could procure for the skins of animals, not only whisky, rum, and brandy, but the still more tempting commodities of powder, shot, and muskets, which had come to be indispensable necessaries for all the tribes in the state of mutual war which was their ordinary existence. ( A corruption of manners, and a decay of many primitive feelings and customs, must in this way have taken place among these Indian tribes in the long interval between the time of their first dis- covery and the time when their modes of life W T < made, /as they at last were, the subject of intelligei and highly interesting study on the part of the fathei It is happily still possible, however, in some cases, e.g. in that of the Iroquois, to supply the informatioi which has been thus lost, perhaps for ever, regarding their religious feelings, by inference from accounts kindred tribes who have always remained outsidi European influences, and have become the objects oi observation since explorers have learnt what to seek for. A further cause of much loss of valuable evidence on our present subject is one which has operated from the very first, and which operates still the expurgation ad usum gregis of such missionary reports as were from time to time made to the ecclesiastical authorities. The college of the Jesuits supervised the Relations des Jesuites, so as to strip them of such contents as were esteemed likely to be prejudicial to the cause of European civilisation and religion, and even works like those of Charlevoix, which are more of the nature of compilations than works of personal observation, were in THE MODE OF HANDLING EVIDENCE 35 treated in a similar manner. The same thing is true of the reports made by missionaries of Protestant churches, a notable example being Williams's Fiji and ike Fijians, which is a small work as we have it, but the materials for which were furnished to the editor in enormous MSS., the labour of long years, and then boiled down to make them suitable for ecclesiastical purposes. We learn from the published text how frequently accounts of native practices or beliefs, as prepared by the author, were omitted by the editor as being disagreeable and disgusting. In this respect, indeed, ordinary books of travel are not one whit better than ' missionary records. They are, as a rule, publishers' ventures, scrupulously dressed for the drawing-room table, and very rarely put into final literary form by the travellers whose name they bear as authors. It is impossible to overestimate the amount of evidence which has been lost in this way for the purposes of science. In one case known to me, mention of the system of Thibetan polyandry as the marriage system of a people was omitted in a work of travel, because the editor thought no notice should be taken of so disagreeable a subject. Other causes sometimes lead to a suppression of evidence. Pausanias and Herodotus, for example, often tell us that they knew things regarding the religious and social customs of the peoples they describe, which they could not relate because they had received them from the priests under seal of secrecy. And it is very common for writers to withhold a fact or to describe it erroneously through opmionativeness and through 36 STUDIES IN ANCIENT HISTORY CHAP. ignorance of the meaning of terms. Sometimes the candour of the writer leads him at some stage to disclose that he has substituted his own opinion for the informa- tion he had received. In one account of the natives of Australia the author states that the people he was describing had eight totems, but in a later passage of the same book he owns that he had said eight because in his view they could not do with fewer, but that they themselves had said they had only four. It is very frequently stated in certain works of travel in Africa, that the family system of the people was patriarchal, while the details of it given by the writers themselvc show that in reality the kinship that prevailed wj kinship through women only, and the father was comparative nonentity. Of late years, and especially since the publication of speculations on the history oi society, the confusion has become even worse con- founded. Tribes are represented as being at once exogamous and endogamous, and the marriage system to be monogamy, polygamy, and polyandry all in combination. It is for this reason that I propose in the following work to cite no book of travels or missionary report published within the last twelve years. They are absolutely untrustworthy as sources of evidence, because they have been written under the influence of prevailing theories. Works written before that date see no reason for treating with suspicion. They are for the most part furnished by missionaries who were quite ignorant of the scientific significance of the facts they supplied. Their authors often mention, in describ- in THE MODE OF HANDLING EVIDENCE 37 ing particular forms, that they have no knowledge of anything like them existing elsewhere ; they treat them as purely local customs ; and usually their sole motive for putting them on record is that they seem strange and curious. They have rarely any explanation to suggest of their meaning and origin, and they are in general entirely free from the bias of any theory or hypothesis. It may be laborious to have to pick up the facts here and there in separate books of travel, but there is certainly a countervailing advantage in the circumstance that being presented to us as single phenomena, un- connected by any shade of theory, they may be accepted as unsophisticated statements of reality. On the whole, it is a most fortunate thing that the evidence was so long accumulating before any one thought of making the facts in question the subject of speculation. I propose to select one or two witnesses in each case whoseem, from their length of residence among the people they describe, to have had ample opportunity of becoming acquainted with the facts of the case, and to spell out from their narratives, as a whole, the import of what they have to communicate, when they have not themselves made their communications very distinct. I shall inform the reader in bibliographical and bio- graphical notes how the case of those writers stands in respect to their opportunities for observation and their skill in observing, and I shall confirm or modify their views by reference to the other authorities we have on the same subject. For, as Eobertson says, "When obliged to have recourse to the superficial remarks of 38 STUDIES IN ANCIENT HISTORY CHAP. vulgar travellers, of sailors, traders, buccaneers, and missionaries, we must often pause, and comparing detached facts, endeavour to discover what they wanted sagacity to observe." Besides these sources of evidence, there are various aids to knowledge of which it is proper to say some- thing here. There is, in the first place, a great variety of collections of books of travel, such as those of Pinkerton and De Bry. So far as these books are unabridged in the collection they require no special characterisation ; when they are abridged I shall of course not cite them. Then there are various com- pendious works founded on the accounts given by books of travel, such as Dapper's Africa, and Walckenaer's Collection des Relations de Voyage en differentes Parties de I'Afrique (Paris, 1842), in which a work is sometimes given at full length, but more frequently in abstract or abridgment. Works of this sort are mainly valuable to the student as catalogues of books, simplifying his researches by enabling him more swiftly to determine which books he must read and which he may overlook. I shall have occasion to use Walckenaer in some cases where I could not get access to the works he abstracted from, but I have tested the accuracy of his abstracts in cases where I possessed the books he abridged, and I can confidently say that in my opinion his work is an exceedingly well-executed one. Then there are compilations of other kinds. There are the ancient lexicons of Suidas and Festus, which must be used with caution, yet on many points they are our in THE MODE OF HANDLING EVIDENCE 39 only authorities. Of modern compilations, one of the best known is Schoolcraft's History, Condition, and Prospects of Indian Tribes, published by authority of Congress in six huge quarto volumes. Writing some twenty years after Dr. Galletin, who was a most careful compiler, and who drew special attention to the law of exogamy, Schoolcraft does not make a single allusion to the subject in the course of his six volumes, except in a single footnote towards the end of the fifth. Omissions like that are characteristic of this book, which is a singularly unsatisfactory one. I entirely agree with the judgment formed of it by one of the author's own countrymen. Mr. Francis Parkman says, in his Jesuits in North America in the Seventeenth Century (Boston, 1875) : " Of recent American writers no one has given so much attention to the subject as Mr. Schoolcraft ; but in view of his opportunities and his zeal, his results are most unsatisfactory. The work in six large quarto volumes, History, Condition, and Prospects of Indian Tribes, published by Government under his editorship, includes the substance of most of his previous writings. It is a singularly crude and illiterate production, stuffed with blunders and con- tradictions, giving evidence on every page of a striking unfitness either for historical or philosophical inquiry, and taxing to the utmost the patience of those who would extract what is valuable in it from its oceans of pedantic verbiage." 1 A work of a similar character is that of Mr. Brough Smyth on the Aborigines of 1 Parkman, Jesuits in North America, p. 80. Boston, 1875. 40 STUDIES IN ANCIENT HISTORY CHAP. Australia, issued under authority of the Government of Victoria. With regard to the statements of this work concerning flints, tools, and weapons, I express no opinion, but I am perfectly satisfied, after careful study, that it is absolutely misinformed respecting the structure of Australian society. The authorities ar< unintelligent, and often give out as facts what are only their own speculations, or their own modes of reconciling the facts with the speculations of other persons, like Mr. Lewis H. Morgan, when they have never actually seen the things they describe, but know of them only by letters received from clergymen on the subject. There is one compilation of this order, however, to which I shall occasionally" refer, and which strikes me as on the whole a valuable and carefully executed work, viz. Bancroft's Native Races of the Pacific States of North America. The passages cited from the authorities are often of such length as to enable the student to judge of the matter in hand as well as if he had the book before him. As an example of this, I may point to the account given of Muller's ingenious hypothesis of the development of the Mexican war-god. I shall have occasion to make use of this work without having been able in every instance to verify the information it supplies by reference to the original authorities from which that information has been taken. I had hoped to be able to carry on throughout the plan of personally sifting all the evidence, but circumstances have lately made it apparent that this would be impossible for me now. in THE MODE OF HANDLING EVIDENCE 41 I regret exceedingly being obliged to depart from this course, especially as in several cases where I had access to the books cited by Mr. Bancroft as authorities, or had either knowledge of their authors or pretty full bibliographical information about them, it has appeared to me that the value of the authorities has not been justly estimated by Mr. Bancroft's compilers. Of course I shall not adduce from his work what for any reason I count bad evidence ; and where I cite his work at all without having verified its information, I shall draw the reader's attention to the second-hand character of the evidence. Lastly, we have various compendious works in which more or fewer of the facts are gathered by the labour of one man and strung together by some thread of theory or principle of description permitting of some degree of orderly arrangement. Such books, of which Waitz's Anthropologie is a good example, may be of value to the student as catalogues of references to guide his reading, but they are almost uniformly compilations at second hand, and should never be trusted for a fact. Mr. Lewis H. Morgan's Systems of Consanguinity, published under the auspices of a great institution, is a work of much apparent importance, but I shall hereafter show, on evidence putting the matter beyond all dispute, that the classificatory system set forth in that book is not a system of consanguinity at all, but a system of modes of salutation. It is almost incredible how the author should have fallen into such a mistake, considering that he borrowed from Lafitau (without 42 STUDIES IN ANCIENT HISTORY CHAP, in acknowledgment) his account of the Iroquois, from which he started his inquiry, and that Lafitau distinctly mentions that among the Indians the expression " my brother," "my father," "my child," were conventional modes of salutation severally addressed to people in proportion to the distinction in which they were held. They indicated not degrees of relationship, but simply degrees of respect. 1 1 " The savages, however, commonly do not like to hear themselves called by their name, and to ask them what it is, is a sort of affront which would make them blush. In speaking to each other they all use the names of relationship, brother, sister, nephew, uncle, observing exactly the degrees of subordination and all the proportions of age, unless there be a real relationship by blood or adoption, for then a child will find himself sometimes the grandfather of those who, according to the order of nature, might easily be his (grandfather). They practise the same civility with regard to strangers, to whom they give, in speaking to them, names of consanguinity, as if there was a real tie of blood, nearer or more distant, in proportion to the honour that they wish to do them, a custom which Nicolas of Damas relates also of ancient peoples of Scythia " (Lafitau, vol. i. pp. 75, 76). CHAPTER IV DEFINITION OF TERMS IT will be convenient in entering on an inquiry into the earlier history of one great branch of marria,ge law to define with some precision what marriage is ; to point out its principal varieties, and the tests available for determining, in any case, whether what, in common parlance, is called marriage is properly so called. Let us put on one side that strict view which re- presents marriage as a sacrament establishing an indis- soluble union of a man and woman as husband and wife ; and on the other hand, that lax view which has permitted the word " marriage for a month," to obtain currency in certain seaports in the East ; and let us see if it be possible to find between these some middle term where all the essential ideas of wedlock shall be found united. 1 1 It is hardly necessary to say that the term marriage is absolutely inapplicable to the consortships of individuals of different sexes of any other species than man. To speak of a gorilla and his " wives " is a mere literary license ; and although much light may hereafter be obtain- able from a study of such consortships in what is called the brute creation, in elucidating the variety of marriage systems among men nothing but 44 STUDIES IN ANCIENT HISTORY CHAP. Marriage, then, as a general term, denotes the union of a man and woman in the legal relation of husband and wife, as the same may be defined by the local law or customs. Where there is no law on the subject no custom of the country establishing any requirements as to the constitution of the relation of man and wife, its endurance, and the rights and obligations it confers and infers there is neither marrying nor giving in marriage. When we study the various marriage laws and customs of the world, we find that no two systems are exactly agreed. They differ more or less as to who may marry at all ; as to whom a man may marry when free to marry ; as to the manner of constituting marriage where it is permissible ; and as to its effects the mutual rights and obligations of husbands and wives. Lastly, religion almost everywhere appears consecrating the local usage as to marriage, whatever it may be. The formula all the world over nearly has come to be that marriage is a divine institution. Even marriages by capture are, in the Bible and Koran, made the subjects of divine revelation and regulation. (But in this diversity of laws and customs, it is possible to perceive certain main points of agreement, and it is at these points we must look for the essentials of marriage.) Otherwise, and to begin, the essentials may be separated from the unessentials thus : where confusion could arise from calling them, or thinking of them, as marriages. This will presently appear so clearly that I need not here say more the subject. iv DEFINITION OF TERMS 45 there is an undoubted marriage system, any feature of it not present in another undoubted marriage system is not an essential of marriage. The noble Roman Jurisprudence, which did more than all the religions put together to improve and beautify human life, has given us the idea of marriage as the union of one man and woman in a consortship for the whole of life an " inseparable consuetude " of life between husband and spouse, with interests the same in all things civil and religious. That idea, despite all woman's rights movements to the contrary, is that destined to prevail in the world. [Here unfortunately the note breaks off, nor does it seem possible to continue it with any certainty of walking in the author's footsteps. Another note gives brief definitions of some varieties of marriage.] 1. Monogamy. Marriage of one man to one woman, ex- cluding legality of either having another spouse at the same time. 2. Monandry. Marriage of one man to one or more women. Includes monogamy and what is usually called polygamy hereinafter called polygyny. It excludes the idea of there being more than one woman in the marriage union. 3. Polygyny, commonly called polygamy, implies that more than one woman may be in the marriage union, but only one man. 4. Polyandry implies that there are several men in or admissible into the marriage union, but only one woman. [Another note begins : " Where several men and women consort in a union." No name, however, is put down for such a union, and Mr. M'Lennan does not indeed appear to have believed that any unions of the kind were ever in accordance with law in any country. If, however, it can be shown 46 STUDIES IN ANCIENT HISTOR Y CHAP, iv that such a connection was anywhere legitimate, we may reasonably conjecture that he would have called it polygamy, and would have restricted his use of the word to this sense. As a matter of fact, however, " polygamy " is used con- tinually throughout his notes as a synonym for "polygyny," and we have not considered ourselves justified in changing the word. It may be advisable here to reprint some remarks on the words exogamy and endogamy from Mr. M'Lennan's article on the subject in the Fortnightly Review for 1877 : "The former is the law prohibiting marriage between persons of the same blood or stock as incest often under pain of death and the latter the law prohibiting marriage except between persons of the same blood or stock." It is, therefore, impossible for both laws to exist together in a clan. " If a man must not under the pains of incest marry a woman of his own stock or blood, and is forbidden, under the pains of law, to marry a woman of any other stock or blood, it would appear that marriage is forbidden to him altogether. If in any tribe exogamy and endogamy should indeed coexist, then indeed in that tribe marriage would be absolutely interdicted. This is manifest if by " tribe " a tribe of descent or body of kindred is intended. It is equally manifest if by "tribe" a local tribe which may contain portions of several tribes of descent be intended. The phrase applied to a local tribe could only acquire a meaning short of an interdict on marriage on the supposition that the local tribe had come to comprise several clans of different stocks, one or more of which had followed the rule of exogamy, while one or more followed the rule of endogamy. But that would truly be a case of juxta- position, not of coexistence, of the two principles."] CHAPTEK V KINSHIP, TOTEMISM, AND MARRIAGE [AT this point would have followed, according to the author's scheme, a discussion of (1) Early conceptions of kindred; (2) Female kinship, its origin and distribution ; (3) Blood-bond ; (4) Fact of capture. Such are the headings which he left for the chapter, and he also left an indication that it would have corresponded to the eighth chapter of Primitive Marriage. (Strictly speaking, as all familiar with his writings will remember, it is only the first three points to which this applies ; the fact of capture is discussed in the fourth chapter.) How far the views expressed in Primitive Marriage would have been recast and remodelled it is impossible to say, but there is no doubt that their substance would have remained practically the same. It may be hoped that the reader will refer to the original source, but still it may be useful to give here the briefest sketch possible of the views in question. Though most of Mr. M'Lennan's work after the publication of Primitive Marriage was but amplifying, strengthening, and completing the theories and the evidence for them which were therein put forth, yet one important branch of primitive history, which is closely connected with them, is not touched upon in it. We mean, of course, totemism. He came to see later that the account of development given in Primitive Marriage was incomplete without this element, and that it is absolutely necessary to his theory of the origin of exogamy. 48 STUDIES IN ANCIENT HISTORY CHAP. We have accordingly endeavoured to include it in the following sketch, going on the lines of the later work. In the first place, then, the idea of kinship is not an innate idea of the human mind, but has been slowly developed in different forms. The earliest groups of men would have no j]y bond to unite them except the fact that they lived together. So soon as they began to speculate on their relationship they would conceive themselves to be all descended from one common ancestor, and from what we know of savages it is to be presumed that this ancestor would be a plant or an animal. For example, a clan of the Moquis believed themselves descended from the tobacco-plant, and the Ngo-taks in Australia were ducks before they were changed into men. The group takes the name of this " totem," and is marked oi by it from all other groups, believing itself to be homo- geneous. 1 So far the only idea of kindred is " the conception ol Stocks" 2 The way in which the idea was further limited by the perception of consanguinity with the mother, henct developing into the system of female kinship. 3 That this in fact the earliest form of kinship known is shown evidence from all parts of the world. A child took his mother's totem and believed himself related by blood to all others who had the same totem, wherever he might find them. But he did not assume his father's totem, nor recognise any blood-relationship upon that side, the cause being presumably the uncertainty of paternity in so rudimentary a stage of development, and the fact that the mother is most concerned in bringing up the children at first. In the very earliest state of things it may be presumed that simple .promiscuity prevailed between the sexes. "Marri* f ^ was at first unknown ; " 4 and the earliest form of it was, h< maintained, polyandry. " It gives men wives." 5 Its origii 1 Studies in Ancient History, p. 127. N.B. The word " tribe " is used loosely in Primitive Marriage for "stock-group," as well as for "1( group." 2 Ibid. p. 84. 3 Ibid. p. 85. 4 Infra, p. 57. 5 Infra, p. 55. v KINSHIP, TOTEM ISM, AND MARRIAGE 49 can only be ascribed to scarcity of women as compared with men, and this scarcity was due upon his view chiefly to female infanticide. " Under the pressure of want the numbers will adjust themselves to the available food." 3 And the weakest will go to the wall ; as the old are deserted and left to perish, so the girls are got rid of as " useless mouths." Of this form of marriage there are two principal kinds which may with advantage be distinguished. The earlier and ruder form was called Nair by Mr. M'Lennan, from the Nairs of Malabar. The husbands in Nair polyandry are not necessarily of the same clan with one another, and the wife remains with her own -people, and there reeei^eajhe husbands.J Thus she is not in their power, and the children belongto her or to her clan ; to no other, indeed, can they be affiliated, inasmuch as not only the individuality but even the clan of the father is un- certain. The ..niore_advanced form, as Mr. M'Lenimn ^nn aidp.rp.fi it, despite the fact thatjjs peculiar characteristics are especially distasteful-tcupjir^ ideas, was called by him Tin' bet a n ; in. this the Imsbands are of the same family and as _in_ Thibet "gucJ^- ai^dvaiice -ttpuii Naim naturally made whenever a. group nf p.la.np^p.n bring their home, instead of her remaining with her own kin. And it is more advanced not only because it succeeded the Nair system, and imposed a further restriction on the right to marry, but also, and especially, because it enabled the first step to be taken in the change from female to male kinship. For though the individual father still remains undetermined, yet the clan of the father is decided, or at least presumed. This transfer- ence of the child from the mother's to the father's side was effected by different nations in a variety of ways which cannot here be entered into; at the same time, while becoming possible, the change did not become necessary, and as a matter of fact female kinship often succeeded in retaining its hold long after not only Thibetan polyandry, but .even monogamy or polygyny, had been firmly established. It may be as well to observe that it is not to be supposed 1 Infra, p. 80. E 50 STUDIES IN ANCIENT HISTORY CHAP. that the evolution of marriage took place everywhere on exactly the ame lines and passed through precisely the same stages. " I take it," writes the author, " polygyny, monogamy, and polyandry (or its equivalents) must have occurred in every district from the first, and grown up together into systems sanctioned by usage first and then law." 3 Again he says that jjor the pw:pose_of ^m^g_we_may take^the Nair polyandry as the jgrimitive type, jlU-ether-^ossible arrangements grouping themselves-- naturally s~to sayy4hat--the^e"culiar feature of Nair polyandry, the wife remaining in her own house and receiving her husbands there, is not essential to the first stage ; the only really essential point is that the husbands are not of the same totem. And so also the essential point of the second stage, the Thibetan, is that the husbands are of the same totem. For all other details a great margin of fluctuation must allowed. But, owing to female infanticide, polyandry, not polygyny or monogamy, has been the normal method ol development, and thus female kinship has everywhere beei the first system of kinship. Here may most conveniently be introduced a letter of the author's to Mr. Darwin, dated 3rd February 1874.] Your scheme of the development of marriage systei is (1) Polygyny and monogamy ; (2) Polyandry ; (3) Pro- miscuity ; (4) Polygyny and monogamy in recurrence. Jealousy, you conceive, determined the first stage infanticide the second ; polyandry, undermining natural jealousy or regulating it, brought on more 0] less promiscuity ; and finally, a feeling of property ii women growing up in aid of natural jealousy, re- established polygyny and monogamy. You ask me whether I see any fatal objection to looking at polyandry as having preceded promiscuous 1 Infra, p. 53. 2 Infra, p. 57. v LETTER TO DARWIN 51 intercourse, and, indirectly, what I understand by promiscuity and think of the statements often made as to its ancient prevalence. Before submitting to you the few observations I have to make, I should like you to read my little book from p. 162 to p. 170, 1 where there is a very imperfect thinking out of the initial stage. The inquiry is, remember, a human one ; and man not only a creature with natural jealousy, but a combining, conspiring creature. The strongest gorilla may be free to conquer the weaker in detail, and thereafter be supreme in his group ; gorillas not being developed up to the point at which the weaker can by combination for a sexual purpose subdue the stronger. Man alone exhibits this capacity for combination, and, I am sorry to say, our criminal records even in this country in late times show him capable of combinations for sexual purposes several men joining to secure a woman and force her in turn. So far as at the initial stage women were got by capture at the hands of more than one they would be apt to be common to their captors. And now a word as to what I understand by promiscuity. You will see I have guarded myself somewhat against alleging its general prevalence. The import of my reasoning is that more or less of it and of indifference must appear in the hordes or their sections or some of them. I have nowhere defined it, but use it as a general term to denote the general conduct as to sexual matters of men without wives. Now, unless we 1 [Pp. 88 to 93 in the edition of 1886.] 52 STUDIES IN ANCIENT HISTORY CHAP. assume that the nature of man has much changed, we may see in our own time and towns what this conduct has always been. The men in that situation, or very many of them, just do as they can, and are neither over-nice nor over-scrupulous as to the manner. As we go back in time I see no reason for thinking men were more nice or more scrupulous. On the contrary, they were less so ; and I know no more instructive fact- disagreeable as it is, it is of high scientific interest than that one practice (to denote it by the general term I have been using), paiderastia, in many countries became systematised. Thus in Greece the relation between a man and his youthful lover was constituted by a form of marriage after contract between the relatives on both sides. To allege, then, that there was a time when there was general promiscuity would be merely to say there was a time before marriage com- menced before any man had a wife. At that time there would be no uniform behaviour of all men ; their indulgence would be as passion prompted and oppor- tunity offered. We may see in the behaviour of other animals at that stage how various the behaviour may have been. The object, I take it, is to ascertain what from that stage were the normal stages in the evolution of modes of marriage, or marriage systems. In the brute stage we may see analogies to marriage systems, e.g. the gorilla may be said to keep a harem and to be poly- gynous, but it would be a misuse of terms to speak of him as married to his females, or of his females as his v LETTER TO DARWIN 53 wives. ^Marriage began with the first consortships of men and women, jprotected by group opinion. Now I agree with you that from what we know of human nature we may be sure each man would aim at having one or more women to himself, and cases would occur wherein for a longer or shorter time the aim would be realised, and there would be instances of what we may call polygyny and monogamy your first stage ; but, observe, every case of polygyny would cause a case or cases of men without women. That is, supposing you correct in thinking that a policy of female infanticide was later than marriage I incline to the opposite opinion, but it is a point that cannot be well settled there would yet be a disturbance of the balance of the sexes caused by a practice of polygyny ; so that the reasons which move you to conceive there must have been polygyny at the first, are also reasons for conceiving that there was alongside of it polyandry (or its equivalents). Nay, the presumption is that the latter would be on a larger scale than the former, and it certainly would be so, so far as the men were con- cerned; and their training probably was more im- portant than that of the women, so far as the future of marriage was concerned. The first stage, then, if marked by polygyny and monogamy, must have also been marked by polyandry or its equivalents. I take it, polygyny, monogamy, and polyandry (or > its equivalents), must have occurred in every district i' from the first, and grown up together into systems ) 54 STUDIES IN ANCIENT HISTORY CHAP. sanctioned by usage first and then law. But I would put them in this order I Polyandry . . . the more common. Monogamy . . . less common. V Polygyny . . . still less common. I think I can show monogamous systems to be, on the whole, post- poly androus systems; that the normal development was through the forms of polyandry and through the break-down of polyandry. As to polygyny, it is to be observed that it is necessarily everywhere the privilege of the few, not the practice of the commonalty. As time passed monogamy would tend in advancing communities in which the tendency inevitably is towards an equal distribution of the means of well-being among the members of society to become the common practice ; and sentiments springing from it as the common lot would be un- favourable to, and in time lead to the condemnation and prohibition of polygyny. The countries in which polygyny is said to prevail are really only countries in which it is still permitted. As a system it can have had less to do than any other with the history of marriage on the whole. Your most important suggestion, viz. that the policy of infanticide may have been of late adoption, I shall carefully consider. At present I see no grounds for it ; but I am forced by your throwing it out to think of fresh inquiries in one or two fields that have always been a puzzle to me, e.g. the native Australians. Since I wrote my book I have accumulated much LETTER TO DARWIN 55 information about the so-called promiscuity of savages. Perhaps some day we may talk of it. My remarks are put hurriedly before you at what they are worth, and I must send them in all the de- formities of a first expression. * P.S. In re -reading I notice I have not made myself quite distinct, though you will probably catch enough of my meaning. Polyandry, in my view, is an advance from, and contraction of, promiscuity. It gives men wives. * Till men have wives they may have tastes, but they have no obligations in matters of sex. You may be sure polygyny in the early stage never had the sanction of group opinion. They would all envy and grieve at the good of their polygynous neighbour. Polygyny, then, did not at first give men wives. JWife- dom begins with polyandry, which is ; ji contract. If I had time I would re- write this, and try and make it more worth your while reading. I should say I have not _ been on this branch ,of my subject for some time, I have been trying to feel my way back to the state of the primitive groups by a variety of avenues apart from marriage ; notably through the totem and its extensive connections. l [To come to the blood-bond and blood-feud. We have seen that the stock at first considers itself to be homogeneous, all being descended, it is supposed, from a common ancestor, and the sign of this connection being the totem. Now this recognition of blood-relationship always carries with it among savages the duty of avenging the blood of all members of the same clan, that is co say, the blood-feud. And if every STUDIES IN ANCIENT HISTORY CHAP. V member of a totem clan is bound to exact vengeance for every other who has been slain, it is also an imperative duty to him not to shed the blood of any of his own totem kin, any more than he would kill the totem animal itself. To " shed the blood " is a phrase to be taken literally as well as metaphori- cally, and indeed to the savage mind the peculiar associations of blood, which in the Hebrew phrase is the life, actually over- ride the idea of death. At least in cases of infanticide a savage often puts to death one of his or her own kin, but is very careful to do it by some method which will avoid blood- shed. But (apart from infanticide) he who sheds the blood of one of his own kin becomes an outlaw, he loses the right of protection by his own clan, for he has broken the covenant, the blood-bond, which holds the clan together ; nay, he incurs the blood-feud himself with every single member of his clan. This is a most important point to remember and hold fast in connection with Mr. M'Len nan's theory of the origin of exogamy. Exogamy as a law, according to that theory, arose out of a practice of capturing women for wives, and this brings us to the fourth heading set down for the present chapter. It must be enough here to say that such capture in the savage state is an every-day occurrence, and is shown by the symbolism of many civilised peoples to have been a custom of their ancestors. The reader may be referred to Primitive Marriage, chap, iv., as well as to the evidence given in the second part of this volume. If we wish to understand the following chapter on the origin of exogamy, we must remember that the author's theory involves three different details : first, stock-groups held together by the religious regard for the totem ; secondly, the blood-feud, making it impossible to go against other members of the group ; thirdly, a scarcity of women caused by female infanticide, and causing in turn polyandry and a habit of capturing women for \ wives.] CHAPTER VI THE ORIGIN OF EXOGAMY MY hypothesis, so far as concerns the present purpose, is in outline as follows. The primitive groups were, of I were by their members, when consanguinity was first thought of, assumed to be all of one stock. Marriage was at first unknown. In time the special attachments of children to mothers led to the subdivision of the groups into rude family groups of the Nair type, and made possible the rise and consolidation of the system of kinship through women only. Whatever other family, or rather household, groups there were, it is attested by the system of kinship that those of the Nair type largely preponderated, and approximately, for the purposes of thinking, we may assume them all to have been of this type. While things were in this situation a practice of capturing women for wives having its root in a want of balance between the sexes arose, and was followed by the rise of the law of exogamy. It is the manner in which the one might give rise to the other which is now to be investigated. By the joint operation, again, of the system of capture, exogamy, and female --JL- 58 STUDIES IN ANCIENT HISTORY CHAP. kinship, the original homogeneity of the groups was destroyed. They lost their character as stock-groups and became local tribes, each having within it as many gentes of different stocks as there were original stock- groups within reach that it habitually plundered for wives. It is of course an almost necessary inference that many groups disappeared in the struggle for existence. Whatever else may be disputable in connection with this hypothesis, it will be admitted, I think, to be beyond dispute that the account it gives of the presence of gentes of precisely the same stocks in the various local tribes inhabiting an extensive country, like Australia, is correct. Assuming it to be so, we obtain a series of inferences as to the state of the original stock-groups just before the commencement of the processes by which they were finally interfused, and every such inference, it will be seen, throws light on the rise of exogamy. It is found that every gens of any stock is connected with every other gens of the same stock, in whatever local tribes they may be, (1) by the religious regard for the totem, which marks the stock ; (2) by the obligation of the blood-feud, springing out of community of blood. This obligation must have followed the blood from its source wherever it went, as surely as the religious regard must have done so. And unless the totem bond had been fully established in the stock-groups before they became to any great extent interfused in local tribes, it could not have been established at all. It is VI THE ORIGIN OF EXOGAMY 59 the test, and, apart from the memory of individuals, the only test, of blood-relationship among the lower races; and without it, as far as we know, there is absolutely nothing which could hold together, as a body of kindred, persons descended from the same stock-group, but living in different local tribes, or even the same persons living in the same local tribe. We have, then, the inference that the religious regard for the totem, the blood-feud, and of course the system of female kinship without which no commencement of the transfusion could have taken place were firmly established in the original stock- groups, before the appearance of the system of capture or exogamy. When we reflect again on the internal structure of the groups, it becomes apparent that each of them must have become subdivided into so many great families of the Nair type holding on to primitive mothers such as (in magnitude at least) are at a later time and in connection with male kinship derived from common male ancestors ; and that within these great families there would be subdivisions again into smaller groups of mothers and their children, or brothers and their sisters or their children. Now whether we imagine these great family groups, of which the stock-groups were made up, to hold together as settled residents on the same lands, or to be nomadic and separated usually, ranging within the same district of country, we may see that they would tend to become ultimately so many separate bands. The men of each would most conveniently find their wives within their own 60 STUDIES IN ANCIENT HISTORY CHAP. band ; and they would more frequently act together for some band purpose than in concert with the men of other bands for the stock-group's purposes. But the bands, while thus acquiring separate interests and having residences more or less apart, would be firmly united by the bonds of common blood, civil and religious. They would truly be so many septs, all of one blood. If now we imagine some cause to initiate a practice of capturing women for wives in a district occupied by several stock-groups, each subdivided, as above con- ceived, into bands united by a common faith and the law of the blood-feud, we may see instantly one leading result that would follow. There would be no limitation on capture as regards capturing the women of any sub- division of a different and therefore hostile stock-group ; but from the first there would be a positive limitation on the practice as regards capturing the women of any band of the stock-group to which oneself belonged. Of course in attempting any capture, as from a hostile group, the captors would be taking their lives in their hands in the adventure as an act of war. But a capture from one of the kindred bands would be more than an act of war ; it would be felt to be an outrage or a crime ; more than that, it would be felt to be a sin ' a violation of the religious obligation which the blood- feud imposed, for it could not well be accomplished without the shedding of kindred blood. Moreover, all of the stock would be bound to avenge it, and we may well see how from the first it might well not VI THE ORIGIN OF EXOGAMY 61 only be a capital offence, but regarded with a degree of horror. Here, then, in a law prohibiting the capturing of women of one's own stock for wives, we have every note of the subsequent law of exogamy. If we can j. / show that this limitation on the right of capturing women for wives could be transformed into a limitation on the right of marriage, we shall have accounted for the origin of exogamy. ^The difficulties at this point are immense. Instead of its being possible to believe, with some thinkers, that the step was taken at a bound by "a natural confusion" of the two things, it seems almost impossible to see how it could have been taken at all. Let us see if we can ascertain how the change might become possible^) The question is, how the ancient custom of wiving within the kindred (l) went into desuetude, and (2) came to be under the prohibition that originally applied only to capturing women of the kindred. So far as there was an association between capture and marriage, the limitation on the right of capture would operate from the first as a limitation on the exercise of the right of marriage among kindred. If now we conceive, as required by my hypothesis, that the cause of the practice of capture was a scarcity of women, we shall see how the exercise of this right would be further restricted. The kindred bands in a group would be unwilling, and unable even if willing, to furnish one another with wives ; for, on the hypothesis, women were scarce with them. Kindred wives would 62 STUDIES IN ANCIENT HISTORY CHAP. then be unattainable from without, by favour or pur- chase, and we have seen that they would be unat- tainable by capture. So far, then, as the men of a band were in need of women, they would be obliged to obtain them by capture from groups of a stock different from their own. Thus the men would think more of foreign women in connection with wiving than of kindred women, and so marriages with kindred women would tend to go into desuetude. On the other hand, the ideas of marriage and capture thus becoming more intimately associated, there would be a further approach to exogamy. But it is a long way from disuse of an ancient right to the rearing up of an absolute interdict on its exercise. In the present case we may believe that so long as there were in a band women of the men's stock, the men would marry them. Can we feign for ourselves how men could come to be without women of their own stock ? We may believe, to give what mathematicians call a singular solution of the problem, that often, where there was a system of capture, the men of a band might be robbed, in their absence or in open fight, of their women and female children. 1 'Thereafter for these men capture and marriage would mean the same thing. The exercise of the right of marrying kindred women would be for them impossible, and the right itself therefore dead. Capture and marriage would be- come for them synonymous. The women they might 1 See Wallace, Travels on the Amazon [p. 516, also p. 362]; and The Malay Archipelago, vol. i. pp. 144, 145. VI THE ORIGIN OF EXOGAMY subsequently capture being necessarily of some foreign stock, and the children of their mother's stock, there would never again be within the band women of their own stock. Such an experience, lasting for the re- mainder of the lifetime of the men of one generation in a band, might well establish exogamy as the marriage law for the band. Could we imagine that such an experience as this was not uncommon, that it was perhaps frequent in its recurrence, with the bands of the various stock-groups of a country, we should have a condition of things in which, for long periods at least, marriage and capture would be practically synonymous, and whatever limitation applied to the one would apply to the other. Exogamy would become the marriage law. (But it is not necessary to make any so violent a supposition. ) A general cause may be shown to have been in operation which would only require assistance from such experiences as I have referred to, to complete the connection between capture and marriage. This cause is to be found in the absolute change in the relations of husbands and wives that must have followed upon the institution of a system of capturing women for wives. I have called Nair polyandry a mode of marriage because, in a juridical view, any relationship of persons of different sexes resting on contract and approved by public opinion by custom or law is marriage. But it may well have been that the rude men of whom we are thinking matured the idea of marriage for the first 64 STUDIES IN ANCIENT HISTOR Y CHAP. time when the Nair species of polyandry began to decay, and give place to a mode of marriage which put the men in the first place, and women in an absolutely subordinate place in families. Under the Nair system a wife would live in the house of her mother, and under the special guardianship and protection of her brothers and mother's brothers. She would be in a position of almost absolute independence of her husbands, free within the limits of her engagements to show and act upon her preferences, and almost certain to treat her husbands rather as favoured suitors than as lords. On a practice of capture arising all this would be changed. The captives would be the slaves of their captors would be owned by them, and under their protection and guardianship. The new mode of marriage would give a sudden extension to the form of the family resting on monandry or Thibetan polyandry. There would be the cohabitation of husbands and wives, and for the first time the idea of a wife as a subject of her husband or husbands would become general. Now the new idea of marriage which would thus be introduced is the idea that was everywhere destined to triumph that has in fact triumphed among all exogamous races, so far as I know. And it was natural and inevitable that it should triumph. It is easily conceivable how, once men had experience of this new marriage system, unions of kindred on the old model should not only go into desuetude but not be accounted marriages at all. If, then, we conceive that some time after the rise of a practice of capture the name of " wife " came to be vi THE ORIGIN OF EXOGAMY 65 synonymous with a subject and enslaved woman in the power of her captor or captors, and the name of marriage to be applied to a man's relation to such a woman as possessor of her, the origin of exogamy becomes apparent. Since a subject and enslaved wife would, in the circum- stances of the time, be attainable only by capture, marriage would be possible only through capture, and the prohibition which, as we have seen, would apply to capture, would apply to marriage. Marriage with a woman of the same stock would be a crime and a sin. It would be incest. [On the view as to the movement from capture to exogamy stated above, exogamy was in the first instance a prohibition of marriage only between persons of the same blood. There is evidence now forthcoming from Australia which helps the theory at this point, since it tends to show that exogamy is not necessarily anything more, and therefore that it was nothing more at first. The absence of such evidence, however, could not of itself make against the theory, so easy and almost inevitable does it seem that, with marriage thoroughly established, and strictly forbidden between persons of the same blood, the history of the prohibition being unknown, irregular relations should come to be forbidden between persons of the same blood ; especially when, as often happens even with female kinship, marriage has become, more or less completely, a bar to irregular relations. The Australian evidence above referred to is as follows : 1. Speaking of tribes about Port Lincoln in South Australia, Mr. Wilhelmi tells us 1 that they "are divided into two i separate classes, viz. the Matteri and the Karraru " ; that " no | one is allowed to intermarry in his own caste, but only into the other one " ; and that children belong to the caste of the 1 The Aborigines of Victoria, by K. Brough Smith, vol. i. p. 87. F 66 STUDIES IN ANCIENT HISTORY CHAP. mother. Of Mr. Wilhelmi's phraseology nothing need here be said ; it is enough that he conveys to us that the tribes which were made up of Matteri and Karraru were exogamous and took kinship through the mother. As regards marriage their exogamy was strict. " There are no instances," he tells us, "of two Karrarus or two Matteris having been married together." And yet, he adds, " connections of a less virtuous character which take place between members of the same caste do not appear to be considered incestuous." Irregular con- nections, then, did occur between persons whose marriage would not have been tolerated, and, so far as Mr. Wilhelmi could learn, they were not objected to. 2. We are told on the authority of the Eev. W. Julius Klihn l the statement apparently is not in Mr. Kiihn's own language that the Turra tribe, also in South Australia, con- sisted of two great divisions, Wiltu (eagle-hawk) and Multa (seal), the former of which contained ten, and the latter six, separate totems ; that the divisions or sub-tribes were exogamous, but that any totem of the one might intermarry with any totem of the other ; and that children belonged to the totem of their father, and therefore to his division or sub-tribe. Faithfulness in marriage, we are told, was expected of both husband and wife. At grand corrobborees, nevertheless (the account proceeds), " the old men took any of the young wives of the other class [sub-tribe] for the time, and the young men of the Wiltu exchanged wives with those of the Multa, and vice versa, but only for a time, and in this the men were not con- fined to any particular totem." The statement that the men were not confined to any particular totem seems to be made with reference to a theory of Mr. Fison's, which it does not support ; it was made, no doubt, in answer to a special question. For the rest, the statement leaves us to understand that the old men were free in their choice, and the younger men in their exchanges that no exogamous restriction bound them. There is nothing to suggest that they were debarred from women of 1 Kamilaroi and Kurnai, by Lorimer Fison and A. W. Howitt, pp. 285-287. vi THE ORIGIN OF EXOGAMY 67 their own totem who had passed by marriage from their original sub-tribe into the other; indeed, so important an exception, had it been possible to make it, could not have escaped mention. And, at any rate, the men were all free from the restriction which is said to have bound them in marriage as Wiltu and Multa respectively. The practice of the Turra people at corrobborees was, no doubt, a tradition from less advanced predecessors. 3. It now seems worth while to refer to what Mr. Eyre tells us of tribes in the Adelaide district. 1 He says that in most of the tribes the utmost license prevailed among the young, and that there was unbounded license for all on solemn occasions. It is clear that he believed there was no restriction whatever. But Mr. Eyre knew nothing of the marriage law. Mr. Gideon Lang, however, 2 makes a somewhat similar statement, and Mr. Lang was aware that the tribes which had been under his observation were exogamous in marriage. Eeference may also be made to what Mr. Beveridge has said of the tribes of the Eiverina district; 3 and to a fact reported of the Kunandaburi a tribe of the Barcoo river, living within the Queensland boundary by Mr. A. W. Howitt on the authority of a Mr. O'Donnell. 4 It may be suggested, 1 Journals of Expeditions of Discovery into Central Australia in the years 1840-41. 2 The Aborigines of Australia, p. 38. 3 Journals, etc., of the Royal Society of New South Wales, 1884, p. 24. 4 Australian Group Relations, p. 8, reprinted from the Smithsonian Keport for 1883. Jus primes noctis is allowed "to all the men present at the camp without regard to class or kin." If this be received (and a person who had lived for some years among the people could scarcely be mistaken about it), it shows clearly that the exogamy of the Kunandaburi was limited to marriage, and gives weight to all the indications or suggestions of exogamy being so limited which are got from the other cases mentioned. The objection to founding on it is that, while the fact is new for the Australians, no detail is given as to the order of marriage among the Kunandaburi. It may here be said that there is a reason why exogamy, if limited to marriage at first, might remain so limited among 68 STUDIES IN ANCIENT HISTORY CHAP. too, that certain well-known statements about the Kamilaroi need to be carefully considered with reference to the bearing they may have upon the limits of exogamy among that people. Unfortunately, Mr. Howitt gives us the bare fact he has to mention only, and the name of his authority, with the state- ment that he had lived some years among the Kunandaburi. And Mr. Beveridge's knowledge of the marriage law of the Riverina tribes was, no doubt, imperfect. What he says of it is, that the very slightest blood-relationship was a definite bar to marriage. But he knew there was a prohibition which applied to marrying, and that it was strictly enforced. And he assures us that, apart from marrying, there was simply no restriction whatever. He had been for twenty-three years in contact with the Eiverina tribes from 1845 to 1868. Perhaps he proves too much ; a less unmeasured statement could be more easily received. But what he says has to be taken along with the impressions of Mr. Eyre and Mr. Lang, and the more definite information given by Mr. Wilhelmi and Mr. Kiihn. If the foregoing evidence raises a doubt as to the original scope of exogamy, it is enough for the purpose for which it has been adduced. And it seems at least sufficient to raise such a doubt. With a distinct statement from Mr. Klihn that in the Turra tribe men were not debarred from their own totem at the corrobborees, one might go further. For that would leave no room for the suggestion that exogamous feeling, still in its original strength as regards each totem, had, by means of the totems, been weakened between the larger divisions, the Multa and Wiltu, the Matteri and Karraru no room for the suggestion that the facts show us, not exogamy operating within its original limits, but exogamy in a state of decay. As to that, however, Mr. Howitt (who procured the information) appears to have made inquiry as to a much smaller matter whether particular totems of the sub-tribes of the Turra Australians a reason consistent with the theory now submitted. It is that among many, perhaps most, of the Australian tribes a wife is prized chiefly for her services as a drudge. vi THE ORIGIN OF EXOGAMY 69 people were confined to each other at those meetings and he cannot have neglected to satisfy himself upon a question of the first importance, which is plainly raised by the statement which he has published, and in which Mr. Fison and he have, throughout their work, shown themselves to be deeply interested. In speculating on the influence of two such factors as capture and female kinship, it is unavoidable, though the two may have acted concurrently throughout, that the attempt should be made to follow the operation of each separately, combining the results ; or (which comes to ;fche same thing) that the effects of the one should first be traced, and then those of the other added on to them. It was necessary in the preceding essay to deal with the kinship first ; but it may be easily seen that there would be ample time for its development, and for tribes which had grown too large to subdivide in the manner supposed, before capture could have any effects which need be taken into account. Capture may have been practised before there was any thought of relationship ; it may have been practised, more or less, all the time that kinship through females was growing up. And stranger women, captives of a hostile totem, must from the first have been in a worse position than the native-born ; while their position must have grown relatively worse and worse as the growth of kinship gave the latter protectors and helped their numbers to secure them some consideration. For long, their children, being regarded as of some hostile totem, would not be allowed to live ; and we may be guided in some very small measure in judging how they would compare with the women through whom the tribe and its totem were propagated, by observing the low position assigned to captive wives wherever we find capture practised in supplement of a regular system of marriage by contract. But it is unlikely (as the analogy of the case just mentioned shows us) that, by their numbers merely, they could sensibly lower the position of native-born women ; and there appears to be no other effect which, in the state of things supposed, could follow upon their presence in a yo STUDIES IN ANCIENT HISTORY tribe. Men cannot have for wives (even in a polyandrous way) women who are doomed to childlessness; and (though a gradual preparation for foreign wiving would no doubt be going on) not until manners had so far softened, and hostile (that is, different) stocks grown to be so far tolerant of each other that the men of a totem could let the children of foreigners grow up in their midst, could there be a beginning of the competition between native and foreign marriage. We may believe that the children of captive women would come to be spared at length by a sort of tacit agreement between neighbouring tribes arrived at gradually, and no doubt very slowly. At first, and, indeed, for long after it became common to spare them, each tribe might remain of one stock or totem, so far as the men were concerned. The blood-feud would, at any rate, tend to drive the sons of captives to their mothers' relatives. The daughters, such f them as were spared, would succeed to the lot of their mothers and by and by would form a nucleus of women available for the lot of foreign wives who could be had without capture. The main source of supply of such wives, however, would almost neces- sarily be in capture until there was, within each stock, so much tolerance of foreign elements that the sons of its captives or women of foreign stock could continue to abide with it, and their daughters had as good a chance of being allowed to live as those of the native-born. That involves a great relaxation of the hostile feeling between different stocks ; it would change each separate body, from being a stock of a single totem, into a more or less heterogeneous local tribe. It might give time for a long practice of getting wives by capture; and it need not be doubted that, once a preference for foreign wives had become general among men, understandings would be arrived at between tribes or methods devised (such as occur in known examples) with a view to their making captures easy for each other understandings or methods such as might lead in time to contract with the form of capture. "With tribes become heterogeneous, of course, the need for captures might cease ; men might find within their own borders wives enough of different blood from their own wives obtained at length by vi THE ORIGIN OF EXOGAMY 71 friendly bargain, but who would succeed to the subject lot proper to captive women and their daughters. It scarcely need be said that either monandry or Thibetan polyandry might exist along with female kinship. This kinship must have lasted at least till after local tribes had become heterogeneous, if, with exogamous (that is, foreign) marriage, it furnishes the only adequate explanation of the heterogeneity. And, with the totem relationship already founded on it (as, by hypothesis, it was), it could not be: superseded all at once or at the will of single individuals or brotherhoods, nor until the minds of people living together, and even of their neighbour peoples, were generally prepared for the change. Moreover, capture, so long as it was practised to any considerable extent (since it would render fatherhood still in many cases uncertain), would tend to keep it up ; and so also would the liberty of intercourse between people of the same stock, so long as that subsisted. The supposition that a stock-group would subdivide into bands composed of persons specially related to each other, though obviously useful, does not seem to be indispensable to the theory of the essay at any rate, a little of such sub- division suffices for it. Without that, we may see that the lot of native women must have been very different from the lot of captive women, and that one of the former could not be treated like the latter without outrage, and no more is in- dispensable. Nevertheless, the conditions of subsistence would, in early times, almost necessarily make each separate band a very small one, and such subdivision as is supposed might be of frequent occurrence. As to the use made of capture in the essay, (though it should not be necessary,) it seems to be necessary to say that it is assumed that what men are known to have done in a certain case prehistoric men in the same or a similar case would do. Within times known to us, men have practised capture (though they have done so also without necessity, no doubt) when women have been scarce with them, whenever they could not otherwise get wives. And, in particular, men have practised capture (or got their wives after a form of 72 STUDIES IN ANCIENT HISTORY CHAP. capture, which shows that their predecessors had to capture their wives) because they have been exogarnous in marriage. On the theory stated in the essay, men, having begun to capture chiefly because their own women were few, formed in time through their relations with captive women a preference for subject wives, and got them by capture because at first and for long they could get them by capture only ; while the exemption of their own women from the fate of the captive, so far as each stock was itself concerned, formed, when a marriage system founded on capture had come to prevail, a limitation on marriage, which was exogamy in its earliest form. How exogamy may force men into a system of capturing wives is excellently illustrated by the case of the Mirdites. 1 The theory assumes that the desire for subject wives, once it had become general, would have effect given to it in the same way, while the exemption of women living among their own people from the lot of captive wives would make marriage in fact exogamous. The Mirdites get their wives by capture because exogamy is they know not why a law with them. Pre- historic men, be it observed, would be, as regards marriage, in precisely the same position as soon as the reason for their not taking their own women in marriage ceased to be thought of. Exogamy in marriage would then, at latest, be fully established. And after that the limitation upon marriage might easily grow into a prohibition of all connections between persons of the same blood. The occurrence of the form of capture along with female kinship shows, however, that the association between capture and marriage was in some cases not easily or quickly lost sight of. There are some peculiar Australian facts, too, which suggest that among certain Australians, after exogamy had been established for people of the same totem, and local tribes had been made heterogeneous by it, capture of wives was practised so extensively that it even availed to give a wider scope to exogamy in marriage. The principle that if it is wrong to capture a woman it is wrong to marry her will at any rate account for marriage being forbidden (as it is in most 1 Researches in the Highlands of Turkey, by the Rev. H. J. Tozer, vol. i pp. 318 et seq. vi THE ORIGIN OF EXOGAMY 73 of the cases referred to) between persons of the same local tribe, even when they are of different totems, and also for it being forbidden (as it is in one or two cases) between all persons of those neighbour tribes who speak the same dialect. Comity and the fear of consequences (especially the latter) would make capture as impossible in the small Australian local tribe as it would be in a body of people all of one totem; and might make it, even as between neighbour tribes having dealings with each other, much too troublesome not to be very seriously disapproved of. And marriage is forbidden within the limits within which a capture might thus have been deemed an outrage. A statement made towards the close of the essay makes it proper to add (and no more can now be done) that no case of beenah marriage not even an exclusive practice of it by exogamous tribes, the only case of it which is not easily intelligible makes any difficulty for the theory therein sub- mitted. D. M'L.] CHAPTER VII FEMALE INFANTICIDE PLUTARCH, in his dialogue on the question whether the lower animals have the use of reason, ^ makes Gryllus point out that in several ways these animals lead their lives more decently, respectably, and, so to speak, humanely, than man himself. Had the knowledge of Odysseus, one of the interlocutors in the dialogue, equalled his craft, he might have confounded Gryllus \by showing that the practices^ he referred to as odious tin man, Attest the terrible struggles Jor_existerLce--which ibhirhuman xace T^s^passed~jh^ugh, and whieh-^o-other species could have survived ; and are_so_many proofs therefore of man's superior intelligence. Shocking- pven horrible as some of thesepractices areT^Ee human race has probably ow.ed to them itsjreservation. It is, moreover, an ample compensation for the degradation iey imported into our life, that while they are every- where disappearing, the moral and legal principles that 'were simultaneously developed with them, out of the same circumstances, remain, increasing in vigour, the CHAP, vii FEMALE INFANTICIDE 75 glory and triumph of man, separating him by an im- passable gulf from the brute creation. Curiously enough Plutarch makes no reference in this dialogue to the practice of infanticide, though he must have been familiar with it as prevailing in Greece. Yet it was in infanticide that the whole group of dis- creditable practices which he does mention, must have had their origin. None of them, moreover, sets man in so unfavourable a light as compared with the brutes. " Even the tiger," says a Chinese poet, " knows the relation of parent and child. Though by nature cruel, it spares and rears its young. Shall man, who is the spiritual part of all things, be inferior to the tiger ? " The higher he is, the more astonishing in him is so cruel a stifling of the promptings of nature. When, however, we think of the systems of in- fanticide as contrivances for the preservation of the species when imperilled like the cannibalism support- ing life, on rudely constructed floats, of shipwrecked sailors we may admire their ingenuity, while moved with pity for men in the terrible circumstances which constrained them to have recourse to such expedients. What extremities men must have been in when > such systems began to be customary we may see by / attending to the circumstances under which infanticides, / as crimes, occur in modern societies, and then throwing \\ out of view the circumstances now acting which cannot j have affected men in early societies. Such crimes are not infrequent among ourselves. Murders of illegitimate children in particular, or con- 76 STUDIES IN ANCIENT HISTORY CHAP. cealments equally fatal to the children, are painfully common. We have been made familiar of late with systems of baby-farming in connection with the larger towns; and this is but a mode of infanticide. We have had before us also the idea, at least, of systematic child-murder in connection with burial societies. The causes of such crimes are familiar pinching poverty or insupportable shame shame too often in- supportable only because of its seeming to threaten the life through the character. The circumstances attend- ing their commission may vary from case to case, but in every case the cause of the crime is expressible in the same general terms a pressure of circumstances under which the instinct of self-preservation overmasters that which guarantees the perpetuation of the species. Where infanticide is a crime it is of course abhorrent to the sentiments of the people. Where with any people it is not only not a crime, but systematic and approved as a custom of the country, we must believe that the people were so long subjected to a pressure of the general description given above as to have their love of offspring deadened and replaced by sentiments springing from the habit of destroying infant life. No temporary pressure of this sort could, however sharp, effect so radical a reversal of the feelings ^hich man must at first have shared with other animals. JLThe pressure may have varied from time to time in amount, but we must think of it as steadily applied to a series of generations!, of any people in order to understand its establishing among them sentiments so contrary to vii FEMALE INFANTICIDE 77 t nature, as we may say, as those which approve the systematic slaughter of infants. Now there are various peoples among whom in- fanticide is an approved system. They must, therefore, at some stage in their history, have been subjected during the lifetime -of a series of successive generations to a pressure of the sort referred to. Let jis ask more particularly what the nature of the pressure must have been. The slightest survey of the facts excludes the idea that shame can have been a factor in that pressure. The standard of propriety which attaches an indelible disgrace to a woman because of a lapse from virtue is unknown to many of them now ; and there is nothing to indicate that their moral standards were anciently higher than they are now. The pressure then must have been a direct pressure of circumstances threatening the life; and there are but two main sources from which such a pressure could spring Want or War, or Want and War in combination. But when we think of a steady, long -enduring pressure exercised for a series of successive generations on a people, and springing from want or war, two ideas are instantly excluded the idea of want in the form of famine, and the idea of war waged with the sharp desolating effects of modern warfare. Famine is like a sudden inundation. The bulk of a people may perish in the wave of want, but it passes, and immediately for the survivors there is a recovery for a time of the normal conditions of subsistence ; nay, it may be of 78 STUDIES IN ANCIENT HISTORY CHAP. superior comfort, their numbers being now fewer in proportion to the normal supply of food. We may see the truth of this by looking either to Ireland or to Orissa, that more recently had its population suddenly reduced by millions in consequence of famine. The famine which now threatens the Dekkan, my friend, Sir David Wedderburn, ascribes to a year of drought a casualty not infrequent in all oriental countries. A drought in the Dekkan, he says, implies not pecuniary loss to a few, but want and misery to millions who can only just manage in ordinary years to earn a bare sub- sistence. " Under former regimes" he remarks, " when famine visited a district, the people either died or migrated, the population was reduced to a number proportionate to the local means of subsistence, and the risk of scarcity was at an end for one generation at least." Mutatis mutandis, what is here aptly said of famine in the Dekkan may be said of famine every- where. It works its own cure, and its effects on a population are strictly temporary. So also with War. A country may be overrun and desolated, its population decimated ; but Victory must declare itself on one side or the other ; and then comes Peace, and with it plenty and security once more. Want that would yield a steady grinding pressure capable of continuing to act on a people for generations must be consistent with the people continuing to live, if they will but just be very careful. I have seen it food enough for two, but not for three. Food enough got with the utmost exertion for five, just enough to vii FEMALE INFANTICIDE 79 support life, and there are seven in the family ! How well known is this condition of things, even in Merry England, in certain sections of society. Such want as this is distinguishable from famine, as diffused galvanism is from electricity in the thunderbolt. War that would yield a steady pressure capable of continuing to act on a people for generations, is simply the State of Hostility. Whoever is not with you is against you. All who are not of your immediate group are your enemies. It is not so much that there are frequent and sanguinary encounters as that at any moment there may be an encounter. There is a total absence of security. Men search for food stealthily, warily, ever on the watch for enemies armed ; they cultivate their fields armed ; and actual combats are just frequent enough to sustain perpetual distrust and fear. Of the state of hostility I have written briefly else- where, and it would be to no purpose to enlarge on that subject here. Man has long been his own worst enemy, but when we go back in thought to the state of things to which Colonel Lane Fox's wonderful collection of weapons points so distinctly, and think of man as yet naked and without weapons, we may see that for him then there must have been numerous and terrible enemies in the brute creation as well. Indeed, it is to this stage of man's life that we are relegated also when we think of him as subjected to the steady pinching pressure of want. Until, under the stimulus of 8o STUDIES IN ANCIENT HISTORY CHAP. necessity, he provided himself with weapons for attack and defence, his life must have been passed in constant insecurity from failing food and enemies everywhere threatening him in the search for it. But by the time that man had armed himself had pointed the spear and invented the throwing-stick or the bow he was already man in the full possession of his faculties. Kecurring then to question how a system of infanti- cide could arise, remembering that such systems have arisen, and recalling the conclusion we reached, namely, that no such system could possibly have arisen except among a people that had been subjected, in the sense just explained, for successive generations, to the pres- sure of want, or of war, or of want and war in com- bination, let us proceed to consider more particularly how such a system could arise, and the form it would most likely assume. Under the pressure of want the numbers in a group will certainly adjust themselves somehow to the avail- able food. Only so many can live as the food suffices for ; the rest must die. Under the pressure of war the combatants of a group would feel the non-combatant members to be of the nature oljmpedimenta.^ The non-combatants would almost certainly be the persons in the group who consumed food without finding any the aged and the infirm, the very old and the very young. A group that was, as a condition of its existence, vii FEMALE INFANTICIDE 8r nomadic, obliged to pass from one district, on the exhaustion of the food supply, to another, would further find the aged and the infirm, and the very young a burden. They might carry the very young. When an adult ceased to be able to support himself, there would be nothing for it but to leave him behind. No more touching stories are told of savage life than those relating to desertions of this description. We may imagine that the first infanticides were of the nature of desertions also. A child born on the march would be specially apt to be abandoned. Many cases, indeed, are recorded in which in such a case the weakened mother gladly bore the burden of the newly- born, but, lagging a little in the march, received orders to throw the child away. Assuming the first infan- ticides to have been of this sort, they would be casual rather than premeditated ; but we may see how they might prepare the way for the premeditation of such murders. The exposition of infants on the march would prepare the way for their exposition at home. The moment infanticide was thought of as an ex- pedient for keeping down numbers, a step was taken, perhaps, the most important that was ever taken in the history of mankind. But the moment we think of infanticide as a, device forced upon a group of men to adjust their numbers to the available food, or for security against enemies, we may see what forms a practice of it would at first assume. The first victims would be those born deformed, maimed, or otherwise imperfectly suited, if G 82 STUDIES IN, ANCIENT HISTOR Y CHAP. reared, to aid in the quest for food and in war, and since women are less available in the quest for food in war than men, female children would be the next victims ; but it requires some consideration to see how far a practice of female infanticide could be carried out consistently with the motive for the practice the desire for self-preservation. There are groups of men who kill all children born in the group and recruit their ranks by capturing boys and girls from other groups, thus dispensing entirely with the troubles of rearing offspring. In these robber bands the love of offspring is absolutely overmastered. They are of course most exceptional, and were they not so the species must have perished. They are interest- ing, however, as exhibiting one solution that was possible of the problem raised by the state of want and war. The solution that was reached we know was very different from this, for the various races of men have survived the sharp struggle for existence which raised that problem. It was a complicated problem. There were two instincts in conflict, both deeply rooted in human nature. On the one hand the instinct of self-preserva- tion, on the other hand the instinctive love of offspring and the need of women for wives. The rearing of some female children as wives for the men of the group would be indispensable to the continued existence 01 the group. And, so far as war was a factor in the pressure causing infanticides, every man would fee] his security to depend on that of the group and vii FEMALE INFANTICIDE 83 the maintenance of its effective for fighting. This would set a limit to the range of female infanticide, but it would also set a limit to, if not proscribe, any practice of male infanticide in supplement of it ; and hence, supposing the double pressure of want and war to be felt, it would supply an impulse to arrangements economising women or dispensing with them. And we may see that the moment such arrangements were once well established, the range of the practice of female infanticide might, consistently with maintaining the groups effective, receive a considerable extension. We may further see that arrangements of the sort referred to would operate per se as a check on population and tend to supersede the practice of in- fanticide so far as it depended on scarcity of food. Owing to the persistence of custom, the arrangements that had this effect might well long survive the state of things out of which they grew. Put in this point of view, a system of infanticide appears as embodying a policy of despair, developed from point to point, through trials and errors that no doubt were sometimes fatal to the groups making them, but which contributed to forward the thinking out by men of what was the best form of the policy, its best practical expression. We may believe that no animal below the rank of man in the full possession of his reasoning powers could have thought out such a policy, and for the credit of human nature that such a policy would never have been thought out or acted upon except in the most desperate circumstances. 84 STUDIES IN ANCIENT HISTORY CHAP. Let us now ask what traces of itself such a system of infanticide might be expected to leave behind it with the representatives of any people with whom it had long prevailed. As the pressure that instituted the system was removed we should expect the features of the system to disappear in the reverse of the order in which they had appeared. Any practice of male infan- ticide would cease first ; next the practice of female infanticide would be limited in its range till it finally disappeared ; and lastly would remain only the practice of putting to death the deformed or otherwise imperfect infants. The state of the moral sentiment where children are put to death, with public approval, at all, is not easily to be accounted for, except as a remainder of a more general form of infanticide, for births of deformed children are rare. In attempting however as I now propose to do to ascertain the range of the practice of infanticide amongst the various races of men, I shall not assume such a limited practice as that last mentioned to be indicative of a preceding more general practice, unless I find it in connec- tion with the arrangements for economising women, or dispensing with them, that I have referred to. In that case it may, I think, be fairly assumed as a remainder of a previous more general system of which female infanticide was a feature. The arrangements referred to fall under two descrip- tions. Those of the one sort have been classed under the name of polyandry, those of the other under a variety of names one drawn from Hebrew history, vii FEMALE INFANTICIDE 85 one from the scene of recent Turkish atrocities, and another, paiderastia (which let us call it), from the Greeks. Both have the widest possible distribution, and neither can be well accounted for, apart from a practice of female infanticide for long destroying the balance of sexes. Of the facts relating to polyandry a partial induction has already been made, but only a pressing necessity in the interests of science would justify an induction of the other class of facts to which I shall refer only so far as it may appear indispensable to do so. For reasons to be hereafter given, I take first the cases of infanticide occurring among peoples having male kinship. 1 I. INDIA. To pass by the evidence of the former prevalence of infanticide in India, I note the following examples of systems of female infanticide still, or till very recently, existing there. (1) The Klionds. Major Charteris Macpherson says of them in his Memorials of Service in India, p. 132 : 1 [See Studies in Ancient History, first series, pp. 145, 146. When female kinship and exogamy are the law, if the father has to provide for the family, the wife living with him instead of with her own kindred, he may be unable to kill his children, for fear of bringing on himself the blood-feud with his wife's kin. An objection might possibly be raised that this is against a theory which bases exogamy on female infanticide at a time when female kinship was the only kind recognised. It must be remembered, however, that with Nair polyandry the child lives with its mother's kindred, who would consequently have the trouble and expense of rearing it, and would have no scruple about putting it out of the way. The father not being akin to it, they would have no fear of the blood-feud from that quarter.] 86 STUDIES IN ANCIENT HISTORY CHAP. " The practice of female infanticide is, I believe, not wholly unknown amongst any portion of the Khond people, while it exists in some of the tribes of the sect of Boora to such an extent that no female is spared except when a woman's first child is a female, and that milages containing a hundred houses may be seen with- out a female child" (2) The Rajputs. The female infanticide of the Rajputs is well known. It was their custom to rear but one or two females, at the most, in a family, and usually but one. Our Indian Government has long exerted itself to put down the practice, but it still continues in a modified form. From a" recent census of British India it appears that as late as 1870 an Act was passed for its suppression, putting under special police supervision every district where girls were found to be fewer than 35 per cent of the whole number of the children. (3) The Cutch. A similar system prevailed among the Cutch. " The number of female children who are an- nually murdered among this people, Captain MacMurdo supposes, cannot be less than one thousand/ 71 (4) The Todahs. Major Walter Campbell, in The Old Forest Ranger, says of the Todahs : " They destroyed all the female children but one, who was reared to supply the place of the mother." It is well known they are polyandrous. (5) The Hill Tribes round Munniepore. A similar system prevailed among some of these tribes, and Col. 1 Asiatic Journal, vol. ii. p. 42. vii FEMALE INFANTICIDE 87 M'Culloch, who was long Political Agent there for our Government, relates that in one of the villages of the Phweelongmai he found not even one female child. The system is worked among these tribes through the custom of "Namoongee," equivalent to the "Thola" of certain African tribes, of which hereafter and though this custom merely gives certain indications challenging the act of murder, it is readily so worked as to operate in effect as a system of female infanticide. With these cases I may connect that of (6) The, Biludii. - - They put to death all their illegitimate female children, which may be taken as a system of female infanticide in decay. 1 II. CHINA. The Chinese Penal Code contains (section 319) the following provision: "If a father, mother, paternal grandfather or grandmother chastises a disobedient child or grandchild in a severe and un- customary manner, so that he or she dies, the party so offending shall be punished with 100 blows. When any of the aforesaid relations are guilty of killing such disobedient child or grandchild designedly, the punish- ment shall be extended to sixty blows and one year's banishment." Sir George Staunton has a note on this article of the code to the effect that it shows that "the crime of infanticide, however prevalent it may be supposed to be in China, is not, in fact, either directly sanctioned by the government or agreeable to the general spirit of the laws and institutions of the empire." But I cannot see how the article touches the subject, 1 Burton's History of Scindh, pp. 244 and 411. 88 STUDIES IN ANCIENT HISTORY CHAP. for the offence it constitutes appears to be that of slaying a disobedient child that has been allowed to grow up. There is no offence under the first branch of the article if the chastisement causing death is inflicted in a customary manner ; and it is remarkable as to the second branch that the killing, though intentional, is not murder, but punished very lightly. And this is the more remarkable that by the immediately pre- ceding article, the killing of a father, mother, paternal grandfather or grandmother, even "purely by accident" is punished much more severely i.e. "with 100 blows and perpetual banishment to the distance of 3000 lee" It seems to me, then, that infanticide, i.e. the slaying of a child at the birth, is not touched by the code. Any way, it would appear that if it is, the code has never had effect, that custom has been too strong for it. Staunton himself immediately proceeds indeed to admit this, and to apologise for it. " This practice," he says, "must certainly be acknowledged to exist in China, and even to be, in some degree, tolerated. . . . Even the dreadful crime of a parent destroying its offspring is extenuated by the wretched and desperate situation to which the labouring poor in China, to whom the practice of infanticide is admitted to be in general confined, must, by the universal and almost compulsory custom of early marriages, often be reduced, of having large and increasing families ; while, owing to the already excessive population of the country, they have not the most distant prospect of being able to maintain them." This is a sufficient disclosure of the vii FEMALE INFANTICIDE 89 pressure of want in an old community on a scale to cause a system of infanticide to appear, and the infor- mation I have received from credible witnesses is that it has appeared accordingly, and that it is not only, "in some degree," tolerated, but that it is openly provided for by the authorities. I have been told of publicly provided receptacles in certain country districts for the newly slain infants, and Mr. Abel, in his Journey to China (1816-17), says it is one of the duties of the police in Pekin to go round every morning and empty the receptacles of the newly-born cast into them during the night. Moreover, a lady writing a few years ago in the Esperance of Nancy, not only attested to the openness of the practice, but to its being in the main a practice of female infanticide. "There are women here," she wrote, " who carry on quite a trade in this sort of murder. . . . They (the infants) are not indeed given to the pigs, but that is the case farther in the interior." And that the system is mainly a system of female infanticide is further attested by the ode already referred to, 1 " On the Drowning of Female Infants." " Hwang-le-ye says that the drowning of infants, though it is the work of cruel women, yet arrives from the will of the husband." The poet pleads against the practice (see Asiatic Journal, vol. v. p. 575). He says, "I have heard that when female children are killed the state of suffering is beyond com- parison. It cries in the tub of water, long suffering before it dies. ... I would advise my people not to 1 [Supra, p. 75.] 90 STUDIES IN ANCIENT HISTORY CHAP. kill their daughters." And in all the pleading there is not a hint of the practice as unlawful, it is not even said to be wrong; it is objected to mainly as distressing in its incidents and questionable in its economies. Of course the balance of the sexes is destroyed in China wherever this custom prevails, but as to the extent of the disturbance we have no trustworthy infor- mation. Mr. Cameron, in his British Possessions in the Malayan Archipelago, says the Chinese in Singa- pore, in his time, were as fifteen males to one female, and the sexes may be equally out of balance in California and Australia. These facts, however, may be mere incidents of their peculiar system of emigration, and no direct inference can be carried over from them as to the state of the home population. Since, however, poly- andry and paiderastia are distinctive of the life of the Chinese as emigrants, wherever we find them, we have the indirect inference that the system of infanticide has indeed in China disturbed the balance of the sexes up to the point at which such a want of balance establishes arrangements for economising or dispensing with females. 1 III. In the Islands of the Southern Pacific we have a new series of phenomena of the same sort in conjunc- tion with practices which are fatal to the lives of children of both sexes indiscriminately. Mr. Williams says of Vanua Levu (Fiji group), "The extent of 1 [Compare Gordon-Cumming's Wanderings in China, vol. i. p. 195 ; Hamilton's East Indies in Pinkerton, vol. viii. p. 491. Many more authorities might be easily added.] vii FEMALE INFANTICIDE 91 infanticide in some parts of this island reaches nearer two-thirds than one-half." It is reduced to a system with professors in every village, and " all destroyed after birth are females, because they are useless in war, or, as some say, because they give so much trouble " (see Fiji and the Fijians, Rowe's edition, p. 155). Admiral Wilkes represents a somewhat similar state of matters as existing in the Hawaian group (United States Exploring Expedition, vol. iv. pp. 45-75), and, according to Ellis, the numbers there killed are about two-thirds of the numbers born (Hawaii, p. 299). Ellis thinks that the system as practised in Tahiti and the adjacent islands could not consistently with the main- tenance of the population have anciently been carried on on the scale on which he found it (see Polynesian Researches, vol. i. pp. 249-258). The traditions of the people, however, exclude the notion that the system is of recent origin. When the missionaries pleaded against it, they were told that " it was the custom of the country," and the first missionaries have published it as their opinion that not less than two- thirds of the children were murdered by their parents. "In the largest families more than two or three children were seldom spared, while the numbers killed were incredible. The very circumstance of their destroying instead of nursing their children rendered their offspring more numerous than it would otherwise have been." Ellis knew a number of parents who admitted having killed "four or six or eight or ten children, and some even a greater number." 92 STUDIES IN ANCIENT HISTORY CHAP. And the universality of the crime, he says, was no less painful than its repetition by the same persons. Mr. Nott on one occasion told him that during thirty years he had spent in the South Sea Islands he had never, that he recollected, known a female unconverted to Christianity who had not killed a child. He gives some details of the modes of disposing of infants and the causes of killing them. As to the latter, he says pointedly (p. 257): "Their sex was often, at their birth, the cause of their destruction. . . . The circum- stance of its being a female child was often sufficient to fix their determination on its death." The general answer to the question why they preferably killed females was, "that the fisheries, the service of the temple, and especially war, were the only purposes for which they thought it desirable to rear children ; and that in these pursuits women were comparatively use- less." The results show how far the system was a system of female infanticide. In the adult population of the islands, he says, " there were probably four or five men to one woman." It is needless to add that he indicates as a consequence other practices among the people not fit to be mentioned. The practice is said to be common in New Zealand (see Angus's Australia and New Zealand, p. 312) ; and " very prevalent " in Madagascar (Ellis's History of Madagascar, vol. i. p. 154) ; but I am without the details of Australia. Mr. Darwin says that female infanticide is still common there. " Sir G. Grey," he adds, " estimated the proportion of native women to vii FEMALE INFANTICIDE 93 men as one to three ; but others say as two to three " (Descent of Man, 2nd ed. p. 592). A custom called " Thola " prevails among certain tribes in Africa. To " thola " is to transgress, and every child that does this is put to death. Among the people of Kuruman Albinos are put to death, and Livingstone suspected that this was the case also in Londa. Deformed children of course were killed. " e Thola/" says Livingstone, "is ascribed to several curious cases. A child who cuts the upper teeth before the under was always put to death among the Bakuas, and I believe among the Bakwains. Of twins one is put to death." And no doubt in Africa as round Munniepore the females have the worst of it in the working of this code of slaughter. For the rest my knowledge of the system in Africa is too scant to make it worth stating. It is indicated by Moffat as occurring among the Bushmen. In many quarters such a note of it occurs as is indicated by praying for and rejoicing over the birth of male children and caring nothing for female children. I cannot doubt, however, but that facts showing its prevalence will yet be brought to our knowledge. Sir Samuel Baker says of the tribes " from Gonookoro to Obbo " that they do not kill their women prisoners, for which the reason is supposed to be that women are scarce, a fact which could not well be explained apart from a practice of female infanticide ( The Albert Nyanza, vol. i. p. 355). When one reads Andrew Battel's account of the Gagas (Pinkerton, vol. xvi. p. 326), who put to death 94 STUDIES IN ANCIENT HISTORY CHAP. all the children born to them, thus evading the whole trouble of rearing children, and then recruited their ranks by systematic stealing from other tribes of boys and girls of thirteen to fourteen years of age for warriors and wives, we feel we are on territory where infanticide must be common. The tribe was of course Gaga only in name. " In the camp," says Battel, " there were but twelve Gagas, and those were captains." And the same impression is derived from M'Kenzie's account of the Matebele " a warlike people, every able-bodied man of whom is a soldier, and (for whom) every year is a year of war." This people, who can hardly be called a tribe, consisted of a few Zulus and for the rest " a heterogeneous assemblage of every tribe " through which the chief of the Matebele had forced his way north. The middle-aged and full-grown men were Bechu- anas, captives taken in the Transvaal ; the young men Makalala and Mashona captives taken in the country now inhabited by the Matebele, and grown up under a sort of Lacedaemonian discipline accustomed to strife and bloodshed. As they grow up they in their turn become the captors of others. Of course they have no marriage properly so called. " The voice of the infant," says Mackenzie, "the song of the mother, are almost unknown there. Only after some signal service does the chief bestow, as a great reward to the soldier, a captive girl to be his wife, who has no choice in the matter, but is delivered over to her new owner as an ox is given to another man whose deeds have been less meritorious" (Ten Years North of the Orange River, vii FEMALE INFANTICIDE 95 p. 329). In these cases Mr. Spencer may see the rigid logic of infanticide pressed to the extreme tribes rear- ing no children and recruiting themselves from their enemies. Turning to America, we find the Eed-men usually very affectionate to children and careful of rearing them. But they everywhere present us with those practices which elsewhere we have seen as incidents and accompaniments of female infanticide, and which seem wholly inexplicable apart from a long -established want of balance at some time between the sexes. They have, moreover, some customs which seem to be re- mainders of a system of infanticide, and in some quarters we have positive records of the prevalence of the system. For instance, Azara, as cited by Darwin (Descent of Man, 2nd ed. p. 592), says that "some tribes of South America formerly destroyed so many infants of both sexes that they were on the point of ex- tinction." Humboldt relates that the savages of Guiana had a custom resembling " thola," " when twins are produced one is always destroyed" (Macgillivray's Abridgment, vol. i. p. 241), and they kill any child that is " deformed, feeble, or bothersome " ; on which Humboldt exclaims at the seemingly trivial provocations to murder to avoid travelling more slowly, in fact to avoid a little inconvenience. The inadequacy of the motive suggests that the deed was easier in respect of moral sentiments inherited from a prior practice. Among the Dacotahs or Sioux of the Upper Mis- sissippi infanticide is occasional, and " the lives of 96 STUDIES IN ANCIENT HISTOR Y CHAP. female children are held in less estimation than the male children" (Schoolcraft, vol. iii. p. 243). The Indians of south - western Texas are not prolific, a woman having seldom more than three children, which if male are nurtured with care, whilst the females are abused and often beaten unmercifully. Lubbock says infanticide was common with the North American Indians, but gives no authority (Prehistoric Times, 2nd ed. p. 510). [So far the author's MS. It may be observed that the first three regions taken by him India, China, Polynesia are the seats of the three great systems of s infanticide with male kinship. After them he proceeds to different parts of the globe without making any observations on the kinship, which indeed is often mixed ; e.g. in Australia female kinship prevails as a rule, but male is found alongside of it. We will now give from his notes a number of quota- tions to show further the extent of the practice, and to illustrate his views on some points not yet noticed. New Zealand. "Infanticide was formerly very common. It was generally perpetrated by the mother. ... A woman of the Thames destroyed seven of her children; the reason she assigned for such unnatural cruelty was that she might be light to run away if attacked or pursued by the enemy ; this was especially the fate of female children " (Taylor, Te Ika a Mani, 1855, p. 165). Australia. " Infanticide is very common. . . . The first three or four are often killed ; no distinction appears to be made in this case between male or female children" (Eyre, Discoveries in Central Australia, 1845, vol. ii. p. 324). Com- pare Bon wick's Wild White Man and Blacks of Victoria, 2nd ed. p. 48, where Eyre is fully confirmed, and there are some good remarks on the subject. Bonwick also declares that " abortion was not uncommon, especially after a quarrel vii FEMALE INFANTICIDE 97 between man and wife." " One female was pointed out to Mr. Wedge as having destroyed ten out of eleven of her children " (quoted from Dr. Eoss). In cases of extremity a man will kill his child to satisfy his hunger (Journal of Anthropological Institute, vol. i. p. 78). " Infanticide appears to have been very prevalent among the Aborigines before the commencement of this colony. I have been assured by Narrinyeri that at that time more than one -half of the children born fell victims to this atrocious custom. One intelligent woman said she thought that if the Europeans had waited a few more years they would have found the country without inhabitants. She herself had de- stroyed one infant. I know several women who have put to death two or three each of their new-born children. . . . Infanticide is not prevalent among the Narrinyeri at the pre- sent time. Thirteen years ago one-third of the infants who were born were put to death. Every child which was born before the one which preceded it could walk was destroyed, because the mother was regarded as incapable of carrying two. All deformed children were killed as soon as born. Of twins one, and often both, were put to death. . . . This terrible crime is covered up and concealed from the observa- tion of the whites with extreme care." l The Narrinyeri have male kinship. Papuans or Negrittos. Apart from Fiji we have not found any notes of the author's on infanticide among this race, except Turner's statement that Mr. Gordon thought there were some cases of infanticide in Erromango, and that an infant was buried alive with its mother on her death. But the Rev. J. G. Paton says of the New Hebrides : " Polygamy, with all its accompanying cruelties and degradations, univer- sally prevails. Infanticide is systematically practised ; and even the despatch of parents, when they grow old and help- less. Widows are put to death on almost every island, to 1 The Narrinyeri: an account of the tribes of South Australian Abori- gines inhabiting the country around the Lakes Alexandrina, Albert, and Coorong, and the lower part of the river Murray. The Eev. George Taplin. 1874, pp. 10 ff. H 98 STUDIES IN ANCIENT HISTORY CHAP. bear their husbands company into the spirit world. There is not an unmentionable vice hinted at in Eomans i. which is not unblushingly practised in those islands, wheresoever the Gospel has not dawned " (Autobiography, Appendix B). Africa. The Fans kill all sickly children (Win wood Eeade, Savage Africa, p. 244). Among the Wanika infanti- cide of misshapen children is authorised by the chiefs and executed on their orders (Krapf, Travels in Eastern Africa, 1860, p. 193). "The children belong not to the parents but to the mother's eldest brother, who not unfrequently sells them into slavery in times of scarcity " (Lieut.-Col. Playfair, Letter of 9th April 1864. See p. 229). Compare Mungo Park's Travels (Chambers, Edinburgh, no date), pp. 238, 239. In Arebo twins are killed, though in the rest of the Benin territories they are esteemed a good omen (Bosnian's Guinea, Pinkerton, vol. xvi. p. 526). The Galles", a robber horde, " every eighth year carry their wives with them, and expose their children without any tenderness in the woods, it being prohibited on pain of death to take any care of those which are born in the camp" (Lobo's Voyage to Abyssinia, Pinker- ton, vol. xv. p. 7). The Wa-Kurafi in a famine had to sell their children; the same tribe show a preference for boys (Thomson, Through Masai Land, pp. 417, 419). The people of Senjers sell females as slaves, but never males (Krapf s East Africa, p. 68). For infanticide among the Bushmen, Hotten- tots, and Zulus, see infra, pp. 482, 487. Munzinger testifies that among the Bedouins in or near Massua, " If a girl bears a child it is killed by its grandmother, without justice inter- fering." Again : " As among the Marea, getting a maiden with child, whether she be noble or not, must be atoned for with blood ; the two guilty ones and the child are slain by their own brothers without exception. For a widow or divorced wife the man only pays a fine, but the child is buried alive." In Eunama, " The eldest brother takes the place of the dead father. The power of the father goes no further ; the life of the child and its freedom belong to its uncle on the mother's side" (OstafriJcanische Studien, pp. 145, 322, 477. Schaff- hausen, 1864). vii FEMALE INFANTICIDE 99 North America. The Koniagas " prize boy babies, but frequently kill the girls" (Bancroft, i. 81). "Loucheux mothers had originally a custom of casting away their female children, but now it is only done by the Mountain Indians " (Simpson's Narrative, p. 187). Among the Haidahs "abor- tion and infanticide are not uncommon" ; the Nootkas " fre- quently prevent increase of their families by abortion " ; "infanticide and abortion are of frequent occurrence" among the Chinooks (Bancroft, i. 169, 197, 242). The inland tribes of California practise abortion often, and " although children and old people are as a rule kindly cared for, yet so great the straits to which the tribes are reduced by circum- stances, that both are sometimes abandoned if not put to death" (i. 279); and in a note is added that it was not uncommon for the Pend d'Oreilles to bury the very old and very young alive, " because, they said, these cannot take care of themselves and we cannot take care of them, and they had better die." There is no abortion among the Central Cali- fornians, but "it is stated that many female children are killed as soon as born" (i. 390). Among the South Cali- fornians the woman retires to be delivered alone near some stream or hole of water. As soon as the child is born, sheH throws it into the water, whence " if it rises to the surface and '. cries, it is taken out and cared for ; if it sinks there it \ remains, and is not even awarded an Indian burial" (i. 413). / The Lower Californians sometimes abandon or kill a child when food is scarce (i. 566), and there are more men than women among them (Smithsonian Report, 1863, p. 367). Aban- donment of the aged is extremely common in North America. See, e.g. Catlin, Letters and Notes, 3rd ed. vol. i. p. 217 : "The tribe were going where hunger and dire necessity compelled them to go, and this pitiable object . . . who was now too old to travel . . . was to be left to starve. . . . This cruel custom of exposing their aged people belongs, I think, to all the tribes who roam about the prairies, making severe marches, when such decrepit persons are totally unable to go. ... It often becomes absolutely necessary in such cases that they should be left." A particularly horrible story is told in ioo STUDIES IN ANCIENT HISTORY CHAP. the Relations des Jtsuites, 1640, p. 33. A woman killed her brother and made his son kill his sister because she could not save more than one besides herself in her canoe. Similar instances are to be found in the same collection, 1636, p. 29 ; 1637, p. 17; 1641, p. 25. "Us font ordinairement mourir 1'enfant quand la mere le laisse si jeune, croyans qu'il ne fera que languir apres son deceds " (Ibid. 1634, p. 5; cf. 1636, p. 59; 1657, p. 48). The Indians of Central America are said to have gone to extremes in the use of abortives, but this is stated to have been in connection with habits of great licentiousness. " Among the Acaxees (New Mexico), if a woman dies in child- birth, the infant surviving is slain as the cause of its mother's death." In Central America " it is suspected that infant murder has something to do with the rarity of deformed people" (Bancroft, i. 590, 714, 773). Paiderastia occurred among the Apaches. "According to Arlegui, Eibas, and other authors, among some of these nations (i.e. the New Mexicans) male concubinage prevails to a great extent; these loathsome semblances of humanity, whom to call beastly were a slander upon beasts, dress them- selves in the clothes and perform the functions of women." " Gomara says that in the province of Tamanlipas there were public brothels where men enacted the part of women " (often 1000 inmates or so, according to size of village). Paide- rastia was practised by the nations of Cueba, Careta, and other places (Central America). " The Caciques and some of the head men kept harems of youths, dressed as women." Among the same people we also find abortions. " In Tlascala and the neighbouring republics this (Quecholli) was the month of love, and many young girls were sacrificed to the goddesses of sensual delights. Among the victims were many courte- sans." They were volunteers, and before sacrifice they were privileged to insult their chaster sisters. " It is further said that a certain class of young men addicted to unnatural lust were allowed at this period to solicit custom on the public streets" (Bancroft, i. 515, 585, 635, 773; ii. 336). In Mexico sodomites were hanged; in Tezcuco, put to vii FEMALE INFANTICIDE 101 specially horrible death. " In Tlascala the sodomite was not punished by law, but was scouted by society, and treated with scorn and contempt by all who knew him." Bancroft (ii. 467) remarks: "From the extreme severity of the laws enacted by the later sovereigns for the suppression of this revolting vice, and from the fact that persons were specially appointed by the judicial authorities to search the provinces for offenders of this class, it is evident that unnatural love had attained a frightful popularity among the Aztecs. Father Pierre de Gand, or, as he is sometimes known, de Mura, bears terrible testimony to this ; he writes, ' Un certain nombre de pretres n'avaient point de femmes, sed eorum loco pueros quibus abutebantur. Ce pe"che etait si commun dans ce pays que, jeunes ou vieux, tous en etaient inf cote's ; ils y e*taient si adonnes que memes des enfants de six ans s'y livraient.' Las Casas relates that in several of the more remote provinces of Mexico, unnatural vice was tolerated, if not actually permitted. 1 And it is not improbable that in earlier times this was the case in the entire empire." All the Maya . nations had strict laws against paide- rastia ; but Bancroft seems to think they were not enacted till they were much needed. " Las Casas says that sodomy was looked upon as a great and abominable sin in Vera Paz, and was not known until a god (variously called Chin, Cavil, and Maran) instructed them by committing the act with another deity. Hence it was held by many to be no sin, inasmuch as a god had introduced it among them. And thus it happened that some fathers gave their sons a boy to use as a woman ; and if any other approached this boy, he was treated as an adulterer. Nevertheless, if a man com- mitted a rape upon a boy he was punished in the same manner as if he had ravished a woman. And, adds the same author, there were also some who reprehended this abomin- able custom" (ii. 677). 1 See footnote in Bancroft, the point of which is that M. de Pauw states that this paiderastia was common among the Mexicans, while Oviedo testifies that it was prevalent in Tabasco. 102 STUDIES IN ANCIENT HISTOR Y CHAP. In Nicaragua sodomites were stoned to death. " In Guatemala, Las Casas tells us, the men never married until they were thirty, notwithstanding he had previously made the extraordinary assertion that the great prevalence of un- natural lusts made parents anxious to get their children wedded as early as possible." The tradition is that the Quinames, who preceded the Olmecs in Puebla, were addicted to paiderastia, " a vice which they refused to abandon even when they were offered the wives and daughters of the new comers," the Olmecs (ii. 664, 678; v. 198). Peru. "There were sodomites in some provinces, though not openly and universally, but some particular men, and in secret. In some parts they had them in their temples, because the devil persuaded them that their gods took great delight in such people." Paiderastia also prevailed in coast valleys of the province of Camana ; severe measures were taken against it by the Yncas. "A law was issued that if hereafter any one should fall into this habit, his villages should be destroyed for one man's crime, and all the inhabitants burnt." We find it again in the province of Huayllas. " Such a crime had never been known or heard of before amongst the Indians of the Sierra, although, as we have before mentioned, it has been found to exist in the coast valleys." There were " many sodomites among the Chinchas," and the coast tribes (near Tunpiz) at Sullana "practise the infamous crime." "The natives of Manta and its district, particularly on the coast (but not the Serranos), committed sodomy 'more openly and shamefully' than any other nation that we have hitherto mentioned as being guilty of this vice " (Royal Commentaries of the Yncas, i. 59, 245; ii. 132, 154, 425, 441). The Peruvians even after their intercourse with the Spaniards continued to put to death feeble, deformed, or defective children (Eobertson, Hist, of America, 2nd ed. 1778, vol. i. p. 471). Common as unnatural vice is throughout every part of America, and many more cases might be adduced from that continent, it does not seem that it is at all an exceptional phenomenon there. We have already seen it alluded to in vii FEMALE INFANTICIDE 103 Polynesia and China ; indeed it prevails over the whole of the East. In Kamtschatka it is common (Klemm, ii. 207). The ancient Germans had to repress it (Tacitus, Germ, xii.) Africa is the only quarter of the world from which the author had no notes on the subject. South America excluding Peru. Divorce is free to the two sexes of the Guanas, and the women are much given to it. " Cela vient de ce que leur nombre est beaucoup raoins con- siderable que celui des hommes. Cette inegalite* ne vient point de la nature. . . . Elles (les femmes) de*truisent la plupart des filles dont elles accouchent" (Azara, Voyage dans I'Amdrique Mdridionale, Paris, 1809, vol. ii. p. 93). The custom is so strong that the women kill the girls even though offered money for them, their theory being that it is " pour faire rechercher davantage les femmes et pour les rendre plus heureuses." Of the Mbayas Azara observes: "J'ai deja dit qu' elles se prostituaient aisement ; mais ce qu'il y a de plus singulier, c'est qu'elles aient adopte la coutume barbare et presque incroyable de n'elever chacune qu'un fils ou une fille et de tuer tous les autres " (ii. 115). He says they ordinarily preserve only the last born when they expect no more, and if they are mistaken in their calculation, they rear none at all or kill any subsequently born. He reproached them for thus arranging to exterminate their nation, and was answered that men had no business to meddle with women's affairs. The method was abortion. The Guaicurus had once been a proud and fierce nation, but Azara found only one man left with three wives. " L'extermination deplorable de cette courageuse et superbe nation ne vient pas seulement de la guerre con^ tinuelle qu'elle n'a cesse" de faire aux Espagnols et aux Indiens de toute espece, mais aussi de la coutume barbare adoptee par leur femmes, qui se faisaient avorter, et ne conservaient que leur dernier enfant" (ii. 146). The Lenguas also were on the point of extinction, only fourteen males and eight females being left. " La destruction de cette nation provient egalement de ce que toutes les femmes ont adopte la coutume de detruire leurs enfans en se faisant avorter, a 1'exception du dernier, et de la meme maniere que les Mbayas" (ii. 148). 104 STUDIES IN ANCIENT HISTORY CHAP. This practice of abortion, which is so widely spread throughout the whole of America, must have supervened on a practice of infanticide. For killing feeble or defective infants, and for female infanticide, compare Gumilla, ii. 233, 238. The Fuegians, " when pressed in winter by hunger, kill and devour their old women before they kill their dogs : the boy, being asked why they did this, answered, ' Doggie catch otters, old women no'" (Darwin, Voyage of the Beagle^ 18*76, p. 214). Desertion or killing of the old and infirm shows of course the same tendency as that which causes infanticide, the rule being simply that the stronger must survive and the weaker go to the wall. Dobrizhoffer in his account of the Abipones in Paraguay (3 vols., London, 1822) tells us that he has known some who killed all their children. " Mothers spare their female offspring more frequently than the male, because the sons when grown up are obliged to purchase a wife, whereas daughters at an age to be married may be sold to a bridegroom at any price " (ii. 97). The natural result was that there were more women than men among them. Here the usual practice is reversed, but the tendency is the same, to kill the useless and preserve the useful. Such a state of things could only be found in a comparatively advanced and peaceful population, and certainly would not be found among the most primitive and savage groups. Compare Studies in Anc. Hist., first series, p. 75, note. For killing and eating the sick see a Voyage to Congo, in Pinkerton, vol. xvi. p. 152 ; and Nieuhoffs Brazil, in Pinkerton, vol. xiv. p. 876. India. "The number of females (children) sacrificed in Cutch and Guzerat alone (for it is practised in several other provinces) amounted, by the very lowest computation, in 1807 to 3000 annually" (Col. Walker's Eeport). "For many years past none of the Jarejah tribes have reared their female offspring" (Letter of the mother of Jehajee to Col. Walker, 1807). Another letter states the period to be 4900 years that the Jarejahs have killed their daughters. Mr. Duncan, governor of Bombay, was instrumental in abolishing the crime FEMALE INFANTICIDE 105 among the tribe of the Eaj-kumars, in Juanpore, near Benares, in 1789 (C. Buchanan, Christian Researches, pp. 48 ff. and 56). The Padaei in India killed and ate the old (Herodotus, iii. 99). "The same custom," adds Eawlinson, " is said to have prevailed among the Massagetae (i. 216) and the Issedonians (iv. 26); and a similar one is mentioned by Strabo as existing among the Caspians (xi. p. 753) and the Derbices (Ibid. p. 756). Marco Polo found the practice in Sumatra in his own day." Compare Asiatic Researches, vol. x. p. 203, where it is par- ticularly stated that the Battas in Sumatra eat their own relations when aged. Tonquin. Infanticide must have formerly prevailed here, as Eichard speaks of a law "which forbids the exposing or strangling of children, let them be ever so deformed " (Hist, of Tonquin, in Pinkerton, ix. 757). But it appears from p. 723 that exposure was still common enough when he wrote. Japan. Miss Bird denies that infanticide is now practised, but it was till recently. Deformed children were reared or killed according to the father's pleasure. Abortion also was common (Recollections of Japan. London, 1819, pp. 95-97, 222). Ceylon. Infanticide was practised among the Kandyans, "but never, I think, to any great extent. I have, however, seen two persons who, when infants, were buried alive by their mothers and rescued " (Bailey, Transactions of the Ethnological Society, 2nd series, vol. ii. p. 293). The hill tribes of North Aracan expose children born before the marriage of the mother, a sign of infanticide in decay (Journal Anthrop. Inst. vol. ii. p. 239). Siberia. The Yakutz abandon the old and sick, and often sell their children to the Eussians (Bell's Travels in Asia, in Pinkerton, vii. 344). One may feel sure that such a people must have committed infanticide freely enough, though it is not distinctly stated. The Tschuktsches in North-Eastern Siberia kill all children bom with any defect. The son also kills his father as soon as he can no longer be of use to the family, from old age or sick- ness. Indeed it often happens that the sick man begs to be put an end to, and not to be left to die an unheroic, natural 106 STUDIES IN ANCIENT HISTOR Y CHAP. death (Sarytschew, Achtjahrige Eeise im nordostlichen Siberien, ubersetzt von J. H. Busse. Leipzig, 1805, part ii. p. 107). Kamtschafka. Several methods of abortion are practised, and new-born children are throttled and thrown to the wild beasts or dogs. There were particular women charged with killing children. Steller found some who had killed three or more, and had not the least remorse (Klemm's Allgemeine Cultur- GeschicUe der MenschJieit. Leipzic, 1843, vol. ii. pp. 208, 296). Greenland. Many boys are forsaken in youth because it is expensive to provide a boat for them, but still more girls perish from nakedness and starvation (Crantz's History of Greenland, Eng. Tr. London, 1820, i. 1*7 7). Nog ay Tartars. " In the plague and famine (in 1558) the Nogay Tartars came to seek relief from their enemies the Eussians, who bestowed their charity so ill that they died in heaps over the island; the rest the Eussians sold or drove from thence. The author could have bought thousands of pretty boys and girls from their parents for a sixpenny loaf a piece, but had more need of victuals at that juncture " (Jenkinson's Voyages and Travels, in Pinkerton, ix. 387). Mingrelia. " When they have not wherewithal to main- tain them, they hold it a piece of charity to murder infants new- born, as also they do such as are sick and past recovery." The bishops " permit the mothers to bury their new-born children alive" (Chardin's Travels, in Pinkerton, ix. 144, 145). The Svans "were in the habit of saving themselves any little trouble or expense incidental to the maintenance of female children by filling the mouths of their hungry girl- babies with a handful of hot ashes " (Phillips- Wolley, Savage Svdnetia, vol. ii. p. 92). They consequently practised marriage by capture regularly. For Arabia it is unnecessary now to do more than refer to Professor Eobertson Smith's Kinship and Marriage in Ancient Arabia, note 6 on chapter iv. Europe. " The story of the old men ' tired of the feast of life ' is obviously based on a tradition of customs which once existed in the north. Even in comparatively modern times the Swedes and Pomeranians killed their old people. . . . vii FEMALE INFANTICIDE 107 Perhaps a tribe of poor and hungry men would easily fall into the habit of killing the useless members of the family, and the practice may have survived long after the dreadful necessity had ceased. We find a notice of the custom in the Saga of Gottrek and Eolf : ' Here by our home,' says the hero, ' is Gilling's Kock. We call it the family cliff, because there we lessen the number of the family when evil fortune comes. . . . The old folk have free access to that happy spot, and we ought to be put to no further trouble or expense about them. The children push the father and mother from the rock, and send them with joy and gladness on their journey to Odin.' The situation of some of these ' Valhalla cliffs ' is still well known in Sweden. The lakes which stretch below were called ' Valhalla-meres,' or ' Odin-ponds.' ' The old people, after dances or sports, threw themselves into the lakes, as the ancients relate of the Hyperboreans ; ' but if an old horseman became too frail to travel to the cliff, his kinsmen would save him from the disgrace of ' dying like a cow in the straw/ and would beat him to death with ' the family-club.' l ' Similar stories are told of the Heruli in the dark forests of Poland ; ' and among the Prussians 'all the daughters except one were destroyed in infancy or sold, and the aged and infirm, the sick and the deformed, were unhesitatingly put to death/ 2 practices 1 Geijer, Hist, of Sweden, pp. 31, 32. One of the family-clubs is still preserved in a farm in East Gothland. For the Heruli, see Procopius, De Bell Goth., ii. 14 ; and Gibbon, Decline and Fall, c. 39. For instances among Icelanders, Westphalians, Slavs, and Wends, see Grimm, Deut. Alterth., pp. 486, 489, " Die Kinder ihre altbetagte Eltern, Blutfreunde und andere Yerwandten auch die so nicht mehr zum Krieg oder Arbeit dienstlich, ertodten, darnach gekoeht und gegessen, oder lebendig begraben" (Ibid. 488). 2 Maclear, Conversion of the Slavs, p. 166 ; Keysler, Antiq. Septent., p. 148, cites several curious instances of this custom in Prussia, from writers of local authority. A Count Schulenberg rescued an old man who was being beaten to death by his sons at a place called Jammerholz . . . and the intended victim lived as the count's hall porter for twenty years after his rescue. A Countess of Mansfield in the fourteenth century is said to have saved the life of an old man on the Liineberg Heath under similar circumstances. io8 STUDIES IN ANCIENT HISTORY CHAP. as remote from the poetry of the Greek description as from the reverence which might have perhaps been expected from descendants of the Aryan household " (Elton's Origins of English History, p. 90). Dasent (Introduction to the Story of Burnt Njal, p. xxv.) says of the ancient Norsemen : " It was the father's right to rear his children or not at his will. As soon as it was born the child was laid upon the bare ground ; and until the father came and looked at it, and heard and saw that it was strong in lung and limb, lifted it in his arms and handed it over to the women to be reared, its fate hung in the balance. ... If it were a weakly boy, and still more often, if it were a girl, no matter whether she were strong or weak, the infant was exposed to die by ravening beasts, or the inclemency of the climate. Many instances occur of children so exposed." And when Christianity was introduced into Iceland, Thorgeir pronounced that the beginning of their laws should be "that all men shall be Christian here in the land, and believe in one God, the Father, the Son, and the Holy Ghost, but leave off all idol- worship, not expose children to perish, and not eat horse-flesh" (Story of Burnt Njal, ii. 79). A remnant of the feelings which prompted female infanti- cide is to be found in the following extract from Tozer's Highlands of Turkey, ii. 120. It refers to the Christian population of the most civilised part of European Turkey : "At Zagora and elsewhere the value set on women is so small that when a female child is born there is weeping and mourning in the house, so that with regard to one sex they actually realise the custom of the ' Thracian wives of yore.' 1 A girl is considered an expense and unremunerative ; for if she marries she must have a dowry, and if she does not she becomes a permanent burden to her family. In consequence the parents endeavour to get their daughters married extremely early." For the exposure of children among the ancient Greeks 1 Herod, v. 4. vii FEMALE INFANTICIDE 109 and Bomans, and for the prevalence of habits resulting from a scarcity of women, it would be superfluous to give much evidence. It is enough to say that Plato and Aristotle both advocate infanticide, and that Christianity alone stopped it in the Eoman Empire. That this was especially female infanti- cide appears from the prohibition to kill male children and the eldest female (Mommsen's Hist, of Rome, English Translation, vol. i. p. 61). The Vendidad forbids abortion (Fargard xv.), as does the Code of Manu (viii. 37), from which it appears to have been at one time common among the Persians and Aryan Hindoos. The same conclusion may be drawn for the Franks from the Loi Salique, cinquieme texte, ch. xxi. art. 4 : " Celui qui aura donne des herbes a une femme pour qu'elle ne puisse avoir des enfants sera pumj^^-amerrdre de 2 5 CTTteniers." From the~foregoing survey of the facts the following pro- positions stand out clearly as results : 1. When hard pressed by want savages will kill their children without pity. If the want is chronic, they will do so regularly and it will become a system. In the same way they will desert or kill the aged and infirm, and this habit also has often grown up into a system. 2. Such a system will often last on being consecrated by long usage, even when there is no longer any actual pressure of want. But if the want is removed by any cause, so that it is possible to rear children, the practice will often be dis- continued. 1 Hence it is that in so many places we find traces only of such a system in decline, the killing of deformed children or of one of twins, or the burial of a child with its dead mother. 3. If it becomes more profitable to keep children than to kill them, they will be kept ; it is simply a question of demand 1 Thus, if the missionaries taught the people of Aitutaki to give up killing the old, they made it possible to preserve useless mouths by in- troducing the prolific pig. " The pigs I brought to your island on my first visit here," said John Williams to them, " have multiplied so greatly that all of you have now an abundance." 1 10 STUDIES IN ANCIENT HISTOR Y CHAP. and supply. Hence a great system of slavery, as in Africa, checks infanticide because it is worth while to rear children for the purpose of selling them. 4. Therefore female infanticide is the rule, because it is more profitable as a rule to savages to rear up boys than girls. But if by any chance it should become more profitable to keep girls than boys, the female infanticide will be replaced by male, as among the Abipones, or at least be stopped, as in Circassia. o. Abortion has sometimes replaced, or is found along with, infanticide. In such a case the proportion of the sexes must obviously tend to be more equalised. Also any difficulty about the blood -feud would be got over. But it is very seldom that abortion has completely prevailed over infanticide. 6. Where the balance of the sexes is upset and there is a scarcity of women, savages will satisfy their instincts as they_can. Other causes besides infanticide give rise to a scarcity of women. The widely-spread habit of killing widows at the funeral of their deceased husbands, best known in the case of the Indian suttee, 1 but also common in Africa, New Zealand, China, North America, acts in this way. The rule that widows may not re-marry is the same in effect. " Young widows of the higher castes in this country (India) cannot marry again, and they are very numerous ; for matches between old men and children are common," says Francis Buchanan (Journey, vol. i. p. 260). And it is a common rule in India that a girl must marry before puberty, so that she is probably left a widow at an early age or else can never marry at all. So again the virgins dedicated to the sun in Peru were very numerous, and were not available as wives except indeed to the Inca. And religious prostitution in temples, when marriage did not terminate it, also cut off many of the women, as the bayaderes in India. But all these arrangements are found only in advanced civilisations ; what is much more important in savage communities is the polygamy practised by the chiefs 1 In 1803 there were 275 women so burnt within thirty miles round Calcutta ; in six months of 1804 in the same area the number was 115 (C. Buchanan's Christian Researches in India. London, 1811, p. 39). vii FEMALE INFANTICIDE in and great men, who go a long way to monopolising what women there are, and compel the common people to resort to polyandry, etc. Any customs of the above-mentioned kinds help to increase the scarcity of women. The custom, of which several instances out of many have been given, of killing and of eating the old, is further evi- dence of the former existence at some time of a period of great scarcity of food, even if that scarcity does not still exist. Cannibalism in general appears to be probably due to the same cause, and is still more widely spread. Such customs, therefore, point to a condition in which we may confidently infer that the female children would have been got rid of.] CHAPTER VIII CAN EXOGAMY AS THE MARRIAGE LAW BE INFERRED IN ANY CASE FROM THE LAW OF SUCCESSION? I PROPOSE to consider whether exogamy and the law of succession peculiar to female kinship are in any cases so connected that, the one being stated to be present, we may infer the presence of the other. With the Nair family system we have the said law of succession, but clearly under circumstances that exclude an inference as to the law of intermarriage. The husband may be of any stock even, so far as we know, of the stock of the wife but he must be a member of a different family : there is no cohabitation between the husband and wife, who respectively continue to be members of their mothers' households. But let us ask what could prevent the succession of children to their father in the case of the wife being in cohabitation with the husband in his home. It will be said it is prevented in respect of the force of custom and the ancient system of kinship developed in the Nair stage. But how does that operate in cases per- mitting certainty of fatherhood ? By declaring that CHAP, viii SUCCESSION AND EXOGAMY 113 his children, however certainly they may be his children and no one else's, are not of his kindred, but of the kindred of their mother. When the step is taken which destroys the force of this declaration, and a son is permitted to be counted the kinsman of his father, it will at once be apparent that he is nearer of kin than either the father's brother or the son of his sister, and under every system it is the nearest of kin that is the heir. The instant effect, then, of the change in kinship is to alter the succession law. Let us recall one or two cases illustrating the pro- position. There were in New Zealand two sorts of marriage in the one the man entered the wife's family, and was counted of her hapu and totem, renouncing his own to that extent that he was obliged to fight against his own real blood relations. In these cases, the son being of the hapu of his father, there was succession of sons to fathers ; there being an artificial incorporation of the man in the clan of the wife, his son was acknowledged as being his nearest of kin. In the other sort of marriage the wife was procured either by capture or purchase. She went into cohabita- tion with her husband in his clan. The children were counted of her clan, and they were not the heirs of the father. His heirs were in his own clan, who alone were his kindred. Might we infer from the system of female succession in this case, supposing we did not know that the husband and wife were of different clans, that, in fact, they must have been of different clans to permit of such a succession law ? I think we might. I 1 14 STUDIES IN ANCIENT HISTOR Y CHAP. Let us take another case. In Guinea a man had a first or chief wife no doubt of a different totem from himself and a second wife who had come to be the most honoured, who was obtained by purchase and consecrated to his totem or bossum. Being dedi- cated to his totem formally, no doubt with the same formalities as elsewhere in Guinea and in most places in Polynesia were employed in the dedication of a child to its god and determined the clan -connection, the bossum wife became artificially or by a fiction of the same totem with her husband. The totem being taken through the mother, a man's son by that wife would now be his kinsman, and his nearest of kin, and almost certainly therefore his heir. It is not surprising, then, to read that in that case the husband anxiously guarded the fidelity of his bossum wife, while on the whole indifferent to the conduct of the others. She alone had the honour of being strangled at his grave. The children of the other wives, we know, were not his heirs at all, but the heirs of their mother's brothers. I cannot doubt but that we have in the bossum wife an indication of one main way in which was introduced the succession of sons to fathers. It will occur to every one that in Greece and Eome within the historical time, in which we find male kin- ship and the succession of sons to fathers, the wife on her marriage gave up her own gentile or family gods, and adopted those of her husband's family or gens. Why was this uniformly done ? By the ancient law, the children would have been of the mother's clan. vin SUCCESSION AND EXOGAMY 115 Was this a device or fiction by which, the mother joining the father's clan, the kinship between son and father was reconciled with the ancient system of kin- ship, and superseded the ancient succession law ? The fiction would serve another purpose. It has involved in it, of course, that the husband and wife have different gentile gods, so that their marriage would be permissible under exogamy. May we not imagine that fact as steadily visible through the fiction that veiled it ? and the presence of the fiction, as in general use, a strong suggestion, if not a proof, that exogamy was the law ? Eecurring to the subject of inquiry since, where there is cohabitation of husbands and wives, and hus- bands are the heads of houses in which their families are born, the law of succession, in certain cases, cuts off a man's children, and prefers to them his brothers uterine and sisters' children as his nearest of km, we have the unavoidable inference that, in such cases, a man's children are not esteemed to be of his kin ; and since they are of their mother's kin, we have the in- ference that husbands and wives must be of different kins wherever such a succession law prevails, i.e. there is exogamy observed in practice if not prescribed by law. And we may be sure that the law and the senti- ments relating to marriage will correspond with the facts. It seems to me this argument stands proof against any observation on the uncertainty of fatherhood, or, at any rate, will stand proof against any such observa- 1 16 STUDIES IN ANCIENT HISTOR Y CHAP, vm tion in any country where wives are jealously guarded, and death is the punishment of adultery ; but, of course, it would not apply where, as it is said is sometimes the case, men deliberately traffic in the intrigues of their wives. I have not, however, as yet found a case of that description so well attested as to be credible of the bulk of any population. CHAPTEE IX SOME EXAMPLES OF FABRICATED GENEALOGIES ADDUCED TO SHOW THE READINESS OF MEN IN ALL TIMES TO FABRICATE GENEALOGIES. ALL students of history are familiar with the legendary genealogies of Greece. Every deme, gens, and phratria had its own, which was rehearsed at its periodical festivals. None of them had any real element in it of historical tradition, none was quite consistent with itself or in agreement with any other. Yet were the Greeks quite contented with them, as we have been with the genealogies from Adam to Noah, notwithstanding their marvellous nature, and the genealogies deducing the nations of men from Noah and his sons. The motive of most of these genealogies was to account for the origin of the various groups they referred to, and in each case the local distribution of the sections of the group ; the origin was usually found in the local or group-god, or in some man sprung from the Earth, the divine mother of all ; the persons connected in the pedigrees were personified districts, nations, tribes, gentes, towns, mountains, springs, lakes, and rivers, 1 18 STUDIES IN ANCIENT HISTOR Y CHAP. connected as fathers and sons or daughters, or as brothers and sisters, etc., a narrative of the personal adventures of the more prominent of these "persons" completing the pedigree. It is difficult for us now to realise how such genealogies could ever have been devised or have obtained currency ; except indeed as concerns the absence of scruple all the world over in regard to pedigrees, and the readiness of men to credit anything that it does not specially interest them to deny. As to some of the peculiarities of these gene- alogies, we must remember that the whole fabric of society has been changed since they were invented, and, what is even more important, that there has been a com- plete change in religion and the standpoint from which we have to contemplate the past history of mankind. So far as these genealogies trace descents from father to son, they are proofs, notwithstanding their falsity, that at the time they were invented the family system was characterised by descents from father to son. Un- fortunately it is almost impossible in any number of cases to say when they were invented ; but in most cases it can be shown that they were of an antiquity lower than that of the earliest literary records belonging to the people they refer to. When again we study these records, we find that the older they are, such pedigrees as they set forth have the fewer human fathers interposed between the subject of the pedigree and the godfathers of origin. That is, the number of what we can believe to be known human descents diminishes. Once the pedigree gets to a god as father, ix EXAMPLES OF FABRICATED GENEALOGIES 119 or to a supernatural or mythical person allied to the gods, it is free to assume any shape the fancy of its fabricator may dictate. We are entitled, I think, to assume that known descents were a restraint on the free fabrication of pedigrees ; and e converso, that where such fabrication appears unrestrained there were few known descents or none making obstacles to fabrication. The inference from the facts, in short, would seem to be that, at the commencement of the genealogy-manufac- turing period, pedigrees tracing descent from father to son were more or less of a novelty; in other words, "N since we are everywhere forced to believe in a long history antecedent to the genealogy - manufacturing ( period, that the family characterised by descents from ) father to son was not primitive. As regards the ancient Greeks, I have elsewhere shown, 1 to a high order of probability, that most anciently they traced descents through mothers and not fathers. It is a minor point made in my argument on ancient Greek kinship that the majority of the divinities of immemorial Greek worship five out of eight were female : Here, Perse- phone, Athene, Demeter, and Aphrodite, and that the number of the Greek eponymse was also remarkably great considering the disposition of the later Greeks to substitute male for female pedigrees. Among the eponymse are Salamis, Corcyra, JSgina, and Thebe, Messene, Sparta, Athene, and Mycene all belonging to the prehistoric period ; whereas it has been proved that 1 " Kinship in Ancient Greece," in Studies in Ancient History [first series, pp. 195-246]. 120 STUDIES IN ANCIENT HISTORY CHAP. many of the eponymi of the genealogists were invented within historic times. Sparta, for example, is older than Spartus ; Mycene than Myceneus. Mycene as an eponyma is mentioned by Homer ; Myceneus, who sup- planted her, is, as Mr. Grote has pointed out, the creation of post-Homeric Greece. Hereinafter, in dealing with genealogies similar to the old Greek, I shall indicate from case to case what proof we have of kinship through women being earlier of date than the pedigrees traced through fathers. To show the nature of the social and religious changes which make it so difficult to sympathise with the fabricators of these genealogies, and the people who were so well contented with them, is beyond my purpose ; but I propose to enable the reader to realise the enormous scale on which such frauds have been perpetrated. Perhaps, however, the word fraud is too strong for the occasion. Of that I shall leave the reader to judge. In many cases the genealogies are un- doubtedly frauds and nothing else. In other cases their character is more doubtful. The dreams of a poet, the amusing inventions of a skilled raconteur, even the foolish hypotheses of " a great philosopher," may have been prompted by no intention to mislead. Should they ever be generally accepted as true, the explanation may be that they were first remembered as interesting and afterwards credited as being creditable or pleasing to those to whom they were addressed, notwithstanding that they were incredible in themselves and however untrustworthy the original authorities. ix EXAMPLES OF FABRICATED GENEALOGIES 121 As to the readiness of men to fabricate genealogies there can be no doubt. It is constantly being exhibited in our own days. A novus homo desires a noble pedigree. He will surely have it if he will pay for it. Coulthard, of obscure origin, having become rich, found himself de- scended from a distinguished Eoman general. The lineage found a place as authentic in standard works until Mr. Burnett demonstrated its falsity. The excep- tional thing here was the suddenness and completeness of the exposure. Such fabrications have gone on in all time. Plutarch says that in his day some Eoman families traced their genealogies by documentary evidence up to Numa, notwithstanding the complete uncertainty of that mythical person's marriages and family, and that the archives of the city had been destroyed when Eome was sacked by the Gauls. 1 Then, as now, there were men whose business it was to make pedigrees for those who would deduce their lineage from ancient and illustrious houses. Families, like in- dividuals, are born and die ; a new family, a new man ; and for the new man sooner or later a history of respectable antiquity. And if this be so under the modern family system, how must it have been with the gentes under the ancient tribal system, seeking to account for their origin in circumstances altogether 1 Some say Numa had no wife but Tatia, and but one child, Pompilia ; others that he had four sons, the eponymous ancestors of the Pomponii, Pinarii, Calpurnii, and Mamercii. " A third set of writers," says Plutarch, "accuse the former of forging these genealogies from Numa to ingratiate themselves with particular families." 122 STUDIES IN ANCIENT HISTOR Y CHAP. different from those in which the resolution of great tribes of descent into gentes took place ? The Stewards were a great family in the time of David I. of Scotland. What was their origin ? No one knows. Yet they have a pedigree with Bern for first father, Bern with " ears like a bear," and, indeed, the son of a bear by "a certain Princess of Denmark." Bern begat Siward, Earl of Northumberland ; Siward begat ? But here discrepancies begin. Walter, the first Steward of Scotland, was, according to one account, Siward's grandson ; according to another, his great-grandson. Walter begat Alan, the Steward of Scotland ; but Hailes has shown there" is no evidence that this Alan ever existed. 1 There was a Walter, Steward of Scotland, in the reign of Malcolm IV., whose father was said to be Alan ; but Hailes says he can only answer the question, Who was this Alan 1 by asking, Who was the father of Martach, Earl of Marre, in the reign of Malcolm III. ; of Gilchrist, Earl of Angus, in the reign of Alexander I. ; of Fergus, Lord of Galloway, in the reign of Malcolm IV. ; or of Friskinus de Moravia, ancestor of the family of Sutherland, in the reign of William the Lion ; or, to keep in the supposed line of the royal family of Stewart, who was " the father of Banquho, Thane of Lochaber " ? It is implied that none of these men had ascertainable fathers. As to Banquho, his father, by one account, was a son of Core, King of Munster ; by another, he was the son of Ferquhard, son of Kenneth III. ; while by a third 1 Annals of Scotland, vol. iii. pp. 53 ff. Edinburgh, 1797. ix EXAMPLES OF FABRICATED GENEALOGIES 123 account, he was fifth in direct descent from Eth, King of Scotland. Thus is it with old families as with new. When we go back far enough we reach a time when they were new and their pedigrees false and fabricated. There may have been, in the case of the Stewards, a tradition of gentile connection with the royal house of Denmark. There is at least a common Totem mark, and the inhabitants of a petty village in the Outer Hebrides on such a tradition, as I believe, trace back their descent to Danish kings. The readiness with which men have been used to fabricate pedigrees is none the less historically mis- chievous that sometimes the fabricators have really believed in at least the goal to which they worked back. I suppose the inventors of the earlier fabulous race of Scots kings were in good faith after a fashion. That was a case of fabrication prompted by historical confusion, the more interesting that it is connected with a change in succession law. Loarn More, King of Dalriada, was succeeded not by a son but by a brother, Fergus the Great ; but till the sixth century the suc- cession among the Scots was not from father to son, but similar to that established among the Turks in the house of Othman a species of succession referable to polyandry as its origin. The change in the law came under the influence of Christianity in Fergus's time, and had effect in favour of his family ; so that he be- came the father of a dynasty. This Fergus, Mr. John Hill Burton observes, may be identified with Fergus II., the fortieth king of Scotland according to Buchanan and 124 STUDIES IN ANCIENT HISTORY CHAP. the older historians. 1 " A Fergus was still the father of the monarchy, but to carry back the line to a respectable antiquity, a preceding Fergus was invented who reigned more than 300 years before Christ much about the time when Babylon was taken by Alexander. To fill up the intervening space between the imaginary and the actual Fergus, thirty-eight other monarchs were devised, whose portraits may now be seen in the picture-gallery at Holyrood" a fact full of evil suggestions as to the value of the best evidence we have of the series of kings in some ancient Egyptian dynasties. Still more interesting is it, as touching the present purpose, that of the thirty-eight successions seventeen are repre- sented as being from father to son or grandson, the rest being usurpations or successions of brothers or nephews consistent so far as appears with the general succession of sons to fathers. 2 After this example of fabrication the reader will not be surprised to learn that the mother of the Scots, as a nation, was Scota, a daughter of the Pharaoh, whose line, if I recollect rightly, somewhere crossed that of the pious ^Eneas before her race finally settled in the north. 3 1 History of Scotland, vol. i. pp. 286 ff. Second Edition. Edin- burgh, 1873. 2 See Haydn's Dictionary of Dates. 3 When Fenius sent his disciples to learn all languages, it was Cai, a Hebrew, who went to Egypt and to Pharaoh its king. On the dispersing of the Fenians over the world, it was Cai who went with Pharaoh's messengers to fetch Fenius. " And the reward which they got was that Scota, the daughter of Pharaoh, was given in marriage " to Nel, son of Fenius, according to one account, but " ^Eneae filium nomine Nelum," according to another. " Hence the Scuit are called Scoti." The plagues ix EXAMPLES OF FABRICATED GENEALOGIES 125 The succession law, no less than numerous facts I have elsewhere adduced, explain why this Celtic people referred their origin to a first mother. The genealogies of Greece merit attention as being specially instructive. The facts respecting them have moreover been well made out and are accessible to all readers, while presently we shall have to do with facts neither so familiar nor so easily put to the test, the books containing them being somewhat rare. I shall attempt no more than to state the facts in the briefest way, following, on the whole, the authority of Mr. Grote. The pedigree of Hellen, the eponym of the Greeks, Hellenes, is variously stated. One account makes him son of Zeus by a nymph ; another the grandson of Prometheus and Pandora, with Deucalion for his father, and Pyrrha, daughter of Epimetheus, for his mother. One story again runs that he had three sons, Dorus, Xuthus, and ^Eolus. Dividing Greece between them, he gave Thessaly to JEolus, the Peloponnesus to Xuthus, and to Dorus the country lying opposite the Pelopon- nesus, on the northern side of the Corinthian Gulf. Xuthus had again two sons, Achseus and Ion. Thus the .ZEolians, Dorians, Achseans, and lonians are made to appear as descendants of a common ancestor. Though coming on Egypt, Fenius and his friends took to studying with the Israelites, and afterwards put to sea and avoided the Ked Sea disaster. Cai, meantime, crossed the desert with Moses, and afterwards went to Greece and settled in Thrace. (See Senchus Mor., vol. i. p. 21, and footnote, p. 20.) 1 26 STUDIES IN ANCIENT HISTOR Y CHAP. is represented in the above genealogy, which is taken from Apollodorus, as son of Hellen, yet the legends concerning the Solids, far from being dependent on this genealogy, are not all even coherent with it. The name of ^Eolus in the legends is older than that of Hellen, inasmuch as it occurs both in the Iliad and the Odyssey. Odysseus sees in the under-world the beautiful Tyro, daughter of Salmoneus and wife of Kretheus, son of ^Eolus. According to Euripides, again, Ion was the son of Kre'usa by Apollo, the real sons of Xuthus being Dorus and Achseus ; but Xuthus is in this scheme the son of ^Eolus the son of Zeus, and Hellen is left out, i.e. in the drama called Ion; but elsewhere Euripides mentions Hellen as the father of ^Eolus. According to the statement which we find in Dionysius of Halikarnassus, Achseus, Phthius, and Pelasgus are sons of Poseidon and Larissa. They migrate from Pelopon- nesus into Thessaly, and distribute the Thessalian terri- tory between them, giving their names to its principal divisions : their descendants in the sixth generation were driven out of that country by the invasion of Deukalion at the head of the Kuretes and the Leleges. 1 This was the story of those who wanted to provide an eponym for the Achseans in the southern districts of Thessaly. Pausanias accomplishes the same objects by different means, representing Achseus the son of Xuthus as having gone back to Thessaly and occupied the portion of it to which his father was entitled. Then by way of explain- ing how it was that there were Achaeans at Sparta and 1 Dionysius of Halikarnassus, A. J?., i. 17. ix EXAMPLES OF FABRICATED GENEALOGIES 127 at Argos, he tells us that Archander and Architeles, the sons of Achseus, came back from Thessaly to Pelopon- nesus, and married two daughters of Danaus. They acquired great influence at Argos and Sparta, and gave to the people the name of Achseans, after their father Achseus. 1 Enough of Greek genealogies. They were all of a piece. When a tribe or gens sought to account for its existence, it invented either a first man, sprung from the soil, or offspring of a god, or a hero of unquestionable repute ; and having regard to the local distribution of its sections, it invented a genealogy such as would in- clude them all in a common chain of connection with the first father. Hercules is thus, as Mr. Grote observes, as a progenitor, placed at the head of perhaps more pedigrees than any other Grecian god or hero ; and in this respect there was no difference on the Asiatic side of the Greek sea. Hector and the pious JEneas were made the first fathers of numerous tribes and gentes. More than a dozen places claimed that ^Eneas lived and died in them, and founded families. He was claimed as an ancestor in Thrace, in Delos, in Arcadia, the isles of Kythera and Zakynthus, on the Salentine peninsula, and other parts of the south region of Italy, in Sicily, at Carthage, Cumse, Misenum, and in Latium to which enumeration of Mr. Grote's it may be added, that not a few Scotch families trace back to him their ancestors. 1 A History of Greece, etc., vol. i. p. 99. By George Grote. Fourth Edition. London, 1872. 128 STUDIES IN ANCIENT HISTORY CHAP. It may suffice to say of the Roman genealogies, which Sir Henry Maine admits to have been for the gentes generally fictitious, that by especial preference for the same pious ^Sneas, they connect their general name of Latins with him by marrying him to Lavinia, the daughter of Latinus, King of Latium. Let us pass to less familiar countries. The gene- alogies of India make the brain reel by their bewildering complications and audacity. 1 We see them in the process of being fabricated as we pass from the oldest Vedic literature to that of the Puranas. The descents are from father to son as a rule, but "here and there tradition appears too strong for the fabricators, and they have to invent fables to explain how the father was a woman, or at once father and mother, in the genealogy. 2 It would, of course, be the cue of the Brahmans, who held the oral literature in their power to a large extent, to make it square with the most recent proprieties and best approved system of descents. In the early Yedic period the distinctions and attri- 1 In the Vishnu Purana, Vena, son of Anga, is the descendant in the ninth generation of the first Manu Svayambhuva. Muir points out that he belongs to a mythical age preceding by five Manvantaras, or periods of 308,571 years each, the beginning of the existing Manvantara, to which belong the descendants of Manu Vaivasvata ! 2 Manu is sometimes male and sometimes female, and so is Ila. Desirous of a son, Manu performed sacrifices, but owing to an error in the invocation, a daughter was born. She was turned into a man by certain gods, but again into a woman by Mahadeva, when she was got with child by a son of Soma. Then she became a man again. Her son by the son of Soma (according to the Vishnu Purana) is represented in the Mahabharata as having her both for father and mother. ix EXAMPLES OF FABRICATED GENEALOGIES 129 butes of the men of different castes are not yet fully defined, and much of the literature relates to the struggles of families to get recognised as Brahmans ; or of acknowledged Brahman families to wrest from the Kshattriyas the priestly functions which had anciently been exercised by them. "Where a family claimed to be Brahman it would have a motive to connect itself with acknowledged Brahman families by a genealogy going back to a common ancestor ; and Professor Wilson is of opinion that this motive was the main source of the multitude of genealogies and origins of castes, gods, men, and animals occurring in the literature ; while, according to Muir, there was in the majority of cases no closer bond between the families so fictitiously connected than that of a common gentile origin. It seems to me, indeed, that the feeling of stock, which everywhere is disclosed in the origins of caste, was stronger than any other feeling of the time. Various as are these origins, they have all one feature in common. The animal religiously regarded by the writer as the totem of his stock is always the animal next produced after man, and therefore is not to be eaten, which other animals might be. The sheep, goat, horse, cow, etc., have in turn this preference shown to them, and we may infer, indeed, the gotra of the writer from the preference. He is willing to go ever so far back for a common ancestor to connect his own with some already acknow- ledged high caste family, but subject always to fidelity to his totem. The first fathers, or Prajapatis, and their progeny 130 STUDIES IN ANCIENT HISTORY CHAP, vary, of course, in number and origin from case to case. The great Rishis, or " mind-born sons " of Brahma, are, according to different passages in the Mahabharata, six, seven, and twenty-one in number ; 1 they are seventeen in one place in the Ramayana, nine according to the Vishnu Purana, and ten according to Manu's account of the origin of castes. And even the same documents give contradictory accounts of their progeny. 2 Daksha, for example, had but thirteen daughters according to one account in the Mahabharata, while according to another he had fifty. "Daksha, the glorious Rishi, tranquil in spirit and great in austere fervour, sprang from the right thumb of Brahma, arid from his left thumb [note of Exogamy] sprang that great Muni's wife on whom he begot fifty daughters." : These daughters in turn " had valorous sons and grandsons innumerable." They became, in fact, the first mothers of great houses. Did they represent gentes in local tribes, and did the men they are represented as having married represent stock tribes summing up a variety of such gentes ? There are some suggestions of an answer in the affirmative. 4 According to one account, ten of Daksha's fifty daughters were given to Dharma, twenty-seven to Indu, and thirteen to Kagyapa. Now we know something of Ka9yapa. He was the incarnation of the Tortoise, or the god Tortoise. Muir 5 says this incarnation seems to be foreshadowed in the Qatapatha Brahmana : " As to 1 Muir, vol. i. pp. 116, 122. 2 Ibid. pp. 124, 126. 3 Ibid. p. 124. 4 Ibid. p. 126. 5 Ibid. p. 54. ix EXAMPLES OF FABRICATED GENEALOGIES 131 its being called kurma (a tortoise). Prajapati having taken this form created offspring. That which he created he made (aJcarot) ; since he made, he is (called) Kurmah. The word Kacyapa means tortoise; hence men say all creatures are descendants of Kayapa. This tortoise is the same as Aditya." In the Kamayana, Kacyapa is the grandson of Brahma, father of Vivasvat the sun, and grandfather of Manu ; * while, according to another account, Kayapa and Vivasvat are contem- poraries, and not related as father and son ; and Kacyapa marries eight (not thirteen) of the sixty daughters of Prajapati Daksha, one of these eight wives being Manu ! In these contradictions does there not seem to be indicated an assortment (not well defined, it is true) of families, gentes, or gotras of the Tortoise stock con- necting themselves with one another through descent from the Tortoise as first father, and the Kishi's daughters as first mothers ? For the filiation through first mothers there may have been tradition, or it also may be pure invention ; the only substratum of fact being the common totem stock, and the recognised gentile subdivisions of that stock in a variety of local tribes. Any number of Greek genealogies occur in which such would appear to have been the only facts guiding the inventor. The statement that the gotras were gentes with totems, and that that fact can be inferred from the Vedic Hymns, can, I think, be easily substantiated. For this purpose it will be convenient to deal with a i ii. 110, 2, ed. Schl. 132 STUDIES IN ANCIENT HISTORY CHAP. few of the principal totems under separate heads, premising that everywhere, as men -gods appear to supersede the totems, the first step in the success of the missionaries is to get the tribesmen to identify their totem with the man-god. Of the tortoise we have already seen enough. Let us consider next (1) THE HORSE. There is a controversy as to whether Agni was not a horse ; he was certainly a goat, as we shall see. MM. Bohtlingk and Eoth say that, in certain Yedic passages, " he (Agni) himself appears as a red horse." An Agni Hippius should no more create wonder than a Hippius Poseidon. (See Miiller's Rigveda Sanhitci, pp. 14-18 ; and see p. 27.) In the Pad ma Pur ana, Krishna in the form of a horse is represented as rescuing the Vedas when " the worlds " were burnt up (Muir's Texts, iii. p. 28, 2nd ed.); and in the Vishnu Purana we have the Sun as a horse teaching a horse- tribe men called Vajins (i.e. horses), from being instructed by the Sun-horse (Muir's Texts, iii. p. 51 ; and see p. 52, 2nd ed.). The Sun (Aditya) appears again as a horse in the ^atapatha Brahmana (ii. 3, 6, 9 ; iii. 5, 1, 19-20, trans, in Sacr.Bks. of the East, i. 447; ii. 115 ; and see Muir, vol. i. pp. xii. and 12, where the horse is also identified with Yama and Trita). Since, as we shall see, there was a Brahmanic gotra named from the horse, we may well believe that these partial contributions to ancient Indian literature were made by men of the horse stock. (2) THE SWAN. It is said in the Bhagavata Purana (ix. 14, 48) that at one time there existed but one ix EXAMPLES OF FABRICATED GENEALOGIES 133 Veda, one God Narayana, one Agni, and one caste. This, we learn from the Commentator, was in the Krita age, and the one caste, he tells us, was called " Hansa " the Swan. The Hansas again are, in the same Purana, said to be one of four castes or tribes existing in a district exterior to India (v. 20, 4) ; and finally, we learn from the Lin'ga Purana (i. 17) that Hansa was a name of Brahma himself i.e. Brahma was called the Swan. How this god, reputed among some tribes to have been the Creator, came to be so named is explained at length in the last-mentioned Purana. When he and Vishnu had grown hot in controversy as to which of them had made all things, there suddenly appeared before them a luminous Lin'ga " encircled with a thousand wreaths of flame, incapable of diminution or increase, without beginning, middle, or end, incomparable, in- describable, undefinable, the source of all things." What happened on this appearing Brahma thus recounts : "Bewildered by its thousand flames, the divine Hari (Vishnu) said to me (Brahma), who was myself be- wildered, c Let us on the spot examine the source of (this) fire. I will go down the unequalled pillar of fire, and thou shouldst quickly proceed strenuously upwards.' Having thus spoken, the universal-formed (Vishnu) took the shape of a boar, and I immediately assumed the character of a sivan. EVER SINCE THEN MEN CALL ME HANSA (SWAN), for Hansa is Viraj. 1 1 Viraj appears to be the first -begotten of the male and female divisions of the Procreator (Muir, i. 36). We shall presently see that according to another set of the Vedic writers Viraj was a cow ! CHAP. 134 STUDIES IN ANCIENT HISTORY Whoever shall call me ' Hansa Hansa ' shall become a Hansa." There follows an account of their respective expeditions to explore the Lin'ga, which occupied them for a thousand years. The one found no top to it, the other no base. Bewildered, they both bowed to it, saying, " What is this ? " in answer to which the Lin'ga is reported to have said, " OM." It is reasonable to conclude that we have a Swan- tribe in the Indian Kansas : the tradition that Brahma was a Hansa is not likely to have originated except with Swans. (3) THE GOAT. I appeal to the following passage from the Taittiriya Sanhita (Black Yajur-veda, vii. 1, 1, 4 if.) as conclusive evidence of the soundness of the views I propound so far as the Vedic races are con- cerned. If any one will furnish a different and satis- factory explanation of the passage, I shall abandon my hypothesis. " Prajapati (the Procreator) desired ' may I pro- pagate/ He formed the Trivrit (stoma) from his mouth. After it were produced the deity Agni, the metre Gayatri ... of men, the Brahman ; of beasts, the goats. Hence they are the chief, because they were created from the mouth. From his breast, from his arms, he formed the Panchadaca (stoma). After it were created the god Indra, the Trishtubh metre ... of men, the Kajanya (Kshattriyas) ; of beasts, the sheep. Hence they are vigorous, because they were created from 1 Muir's Sanskrit Texts, vol. i. pp. 158, 498 ; vol. iv. pp. 388-391, 2nd ed. ix EXAMPLES OF FABRICATED GENEALOGIES 135 vigour. From his middle lie formed the Saptadaga (stoma). After it were created the gods (called) the Vicve-devas, the Jagati metre ... of men, the Vaigya ; of beasts, kine. HENCE THEY (KINE) ARE TO BE EATEN, because they were created from the receptacle of food," etc. etc. Along with Qudras, in the lowest place, was produced the horse. The narrative is that Agni, the Brahman caste and the goat, were first created ; next Indra, the Kshattriya caste and sheep ; thirdly, the Vaigya caste and kine ; lastly, the Qudras and horse. And the kine, as having come from the middle, were to be eaten; which, by implication, goats and sheep were not to be ! If the reader will look at Muir's Texts, vol. i. 16, he will see that in another account (i. 26) kine were the first creatures produced after men, and it is familiar that in later times the cow came to be in India the most sacred thing on earth, next to a Brahman (see Manu, c. xi. 60, 79, 80) not to be eaten or injured, while goats and sheep might be. What, then, is the explanation of this? It is that the cow-stock came slowly into the first place ; that the contributories to the Vedic literature, even subsequently to the establishment of castes, were still so far in the Totem stage as to retain their Totem preferences ; that men of the goat, sheep, horse, and serpent tribes were contributories to the Vedas, as well as, or even more prominently than, men of the cow, ox, or bull tribes. It is in accordance with our hypothesis that Indra should be identified with the horse by men of the horse-stock, similarly that the 136 STUDIES IN ANCIENT HISTORY CHAP. sheep-tribe, taking him up, should make of him a ram as Mr. Muir assures us some Vedic writers did. As with Indra, so with Agni, and the other gods specula- tively produced ; the god, whoever he was, that was put in the first place by a tribe, was identified with its totem. On this view Agni, being represented as pro- duced along with Brahmans and goats, may be believed to be, in the writer's opinion (clearly a man of the goat- stock), foremost of the gods. He should therefore be a goat. Accordingly it did not surprise us when we found that Agni, as connected with the creation, was a /le-goat, and, in a procreative view, a she-goat } " the unborn female," the mother, we presume, of all creatures. 1 The goat, we shall see, gave its name to a Brahmanic gotra. (4) THE Cow. Every one knows that the cow is now religiously regarded almost universally in India is the creature next most sacred to a Brahman. It was not always so, however. It was the totem of one stock only, and had to fight its way up slowly to its present prominent position. It would take a book to treat fully of the worship of the cow. The reader who would understand its position in India may consult Muir's Sanskrit Texts at the points under noted : Part I. (2nd ed.) pp. 96, 217, 285, 287, 325, 374, and 390. At p. 217 he will find that Manu and 1 See Muir's Sanskrit Texts, vol. i. p. 16, 2nd ed. ; and vol. iii. pp. 166, 310, 311, 2nd ed. That Agni should turn out to be a goat was a prediction made by me on reading the above passage. The passage showing him to be a goat was then untranslated in Mr. Muir's work. ix EXAMPLES OF FABRICATED GENEALOGIES 137 Yama were descendants of the divine cow Viraj, 1 and at p. 325, "the cow with thrice seven names," too sacred to be divulged; Part III. (2nd ed.) pp. 248, 249, 253-254, 264, 266, and 282 ; Part IV. pp. 59, 63, 70, 77, and 145. Perhaps the most striking manifestation made by the celestial cow is that noted in Part I. p. 391, from the Mahabharata. The cow being incensed, discharged firebrands from her tail, arid created from that member and her sweat, etc., eleven tribes of armed men to fight for her ! (5) THE BEAR. The King of the Bears occurs in the ancient literature of India. He occurs also in Lapland, and every student knows him in Arcadia. Krishna appears in the Mahabharata as married to Jambavati, daughter of the King of the Monkeys, a lady who in the Vishnu Purana is daughter of the king, not of the monkeys, but of the bears. Jambavat, the lady's father, appears again in the Bhagavata Purana, and there he is not only the King of the Bears, but a celestial personage. Hari having gained a victory, the gods assemble to do homage to him, and celebrate his triumph, which is proclaimed by Jambavat. " Jam- bavat, King of the Bears, swift as thought, proclaimed this victory, the occasion of great festivity, with sound of kettledrums, in all the regions ! " a proceeding competent to a celestial only, we should say. 2 The Bear, we shall see, gave its name to a Brahmanic gotra. 1 The reader will see ante, p. 133, that it was the swan (Hansa) that was Viraj. 2 Muir's Sanskrit Texts, vol. iv. pp. 146 and 189. 138 STUDIES IN ANCIENT HISTORY CHAP. As proofs and illustrations of my propositions the above examples should suffice. We must not, how- ever, dismiss the subject of totemism in India without reference to the early history of Vishnu. In the Eig- Veda, according to the Commentary, he is a representa- tion of the sun, with powers derived from Indra, is not as yet among the Adityas, and, so far from being the Lord of Creation, is not even a god of the first rank. He and Brahma, indeed, as Muller observes, properly belong to a secondary, post-Vedic formation of the gods. 1 In the Brahmanic period we see him strongly impressing the popular imagination, and the germs of those legends appearing that reached their full develop- ment in the Epics and Puranas, and through which he attained a first rank, nay, even became the supreme god, as he appears in some parts of these works. These legends relate to his incarnations, of which, in the generally received account, the first was in a fish, the second in a tortoise, the third in a boar, and the fourth in a man-lion. The fish legend, among other details comprised in the form it finally assumed, re- presents the fish as instructing Manu in all wisdom. The legend wanting this detail is in the Mahabharata, and there the fish is Brahma ; and we have its original in 1 Vishnu and Brahma may have been tribal gods for any length of time. The meaning of Miiller's statement must be that they were of low rank in the group of tribes that comprised the chief contributories to the Veda. Probably they rose into importance, like other gods, with the tribes that possessed them. In what follows we have a hint of coalitions of tribes, which would explain their advancement. The history of Vishnu is ably traced in Muir's Sanskrit Texts, vol. iv., and in Chamb. Encyc., s.v. " Vishnu." ix EXAMPLES OF FABRICATED GENEALOGIES 139 the White Yajur-veda (Muir, i. 180 ; Qatap. Brahm. i. 8, 1), where the fish represents no god in particular, and the legend is introduced merely to explain certain sacrificial ceremonies. The legend of the tortoise-incarnation of Vishnu, again, is post-Vedic, while the idea of the Lord of the Creation becoming a tortoise is Yedic. It occurs in the Yajur-veda. In the Kamayana and Lin'ga-Purana (Muir, iv. 33, 39) it is Brahma, not Vishnu, who, as Creator of the Universe, becomes a boar. This belief first appears in the Black Yajur-veda, and there it is the Lord of Creation who is the boar, and not either Vishnu or Brahma. The original legend of the incarnation, moreover, represents it as cosmical ; it is emblematical according to a later conception ; while a third form of the legend has Vishnu for some time incarnate in the boar. During the avatara the gods, their very existence being threatened by an enemy, implored the aid of Vishnu, who "at that period was the mysterious or primitive boar." He slew the invader, which was but one of his many exploits in this character. As a man- lion he was of fearful aspect and size ; as a boar he was gigantic ; as a tortoise he was gigantic ; as a fish he filled the ocean. 1 In his fifth and subsequent avataras he was incarnate in men -gods, such as Krishna and Buddha, whose histories have been traced, the intention of the incarnations being obvious, namely, to effect a compromise with other religions, and, if possible, draw 1 Muir, i. 206. Will any one venture to suggest that Vishnu, a man-god who had an avatar as a tortoise, has degenerated into a totem of the Delawares ? 140 STUDIES IN ANCIENT HISTORY CHAP. their adherents within the fold of Brahmanism a policy that altogether has been highly successful. Was this the policy of the earlier incarnations ? We at once recognise the fish and man-lion totem gods, and can see how the policy that dictated an avatara in Buddha, and is now suggesting an avatara in Christ, to reconcile Brahmanism with Christianity, should have dictated an incarnation in the fish and man-lion. What, then, of the tortoise and the boar ? We say they were totem gods, and their avataras dictated by the same policy. Of the tortoise in mythology, except in this case, the present writer is almost ignorant ; 1 but he is a totem in America, and figures, as does the turtle, on coins of ^Egina of ancient date, ranging from 700 B.C. to 450 or 400 B.C., and was presumably a Totem god. Of the boar there is no doubt. He is worshipped now in China, and was worshipped among the Celts ; is a totem, and figures on the coins of many cities, and the crests of many noble animals with whose genealogies legends connect him. 2 Since the Vedic legends show 1 The Greeks had a few tortoise names and one nymph, Chelone, who was turned into a tortoise for not attending the nuptials of Jupiter and Juno. 2 For pig- worship in China, see American Expedition to Japan, p. 161. New York, 1856. Of the sacred pigs, in sacred styes at Canton, the writer says : " It was something of a curiosity, though somewhat sadden- ing in the reflections it occasioned, to behold the sanctified pork and the reverence with which it is worshipped." For Celtic pig-worship, see Transactions of the Ossianic Society, vol. v. p. 62, 1860. The Celtic legends of Ireland, Wales, and Scotland are pervaded by " the primitive, mysterious boar," and the Irish scholars connect him with the sacred swine of the ancient Celts, who, they suppose, had a " porcine worship which was analogous to, if not identical with, the existing worship of ix EXAMPLES OF FABRICATED GENEALOGIES 141 the fish, tortoise, and boar to have been earlier than Vishnu ; to have had to do with the creation with which he only lately came to be connected ; and since we have the key to the fictions by which each of them was at the later time made out to have been Vishnu, and so robbed of its primitive character by him, 1 we cannot doubt but that we possess in this case so many illustrations of the manner in which Zeus, Poseidon, Demeter, Athene, and others of the Egyptian and Greek gods superseded the totem gods of the earlier time, derived names from them, and came to be worshipped under their forms. As to ancient polyandry and female kinship among the Vedic Aryans, I must refer the reader to the argument on the Levirate. I shall here, however, adduce in conclusion some further evidence as to their having come through the totem stage, some of which also bears on their polyandry. The reader will find in Miiller's History of Sanskrit Literature (pp. 380-385) a list of the names of the most ancient Brahmanic gotras. A fact bearing on the ancient polyandry, it seems to me, appears ex facie of this list ; Vishnu in his avatar as a boar." Their boar, they may rely on it, was much more ancient than Vishnu, and worshipped over a wider area. He occurs on coins of various cities of Gallia, Hispania, and Britannia ; of Capua in Campania, Arpi in Apulia, Paestum in Lucania, JStolia in genere ; of ancient Athens, of Methymna in Lesbos, Clazomenae in Ionia, Chios in Ionia, and on several other classical coins all of date B.C., besides being figured on many ancient sculptured stones. 1 An instructive fact is that in Fiji two gods, who will naturally hereafter turn into men-gods, lay claim to the Hawk. 142 STUDIES IN ANCIENT HISTORY CHAP. and it is a fact which also bears directly on our present subject. Of the forty -nine gotras, only one has a single ancestor for certain. In two cases the ancestors may have been one or three. The remaining forty-six have all a plurality of ancestors. The significance of this appears when we recall the hints we have of the ancient form of the family in India the five Pandu brothers married to one wife ; the ten brothers, called Prachetas, that were married to Varkshi, the daughter of a Muni ; and " Jatila, of the family of Gotama, that most ex- cellent of mortal women, who dwelt with seven saints," all commemorated for us in the Mahabharata. 1 Another fact which only Sanskrit scholarship could deduce from it, is that the bulk of these gotra names are names of animals and plants. There were so many indications in the Vedas that this must be so, and the marriage law, which was exogamy, gave such promise of the preservation from a high antiquity of the ancient gentile names, that I thought it worth while when I began investigating totemism, to ask the late Professor Goldstticker to examine the list to see how far the etymons could be ascertained. The following is Professor Goldstiicker's report. It was put into type in 1869, from his MS., revised by him in proof, and is here printed from his re visa!. 2 " Amongst the names of gotras and their ancestors, 1 See Westminster Review, April 1868. The article therein on the Mahabharata is from the pen of the late Professor Goldstiicker. 2 I had hoped that the distinguished scholar would have completed for me an examination of the ancient stock-names of India, a task of love which he readily undertook. But it was not to be. ix EXAMPLES OF FABRICATED GENEALOGIES 143 there are several which unquestionably are names of animals or plants. Others may or may not designate such objects, since names of colours or qualities, for instance, also pass into those of animals or plants. " A third class, however, belongs to a different cate- gory, e.g. Jamadagni, ' fire - eater ' ; Sdvarni, from savarna, of ' the same colour ' ; Dlrghatamas, ( having great darkness'; and a fourth contains names which, with our present means, would not yield an intelligible etymon e.g. Ydska, Jaimini, Bddhaula, Garga, Kdnva, Gdlava, Kata, etc. 1 "The subjoined names are those the etymon or sense of*which, in my opinion, is safe. " N.B. In the list which Max Miiller quotes from A9valayana, the gotra names are given in the plural, e.g. Biddli, Qunakdh, this being the way in which the descendants of a celebrated man, collectively, are designated. In the following enumeration, of course, the base alone of such plurals had to be considered : Page 380. "1. Vatsa, a calf. 2. Anupa (ancestor name), a buffalo. 381. 4. Mauka, a descendant of Muka, dumb; or a fish. 5. Qyaita, a descendant of Qyeta, white ; or perhaps a hawk. 7. QunaJca, a dog. 8. Gotama, an excellent bull. 1 [In the St. Petersburg Diet. Garga is said to mean a bull, Gdlava and Kataha are names of trees. K.] 144 STUDIES IN ANCIENT HISTOR Y CHAP. Page 382. 14. Prishadaqva, a spotted horse, or having speckled horses. Ashtddanshtra, having eight large teeth (perhaps the name of an animal). 15. Riksha, a bear. 383. 22. Harita, green ; a lion. Pin'ga, tawny ; a rat. 25. Kapi, a monkey. 27. Qyflvagwi, a brown horse, or possessing such a horse. 384. 31. Dhanamjaya, the name of a Naga or snake-demon. 32. Aja, a goat. * 41. Kaqyapa, a kind of deer or fish. 1 385. L. 9 (below), Pain'galdyana, descendant of Pin' gala, tawny ; a monkey ; an ich- neumon; a kind of serpent; a small owl. L. 8 (below), Vdlmiki, from Valmlka, an ant-hill. 381. 11. SomardjaJci, descendant of Somardja, probably the soma plant. 382. 19. Mudgala, probably the same as Mud- gar a, jasmin. 383. 22. Darbhya, descendant of Darbha, the sacred grass (Poa cynosuroides). 23. Putimdsha, a sort of kidney-bean. 29. Kuyika, the sal tree, or from Kuga, Poa cynosuroides. 1 [This word also means a tortoise ; see above, p. 130, and B. and E. s.v. R] ix EXAMPLES OF FABRICATED GENEALOGIES 145 Page 384. 39. Venn, a bamboo. 383. 22. Can'kha, a univalve shell. 384. 38. Renu, dust, but also the name of a medi- cinal plant ; the former meaning is the usual one. 383. 24. Ajamllha, descendant of Ajamilha, goat's urine." Several of the names here appearing, e.g. the goat, bear, bull, horse, and soma, are names of stocks one would infer must have existed, from studying the pre- ferences, already referred to, shown to certain animals, etc., by the contributors to the Vedas. It is confir- matory of the view that the preferences were really totem preferences that the etymological evidence showed, after such inference had been made from such a study as the sole means of information, that among the ancient Vedic peoples stock-groups named after the animals did really exist. But with totemism and exogamy we should expect female kinship as an accompaniment, and descents traced through mothers instead of fathers. Let us now take an example or two from the races of Africa and Asia. The Berbers are a very ancient people. Herodotus makes some mention of one division of them, the Kabeyles. Another great division inhabits Morocco, and many less important are widely distributed over Africa. Ibn Batouta, early in the fourteenth century, found Berbers on the African coast opposite Aden. 1 1 Voyages d'Ibn Batouta. Texte Arabe, accompagn d'une Traduc- L 146 STUDIES IN ANCIENT HISTOR Y CHAP. He also found them established at several points in the Soudan. They are a flourishing race, and have been long under Islam, without, however, having fully adopted Mohammedan law. About the origin of the name Berber there cannot be two opinions. The Arabs call them Breber or Berbers. According to all appearance they borrowed the term from the inhabitants of the Koman towns, who treated as barbarians every people who did not speak Latin or Greek. They are called at present Amazirg in the empire of Morocco, and Kabeyles in Algeria. They are undoubtedly the same people who passed anciently under the general name of Libyans and Moors, if indeed they ever had a common name. The Greeks who knew them in the East gave them the name of Nomads, which was converted by the Eomans into Numidians. Their language has not the least analogy to Arabic, and their customs also differ from those of the Arabs. On the principle upon which the Greeks invented Hellen, and made of him a son of Deucalion, the Berbers, under the influence of ideas derived from Mohamme- danism, invented a first father Berr and connected him with Noah. We have a pretty full account of the genealogies framed to support this derivation from Ibn Khaldoun, a good Mussulman, who wrote in Arabic in the sixteenth century. 1 We have Ibn Khaldoun's work tion par C. Defremery et le Dr. B. E. Sanguinetti. 4 vols. Paris, 1853-58. Vol. ii. p. 180. 1 Histoire des Berbfrres. Ibn Khaldoun. 4 vols. Trad, par le Baron de Slane. Alger, 1852. ix EXAMPLES OF FABRICATED GENEALOGIES 147 in a French version by the Baron de Slane, who enables us by some notes to understand the deficiencies in the genealogies. He shows that one of the names given to a main branch of the Berbers, El Bko, is of Arabic origin, and that Arabic was already widespread in Mauritania before the Berber savants began to fabricate their genealogies. The genealogists again differ widely from one another, sometimes as to their starting-points, other times along the line. Madjhis and Bernes, two eponymous persons, are both called sons of Berr, but the genealogists are not agreed whether there was not more than one Berr. Bernes was, of course, the eponymus of the Beranes, of whom there were sundry tribes, whose first fathers should of course be sons of Bernes, but there must have been some obstacle in tradition to this being made out, for some genealogists refer two of the tribes, the Sanhadj and Lant, to a common mother and an unknown father; while Ibn-el-Kelbi, one of the genealogists, declares that the Sanhadj a and the Ketama did not belong to the Berber race at all, but were branches of the population of Yemen. The other tribes are run back, purely by an effort of the imagination, to first fathers of a very early date, the genealogies being here and there supported by the citations of morsels of ancient literature, which, De Slane points out, are in very bad Arabic, showing that they were fabricated by ill -educated Berbers, with a view to glorify the nation by giving them an Arab origin. As a specimen of the mode of working, it may be mentioned that one of the genealogies starts not 148 STUDIES IN ANCIENT HISTORY CHAP. from a Berr, but from a Berber, as eponymus. This Berber was son of Temla, son of Mazigh, son of Canaan, son of Ham, son of Noah (see Ibn Khaldoun, vol. i. pp. 167 ff.) Ibn Khaldoun's contribution to the subject is to show that all these genealogies, as given by the Berbers, are erroneous, except so far as that the Berbers "are the descendants of Canaan, son of Ham, son of Noah, as we have already announced in treating of the great divisions of the human family" (vol. i. p. 182). Can it be doubted that if the Berber genealogists had gone to work under the influence of the cos- mogonies of Greece or Kome instead of Islam, their pedigrees would have run back to Deucalion, Numa, or the pious ^Eneas, or some other celebrity ? Here again, as in the case of Greece, investigation discloses the fact that the Berbers anciently traced their descents through mothers and not through fathers. They to some extent do so to this day in some parts of the Soudan, as Dr. Barth attests (vol. i. pp. 337, 340 ff. ; and see vol. ii. pp. 273 ff.) The Kanuri, for example, call people in general, but principally their kings, after the name of their mother, and the mother's tribe is almost always added in the chronicle of descents. Again the Bedjahs, of Berber origin, according to Macrizi, " comptent leurs genealogies du cote des femmes. Chez eux I'heritage passe aux fils de la sceur et a celui de la fille, au prejudice des fils du mort." l 1 Macrizi's History of Nubia as translated by Quatremere. " M^moires ix EXAMPLES OF FABRICATED GENEALOGIES 149 The succession law again in various Berber states gave the throne not to the king's sons, but to his sister's sons. Ibn Batouta 1 found that law of succession in force at Tacadda, in the fourteenth century. Dr. Barth 2 says that he found the law of succession mentioned by Batouta " in full operation even at this present moment " among the Berbers of Tacadda, and that "it must be supposed to have belonged originally to the Berber race, for the Azkar, who have preserved their original manners tolerably pure, have the same custom." Barth adds that it may be doubted whether in the mixed empires of Ghanata, Melle, and Wulata, it was not the Berbers who introduced this custom. We may be sure, however, it belonged to the black natives as well as the Berbers. He mentions one noble Berber tribe that counted the custom shameful, as indicating mistrust of the wife's fidelity, " for such is certainly its foundation." Batouta also mentions, as a peculiar fact, that the pro- tection given to caravans by the Berber women might be more trusted than that given by the men, 3 a fact attesting the high position of women, which is true to this day of the Kabeyles, as General Daumas attests in his work on Algeria. 4 We can see in this case how Geographiques et Historiques sur 1'Egypte, etc., recueillis et extraits des MSS. Copies, Arabes et de la Bibliotheque Imperiale." 2 vols. Paris, 1811. 1 Vol. iv. p. 442. 2 Travels and Discoveries in North and Central Africa, 1849-55, vol. i. p. 337. 5 vols. London, 1857. 3 Vol. iv. p. 437. 4 Mceurs et Coutumes de TAlgerie. Paris, 1855. See pp. 225-227. That we have here a note of the system of kinship through women only 1 50 STUDIES IN ANCIENT HIS TOR Y CHAP. misleading it would be were we to trust the patriarchal genealogies of the Berbers for an inference as to their ancient condition, manufactured as the genealogies were on a foreign type, after their conversion to a religion which must have largely remodelled their manners and laws. I have made several attempts to obtain the etymons of the Berber tribes as given by Ibn Khaldoun, to see whether they were named after animals or plants, but hitherto without success. I am not without hope, however, that the tribal names will soon be examined. Barth has interpreted one or two. Among the " sections," i.e. sub-tribes, of the Kel-owi, for example, are the Kel-ulli, or people of the goats, and a sub-tribe of the same name occurs among the Awelimmiden, which looks towards totemism, exogamy, and female kinship ; but the facts are too few to justify more than a surmise. We have two facts from Macrizi (l) "Among the Berber Bedjahs there are some [it must mean a gens] qui s'arrachent les dents de devant, ne voulant pas, disent-ils, ressembler a des anes." 1 The mutilation would indicate a people of the cow stock. (2) " Chacun d'eux adore ce qui lui plait, une plante, un animal, ou un mineral,"' which is indicative of totemism imperfectly observed. we may see in the case of the Beni Amer. Among them " who ever recommends himself to the protection of a woman is safe, much safer than under the protection of a man." See Ost-Africanische Studien, p. 327. By Von Werner Munzinger. Schaffhausen. 1864. 1 Quatremere, I.e. vol. ii. p. 122. 2 Ibid. vol. ii. p. 188. ix EXAMPLES OF FABRICATED GENEALOGIES 151 The Arabs have been divided into the ancient and modern. The ancient Arabs are gone, leaving no traces behind, so nothing need be said about them. The modern Arabs are divided into the pure and the naturalised ; and we have genealogies for them tracing them up to Adam. The pure Arabs are the descendants of Kahtan. the Joktan of Genesis x. 28 ; the naturalised are the descendants of Ishmael, son of Abraham, who (i.e. Ishmael) it was that, with the help of the angel Gabriel, built the original temple at Mecca. The descendants of Joktan settled in the south ; those of Ishmael in the 'Hejaz. Those of both stocks that settled in towns became in a measure what we call civilised, and fell away from their primitive customs. The wandering Arabs, who kept to life in the deserts and who are called Bedouins, are believed to have preserved for us these customs in comparative purity, notwithstanding their acceptance of Islam. Of course an eponymus has been found for the whole race Ya'rab, son of Yoktan, after whom the people were called Arabs. Ya'rab's grandson, again, was Saba, after whom the southern Arabs were called Sabseans. Himyar was one of Saba's sons, and for 2020 years the government of the Sabaeans rested with the family of Himyar, whence it came about that the Sabaeans were also known as the Himyarides. Belkis, or Balkis, queen of the Sabseans, was of Himyar's line, if indeed she was of human origin. She has of course been identified with the Queen of Sheba who visited Solomon. How came it that there was a queen of Sheba under the patriarchal system? 152 STUDIES IN ANCIENT HISTOR Y CHAP. Joktan, Ishmael, and the Queen of Sheba are not the only biblical personages referred to in the Arabian genealogies. What the Arabs were to the Berbers, the Jews were to the Arabs. One of their genealogical epochs is marked as the time when Nebuchodonosor ravaged Arabia. On this Percival remarks, " On pourrait soupgonner que, comme d'autres recits du meme genre, dans lesquels figurent des personnages bibliques, il a ete je ne dis pas forge, mais arrange au temps ou les Arabes ont commence a etudier les livres desJuifs." 1 I follow Percival in saying that the Arabs have no writing older than the Goran, except some morsels of poetry of date about the birth of Mahomet, and some inscriptions on stones in Yemen, the sense of which has not yet been ascertained. It was some time after Islam that the first attempts were made to give form to the uncertain oral traditions of the people, and the books in which these attempts were made have come down to us only in disorderly fragments. These traditions, or legends, again are often contradictory and usually mixed with fable ; only as one approaches the era of Mahomet, " on voit dans les traditions s'effacer peu a peu la teinte fabuleuse, et apparaitre la caractere historique de plus 1 Essai sur VHistoire des Arabes avant I'Islamisme, pendant VEpoque de Mahomet, et jusqu'a la Reduction de toutes les Tribus sous la Loi Mussul- mane. Par A. P. Caussin de Percival, Professor d'Arabe au Coll. Koy. de France, etc. Three vols. Paris, 1847-48. Vol. i. p. 31. [This is our chief authority for the history of Arabia before Islam. When the author had, after years of labour, become familiar with all the Oriental MSS. in Paris, he was promoted to the chair of Belles Lettres at Rouen, and so lost to research.] ix EXAMPLES OF FABRICATED GENEALOGIES 153 en plus prononce." 1 Notwithstanding this, Percival has taken the trouble to form for us genealogical tables for the people, going back to the earliest times, and I fancy we can see in what he has done the modus operandi of the ancient (honest) fabricators of genea- logies : " Je les [the fables, etc.] ai exposes, en cherchant a les concilier, a les expliquer, ou plutot, j'ai fait un choix de celles qui m'ont paru meriter le plus d'interet." 2 It is as if he thought it possible that the genealogies, which exhibit the relations of the various tribes and trace them all back to Adam, might have some founda- tion in the traditions of the people, instead of being borrowed from Hebrew and other sources. As Professor Eobertson Smith has pointed out, on the authority of Noldeke, the Arabs had no such wonderful historical memory, since already by Mahomet's time they had no longer any trustworthy traditions of great nations who flourished after the time of Christ. 3 But indeed Percival himself admits that his materials fail him almost absolutely even in Adnan's line, about 130 B.C. I propose first to show the nature of the uncertain- ties surrounding the Arab genealogies, and next to show the numerous traces they contain of descents through women, and generally of female kinship ; and lastly, the strong reasons we have to believe that the Arabs 1 I.e. vol. i. pp. vi. vii. 2 Elsewhere, speaking of the tables he formed, he says, " Avec d'aussi faibles elements pour reconstituer, on ne peut aspirer de parvenir a la ve*rite. Peut-etre n'est-il pas impossible d'atteindre a la vraisemblance." 3 Journal of Philology, vol. ix. p. 80. 154 STUDIES IN ANCIENT HISTORY CHAP. before Islam were in the Totem stage, and the filiation of their groups purely gentile. I. A profound uncertainty is admitted to exist as to the pure Arabs, or Sabaeans, the descendants of Joktan, our " information " respecting them consisting only of vague traditions, lists of kings which do not agree with one another, and which present manifest lacunes, and genealogies broken and incredible. 1 We have an example of the uncertainty in the case of Chourahbll. He is " fils d'Amr-Dhou-1-Adhar, ou fils d'Amr, fils de Ghalib, fils de Mentab . . . fils de Sacsac ; ou enfin fils de Malik, fils de Kayyan . . . fils de Himyar." : Here are three choices for his parentage. His name varies like his genealogy. There are in fact four names for him, which it were needless to quote here. A similar uncertainty exists as to the father of Balkis ; there are at least two choices, while they say her mother was not a woman at all, but a being of the order of the genii. 3 As to the Arabs of Ishmael's stock, the irruption of the troops of Nabuchodonosor II. into the Hedjaz again is said to have occurred in the time of Adnan ; but no computation can put the birth of Adnan farther back than 130 B.C., so that the irruption must have taken place, if at all, four centuries before his time. 4 One can see that no proper materials exist even for tracing the descent from Adnan. 5 Percival, trying hard 1 Percival, I.e. vol. i. p. 47. 2 Ibid. p. 74. 3 Ibid. p. 75. 4 Ibid. p. 183. 5 Percival, p. 185, and see p. 196, vol. i., for proof of uncertainty in this genealogy ; also p. 208, vol. i., for uncertainty in the genealogy of Codhaa ; p. 216 for complete uncertainty as to the origin of the Khozaa ; ix EXAMPLES OF FABRICATED GENEALOGIES 155 to make something of his materials, and I think believing in them somewhat, finds himself forced to choose between Ishmael and Maadd, son of Adnan, as to which of them must be relegated to myth. He decides in a way against Ishmael "Dans ces faits doubles, Ismael est sans doute un mythe : Maadd est probable- ment la realite." The "double facts" are products of his own ingenuity, and there is probably as much reason to believe in Maadd as in Ishmael. If the credibility of this genealogy is to depend on the credibility of the main facts related in connection with it, there is an end of it. The stories of the building of the temple at Mecca, and of the visits of Abraham to his son's wives in Arabia somehow Ishmael was always out when his father called belong to a most interesting, though utterly incredible, order of fable. II. The number of descents traced through women is great, and can be explained only through the force of some traditions so strong that they could not be resisted. The mothers, moreover, are not only, as in the case of the Berbers, prominent in the genealogies, but they give their names to their lines. (a) I take first the case of Chakika. The tradition here might well be strong, for the facts were still also p. 228, vol. i., as to the genealogy of Holayl, ancestor of Mahomet. For the easier testing of these uncertainties, of which I shall say little more, I give here the following further references : id. vol. ii. p. 27, Amr's genealogy ; id. vol. ii. p. 54, Neman's ; and see note, vol. ii. p. 74, as to disagreement of authors ; and p. 74 itself for Percival's principles of reconstructing genealogies from his materials ; and see further id. vol. ii. pp. 135, 186, and 267. 156 STUDIES IN ANCIENT HISTORY CHAP. recent when the genealogists set to work. Moundhir III., who flourished 513-562 A.D., is called by Pro- copius, "descendant of Chakika." 1 He is also so de- signated by Theophanes ; and it is admitted, as sure, that this designation was borrowed from the Arabs of the time. Moreover, as Percival tells us, many Eastern historians declare that this Chakika was the head of the house of the Moundhirs ; was the mother of Noman le borgne, the father of Moundhir L, so that all the Moundhirs were " descendants of Chakika." Percival's attempt to set aside this view rests wholly on a conjec- ture of his own, which I shall consider presently. (b) Moundhir III. was not only called " descendant of Chakika," but he was also called by Arab historians " son of Ma-essema," after Ma-essema, his mother. To name men of the same name after their mothers might be convenient where they were of the same father (or of fathers of the same name) by different mothers, but there is here obviously something more than that. The Moundhirs are named after their mothers, and grouped as descended from a first mother. The son of Ma-essema is also the " descendant of Chakika." Why not of Noman le borgne ? (c) Percival's conjecture, to set aside the conclu- sion that all the Moundhirs were of the issue of Chakika, was that Chakika was the mother of Noman II., and the first wife of Moundhir L, and that this Moundhir also married Hind, daughter of Harith, so that in the princely family there were two lines ; that 1 Percival, I.e. vol. ii. p. 77. ix EXAMPLES OF FABRICATED GENEALOGIES 157 of Chakika, mother of Noman II., and that of Hind, mother of Moundhir II. But in another place we find that this Hind was the wife of Moundhir III. (562-574 A.D.), and mother of Amr III. 1 Moundhir, son of Ma-essema, left several sons, of whom Amr, the eldest, succeeded to the sovereignty. The Arab authors, it is said, ordinarily call him " son of Hind." It is not conceivable that Hind, wife of Moundhir III., can have been also wife of Moundhir I. and mother of Moundhir II. ; and it is nowhere stated that Moundhir II. was called by the Arabs " son of Hind," as Amr III. was. The conjecture, therefore, on which Percival tries to avoid recognising the race of Chakika seems without foundation. But in considering it we have found another case of a king with a designation taken from his mother. I may as well add here a few other examples of Arabs being called sons of their mothers : 1. Among the Khazodj, of the branch of Middjar, we find "Auf and Moadh, sons of Afra." Afr& was their mother ; their father's name was Harith. 2 2. In the pedigree of Amr, son of Colthoum, are given the names both of his mother and grandmother. His mother was Layla, and the mother of Layla was Hind, daughter of Badj. 3 3. The Benou-Maddjar, a Khazradjite branch, were commanded by an old man called "Amr, son of Zholla." Zholla was his mother ; his father was Mo&ura. 4 (d) There are some cases of tribes being named 1 Percival, I.e. vol. ii. p. 115. 2 Ibid. vol. iii. p. 2. 3 Ibid, vol. ii. p. 373. 4 Ibid. vol. ii. p. 655. 158 STUDIES IN ANCIENT HIS TOR Y CHAP. from women, and they are of date sufficiently recent to have possibly left a trustworthy tradition that they actually occurred. 1 1. Mention is made of the arrival in the Hejaz of two Arab tribes, "Aus and Khazradj," about 300 B.C. The authors of these families were two brothers. Their mother was Cayla. Aus and Khazradj (the brothers) are often designated Ibna- Cayla, children of Cayla, which denomination extends also to the tribes of which they were fathers. Here though first fathers were found for the tribes, yet the tribes took their name from a first mother, whose place in tradition was too well settled to be disturbed by the genealogists. 2. The tribe of Mozayna, settled in the Hejaz, are said to be descended from Amr, son of Odd, son of, etc. They were called Mozana, from the name of their grandmother Mozayna, daughter of Kelb, "son of Walra," wife of Odd, and mother of Amr. 2 3. One of the four Maaddique tribes, named Ad- wan, was so named after Ad wan, wife of Modhar. 3 4. The Beni Chindif are expressly said to bear the name of their mother, "not mentioning their father Ilyas." 4 It would be tedious to adduce more of this sort of evidence here. The genealogies abound in indications of the system of female kinship. 5 The position of 1 Percival, vol. ii. pp. 646-647. 2 Ibid. vol. ii. p. 527. 3 Ibid. vol. ii. p. 259. 4 Abulfeda, p. 196, 1. 4, cited by Professor Eobertson Smith, Journal of Philology, vol. ix. p. 88. 5 A striking note of this is the frequent appeals for protection, and ix EXAMPLES OF FABRICATED GENEALOGIES 159 women is exceedingly high and influential, the position of the maternal uncle resembles that of the mother's brother among the Iroquois or ancient Germans, and tribes are not only named from women, but fre- quently they are governed by them. The details of this evidence must remain over for handling in its proper place, and so must the evidence as to ancient Arab polyandry of both the Nair and Thibetan types. Professor Smith, who has lately been studying this subject, adduces some evidence to show that many centuries after Mahomet polyandry practically survived, "at least in South Arabia, in a grosser form than that described by Strabo." 1 Strabo is an authority for polyandry of the Thibetan type only as occurring in Arabia. Such marriage customs must necessarily have been accompanied at some time by the system of female kinship. It is, of course, needless to say that the polyandry again (and exogamy, which also pre- vailed) had its roots here as elsewhere in systematic female infanticide; and that this savage usage again had its roots in the smaller importance of women in hordes struggling for existence. As Percival puts it, the Arabs were "pousses a cet acte barbare [the slaughter of female children], les uns par la misere qui leur faisait craindre de partager leur nourriture avec un 6tre incapable de les aider; les uns par une fierte frequent adjurations in the name of mothers. The student of African tribes knows this as an infallible test of the system of kinship. Where men say, " my mother," instead of " O my father," kinship is invari- ably found to be traced through the mother only. 1 Journal of Philology, vol. ix. p. 87. 160 STUDIES IN ANCIENT HISTORY CHAP. feroce et un sentiment exagere de 1'honneur," 1 i.e. the motive to kill their female infants connected itself, on the one hand, with the scarcity of food and the need of warriors, and on the other, with the fear of attacks in which their girls if allowed to live would be cap- tured by other tribes as wives. III. I pass now to the subject of totemism among the Arabs. We saw in the case of Greece how the genealogies were truly designed to account for the origin of local tribes and their subdivisions, and the origin of tribes of descent, and of the distribution of the sections of such tribes in different local tribes. We shall find that the same was the purpose on the whole of the Arabian genealogists. They are shown by Sprenger to have ascribed gentile unity to mere [political] confederations. 2 We shall presently see reason to believe that their local tribes, instead of being each composed of the descendants of one man, in which case the tribesmen would be of one stock, were com- posites of various stocks, like the local tribes of the Mohawks or Delawares ; and that the stocks composing such tribes were mostly named after animals, and had been brought into conjunction in the manner I have elsewhere explained 3 by the joint operation of exogamy 1 See Percival, Lc. vol. i. p. 357, and pp. 373-374 ; and see Ibid. vol. iii. p. 2. In the first oath of Acaba, administered by Mahomet to his disciples, the disciples swore not thereafter to kill their children. 2 Sprenger I have not seen. The statement is made on Professor Kobertson Smith's authority (Journ. Phil., voL ix. p. 81). 3 Studies in Ancient History [first series, pp. 60, 128]. ix EXAMPLES OF FABRICATED GENEALOGIES 161 and the system of female kinship. I had brought together numerous indications of this from the pages of Percival and other writers which seemed conclusive to my mind, though they might appear faint to others, and may, therefore, be omitted here. Since I did so, Professor Smith's paper, to which I have already more than once referred, has been published, and a good commencement made of a research in the Semitic field, which I trust will be followed up. It is only by skilled labour of the sort by many workers in many fields that the truth can be surely ascertained. Professor Smith, premising that a very great number of tribes or families named after animals is to be found among the Arabs, writes as follows: "The following examples are gathered from the Lubbu-l-lubab (Suyuti's dictionary of gentile names), and make no pretence to completeness. " Asad, lion ; ' a number of tribes/ Aws, wolf ; ' a tribe of the Angar,' or defenders. Badan, ibex ; ' a tribe of the Kalb and others.' Tha'laba, she -fox ; e name of tribes/ Jwrdd, locusts ; ( a sub- tribe of the Tamim/ Beni Hamama, sons of the dove ; ' a sub- tribe of the Azd/ Thawr, bull ; 'a sub-tribe of Hamdan and of 'Abd Manah/ Jahsh, colt of an ass ; ' a sub-tribe of the Arabs/ Hida ', kite; c a sub-tribe of Murad/ Dhib, wolf; c son of c Amr, a sub-tribe of the Azd/ Pubey'a, little hyaena ; ' son of Qays, a sub- tribe of Bekr bin Wail, and Dubey'a bin Kabi'a bin Nizar bin Ma'add.' Dabba, lizard ; ' son of Udd bin Tabicha bin Ilyas bin Modar' (eponym of the Ben! M 162 STUDIES IN ANCIENT HISTORY CHAP. Dabba or sons of the Lizard). Also the ancestral name of families in Qoreysh and Hudheyl. Dibab, lizards (pi.); 'son of Amir bin Ca^a'a/ Dabab, a subdivision of the Ben! Harith and of the Qoreysh, is perhaps the same thing. 'Oqdb, eagle; ' a sub-tribe of Hadramaut.' 'Anz, she-goat; 'son of Wa/il, brother of Bekr.' The tribe of the ' Anaza, whose eponym is represented as the uncle of Wa'il, are probably not different in origin. Ghordb, raven ; ' a sub-tribe of the Fazara.' Qonfudh, hedgehog; 'a sub-tribe of Suleym.' Kalb, dog; 'a sub-tribe of Qoda'a and of the Beni Leyth and of Bajila.' Kuleyb, whelp ; c a sub-tribe of Tamim and of Choza'a and of NachaV Kildb, dogs (pi). Two eponyms of this name are given. The Ben! Kilab, who are Qaysites, are quite distinct from the Kalb, who are Yemenites. Leytli, lion. Two eponyms of this name. The Beni Leyth have been mentioned under Kalb. Yarbu', jerboa; 'a sub-tribe of the Ben! Tamim and of the Hawazin and of the Dhubyan.' Namir, panther ; ' a sub-tribe of Eabfa bin Nizar, and of the Azd, and of Qoda'a.' Anmar, panthers; 'sub- tribes of the Arabs.' Anmar, son of Nizar, is the eponym of a Ma'addite tribe that settled in Yemen. Anmar is also a son of Saba', the eponym of the Sabaeans (Tabarl, i. p. 225, 1. 9). To the same source belong, no doubt, Numara, 'a subdivision of the Lachm and others/ and Nomeyr (little panther) among the Qaysites. "In these and numerous other cases the animal name is undisguised. In some cases we find ix EXAMPLES OF FABRICATED GENEALOGIES 163 termination dn, which is noteworthy, because the same thing occurs in Hebrew gentilicia. Thus : " Zabydn (from ^&> gazelle), a subdivision of the Azd'; " Wa'lan (from Jcj ibex), ' a subdivision of Murad '; " Labwdn (from x^A lioness), { a subdivision of Ma'aiir.' "Finally, I add what seems to be the case of a mongrel. The Arabs have many fables of the Sim' (f<+*), a beast begotten by the hysena on the wolf, and so we find Sim 1 , ' a subdivision of the Defenders (the Medinites).' Here we seem also to have the form in an, for Sam'an is a subdivision of the Tamim." I have quoted so far to include all the subdivisions of the Azd for which animal names have been found. This tribe then included doves, wolves, panthers, and gazelles. It also (Journ. Philo., I.e. p. 81) included a tribe of the sun stock ; and a sub-tribe called Ghanm (not translated), but said to be the name of a god. Every sub-tribe, we may be sure, had its own god. Percival (vol. iii. p. 255) mentions Dhou-1-Caffayn as adored by a fraction [sub-tribe] of the Azdites. Were the animals after which so many gentes of the Azd were named their totems or totem gods ? Professor Robertson Smith shows some reasons for believing that they were. The student of primitive races can scarcely doubt it. 1 1 It will be seen that the Ibex occurs as a tribe [sub-tribe] of the Kalb, i.e. of the Dog tribe. It also occurs in other tribes. This might 1 64 STUDIES IN ANCIENT HISTOR Y CHAP. The genealogists refer all these names to some ancestor who bore the name ; but Professor Eobertson Smith declares this to be inadmissible, because often the tribal names have a plural form. " Panthers, dogs, lizards are originally the names of tribes, each member of which would call himself a panther, a dog, a lizard. The idea of an ancestor bearing the plural name is plainly artificial, invented in the interests of a system." l In one case Professor Eobertson Smith has been able to show the ideas of god, animal, and ancestor all in connection. The Qaysites, or Beni Qays, trace their genealogy to 'AyMn, son of Modar. Qays, it appears, was a god. As to 'Aylan, some say he was son of Modar [as above]. " Others say that 'Aylan was his horse, others that he was his dog." The con- fusion occurs at the link in the genealogy where the ancestor is a god ; and Professor Eobertson Smith suggests that the two-fold animal interpretation of 'Aylan must be referred to, there being in the great Qaysite tribes gentes of the horse and dog stocks respectively. The suggestion is ingenious, and most probably correct. Many facts of the same kind must lie, he thinks, behind the genealogies in their present seem inconsistent. But in numerous instances Samoa alone furnishes several such instances a local tribe or town is found named after its chief division, the others being named again as of the town, and as if they were divisions of a division. 1 Of. with the plural names of ancestors in India, ante, p. 130. If what I have there said of the plurality of the ancestors will not hold, we have here an alternative view in favour of the totemism of the Vedic Aryans. ix EXAMPLES OF FABRICATED GENEALOGIES 165 form. Sprenger (Geog. Ar., p. 225) is said to have shown that Kinda, ancestor of the Kindites whose real name was the Bull is a mythical character. None the less he figured as ancestor of a line of seventy kings, ending in the time of the prophet. To sum up, we have now seen that the genealogists of Arabia were entirely without records ; that while they freely pass from son to father up to Noah in the earliest times, they have to contend in the times nearer to them with numerous traditions pointing to women as the heads of families, and to descents traced through mothers ; that there is reason for believing that in the earliest times descents were really so traced among the Arabs ; that their gentes were of the totem type and their local tribes composites of different stocks, not- withstanding which, they assign to them a common father. If the Jews preserved their genealogies with great care, how have we two versions of the genealogy of the putative father of Jesus, which differ so much from one another? They disagree as to Joseph's father; as to several of the ancestors ; and even as to the number of the descents (or generations) from Abraham to Joseph. The pedigree in Matthew goes no farther than Abraham ; in Luke it is traced back to Adam. Both declare themselves to be pedigrees of Joseph. Adam was "the son of God," according to Luke. But already in Genesis there are what seem to be 166 STUDIES IN ANCIENT HISTORY CHAP. conflicting versions of the origin of men. (1) The earth having, in obedience to command, produced all living things, God next created man "male and female created he them." This account is repeated in Genesis v., "male and female created he them, and called their name Adam, in the day when they were created." (2) God made a man of the dust of the ground, and thereafter, having made the animals, he directed the man to give them names. Then per- ceiving that it was not good for the man to be without a female companion, he threw him into a sleep, and abstracting one of his ribs, made of it a woman. According to the first account there was no bond of blood between the first man and the first woman to prevent their freely intermarrying. According to the second, the woman was " bone of the man's bone and flesh of his flesh." In this there is no hint of exogamy. In several cosmogonies skill is shown in producing the first man and woman so that they should not be so related, or in producing the first children so that they should not be uterine relations. Mr. Cameron, in his excellent work on Our Tropical Possessions in Malayan India (London, 1865), draws attention (p. 113) to one exogamous tribe, who give the follow- ing account of the origin of the race. God having made a man and woman, the woman in time became pregnant, not however in her womb, but in the calves of her legs. "From the right leg was brought forth a male, and from the left a female child. Hence it is that the issue of the same womb cannot intermarry. jx EXAMPLES OF FABRICATED GENEALOGIES 167 All mankind are the descendants of the children of the first pair." The story of the fall follows. The Talmud, which is quite as authentic and credible as Genesis, represents, in agreement with the second of the above accounts of the origin of man, that the first man was for some time without a wife. During that time he suffered much from " succubus," and became the unwitting father of a variety of demons. That is the origin of demons. Next he wedded Lilith, 1 of the serpent stock, whose love " turned into hate " led to the fall; the god of her stock helping to avenge Adam's desertion of her for Eve. That is the origin of the fall. Between Adam and Noah intervene but a few genera- tions of marvellous men, some of whom lived more than nine hundred years. The story of the universal deluge follows. God repented of having made men because of their wicked- ness, and determined that life on the earth should be destroyed and recommenced. After the deluge God was so gratified with the smell of a sacrifice of animals Noah made to him that he established the rotation of the seasons ; seed-time and harvest ; summer and winter ; and came under a promise not to destroy the world again, placing in the clouds the rainbow in token of that covenant. Thus are explained the origins of the seasons and of the rainbow. A rapid step forward in the narrative, and 1 Lilith is mentioned in Scripture, but always in our version as the screech-owl. 168 STUDIES IN ANCIENT HISTOR Y CHAP. we reach the Tower of Babel, the dispersion of man- kind, and confusion of tongues, for till then "the whole earth was of one language." The scattering of men in nations arid the varieties of speech are thus accounted for. Such is the setting in which we find the ancient genealogies of Israel. As a collection of " origins " the narrative is unsurpassed even by the marvellous history of Te-Ika-a-Maui. If the reader will look at Arrowsmith's Bible Atlas it is the edition of 1835 I have before me he will see on the first map (right-hand corner) the Eastern Hemisphere with the world as known to the genealogist of Genesis marked upon it in red. The" extent of land is much less than that of Australia, and the larger maps show that even of this skirts or fringes only were known. More striking still is the map of the world, as known to these genealogists, given in Nott and Gliddon's Types of Mankind, p. 552 j 1 where also the ethnological divisions of men in Genesis x. are translated and found to be district and tribal epony- mous names. It is indisputable that this is a true account of that settlement of the world by the families descended from Noah, "according to their tongue and in their nations." Can it be doubted that, had the genealogists known more districts and nations, they would have had to hand the necessary descendants of Noah from whom to name them ? It can be shown to a high order of probability that 1 Philadelphia, 1854. ix EXAMPLES OF FABRICATED GENEALOGIES 169 the people who had these genealogies had anciently the system of kinship through mothers only, and accordingly could not have had genealogies with descents from father to son. (1) There is a suggestion of female kinship in Genesis ii. 24 : " Therefore shall a man leave his father and mother, and cleave unto his wife/' etc. This points to what in Ceylon is called Beena marriage. The man enters the family of the woman instead of the woman entering his family ; and by consequence, the children, where this arrangement prevails, belong to the house and family of their mother. Laban, it will be remembered, claimed his daughter's children as being his : " These daughters are my daughters, and these children are my children," etc. 1 Abraham's commission to the servant entrusted with procuring a wife for Isaac shows that Beena marriage was then common. The servant asks whether, in the event of the maiden declining to leave her home, he is to conduct Isaac to her, as if it was the most natural and probable thing that she should decline. The negative answer turned on no slight consideration. It was that Isaac's descendants were to inherit Canaan, and that there he must remain. In the course of the 1 Elsewhere he upbraids Jacob with carrying them off " like captives taken by the sword." This points to some experience of marriage by capture. We shall hereafter see reason to believe that marriage by capture introduced a new idea of marriage, which indeed in time came to be the only idea the word contained. It was that of consortship between husband and wife in the house of the husband or his kindred, with the wife in subjection to the husband instead of the husband being in a way the subject of the wife's father or mother's brother. i;o STUDIES IN ANCIENT HISTORY CHAP. negotiation for the marriage, the woman had to be expressly consulted as to whether she would go. (2) Numerous marriages mentioned in the Old Testament can only be explained as permissible on the supposition that kinship through women only still regulated the right of intermarriage. The first Old Testament facts as to marriage occur in the genealogy of Abraham. No wives are mentioned in the descents from Seth to Noah or from Shem to Terah. And the first marriages we are told of are, as we should say, incestuous. Abraham married Sarai his sister, and Noah married his niece. With regard to these marriages, I s have elsewhere expressed the opinion that the women were not relatives in a full legal sense of their husbands. 1 Abraham's plea to Abimelech seems to justify this : " And yet indeed she is my sister ; she is the daughter of my father, but not the daughter of my mother, and she became my wife." It seems implied that had they been connected through the mother they could not marry. Nahor's case is similar. The niece he married was a brother's daughter. Had she been a uterine sister's child, my plea is he could not have married her. So in the case of Amram, the father of Moses; he married a father's sister, but she was connected with him on the male side only, and therefore not in the full sense a relative. Still later we can see this distinction having effect in the story of Amnon and Tamar. She was his half-sister by the same father, yet they were 1 Studies in Ancient History [first series, p. 121]. ix EXAMPLES OF FABRICATED GENEALOGIES 171 marriageable. " Speak to the king ; for he will not withhold me from thee" (2 Sam. xiii. 13). 1 It would be an answer to this if a marriage between near kindred connected by both father and mother could be adduced. But we have none such. Abraham, in- deed, directs his servant to go to his country and to his kindred and take a wife for Isaac. But ex hypothesi Isaac was not kindred of Abraham, and might marry in Abraham's kindred. Again, suppose Abraham and Nahor to have had the same mother, yet there would not be full kinship between Abraham and Bethuel, for Bethuel would be of Milcah's kindred. We know not who was Bethuel's wife ; but his daughter would be of her mother's kindred as concerned the law of marriage. So as regards Jacob's marriage. We do not know, but we may believe, that Laban was Rebecca's brother uterine ; still as his daughters would be of their mother's stock and not his, Jacob, supposing him of Laban's stock, would be free to marry them. (3) The independent indications of female kinship are numerous. (a) To any one acquainted with the usages of exist- ing races in the stage of female kinship, one fact alone 1 Lafitau (Mwurs des Sauvages Ameriquains compares aux Mceurs des Premiers Temps, vol. i. p. 548. Two vols. Paris, 1724) ingeniously suggests that Sarai may have been Abraham's cousin merely. The indications of a system of addresses similar to that prevailing among the Iroquois are indeed numerous, but the plea, as we have it, is in a form to exclude the suggestion. Lafitau very justly infers from the plea that marriages of half-brothers and sisters by the same father were permitted in Egypt. 172 STUDIES IN ANCIENT HISTORY CHAP. would be conclusive. The " price " of Kebecca was paid to her mother and brother. (6) The duties of revenge and protection lay with the kindred of the mother. Absalom, having revenged the rape of his uterine sister, fled to the kindred of his mother (2 Sam. iii. 3 ; xiii. 37). See also Judges viii. 19, and Genesis* xxxiv., to which Professor Eobertson Smith has called attention. It may be thought that these facts may be explained by the polygamous marriages prevalent among the Hebrews, but as I have elsewhere shown, though poly- gamy tends to sustain the system of female kinship, once it has been established, it cannot possibly originate such a system. (See Fortnightly Review, 1877. " Le- virate and Polyandry.") (c) We have, I think, an undoubted indication of kinship through women only, as the ancient system of the people, in the institution of the Levirate. On this subject, however, I must beg the reader's attention to an argument. Polyandry having been traced, on the testimony oj witnesses, as still existing over a wide area, as having till recently existed in various places from which it has now died out, and as having existed among several of the ancient nations, it became a question whether to regard it as abnormal or as normal in the development of marriage. If the area of its former prevalence could, on any good evidence, be extended much beyond the limits fixed for it on the testimony of witnesses, it ix EXAMPLES OF FABRICATED GENEALOGIES 173 would unquestionably have to be regarded as normal, as a form which the marriage systems of the world, speaking broadly, everywhere at some time or other passed through. But where was such evidence to be looked for ? Assuming that polyandry had anywhere prevailed formerly, it seemed to me that we might expect evidence of the fact first, in the local laws of inheritance, and secondly, in the perpetuation of practices and moral sentiments derived from polyandry. And it seemed worth while to consider whether any such evidence was to be found. On a discrimination of the cases, existing or recorded, of polyandry, it appeared that it presented itself chiefly under one or other of three forms the Nair, the British, and the Thibetan. In British polyandry a father or fathers and the sons of the house lived in consortship with the same woman. In Nair polyandry the several husbands of a wife are unrelated to one another by blood ; in Thibetan the several husbands are brothers. British polyandry, though common, seemed abnormal ; but on evidence that seemed sufficient, the conclusion was reached that Nair polyandry was a preparation for Thibetan ; Thibetan polyandry an advance from Nair ; and that both these forms might be normal. The inheritance law is the same for both forms, with a difference. In connection with Nair polyandry brothers, speaking broadly, succeed as heirs to one another, and the last surviving brother is succeeded by his sisters' children the succession law proper to the system of female kinship. In connection with Thibetan poly- 174 STUDIES IN ANCIENT HISTORY CHAP. andry brothers succeed as heirs to one another, and the last surviving brother is succeeded by the eldest son of the brotherhood, this son being, like the rest of the children, accounted as in a special sense the son of the eldest brother the rudest form of succession law proper to a system of male kinship. In the special case of a Nair family being decomposed into sub-groups com- posed of a man and his favourite sister and her children, it would seem as if these children alone succeeded to the man's movable estate direct, a fact illustrating the operation of the motives that at a later stage estab- lished the succession of sons to fathers. But in the case of heritable estate, its administration or possession in trust for the family was always among the Nairs in the oldest male of the family. Here, then, was a peculiar law of inheritance con- nected with polyandry, brothers succeeding in preference to sons ; and, in connection with Thibetan polyandry, an equally striking fiction by which the eldest brother in a group of brother-husbands was accounted to be, in a special, sense, the father of all the children. The Thibetan inheritance law is, of course, easily intelligible from two points of view. First, the succession of brother to brother originating in the Nair stage, 1 and necessitated by the system of female kinship, would simply, because it was established, long persist in the Thibetan stage ; and, again, were there no other reason 1 In the Nair stage the right of administration was in the oldest male. His successor was the next oldest ; all of a generation were counted to be brothers, and usually, where daughters had separate houses, the succession would open to the next brother uterine. ix EXAMPLES OF FABRICATED GENEALOGIES 175 for it, the copartnery of the brothers in marriage would exclude the idea of their children succeeding in prefer- ence to any of them. The fiction which accounted the children as specially belonging to the eldest brother is probably attributable in part to his superior position as ruler of the house and administrator of the family pro- perty, and in part to his being the first to marry and beget children. The name of " father " or protector, unconnected at first with the idea of begetting father, had been taken over from the Nair into the Thibetan family system ; and it is consistent with all the brothers being " fathers " to the children, and the children being to them severally " sons and daughters," that the children, in a special sense, should belong to the eldest brother. Could such an inheritance law as we find with poly- andry be derived from any form of the family founded upon monandry ? It seemed indisputable that it could not ; that such a law, wherever found prevailing, must be taken as a proof of the former existence there of polyandry. Accordingly I concluded that wherever the law of inheritance constituted the brothers of a deceased person his heirs in preference to his sons, we had evidence of the former existence of polyandry. What, then, of the fiction which made the children to be accounted the children of the eldest brother ? It, of course, at once suggested the Levirate. Was the Levirate to be accounted a remainder of polyandry ? There are two main features of the Levirate (1) the obligation laid on a brother to marry the widow 176 STUDIES IN ANCIENT HISTORY CHAP. of an elder brother who has died childless ; (2) the purpose of the obligation, namely, to " raise up seed upon the inheritance " of the deceased brother. Judg- ing by these signs, we infer the existence of the Levirate wherever we know that a man must marry the widow of his next elder brother deceased without issue, and that the children of this marriage are ac- counted the children of the deceased brother. And going on these tests, we find numerous cases of the Levirate. It happens, however, that, except in regard to the two main cases of the Indians and Hebrews, we have not full accounts of the Levirate as an operative law. The obligation laid upon brothers by old Hebrew law to marry the widow of a brother deceased without issue is a remarkable exception to the spirit of the Levitical prohibition of marriages between persons related by blood or affinity. The earliest recorded example of it is in the story of Judah and Tamar. Tamar had successively been wife to Er, Judah's eldest son ; and to Onan, who was disinclined to perform the duty of the Levir, as the children he might have by Tamar would be accounted the children of Er. There remained another brother, Shelah, to whom Tamar was entitled as husband, but he was a youth, and for pru- dential reasons his marriage to the widow was post- poned. After a time the widow, who had meanwhile been staying in her father's house, felt aggrieved at this postponement, and planned an incident through which she became the mother of twins to Judah himself. ix EXAMPLES OF FABRICATED GENEALOGIES 177 Judah, on hearing that she was with child, ordered her to be brought to him to be burned ; but ascertain- ing the true state of the facts, he confessed himself in the wrong. "She hath been more righteous than I, because I gave her not to Shelah my son." The posi- tion of the issue of her connection with Judah is shown by the event. Her son became the head of the family, obviously succeeding in the name of Er. Now the story of Tamar connects itself with inherit- ance law only through this fact. The son of the Levirate union succeeded to the inheritance of Er, and cut out Er's brother Shelah from that inheritance ; and he would have done the same thing even had Shelah been his father. For the rest, the story is simply this that the woman having married an eldest son, was in law entitled, so long as she was childless, to have the other brothers of her husband as husbands in succession. The brothers, it may be, were entitled to claim her ; but Onan at least was disinclined to make such a claim. Regarding the duty laid on him as disagreeable, he failed to discharge it ; " and the Lord slew him." It is similarly as a duty that in Deuteronomy xxv. 5, in the case of brethren dwelling together, and one of them dying childless, a brother of the deceased (obviously the next brother) is required to marry her, and " perform the duty of an husband's brother unto her." "And it shall be" says the text, " that the first born which she beareth shall succeed in the name of his brother which is dead." Here we have a reference to inheritance, and its meaning is obvious. The child of the Levirate N 1 78 STUDIES IN ANCIENT HIS TOR Y CHAP. marriage succeeded in the name and as the son of the deceased brother to his putative father's property. The Levir gained nothing by the transaction. If by taking the widow he had got the estate and united it to his own, for the benefit of his children at large, he would have gained ; but he did not get the estate. The estate was held apart from his, and lost to his family, through its destination to a child counted the child of his dead brother. His business was merely "to raise up the name of the dead upon his [the dead's] inheritance." In the circumstances, and the Levirate having much about it revolting to sentiments that had grown up with monandry, and which have since destroyed the Levirate, it is no wonder that the law came to provide a process by which the Levir might evade the obligation. At first he had no choice : the woman was his wife de jure and without form of marriage a fact easily compre- hensible if the Levirate had its origin in Thibetan polyandry. Afterwards under the growing influence, no doubt, of ideas of propriety derived from a practice of monandry a formal marriage between the Levir and the widow became indispensable. The case of Kuth and Boaz, I may say, is not, strictly speaking, a case of Levirate at all. "Where a kinsman, other than a brother of the deceased husband, took the widow, he took her not as Levir, but as a Goel, or redeemer of the inherit- ance of the dead ; and there was no law requiring him to marry her as a condition of the redemption, though no doubt he usually did so. At the same time we may see, from the opening chapter of the Book of Ruth ix EXAMPLES OF FABRICATED GENEALOGIES 179 (verses 11-13), the position of the Levir. The widow could claim her husband's brother, the Levir, as husband, even were he unborn when she became a widow. The Indian case is on " all fours " with the Hebrew. On a man dying childless, the law provided for the begetting on the widow, by his brother, or, in the event of the brother being incapable, by some other relation authorised to act for him, of at least one son. There is no mention of marriage in the Code of Manu as entered into between the Levir and the widow ; and possibly, as in the Hebrew case in the oldest times, the widow was pro re his wife de jure. His relations to her, however, were regulated by both law and religion. " Sprinkled with clarified butter, silent, in the night, let the kins- man thus appointed beget one son, but a second by no means, on the widow or childless wife" (Manu, ix. 60). Here again the duty of the Levir appears pure ; and he took no advantage from its performance, for the law provided as follows : " Should a younger brother have begotten a son on the wife of his deceased elder brother, the division of the estate [the estate, that is, of the father of the two brothers] must then be made equally between that son, who represents the deceased, and his natural father : thus is the law settled." So that the son of the Levirate union carried away the deceased's estate in the name of his putative father from the family and children of his real father. Meantime, and till this heir was born, the estate of the deceased if already there had been a partition made of the father's property i8o STUDIES IN ANCIENT HISTORY CHAP. was vested in the widow. (See the Vivada Chinta- manij pp. 261 and 289.) The Levir's obligation, then, appears to have been unconnected either with marrying the widow or inheriting his brother's property. The Levirate was, in fact, a process for cutting off the Levir and his family from the succession. It is needless to say that in the Indian case, as in the Hebrew, the Levirate is seen from the first as in decay under the influence of sentiments growing up with the practice of monandry. Already it had fallen into desuetude among the twice-born classes, and was in use only among the servile classes; but the same text that declares it a practice " fit only for cattle " discloses that, anciently, even the twice-born classes had had the practice. The Levirate, as seen in these two cases, is obviously one and the same thing; its effects and intention are the same. By the obligation laid on the next brother, it cuts off that brother and his family, and the brothers generally, from the inheritance of the deceased brother, and gives it to that brother's reputed son. It is an institution, then, which from its nature must have been posterior to the law of succession of sons to fathers being firmly established, so firmly that even a fictitious son was preferred to a real brother. In both the Hindu and Hebrew cases there are notes of Thibetan polyandry, and they are the same in both. The appointment of a brother to perform such a duty as the Levir's would be nowise abhorrent to a people who had had recent experience of a practice of Thibetan polyandry ; and moreover, among a people who had had ix EXAMPLES OF FABRICATED GENEALOGIES 181 such an experience, it would not be surprising to find the fiction that a child begotten by one brother was truly the child of an elder brother deceased. It will, I think, at least be conceded that the origin of so strange an arrangement must have been the same or similar in the Hebrew case and in the Indian. But in the Indian case we can prove, aliunde, the ancient prevalence of Thibetan polyandry. One text in the Code of Manu seems to me conclusive on this point (ix. 182): "If among several brothers of the whole blood, one have a son born, Manu pronounces them all fathers of a male child by means of that son, so that if such nephew would be the heir the uncles have no power to adopt sons." But I shall not dwell on this text, as we have in the case of Draupadi the fact independently proved. It is familiar that in the great epic, the Mahabharata, the heroes, the five Pandava princes, had but one wife between them Draupadi. The authorities hold that the Brahmans who compiled this epic from old materials found the tradition of this marriage so strong that they could not suppress it ; and that, since the marriage was repugnant on the whole to Vedic, and altogether to post-Vedic ideas, the story must be referred to the pre- Vedic period. The father of Draupadi (as I have said elsewhere) is represented by the compilers of the epic as shocked at the proposal of the princes to marry his daughter. " You who know the law," he is made to say, " must not commit an unlawful act, which is contrary to usage 182 STUDIES IN ANCIENT HISTOR Y CHAP. and the Vedas." The reply is, "The law, king, is subtle. We do not know its way. We follow the path which has been trodden liy our ancestors in succession." One of the princes then pleads precedent : "In an old tradition it is recorded that Jatila, of the family of Gotama, that most excellent of mortal women, dwelt with seven saints ; and that Varski, the daughter of a Muni, cohabited with ten brothers, all of them called Prachetas, whose souls had been purified by penance." The tradition being too stubborn for the Brahmans, they thus tried as much as they could to palliate it. It is a clear tradition of Thibetan polyandry ; it is confirmatory of the supposition s that what seem notes of that kind of marriage found in the Code of Manu are truly notes thereof; and accordingly it serves to show that what seem notes of polyandry in the Levirate in India are truly notes thereof. But having thus connected the Levirate and pre-existing polyandry in India, we cannot refuse to connect the Levirate and pre-existing polyandry among the Hebrews. The general inference, of course, is that the Levirate, wher- ever found, is a remainder of Thibetan polyandry. 1 The foregoing argument differs from that which I stated in Primitive Marriage fifteen years ago in two particulars only, and neither of them affects its force. I was not aware of what Selden had pointed out, namely, that the author of the Book of Kuth and 1 Since the above was written, Professor Eobertson Smith has drawn attention to express mention of Hebrew polyandry in the eighth century B.C. : Amos ii. 7. ix EXAMPLES OF FABRICATED GENEALOGIES 183 Josephus were both in error in regarding the case of Boaz and Kuth as a case of the Levirate. Accord- ingly, I assumed the Levirate to be the counterpart of a right of succession, an error possible only through inattention to the operation of the law in the un- questionable cases of the Levirate. Owing to this error, again, I connected the Levirate and the law of inheritance preferring brothers to sons, as if they had something more in common than being both of them remainders of polyandry. I may be permitted to say that I do not owe the knowledge of these errors to criticism. But if the Levirate is a remainder of Thibetan polyandry, it is at one remove more a remainder of Nair polyandry, and therefore a proof of the ancient prevalence of the system of female kinship among the people. I now pass to the evidence of totemism among the Hebrews. In the papers I published on totemism in 1869-70, I drew attention to some of this evidence, to which I need not here further refer. 1 The only evidence since brought forward is that adduced by Professor Eobertson Smith. He has found gentes of the Hebrews and their congeners in some number to be named after animals, and in several cases has shown the local tribes to have 1 "The Worship of Animals and Plants," Fortnightly Review for October and November 1869, and February 1870. [Reprinted at the end of this volume.] 1 84 STUDIES IN ANCIENT HISTOR Y CHAP. been composites of gentes of different stocks named after animals, which cannot, that I know, be explained independently of totemism, exogamy, and female kin- ship ; and, not to attempt to produce the evidence here, he has drawn special attention, in connection with their totemism, to the singular evidence given by the prophet Ezekiel. "Our most definite information as to animal wor- ship in Israel is derived from Ezek. viii. 10, 11. There we find seventy of the elders of Israel that is, the heads of houses worshipping in a chamber which had on its walls the figures of all manner of unclean creep- ing things and quadrupeds, even all "the idols of the house of Israel. ... It appears also, that though the prophet in vision saw the seventy elders together, the actual practice was that each elder had his own chamber of imagery (ver. 12). We have here, in short, an account of gentile or family idolatry, in which the head of each house acted as priest. And the family images which are the object of the cult are those of unclean reptiles and quadrupeds. The last point is important. The word ppoj is, in the Levitical law, the technical term for a creature that must not be used as food. That such prohibitions are associated with the totem system of animal worship is well known. The totem is not eaten by men of its stock, or else is eaten sacramentally on special occasions, while con- versely to eat the totem of an enemy is a laudable exploit. Thus in the fact that the animals worshipped were unclean in the Levitical sense, we gain an ad- ix EXAMPLES OF FABRICATED GENEALOGIES 185 ditional argument that the worship was of the totem type. And finally, to clinch the whole matter, we find that among the worshippers Ezekiel recognised Jaaz- aniah the son of Shaphan that is, of the rock badger (E.Y. coney), which is one of the unclean quadrupeds (Deut. xiv. 7 ; Lev. xi. 5), and must therefore have been figured on the wall as his particular stock-god and animal ancestor." 1 1 Professor Kobertson Smith adds : " It so happens that the totem character of the Shaphan, or, as the Arabs call him, the wabr, is certified by a quite independent piece of testimony. The Arabs of the Sinai peninsula to this day refuse to eat the flesh of the wabr, whom they call ' man's brother,' and suppose to be a human being transformed. . . . The close connection which we have found to exist between Arab tribes and southern Judah, and the identity of so many of the stock- names among the two, give this fact a direct significance." PART II SECTION I THE PACIFIC ISLANDS AND AUSTRALIA CHAPTEE X HERVEY GROUP, OR COOK'S ISLANDS THE Hervey group, or Cook's Islands, consists of seven inhabited islets, lying between the 19th and 22nd parallels of S. latitude and 157th and 160th of W. longitude: (1) Earotonga, (2) Auau or Mangaia, (3) Aitutaki, (4) Atiu, (5) Mauki, (6) Mitiaro, (7) Manuae. The inhabitants of these islands seem to have direct connections, on the one hand, with the Friendly Isles, the Samoan and the Sandwich group to the west, with the Society Isles to the N.E., and with New Zealand to the S.W. ; and Mr. Gill (p. 166) shows reasons of a philological sort for believing that the first three original tribes of Mangaia came from Hawaii, and that at least one tribe in the same island was of Tongan origin. Many of the myths prevailing in the Hervey group are the same as or mere variants of myths prevail- ing in New Zealand, and several of the leading gods of the Hervey Isles may be traced throughout the whole series of islands mentioned. Our chief information respecting the Hervey group has been drawn from Myths and Songs from the South Pacific, by the Eev. 190 STUDIES IN ANCIENT HISTORY CHAP. W. Wyatt Gill, who was for twenty-two years a mis- sionary in these islands. Little of it is of the nature of direct statement, and the view of Mangaian society which follows may be taken as resulting from the piecing together of statements and allusions scattered through the Myths and Songs. As to the structure of Mangaian society for to the people of these islands we shall mainly confine ourselves we learn that the population was under something like kingly government, the authority being invested in the chief of some tribal league that had proved victorious in war. When there was war there was no king, but the leaders of tribes in council on the one side guided their forces against the tribes leagued on the other side. Mr. Gill uses the word tribe in a way to cause very considerable perplexity : sometimes the word clearly means a local tribe, as at p. 290, where the tribe of Tongans included the clans of Teipe and Tongaiti, but usually the word means clan and no more. The tribe of Tane, for instance (p. 175), is called the clan of Tane at p. 278, in which we find the "clan of Tane was cut up by the shark worshippers"; and again we find, "Tiairi is filled with the tribe of Tane"; Mr. Gill explaining in a footnote that Tiairi is a warrior's paradise, "in which the clan of Tane is supposed to have a large share," so that on this page the words clan and tribe, which occur each twice, are clearly inter- changeable. The word family is also sometimes used in a sense broad enough to include clan. The family, properly so called, seems to have rested x HERVEY GROUP, OR COOK^S ISLANDS 191 on monogamous, possibly occasionally polygamous marriage, but the clan does not appear to have been formed of the members of families definitely according to the law of descent through either males or females. The test of clanship was before everything else religion, the having the same god; and in the devotion of children to the gods, sometimes the god of the father of the child was preferred, and sometimes the god of the mother. Usually, says Mr. Gill, the father had the preference, but occasionally, when the father's family was devoted to furnish sacrifices, the mother would seek to save her child's life by getting it adopted into her own tribe, the name of her own tribal divinity being pronounced over the baby. As a rule, however, the father would stoically pronounce over his child the name of his own god Utakea, Teipe, or Tangiia which would almost certainly ensure its destruction in after years. According to this statement (p. 36), it was a matter of arrangement whether the child should belong to the father's or the mother's clan ; the state- ment implying, as does indeed every hint on the subject scattered through the w^hole Myths and Songs, that the father and the mother were of different clans, and that exogamy was the practice at least, if not the law. An illustration which Mr. Gill gives of the way in which arrangements as to the clan of a child were come to, contradicts his statement that the child belonged to the mother's clan only as a means of saving it from being sacrificed, and shows that the name of the mother's god was sometimes pronounced over the child, even at the 192 STUDIES IN ANCIENT HISTORY CHAP. risk of the child being sacrificed. " A deacon still liv- ing," says Mr. Gill (p. 37), " told me that his god was to have been Teipe, but when half-way to the Marae of that unfortunate god his father resolved to break his promise to his wife, and actually turned back and presented the knife to his own god. ' Had my father not done so, I should long since have been offered in sacrifice.'* It is inconceivable that a mother should desire the child to be of her clan in such a case, except under the force of a feeling which must have been in common action. Any way the clan was not composed wholly of persons connected through fathers only or mothers only, like either the patriarchal clan or the totem -clan of Australia or America. Kinship is seen in a state of transition. That it was formerly traced through women only may be believed on all the facts of the case. That it was in the course of becoming a system of male kinship only is equally apparent. At present let us notice that while the blood-bond was in uncertainty, owing to the transitional stage of kinship, the totem, here elevated to the rank of a god, determined all the obligations which we have seen in so many cases were determined by blood simply. Your clansman is your kinsman in what has become an indefinite sort of way owing to the break up of the ancient simple rule of descents. He is your fellow-worshipper, however, and it is as such you are bound to him. "The greatest possible sin in heathenism," says Mr. Gill (p. 38), "was ta atua, i.e. to kill a fellow -worshipper by stealth." " In general it might be done in battle ; otherwise such x HERVEY GROUP, OR COOK'S ISLANDS 193 a blow was regarded as falling upon the god himself the literal sense of ta atua being god-striking or god- killing. Such crimes were generally the consequence of ignorance," and accordingly special arrangements were made for securing that every one should have full know- ledge of his fellow-clansmen. The public naming of the young people was a religious act, superintended by the principal king of Mangaia, assisted by the priest of Motoro, at the shrine of the god to whom the young people belonged. " Namings," which were always followed by feasting, were held at intervals : of course, bringing all of the clan together, they made them acquainted with one another. We shall presently also see that tattoo marks were probably a better safeguard against ignorance than even the "namings." It will serve to show the completeness of the bond between the fellow-worshippers that in the Myths and Songs the name of the god is frequently used instead of the name of the clan, as in the Axe dirge " twice has the god Turanga thus served our clan/' "where," as Mr. Gill observes, "the god Turanga is put for the Tonga tribe." It is also worth noting that the bond of common worship which held them together in life still held them united after death. On the mysterious Bua- tree, which rises in the world of spirits to receive the dead, each clan had a special branch reserved for itself, and on to which its members would be doomed to step. The worshippers of Motoro had a branch to themselves ; the worshippers of Tane another ; and there were as many branches as there were principal gods in Mangaia o 194 STUDIES IN ANCIENT HISTORY CHAP. one great branch, however, sufficing for the whole batch of the lesser Tanes. The facts which suggest that kinship was anciently through women only are not numerous, but are suffi- ciently striking. In the performance of death-dirges the clan of the father of a deceased warrior took precedence of the clan of the mother in performing the eva; but the mother's clan separately performed an eva specially composed for the dead ; and in some evas performed by the father's clan will be found episodes in the history of distinguished members of the mother's clan, as specially connected with the dead, and as specially qualified to stimulate a cry for vengeance (see pp. 274, 276). In some cases the dead were buried not with the father's deceased relatives but with the mother's, but whether this was only in cases in which the dead was of the mother's clan does not appear. Eecur- ring to the mythology of the tribes, we find that the three tribes of the Ngariki, literally the royal house, the kingly tribe, represented as being the three original tribes of Mangaia, deduced their descent from a common mother, Tavake, and her three illegitimate sons, Eangi, Mokoiro, Akatauira ; while all the tribes throughout the Hervey group trace their origin back each to one of a series of gods, who were the offspring of Vatea and Papa, and ultimately of a woman of flesh and blood, Vari-mate-takere, or " The Very Beginning.' 7 We now proceed to consider who were the gods from whom the various clans derived themselves, and through the worship of whom they were as clans con- x HERVEY GROUP, OR COOK'S ISLANDS 195 stituted. The woman Vari having formed from a bit of her right side, which she plucked off, a man Vatea, tore off another bit from the same side, and it became Tinirau. She then formed a third, fourth, and fifth child from her left side, who do not appear to be specially connected with the tribes, and finally a sixth, a girl, who is represented as remaining with her mother. Vatea, her first produced, and the father of gods and men, was half -man and half -fish. The one eye was human, the other a fish eye ; his right side had an arm, the left a fin ; one foot was proper, and the other half a fish-tail the fish incorporated with the human form being a porpoise. His brother Tinirau had also a fish- form, the fish which was compounded, as in the case of Vatea, with his human part being of the sprat kind. As already mentioned, Vatea, marrying Papa or founda- tion (the daughter of Nothing-more and his wife Soft- bodied), begot twin children, Tangaroa and Eongo, then Tongaiti, then Tangiia, and lastly Tanepapakai. It is needless to say that each of these sons of Vatea and Papa represents a clan in Mangaia, and that the story which thus connects them by descent from the common parents is a mere illustration of the kind of fiction by which nations try to establish ties of consanguinity between all their sections. In point of fact, Mr. Gill is probably correct in surmising that the so-called three original tribes, the Ngariki, who claimed to be earth- born and descendants of Eongo, were truly the first inhabitants of Mangaia, and that the tribe of Tongans were the first immigrants reaching the Hervey Isles 196 STUDIES IN ANCIENT HISTORY CHAP. from Tonga, the tribe of Tane coming from Tahiti at a later time. Be this as it may, it would be very curious if there had been in Mangaia or in the Herveys any general agreement as to the genealogy of the tribes deduced from Vatea and Papa. We find there was no such agreement ; one account makes Eongo to be the elder born of the twins, and not Tangaroa. Another account makes Tongaiti to be not the third son of Vatea and Papa, but the husband of Vari. Also Vatea and Tongaiti are found disputing as to the parentage of the firstborn of Papa (see pp. 10 and 45). Thus it will be seen that the tribe of Tongans claimed^ precedence over their hereditary foes, the descendants of Eongo. At Atiu, and also at Eaiata, Eongo was even represented as a son of Tangaroa. The scheme of descents of course covers the Herveys, each god being represented as settling in the island where his clan was strongest. Thus Eongo is said to have settled in Mangaia, and Tangaroa in Earatonga and Aitutaki. One or two further observations as to the composi- tion of the clan seem called for before disclosing the nature of the clan-gods. We find in Africa not a few examples of the clan of the children being matter of arrangement between husband and wife, or their relations, before the marriage a common adjustment being that the boys shall belong to the totem of the father, and the girls to the totem of the mother. This arrangement is the same that is so familiar in marriages between Eoman Catholics and Protestants. One can see how, as the religious regard for the totem developed x HERVEY GROUP, OR COOK^S ISLANDS 197 into regular worship with a ritual, the man should desire his children to be all of his own religion, and, the family system permitting it, should even insist on their being so. Thus religion is seen to come in aid of property considerations the desire of a man to find heirs to his wealth in his own house to subvert the system of kinship through females only, and establish exclusively a system of kinship through males. In the dirge of Koroa for a pet-child, given by Mr. Gill (p. 215), we find the following : Thy god, pet-child, is a bad one. Ah ! that god, that bad god ! Inexpressibly bad, my child. The god Turanga is devouring thee Although only partially his own. I am disgusted with the god of thy mother. One can understand also how, as in the case of the Greeks and the Eomans, and in numerous other cases, on the family system settling to the patriarchal and agnatic type, not only the children should be all of the father's faith, but even their mother should be required on her marriage to forsake the god of her fathers, and accept as her god that of her husband and his family. But what shall we say of a system where religion over- rode other considerations, and the clan of a child was liable to be determined by a predilection of the mother for some one of the more eminent gods, not necessarily her own ? Yet this appears to have been the case among many of the Polynesian tribes. It is not said that it was the case in any tribes of the Hervey group ; 198 STUDIES IN ANCIENT HISTORY CHAP. but the Rev. John Williams (Missionary Enterprises, London, 1840, p. 142 l ) says, in his general account of the religion of the Polynesians, that it was very general. " Many mothers dedicated their children to one of these deities (class deities, such as the god of the fisherman, god of the husbandman, etc.), but principally to Hiero the god of thieves, and to Oro (identified by Mr. Gill with Kongo) the god of war. Most parents, however, were most anxious that their children should become grave and renowned warriors. This appears to have been the very summit of a heathen mother's ambition, and to secure it numerous ceremonies were performed before the child was born, and after its birth it was taken to the Marae and formally dedicated to Oro." It is obvious that a clan based on the notion of its members mainly being fellow-worshippers, and recruited by dedication of children, occasionally on the choice of the mother to her clan-god, occasionally on the choice of the father to his clan-god and occasionally, let us suppose, on the choice of both parents to the god of some clan to which neither of them belonged would more resemble what among Dissenters is known as the congregation than any clan founded upon blood-relationship that we have hitherto had cognisance of. We shall have reason to see that the clans or tribes of the Hervey group maybe suspected to be more or less of this nondescript kind. The common mother of all beings, whom for brevity we have called Vari, had no temple (Marae), no worship- 1 Williams, I.e. p. 146, says infanticide was unknown at the Hervey group or the Samoas. Turner says abortion prevailed at the Samoas. x HERVEY GROUP, OR COOK^S ISLANDS 199 pers, and no clan ; nor had Vatea, half-man, half-porpoise, any Marae or clan. Some of the sons, however, of Vatea had their temples and clans. Eongo had for worshippers the so-called three original tribes who claimed to repre- sent him as descendants of his three illegitimate grand- sons, children born to him by his own daughter, Tavake. Of course this is mere fable, worked into a genealogy to explain some connection of consanguinity acknow- ledged as existing between these three tribes, who had the common name of the Ngariki. The three brothers to whom they trace their descent go in the account for nothing. What we know for certain respecting them, on Mr. Gill's authority, is that the gods they worshipped were Kongo and Motoro. It will be remembered that the earlier creations of Vari were not of perfect human form. Mr. Gill says of Tangaroa and Kongo, the first-born twin children of Vatea, " that these boys were the first beings of perfect human form, having no second shape" (p. 10) ; but the correctness of this as to Kongo at least may be doubted. Kongo's special title, and in fact the meaning of his name, is "The Kesounder" (p. 14), and in the Pantheon of Mangaia he had but one representation, and it is con- sistent with this title. " On entering (the god-house of the king) the dwelling-place of the chief divinities of Mangaia, the first idol was Kongo in the form of a trumpet-shell (Triton variegatus) ; " so that the form of the god absolutely answered to his description, and to the fishy character of his ancestry, and of the whole Pantheon, as we shall presently see. Kongo had a large 200 STUDIES IN ANCIENT HISTORY CHAP. stone image in Mangaia, but I can find no account of the shape of it. Three small rocks united at the base close to the Marae of Kongo were the symbols of his reputed grandchildren, the ancestors of the Ngariki. 2. Tongaiti, reputed the third son of Vatea, was an object of worship in the Hervey group, his visible form being the white-and-black-spotted lizard (p. 10). It was worshipped under the name of Matarau, the sharp- sighted, and had a regular Marae at Taruarua. 3. There was another lizard god, Teipe, who also had a regular worship. The Teipe clan appears in union with that of the lizard-god, Matarau, as forming to- gether the Tongan tribe. It does not appear in what respect the lizard Teipe differed from the other. In one place (p. 306) the whole of the Teipe clan is represented as being in hiding in time of war with the ancient tribe of the Ngariki, inside a grand and almost inaccessible cave named Erue. 4. The tribe Amama were worshippers of Tiaio, under the double form of shark and eel. The shark- god, like the lizard, seems to have been the god of more than one clan, unless the different names under which he is referred to are merely various titles of one and the same god. It is interesting to note that Mr. Gill in his introduction states, that a large portion of his work was derived from Tereavai, the last priest of the shark-god Tiaio. 5. Tuna was the eel-god ; but whether there was an eel-clan is not so clear. We shall consider this matter presently. x HERVEY GROUP, OR COOK'S ISLANDS 201 6. Motoro, called the living -god, was in Mangaia the next most important god to Kongo, and as he and Tiaio both have certain claims to being men-gods, while in connection with the animal forms, it is worth while attending for a moment to their histories. Until 1824, Kongo and Motoro were conjointly wor- shipped as the supreme deities of Mangaia. The Ngariki, and the kings, who belonged to their clan, invariably worshipped them ; and all the other clans were con- tinually laid under contributions in that worship, which involved frequent human sacrifice, the victims being chosen from the other clans. The history of Motoro is as follows : Kangi, the eldest of the three grandsons of Kongo, and therefore of the ancestors of the Ngariki, had for his god in the day, Kongo being his god in the night, the god Tane, with whom he was dissatisfied ; and he applied to Tangiia at Karotonga to give him one of his sons as a god. Mr. Gill explains that this Tangiia was a real man who lived some five hundred years ago, and is not to be confounded with the god Tangiia, the fourth son of Vatea. Be that as it may, it appears that no less than three sons of Tangiia of Karo- tonga became gods at Mangaia. We have a brief account of the deification of one of them only, Motoro. Tangiia, on the request of Kangi, sent Motoro to him to be his god, and with him two of his elder brothers as an escort. The elder brothers on the voyage, quarrelling with Motoro, threw him into the sea, where he miserably perished. It was well known, says Mr. Gill (p. 27), that Motoro's body was devoured by sharks, but then it 202 STUDIES IN ANCIENT HISTORY CHAP. was asserted that his spirit floated on a piece of hibiscus (in the native language called Au, whence it is, say the sacred men, that this word came to mean reign or rule), " over the crest of the ocean billows until it reached Mangaia, where it was pleased to inhabit or possess Papaaunuku, and driving him to a frenzy, compelled him to utter his oracles from a foaming mouth." The god thus arrived on a piece of hibiscus at the court of Eangi, was just what that king wanted. The king recognised him at once as his own god, and Papaaunuku and his descendants as his priests. It appears (p. 25) that the name Motoro may have a phallic signification, and we learn from Mr. Gill (p. 33) that phallic worship de- veloped itself in the Hervey group. We are brought into contact with the hibiscus again in the story of the deification of Tiaio. Mr. Gill says (p. 29) that Tiaio' s history is well known, and that he was led through pride of successful exploits to defile the favourite haunts of the gods, by wearing some beautiful scarlet hibiscus flowers in his ears. For this he was slain by a blow on the head by Mouna, priest of Tane, the man-eater. " The blood of Tiaio mingled with the waters of the brook running past the Marae of Motoro, and eventually mixed with the ocean. Thence- forth that stream was held to be sacred, and it was fabled that a great fresh-water eel, Tuna, drank up the blood of the murdered king, whose spirit at the same time entered the fish. Tuna made its way to the dark, deep fissure running underneath the rocks into the sea. The indomitable spirit of Tiaio having thus succeeded x HERVEY GROUP, OR COOK^S ISLANDS 203 in reaching the ocean forsook the form of the eel, and took possession of the large white shark, the terror of these islanders. The new divinity had a Marae set apart for his worship, close by the more sacred grove of Motoro, and but a few yards from where he fell by the hand of the jealous priest." Now there is abundant evidence in the Myths and Songs of shark worship as widely distributed throughout Polynesia, and of the shark-god pre-existing therefore as a god before this pretence of the spirit in him being that of the man and king Tiaio. The man-god, in short, is the creation of priestcraft and perhaps family affection, and was only possible through conjoining the man-spirit with that of a god already recognised as existing. The same remark applies to the eel, in which Tiaio is represented as having temporarily taken up his residence. The myth of the cocoa-nut tree, as given by Mr. Gill, is proof of this, while it illustrates the fictions through which a fusion of savage deities, and so of clans, was made possible in primitive times. Ina was surprised one day to see an eel she had been looking at assume the appearance of a handsome youth. " I am Tuna-eel, that is, the god and protector of all fresh-water eels. Smitten by your beauty, I left my gloomy home to win your love. Be mine." From that day he was her attached lover as a man, becoming an eel only when he left her ; but an occasion came for their parting for ever. Tuna, foretelling a mighty flood, told Ina that he would visit her in his eel-form at her house, when she was to cut off his head, and bury it at the back of her hut. A 204 STUDIES IN ANCIENT HISTORY CHAP. flood came, and Ina did to Tuna as he had directed ; and from the two halves of the buried head sprung the two principal varieties of the cocoa-nut, which it would surprise no one to find religiously regarded (pp. 77, 78). " In the year 1855, at the place indicated in this story, an enormous eel, measuring 7 feet in length, was caught by daylight in a strong fish-net. In heathenism this would have been regarded as a visit of Tiaio, and the dainty morsel allowed to return under rocks un- molested. As it was, it furnished several families with a good supper" (p. 79). From this we make two inferences : (1) that in heathenism the clansmen did not eat their totem in Mangaia, a fact which may be proved independently ; and (2) that the possession of the eel by Tiaio was not, strictly speaking, temporary, but that the promoters of the worship of this man-god maintained that it was truly his spirit that was in both the eel and the shark. One word in conclusion : it was from the hibiscus that the inspiration reached the first priest of Motoro, and it would be agreeable with all we know that the hibiscus was a plant god or totem before the thought was formed of displacing the spirit in it by that of a man. As illustrating the ancient pre-eminence of women in the Hervey group, it may be noted that Makea, of the family of Karika, king of Earotonga, worshipped a goddess as his special divinity. We also have a note of totemism in the formula contained in the mythical story x HERVEY GROUP, OR COOK'S ISLANDS 205 of the submission of Tangiia to Karika. Tangiia, ac- cording to one formula literally, Yours is the long- legged, or man belongs to you made over the political supremacy. According to another formula literally, Yours is the short-legged, or the turtle belongs to you he made over the spiritual supremacy. The turtle, according to Mr. Williams, as the most sacred fish, was considered the emblem of supremacy in religious affairs (Williams, I.e. p. 51). Another indication of the position of woman is this. Mr. Williams (I.e. p. 56) points out, that while in the Tahitian and Society groups females had a share of their father's possessions, no share of these went at Rarotonga to the daughters, on the ground, as they alleged, that their person was their portion. Whatever may have been the explanation, there is no doubt that the women of Earotonga were in a position to be fastidi- ous in the choice of husbands, and, as Williams says, however great a man's possessions might be, no woman would have him unless he were personally attractive to her, which shows a marked independence in regard to marriage. CHAPTER XI FIJIAN GROUP THIS great group of islands was discovered by Tasman, the Dutch navigator, in 1643, after which they were un visited till Cook's time. It was not till 1806 thai they began to be visited by traders, and the first appearance of the missionaries in the islands was as recent as 1835. The islands have now been annexed to the United Kingdom. The natives were speedily converted first, and slowly extinguished afterwards. Comparatively few of the natives remain, and our chance of knowing well what were their laws and customs is perhaps gone for ever. The chief authority on the manners and customs of the Fijians is undoubtedly Mr. Thomas Williams. 1 Mr. Thomas Williams seems to have been a man of more than ordinary intelligence, but absolutely untrained in juridical notions Like other missionaries, too, he employed a very loose ter- minology in describing Fijian society; "tribe" means almost anything in the way of a group, and he has no term for clan. Again, he seems not to have attended to the jus connubii, and found nothing remarkable in it; at any rate he never once 1 Fiji and the Fijians, edited by George Stringer Rowe. Vol. i. The Islands and their Inhabitants, by Thomas Williams. Vol. ii. Mission His- tory, by James Calvert. London, 1858. There is another edition dated London 1870, "extended with notices of recent events by Sir James Calvert." The references are all to vol. i., except when otherwise men- tioned. CHAP, xi FIJIAN GROUP 207 mentions it. It may be, however, that he is not to be charged with this oversight. The history of the book is curious. A missionary, Mr. Calvert, returning to England in 1856, brought with him "a copious manuscript on The Islands and their Inhabitants" which had been prepared with great care and skill " by Mr. Williams, a brother-missionary." The MS. was handed to Mr. George Stringer Eowe, who " re- wrote most of what was supplied to him, he apparently having no special knowledge of the subject." ] So the picture is not at first hand ; it is moreover professedly incomplete. No doubt it is Mr. Eowe who at p. 132 says, " But here, even at the risk of making the picture incomplete, there may not be given a faithful representation" of the licentious sensuality of the Fijians, i.e. practically of the relations of the sexes among them. It is Mr. Eowe as editor who signs the note at p. 214, intimating that "much detail and illustrative incident" fur- nished by Mr. Williams have been omitted. Fortunately this note refers to cannibalism specially, though it is appended to the chapter on manners and customs. We may assume that Mr. Eowe's discretion was exercised throughout to make the book a good readable and saleable record of missionary work. But with what a minute omission might evidence of the highest scientific value disappear. May not even the vague term "tribe," used in all sorts of senses, have been an editorial " neatness " of expression, substituted for clumsier phrases of Mr. Williams struggling to be accurate in his way " with great care and skill " ? However, the MS. has long disappeared, and we must make what we can of the re'chauffe'. One point in its favour is that it belongs to the pre-speculation period, and so far may be trusted. So do our other chief sources of informa- tion as to Fiji Erskine's Pacific (Lond. 1853), and the Narra- tive of the United States Exploring Expedition (Lond. 1845). It has been said that Thomas Williams is our chief autho- rity as to Fiji, and that he used a vague terminology. If, then, 1 Some additions were also made of facts which had " transpired since Mr. Williams gathered and arranged the fruit of his own personal observa- tions and inquiries." 208 STUDIES IN ANCIENT HISTORY CHAP. we would be instructed by him we must reach his meaning by comparing his various statements, and by sifting and analysis of statements where we have no means of reducing them by comparison to their real elements. It is a tedious process for reader and writer alike ; but there is no way of helping it. To condense into a brief statement one's views of what he had really to say without exhibiting the grounds for them, might make one sometimes seem to be flying in the very face of our authority unless, indeed, the reader was to take the trouble involved in examining the evidence for himself. There are no fewer than 225 islands and islets in the Fijian group, about 80 of which were inhabited when Williams wrote. Some of these were very small, with but one village and from 50 to 100 inhabitants (p. 5); others larger, with 200 to 400 inhabitants; others again, from 15 to 30 miles in circumference, had from 1000 to 7000 in- habitants. Mbau, " a small island scarcely a mile long, joined to the main Viti Levu by a long flat of coral, which at low water is nearly dry and at high water fordable " (p. 5), was when "Williams wrote the chief political power of Fiji. The town covered a great part of the island, and it is noted that its tall temples (so that there were a number) helped to give it a striking appearance. " Its inhabitants comprised natives of Mbau and the Lasakau and Soso tribes of fisher- men " (p. 7). This little island was when Williams wrote the centre of political power in Fiji. As to the larger islands, Mr. Williams estimated that Kandavu had from 10,000 to 13,000 inhabitants; Vanua Levu (the great Land) 31,000; Na Viti Levu (the great Fiji) at least 50,000. I. TRIBES AND CLANS We may omit notice of Williams's speculations as to what the government of Fiji was " before the last hundred years." What it was in his day is described as follows : " There are many independent kings who have been constantly at war with each other ; and intestine broils make up, for the most part, the xi FIJIAN GROUP 209 past history of Fiji. Still, though to a much less extent, civil dissensions abound, and it is not vmcommon for several garrisons on the same island to be fighting against each other. The chiefs have been ever warring among themselves " (p. 18). Here are kings : of what ? Here are chiefs : of what ? The kings, of course, had kingdoms. Mbau was the capital of the chief kingdom. 1 There were kingdoms subject to it but only nominally. " The other leading powers (king- doms) were Eewa, Somosomo, Verata, Lakemba, Naitasiri, Mathuata, and Mbua" (p. 20). But what sort of kingdoms were there that so many of them could exist, amongst a total population variously estimated at from 133,000 to 150,000 ? and besides the leading kingdoms above mentioned, there were many smaller ones. In fact a single village might constitute a kingdom (see p. 24), and its headman be called king. The kingdoms clearly were only local tribes or village com- munities. What, then, were the " tribes " ? This is more puzzling. The puzzle faces one early in the book, and confronts one steadily to its close. Of the small island of Mbau (scarcely a mile long) we are told that "its inhabitants comprise natives of Mbau and the Lasakau and Soso tribes of fishermen." There were therefore two tribes of fishermen. Were the " natives of Mbau " other than the fishermen in tribes also ? At p. 20 we have this information : " The name of the tribe from which the kings of Mbau are taken is Kamba. The four chief personages or families in this state are the Eoko Tui Mbau, the Tuni-tonga, the Vusar-andavi, and the Tui-Kamba." Now Tui means king (see p. 33), and we have here four persons named. They represent families, of which they are chiefs. But one of these families is Kamba, and in the first branch of the statement Kamba is a tribe. Therefore tribe seems to be equi- valent to family, in this statement. But this may be mere matter of style on the part of the editor, and we must not be too sure. We soon find reason for hesitation. At p. 3 3 we have the following statement : " Tribes, chief families, the houses of chiefs and the wives of kings have distinctive appellations, to 1 Williams, vol. i. p. 105. P 2io STUDIES IN ANCIENT HISTORY CHAP. which great importance is attached." Unfortunately this is not illustrated, except it be counted an illustration that we are told that the hereditary title of the king of Mbau was Vu-ni- valu (the root of war). Had examples in each of the enumerated cases been given, we might have known (what the statement leaves doubtful) whether " tribes/' " chief families," and " the houses of chiefs " were synonyms or different things. But for wives of kings being enumerated in the list, it might be supposed that these terms were here synonymous. 1 Let us try again. There is to be a great feast, and this is the manner of it : " The food prepared by each tribe and family [of the entertainers] is presented for inspection, and in some cases collected and piled before the house of the king. . . . The food having been divided into as many portions as there are tribes [among the visitors], the tui-nara [a sort of master of the ceremonies], beginning with the first in rank, shouts out 'The share of Lakemba,' or whichever [tribe] may take pre- cedence. This is met by a reply from that party, ' Good, good,' or 'Thanks, thanks,' and a number of young men are sent to fetch the allotted portion. The tui-nara goes on calling the names [of the tribes] in succession until his list is ex- hausted. . . . When each tribe has received its share a re-division takes place, answering to the number of its towns ; these again subdivide it among the head families, who in their turn share what they get with their dependants, and these with the individual members of their household, until no one is left without a portion " (p. 148). Here as regards the visitors tribe = kingdom = local tribe comprising as many towns or villages ; town or village again appears as comprising and represented by several head families ; while the head families are exhibited to us as having dependants, and these as having households, so that each head family gives 1 By houses of chiefs, moreover, their dwelling-places may be meant. The name of the house of the king of Somosomo is separately given it was JSTasima. It is stated that the chief wife of that king would have been saluted as the lady from Nasima. The names of one or two oth houses of kings are mentioned in Jackson's narrative appended to Erskin Southern Pacific. xi FIJIAN GROUP in us, besides the family of a headman, a series of other families connected with him, using family in the sense of household. As regards the entertainers, it is obvious that tribe is used in a different sense, since it necessarily refers to some portion or subdivision of the local tribe where territory is spoken of as a kingdom. Nevertheless this account of a feast gives us in a rough way (and to a certain extent) the constitution of society among the Fijians. The head family with its dependants formed the unit of this town or village population in general (for possibly one such body sometimes had a town to itself), several head families with their dependent households made up a town or village population, and an aggregation of connected towns made up the kingdom. It cannot be doubted that head families so different from each other were regarded as being of different kinships, and it is at once suggested that each head family and its dependants formed a clan a point upon which no doubt will remain if it appear (as it will do by and by) that there was a religious connection between them. There is nothing to show whether the same families appeared in the several towns of a district, but the chief families spoken of at p. 33 must have been, at any rate, of the same sort as the head families which appeared in each town population; while in Mbau, where a single town covered a great part of the island, the head families of the town are necessarily the same with those which are spoken of (p. 20) as the chief families of the king- dom, and whose headmen or chiefs were its chief personages. The kingdom of Lakemba is mentioned in the passage just commented on, and some casual notices in Mr. Calvert's part of the book (for the form of which also Mr. Eowe is responsible) throw further light upon the constitution of the town popula- tions of Lakemba. Of Yaudrana, " the most populous town on Lakemba " (but not the king's town), we are told at one place (vol. ii. p. 117) that " Lua, the head chief of the town, with three other of the principal men [chiefs] and a few others, met in a heathen temple as representatives of the four tribes be- longing to that settlement." It is quite clear that by settlement the town is meant, and that the four tribes spoken of constitute the town population ; so that the head family to which, no 212 STUDIES IN ANCIENT HISTORY CHAP. doubt, the chief or " principal man " belonged, with its depen- dent households, is here spoken of as a tribe, a term which would be strictly correct if they formed a tribe of descent or clan. Again, the king of Lakemba having resolved to turn Christian, to the great delight of the majority of his subjects, " a meeting of the principal chiefs and people [probably of the whole island] was held in the king's house. . . . Among other things it was agreed that the common people [dependent households] should be respectful to their own chiefs and to the king, and that all should be industrious " (vol. ii. p. 139), where again the people appear grouped together as if they were clans under their respective chiefs. The opening of a new chapel in the town of Lakemba, the king's town, celebrated by a gathering of the population, at which Wetasau, a chief of the town and the next in rank to the king, presided, is described further on (vol. ii. p. 149). " In the afternoon we assembled again to receive the contributions of the people, who entered the spacious chapel according to their tribes. The king, leading the way with a few of his principal men, presented his free-will offering and sat down. Then the people, each tribe accompanied by its chief, chanting as they moved slowly onwards, brought their gifts." Here also we have the head family and its dependent household spoken of as a tribe, and we see the tribe acting together in worship under its chief, and offering its gift as a tribe ; so that the population was made up of bodies under chiefs closely united together, which had every appearance of being clans. That they were so will appear more clearly when the religious notions of the people are examined. It will be easy for the reader to discover for himself many proofs that such bodies were the units of the town populations, and that the term tribe is applied by Williams equally to them and to the local tribe which, taken together, they constituted ; that is, the whole population which was under the same govern- ment. Where tribe is used by him to denote something more than the former, but less than the latter, the meaning is always uncertain ; but it may mean the aggregate of those bodies con- tained in the several towns of a kingdom, which were united by relationship and worship. In Mr. Calvert's narrative the xi FIJIAN GROUP 213 term appears not to be applied to the whole of a town population. II. TOTEMS The religions system of the Fijians presents the clearest signs of a totem origin. Most of their gods were, plainly, developed totems. Certain " birds, fish, plants " were supposed to have deities " closely connected with or residing in them " ; one god was " supposed to inhabit the eel, and another the fowl, and so on, until nearly every animal became the shrine of some deity" (pp. 219, 220) ; and while these were real gods, and had temples and priests whom they inspired, through whom they thus communicated with their worshippers and were propitiated with offerings and sacrifices, the " shrine " itself also received worship from the people, the animal which a god inhabited was never eaten by the worshippers of that god, and it comes out casually that some at least of the worshippers regarded them- selves as descended from that animal. It appears, too, that those who worshipped the same god regarded each other as relations, and treated each other accordingly, even when the districts to which they belonged were far apart, and there was scarcely any intercourse between them. There were gods who presided over districts and islands, and gods who presided over tribes and families, " their influence never reaching beyond their own special jurisdiction " ; and while a superiority of rank is claimed for Ndengei, the serpent, each district is said to have, with this exception, contended for the superiority of its own divinity. As each " tribe and family " had its own god, it is clear (though that, indeed, could not have been doubted) that the head families which were found in every town were of different kinships, and that the tribe consisting of a head family and its dependent households had its own god, and therefore that it was a clan, the successor, with such change as time had brought about, change especially in the relation of the chief and his family to their people, of the totem kindred of an earlier period. The higher chiefs claimed almost to be gods ; and it may be that it was for chiefs and their families only that a claim of descent from the totem could be preferred 214 STUDIES IN ANCIENT HISTORY CHAP. in Williams's time. The benignant totem protector of the kindred at any rate appears to have been completely lost sight of. Fijian religion afforded no conception so respectable. We are not told anything definite of the district or island divinities, but it may be believed that the clans of a district, while each holding by its own god, allowed the superiority, or the supremacy, of the god or totem of its ruling clan, which would thus be regarded as presiding over the tribal territory. The same deity was worshipped under different names in different localities, the " shrine," that is, being the same ; the god supposed to inhabit it was variously designated. " At Lakemba," we are told, " Tui Lakemba, and on Vanua Levu Eavuravu, claim the hawk as their abode. Vivia and other gods the shark" (p. 219). Mr. Williams gives casually (p. 21) a curious tradition which shows how a god might continue to be thought of as identified with the animal out of which he had been developed, and as hampered by his animal form. One of the chief gods of Somo- somo was a rat, and he had a good, loud-sounding, man-god name, Ng-gurai. Having resolved to pay a visit to Mbau, Ng-gurai " entered into a rat, took his club [like a man-god], and started " ; but, being a rat, he narrowly escaped devouring by the way, and fared so badly as a rat that, as a result of his adventures, Somosomo became tributary to Mbau. The people could not rid themselves of the impression of his rat nature. In another district tradition makes the rat determine that men should die. " Ea-Vula (the moon) contended that man should be like himself disappear awhile, and then live again. Ea Kalavo (the rat) would not listen to this proposal, but said, 'Let man die as a rat dies.' And he prevailed" (p. 205). But Mr. Williams shows us more directly how the animal which was the " shrine " of a god was regarded by his followers. "The land-crab," he tells us (p. 220), "is the representative of Eoko-Suka, one of the gods formerly [i.e. before the conversion to Christianity] worshipped in Tiliva, where land -crabs are rarely seen, so that a visit from one becomes an important matter. Any person who saw one of these creatures, hastened to report to an old man who acted as priest that their god had favoured them with a call. Orders were forthwith given that xi FIJIAN GROUP 215 new nuts should be gathered, and a string of them was formally presented to the crab, to prevent the deity leaving with an impression that he was neglected, and visiting his remiss worshippers with drought, dearth, or death." 3 Similarly in the Hervey group, the appearance of a large eel even in the nets would have been regarded as a visit from the eel-god Tuna, who was a man-god as well as an eel. Again : " The heathen sailors are very superstitious. . . . The common tropic bird is the shrine of one of their gods, and the shark of another ; and should the one fly over their heads or the other swim past, those who wore turbans would doff them, and all utter the word of respect. . . . Canoes have been lost because the crew, instead of exerting themselves in a storm, have quitted their posts to soro [i.e. give atoning or propitiatory offerings] to their god"; for example, to the shark, if they happened to see one (p. 89). Not all the divinities were gods, however. In the island of Vatulele the divinity, or one of the chief divinities, was a goddess. 2 Mr. Hunt, we are told (vol. ii. p. 256), went to see the place of her residence with one of the chiefs of the people who worshipped her. " The objects of the superstitious veneration of these poor creatures," he says, " are nothing more than a number of red crustaceous fishes, larger than a shrimp. There is abundance of them in Fiji. . . . The mother of the fish is said to be of immense size, and to reside in a large cave by herself, and her children leave her when called by their name, which in Fijian is Ura." He pro- ceeds to describe the cave and its surroundings, and then bears this testimony to the sincerity of the faith that so amused him : " The chief stood at the mouth of the cave and called with all his might, ' Ura, Ura, come, that the chief from England may 1 There were no idols. There was nothing for the people to worship except the "shrine" of the god. The god communicated with his people by inspiring the priest, which he did when properly propitiated with sacrifices and gifts. As among peoples merely in the totem stage, " gods are supposed to enter into some men while asleep." 2 Mr. Williams tells us little about goddesses, but there must have been something to tell ; for he says (p. 110) that "the arithmetical skill of goddesses is an article of Fijian faith." 216 STUDIES IN ANCIENT HISTOR Y CHAP. see you.' There was no answer, however, and but few of the fish appeared. They took no notice of their worshipper. ... I tried to convince him of the folly of considering such things as these to be gods." 1 As to the other unmistakable signs of totemism already briefly referred to : 1. The worshipper of a god was prohibited from eating the species of animal to which the god belonged, or which he inhabited. " He who worships the god dwelling in the eel must never eat of that fish, and thus of the rest ; so that some are tabu from eating human flesh, because the shrine of their god is a man " (p. 220). In this paradise of totems, the men- gods, to rank as gods at all, had to do as the earlier gods did, and to class in all respects with them. The observance of the tabu against human flesh must have been a real trial of faith among a people who relished that food so much. 2 1 The following is from Jackson's narrative of a residence in Fiji, appended to Erskine's Western Pacific (p. 434) : " One day while I was at a place called Vusaratu, the natives gave me some eels to eat, and asked me if we had any in ' Papalangi ' (white man's country). When I said we had, they asked me if there were any king eels amongst them ; I answered, No ; when they straightway conducted me to a fresh-water hole with a temple erected at one end. In this hole there was an immense- sized eel \ his body at the thickest part was as big round as a stout man's thigh, and his head was enormously large and frightful, but his whole length I could not tell ; they said he was two fathoms long. I inquired the meaning of the temple ; they said it was his, and that he was a kalvu (or spirit). I thought I would prove the veneration they held him in, so I pointed my musket at him and cocked it ; they seemed extremely agitated, and begged me to desist, and then ran off to fetch some cooked breadfruit to propitiate him for the insult offered, which he took from their hands. They told me that he was of great age, and that he had eaten several infants, which they had given him at different times children of prisoners taken in war." The king eel may have been of a different variety from the eels given to Jackson to eat, or these may have been given him by persons who were not worshippers of the eel-god. It will be seen immediately that such worshippers would have been prohibited from eating the eel. 2 Mr. Williams tells us that besides birds, fish, and plants (he might xi FIJIAN GROUP 217 2. Having the same god implied being of the same stock, i.e. being relations. Several casual proofs of this occur. " It is remarkable that the people of Ono, the most distant island, say that they originally belonged to this locality [place near Na Sava, Vanua Levu] ; and it is still more remarkable that there exists a dialectic similarity between these extremes ; and the inhabitants of each are tauvu, worshippers of the same god, and in virtue of this may take from each other what they like, etc." (pp. 253, 254). Does this suggest an ancient community of property in the clan ? It establishes that relationship was admitted between those who were fellow-worshippers, beyond a doubt. Take another case, also quite casually mentioned like the preceding, which is entirely disconnected from Williams's exposition of the social structure and religion of Fiji. It puts the relationship of fellow - worshippers beyond doubt. In Gal vert's Mission History (vol. ii. p. 94) we find the following : " The people of this island (Vanuambalavu) and the Oneatans were related, and had the same gods ; and, therefore, according to Fijian custom, enjoyed the privilege of pilfering each other's goods with impunity." l have added animals, since the rat was the shrine of a god), " some men " were supposed to have deities closely connected with or residing in them (p. 2 1 9), but he does not explain or illustrate the statement. But that he did not pretend to have fully mastered the mysteries of the system of faith about him, appears from the following statement : " Some priests are tabu from eating flesh. The priest of Ndau Thina has assured me, that neither he nor those who worshipped his god might eat it, nor might the abomination be brought into the temple. Probably the shrine of Ndau Thina is a man, and hence the prohibition. To the priest of second rank in Somosomo, I know that no greater delicacy could be presented than hashed human flesh " (pp. 231, 232). We are told in another place (p. 218) that Ndau Thina was a god who stole women of rank and beauty by night, or torch-light. It may be taken that he was not one of the totem- gods, and that he was known only by description and as inspiring his priest. 1 Ratu Nggara, king of Rewa, while at war with Mbau, which had become Christian, being urged to become Christian, refused, saying, " If we 2 1 8 STUDIES IN ANCIENT HISTOR Y CHAP. 3. There remained among the Fijians even a claim of descent from the totem that is, from the " shrine " inhabited by the god whom they followed. Toki, a chief of Ravi-ravi, we are told (p. 123), " used to speak of himself as the offspring of a turtle, regarding all other chiefs as the progeny of inferior fishes," the meaning of which is unmistakable. We are told elsewhere (p. 24) that " in some instances, Fijian monarchs claim a divine origin, and . . . assert the rights of deity, and de- mand from their subjects respect for those claims." This, Mr. Williams adds, "is easily yielded, for the pride of descent which runs so high among the chiefs is equalled by the admira- tion in which their lofty lineage is held by the people, who are its sincere and servile worshippers." Claiming a divine origin is, of course, claiming what Toki claimed, a descent from the totem, and it may be inferred from Toki's words that they all did as much as that. The assertion of the rights of deity is probably what occurred only "in some instances." It may well be, too, that the claim of the chiefs was the more readily allowed because the people claimed descent from the totem for themselves, but as to that there is no evidence. 1 Mr. Gill got his information as to the gods of the Hervey group from the priest of the shark-god. It is not said who instructed Mr. Williams, but it may be suspected that it was an ex-priest of Ndengei, the serpent, who has supreme rank assigned to him in the theogony as taught to Mr. Williams. As a matter of fact, the serpent whose clan we must suppose all lotu [become Christian] we must give up fighting ; as it will not do to pray to the same god and fight witli each other " (vol. ii. pp. 85, 86), which also shows that it was proper for followers of the same god in different districts to be friends with each other, to treat each other as relations and never as enemies. The people of Namuka and Mbau also, according to Mr. Seemann (Viti, London, 1862, p. 229), had the same gods, in consequence of which " the people possess mutual rights similar to those of the Yasus, visitors being allowed to take whatever articles they choose." 1 " When at Lakemba I was told by Moses Vakaloloma that, in their heathen state, they did not address their little ones as children, but would say, ' Come here, you rats ' " (p. 1 7 7). xi FIJIAN GROUP 219 to have been at one time very powerful had when Williams wrote but few worshippers : " Except about Eakiraki he has scarcely a temple " (p. 217). But here is a sample of the boasting of the serpent-priest, inspired by and speaking for his god : " Great Fiji is my small club ; Muaimbila is the head; Kamba is the handle. If I step on Muaimbila I shall sink it into the sea, whilst Kamba shall rise to the sky. If I step on Kamba it will be lost in the sea/' and so on (p. 225). The reader will find, however (p. 231), a suggestion that the chief of the serpent clan had sufficient authority to regulate the worship of the god. Human sacrifices were the delight of Ndengei, and the chief in disgust stopped them. What Mr. Williams has to say of the other gods would certainly not prepare us to find undisputed supremacy allowed to Ndengei. " The rank of the gods below Ndengei," he says, " is not easily ascertained, each district contending for the superiority of its own divinity" (p. 219). " Tokairambe and Tui Lakemba Eandinandina," he adds, "seem to stand next to IsTdengei." Now Tui Lakemba was a hawk-god of Lakemba (p. 219). Did the people of the hawk admit the inferiority of the hawk to the serpent ? One must doubt it much. And would Toki, the Eaviravi chief, who "used to speak of himself as the offspring of a turtle, regarding all other chiefs as the progeny of inferior fishes," have admitted the serpent to be superior to the turtle ? An account of the origin of the human race, given by a chief from the Kauvandra district, assigned a more important part to the hawk than the serpent. The hawk having produced two eggs, Ndengei hatched them, and they yielded two human infants, a boy and a girl (p. 251). Since hatching goes for nothing, mankind is here represented as of the hawk stock. 1 1 The priesthood was hereditary, but a man who could " shape well " might get himself accepted as the priest of the god by whom he professed to be inspired (who, perhaps, could only be one of the newer non-totem gods). The priest took rank from his god, from the number and power, that is, of the followers of the god. There were priestesses also (it has appeared that there were goddesses), but all we learn of them is that few were of sufficient importance to have a temple. We are told of "strangers" 220 STUDIES IN ANCIENT HISTORY CHAP. We have now seen that the population of Fiji (1) was comprised in local tribes, variously called by Mr. Williams kingdoms and tribes ; (2) that the local tribes were composed of various clans, spoken of sometimes as tribes and sometimes as head or chief families, and their dependents ; (3) that eacl of these clans had its god of the totem type an animal, fish, reptile (but sometimes perhaps a man) ; (4) that the clansmei dared not eat of the animal, etc., which represented their god and (5) that being " fellow- worshippers " meant being clansmei and relatives, and, as appeared in the case of the chief of th( turtle stock, being of the stock of the god. We must wishing to consult a god ; and we are told also that, while every island had its own gods, and each locality its own superstitions, " almost every individual had his own modifications of both." It may be believed that there was much disposition to propitiate any god who was reputed power- ful at any rate, when not the god of a particular tribe ; and this indeed was the secret of the conversion of the islanders to Christianity. In their conversion, though the clan seems to have generally followed the chief, individuals did not wait for him. The Fijians, we are told, " reverence certain stones as shrines of the gods, and regard some clubs with superstitious respect." Of the clubs we learn nothing further ; the stones were phallic. "We are told that one of them was the abode of a goddess. But possession of them seems some- times to have been contested. A man who was inspired by Tanggirianima said, " I and Kumbunavanua only are gods. I preside over wars, and do as I please with sickness. But it is difficult for me to come here, as the foreign god fills the place. If I attempt to descend by that pillar, I find it preoccupied by the foreign god. If I try another pillar, I find it the same. However, we two are fighting the foreign god ; and, if we are victorious we will save the woman." Who Tanggirianima was does not appear, but the phrase, "a man who was inspired by Tanggirianima," suggests that this person was not one of the hereditary priests, and there- fore his god was not one of the ancient gods. Besides the gods proper (Kalou Vu) Mr. Williams says there were deified heroes (Kalou Yalou), into whose number admission was not difficult for any one who found a priest to take him up. Of the non-totem gods, one came from the centre of a large stone, and may therefore have been phallic ; another was a giant in human form, sixty feet high ; another was the one-toothed lord, a man with wings. xi FIJIAN GROUP 221 inquire what was the system of kinship, and whether there is reason to think that anciently the Fijians were organised in clans on the totemic principle, such as we find among the Iroquois. III. KINSHIP The subject of kinship is not touched upon by Mr. Williams. As it is only owing to accidents in his narrative that we know that there were clans in Fiji, it is needless to say we are not told whether a child was held to belong to his father's or his mother's clan, or whether fathers and mothers There were also the wooden-handed god ; the eight-armed, the two-bodied, the miracle-spitter ; the leper ; a war-god, "worshipped at Na Vanindoaloa" ; the adulterer, the woman-stealer, the rioter, the brain-eater, the murderer, and " a host besides of the same sort." We have found it stated casually as to one of these gods, that he had a priest and a temple ; and we have just seen that another is mentioned in connection with the pillars (pp. 216-227). Two others of the non- totem gods at least are mentioned as having priests and regular worship Nva-Kandiote, the war-god of Na Vanindoaloa (whose priest was entitled to all one-eared pigs born in that district), and Kanusimava (spit-miracles). No doubt they all had. On the occasion of the priests being consulted at Somosomo as to whether the "tribe" should go out to battle, we are told (p. 226) that "a long list of deities " was enumerated by the chief priest, and all the priests who were present shared in the offering more or less. The oracle was propitious, but the priest of the miracle-spitter being dissatisfied with his share a very small one brought out an oracle of his own next morning and stopped the war. We find Thakombau (vol. ii. p. 313) going to consult the priest at his own " small family temple " before going to the chief temple of Mbau for the general consultation. The family priest promised Thakombau the protection of his god, but would not go further. " Yes, you have always protected us," the chief replies ; " that we expect. But now we require the destruction of our enemies." Wilkes's United States Exploring Expedition (vol. iii. p. 84) refers to the totem gods as " the tribal gods that have no authority except over the tribe." He calls them " the distant relatives of Ndengei," and says "they are all benignant," as totems ought to be. According to Wilkes (ibid. p. 83), some say Ndengei had but one son, Mautu, the bread-fruit ; others say that he had two sons in the form of men. 222 STUDIES IN ANCIENT HISTORY CHAP. were usually of the same or of different clans. If we can 01 his testimony learn anything on these subjects it must be argument on, and inference from, casual statements made him on other subjects. That kinship had shifted in Fiji from the female to th( male side when Williams wrote, might be confidently inferre< from the terrible prevalence of infanticide chiefly fei infanticide to which he and others bear witness ; a prevalence which, considering the operation of the law of blood-feud, would be absolutely incompatible with the system of kinship through women only. 1 And it is in connection with infanticide that we find the only express statement made by him from which a sure inference as to the system of kinship in Fiji can be made. A reason for killing a child is suggested as follows : " Perhaps the parents belong to two tribes that are at enmity, in which case the mother, rather than multiply the foes of her tribe (clan), will destroy her progeny" (p. 180). This implies that the child belonged to the father's clan, and not to the mother's. 2 But that kinship was formerly counted through women only may be inferred from the following con- siderations : (1) " Bank," Williams tells us, " is hereditary, descending through the female " (p. 32). Though this is illustrated and 1 See ante, p. 85. As to the prevalence of infanticide, this practice, it is stated (pp. 180, 181), seemed to be universal, on Vanua Levu quite a matter of course ; " the extent of infanticide in some parts of this island reaches nearer to two-thirds than half. ... I know of no case after the child is one or two days old ; and all destroyed after birth are females, because tliey are useless in war, or, as some say, because they give so much trouble. But that the former is the prevailing opinion appears from such questions as these put to persons who may plead for the little one's life. * Why live ? will she wield a club, will she poise a spear ? ' When a pro- fessed murderess is not near, the mother does not hesitate to kill her own child." 2 " It is a common practice to name the first child after the man's father, and the second after the mother's father. In the first case, the friends of the man make the wife a present, and in the other her friends offer the gift to the husband " (p. 176). xi FIJIAN GROUP 223 its origin explained (by the polygamy of the chiefs and the widely different grade of their numerous wives) in reference to the chieftain class only, yet the statement is in the most general terms, and follows immediately the enumeration of the six recognised ranks or classes in Fiji. It is possible, no doubt, to press such a statement too far. But it may, nevertheless, be taken as a note of the system of kinship through women only having formerly prevailed in Fiji, that children should take rank from the mother and not from the father, there being nothing in the marriage system to prevent effect being given to paternity. Polygamy is in the nature of things too rare in any population to be made an explanation of any of its general laws. In Fiji it must have been confined to the chieftain class. (2) "When rule is strictly followed/' says Mr. Williams (p. 24), "the successor of a deceased king is his next brother; failing whom his eldest son, or the eldest son of his eldest brother, fills his place. But the rank of mothers and other circumstances often cause a deviation from the rule." This will be recognised as the succession law (somewhat imperfectly stated) peculiar to or immediately derivable from the family system as founded on Thibetan polyandry and this may be taken as a further suggestion of the prevalence at one time of the system of female kinship. 1 (3) This suggestion receives irresistible force from the system of Vasuing which prevailed throughout Fiji. We have accounts of Vasuing from Erskine, Williams, and Wilkes, and they are substantially in agreement. " The word [Vasu]," says Williams, " means a nephew or niece, but becomes a title of office in the case of the male, who in some localities has the extraordinary privilege of appropri- ating whatever he chooses, belonging to his uncle or those under his uncle's power. Yasus are of three kinds : the Vasu taukei, the Vasu levu, and the Vasu : the last is a common 1 At p. 181 we read of a man, whose brother liad died leaving a son and an infant daughter, taking these children as his own, and that the infant might get proper care, arranging with his wife, who was just then confined, to murder their own baby. 224 STUDIES IN ANCIENT HISTORY CHAP. name belonging to any nephew (sister's son) whatever. Vasu taukei is a term applied to any Vasu whose mother is a lady of the land in which he is born. . . . No material difference exists between the power of a Vasu taukei and that of a Vasu levu, which latter title is given to every Vasu born of a woman of rank and having a first-class chief for his father. A Vasu taukei can claim anything belonging to a native of his mother's land, excepting the wives, home, and land of a chief. . . . However high a chief may rank, however powerful a king may be, if he has a nephew [sister's son], he has a master. . . . Kesistance is not thought of, and objection only offered in extreme cases. A striking instance of the power of the Vasu occurred in the case of Thokonauto, a Eewa chief, who during a quarrel with an uncle used the right of Vasu and actually supplied himself with ammunition from his, enemy's stores. . . . Great Vasus are also Vasus to great places, and when they visit them at their superior's (say the king's) command they have a numerous retinue and increased authority." After an account of the reception at Somosomo of the Vasu levu (great Vasu) of that district, who was from Mbau, Mr. Williams makes a statement which shows that this right of the sister's son ran against every head of a family. 1 " Descending in the social scale, the Vasu is a hindrance to industry, few being willing to labour unrewarded for another's benefit. One illustration will suffice. An industrious uncle builds a canoe, in which he has not made half a dozen trips when an idle nephew mounts the deck, sounds his trumpet shell, and the blast announces to all within hearing that the canoe has that instant changed masters" (pp. 34-37). Here again we may be excused for wishing that we had the very account of Vasuing that Mr. Thomas Williams wrote. What precedes, though seemingly clear and in a literary view excellent, is obviously very incomplete. At what age, and on what conditions, if any, did the nephew acquire his right as Vasu ? The context suggests that the statement is very in- complete. For example, Mr. Williams says elsewhere (pp. 201, 202) : "Some women, it is said, submit to be strangled 1 Perhaps this was only in some localities. xi FIJIAN GROUP 22$ [when their husbands die], that they may prove thereby the legitimacy of their children. This particularly refers to such children as are Vasus." And we are told in Erskine's work l (the statement is Jackson's) that a man's right as Vasu was open to doubt until his mother had proved her fidelity as wife by dying with her husband, and would be denied if she failed to do so. It is hard to see any connection between the two things, but so it is stated, and if correctly stated, the rights of a Vasu were not fully acknowledged (though they might be admitted previously) till after the death of a man's father and mother. It is left somewhat uncertain too against whom they lay. Against "a native of his mother's land," says Mr. Williams, speaking specially of the Vasu taukei ; while by his definition of Vasu taukei the mother had to be a lady of the land in which the Vasu was born. Does this limit the rights of the Vasu to the people of the land in which he himself was born ? Were that the whole matter, the right to take the highest case of which Mr. Williams seems to have been think- ing would in effect be a mere anticipation by the heir pre- sumptive of the powers of his father ; for being born in the land of his father, in which his mother by definition was a lady, his rights would lie against his father's subjects. The case of the Mbau chief already mentioned, however, shows us the right exercised by a Vasu levu against a foreign power, and the same thing occurs in all the other concrete examples of Vasuing which are casually given. These occur in Mr. Calvert's part of the work, relating to mission progress. The king of Somosomo, for example, had two sons. In dis- cussing the advantages of establishing a mission in Somosomo the following consideration had its weight : "The king's territories were very extensive. The [his] two sons were not only of high rank on their father's side, but their mother was a Mbau lady of highest family, which made them Vasus to all the chiefs and dominions of Mbau " (vol. ii. p. 35). Here the point is that the mother belonged to the highest family in the land from which she came, not in the land in 1 Erskirie's Western Pacific, 1853, p. 448. Q 226 STUDIES IN ANCIENT HISTORY CHAP. which the Vasu was born, and the Yasu's rights lay against her kindred and their dependents. Tanoa, again, was the old king of Mbau, and his son Thakombau was acting king. Thakombau took deep offence on a certain account against Eewa. But Eewa was a power- ful state, and there were reasons for not instantly taking ven- geance. " The mother of the old king of Mbau [Tanoa] was a lady of the highest rank from Eewa, and related to most of the principal chiefs of Eewa. Furthermore, Thakombau's rival brother Eaivalita was a high Vasu to Eewa, his mother being sister to the reigning king [for the same reason Tanoa also was in his time Yasu of Eewa]. He would therefore as a matter of course be favourable to his mother's relatives, since the law of the land permitted him to claim and take their property as he saw fit" (vol. ii. p. 175). Here, again, the point is that the mother of the high Vasu must be a lady of the land to which she belonged by birth, and that the Vasu's right lay against her kindred and their dependents. The definitions then are totally wanting in clearness. What a Vasu taukei was, is left especially uncertain ; but the examples show, as far as they go, that the great or high Vasu's power was always exercised over a foreign country, and arose out of his mother having come from the ruling family in that country; and this seems at any rate to have been what is most important in Vasuing. 1 The case last cited appears to 1 Thokonauto, a brother of the chief of Eewa, who has been already mentioned, was Vasu of Mbau, and Wilkes describes his exercise of the right of Vasu against Tanoa, king of that island, when he was the most powerful chief of Fiji. With reference to what follows above it may be said that this Thokonauto sided with the Mbau people (to whom he was Vasu) against Rewa in the war in which the latter kingdom was for a time completely subjugated, and was afterwards set up by them as king of Eewa. While Eaivalita, above mentioned, was Vasu of Eewa, Mara, another of Tanoa's sons, was Vasu of Lakemba, and, according to Jackson (statement appended to Erskine's Western Pacific, p. 458), a third brother, older than Thakombau, was Vasu of Somosomo. All three are styled Vasu levu by Jackson. Who Thakombau's mother was is nowhere men- xi FIJIAN GROUP 227 show that there were duties as well as rights attaching to the Vasu. We read (vol. ii. p. 178) that Eewa was obstinate in the war with Mbau, "resting in hope of assistance from Eaivalita, their Vasu, who had engaged to kill his brother Thakombau on condition that Eewa should become tributary to him on his assuming the government of Mbau." This hope was defeated by Thakombau anticipating his brother and murdering him. And then it was felt that this murder of the Vasu of Eewa would make a reconciliation between the two kingdoms very difficult. The suggestion here is that the murder of Vasu would create a blood-feud between them. The view just deduced from Williams is confirmed by what Erskine says of Vasuing, though probably Erskine's state- ment is too narrow in its scope. He says the right of Vasu tioned in Fiji and the Fijians, but Jackson says (Lc. p. 458) that Thakombau used to boast that " his mother was the greatest woman, as his father was the greatest man in the greatest place of the Fijis, and belonged to Mbau." He is nowhere spoken of as being a Vasu. Wilkes (United States Exploring Expedition, vol. iii. p. 63), after stating that there are three kinds of Vasus, says that " Vasu Togai " is the highest title, and "is derived from the mother being queen of Ambau [Mbau].' 5 But he speaks of Thokonauto of Eewa, who was the son of Tanoa's sister, and therefore not of a queen of Mbau ; indeed he says she was not even queen of Kewa, but the king's second wife the first wife being a descendant of a family which had previously reigned at Mbau as being Vasu Togai of Mbau. This appears to have been a Mbau term, at any rate. Tanoa, as is shown above, was Vasu of Kewa, and after he succeeded his brother at Mbau, he is said to have got into trouble through helping the Rewans. Wilkes (Lc. pp. 63-65) says that the encouragement, and even assistance which he gave them, while the Mbau chiefs were at war with Rewa, caused much discontent among the chiefs, and that when, on a truce being made, he presented the Rewans with a large and much admired canoe, they conspired against and dethroned him. After his restoration, though Mbau and Rewa were frequently at war, "Tanoa takes no part in these contests," Wilkes says, " but when he thinks the belligerents have fought long enough, he sends the Rewa people word to 1 come and beg pardon,' after the Feejee custom, which they invariably do, even when they have been victorious." He was near his end when the incidents above spoken of occurred. 228 STUDIES IN ANCIENT HISTORY CHAP. belonged often to sons of female members of a reigning family, even when married to chiefs of dependent states. Such son is nephew or Vasu "to all the members of his mother's tribe/' and as such has certain rights against them (Lc. p. 215). Williams seems to have thought the " office " of Yasu an invention of some chiefs to extend their influence (perhaps meaning only that they had adapted the right of the nephew to state purposes), and remarks how often they were "griped" by their own contrivance (p. 35). But this sort of thing is not of the nature of an invention. It is an obvious relic of an earlier social phase. When Williams wrote there were signs on all hands of the decay of the ancient rule of succession which preferred brothers to sons; sons had often succeeded to their fathers, and probably in humble life always did so, the old rule lingering longest about the succession to chieftainries and kinships. 1 But there was a still older succession law than that which preferred brothers to sons, and failing brothers, preferred the eldest son of the eldest brother. That was the law which, failing brothers, preferred the sister's son, and we have in Vasuing an obvious relic of that most ancient of succession laws. The curious thing so regarding it, is that the right should lie against the uncle. That it should open against his son would be proper enough. On the failure of brothers the legitimate first heirs to a throne or chieftainry the eldest sister's son was the rightful heir. Over and over we are told by Williams that the kings and chiefs claimed their rank from the gods, and some of them even claimed to rank as gods. Eeligion then would concur with ancient law to point out th( sister's son as de jure the king or chief, and perhaps the god, of her country and clan, whoever might hold the office de facto; and in these circumstances one can imagine a larg< power over the people conceded to the Vasu without disput 1 "On Yanua Levu, on the announcement of a man's death, th< nearest relatives [of the deceased] rush to the house to appropriate all thej can seize belonging to those who lived there with the deceased " (p. 187] an assertion apparently of the right they would have had under tht earlier succession law. XI FIJIAN GROUP 229 So also in the case of the sister's son of the humblest person as well as of a chief; the area of power, and materials for using it upon, contracting as the status lowers. As Williams states the matter, however, the right lay against the mother's brother. He never speaks of it as lying against the brother's son after the death of his father ; so that the heir under the old law was allowed to appropriate in his uncle's lifetime what no longer came to him at his uncle's death. Erskine limits Vasuing to the case of royal families; and Mr. Williams's description is said to apply, or to apply in its fulness, to some localities only. If it existed generally, as described by Williams, in any locality, no more curious relic of ancient times than Vasuing can well be imagined. 1 The most limited view of its prevalence is enough to show us clearly that the totem clans of the Fijians were anciently continued through the mother ; while the succession law suffices to prove that they arrived at male kinship through Thibetan polyandry. IV. EXOGAMY Of the law of intermarriage among the Fijians, not one word is said by Williams, nor by any of our authorities ; and indeed the subject has been commonly passed over by writers about the Polynesians no doubt because until lately its im- portance was not known. The curious failure of Schoolcraft, in his voluminous work upon the tribes of North America, to notice the law which prohibits marriage within the clan, except in a single instance, that, too, mentioned casually in a footnote, warns us that that may happen where exogamy is everywhere prevalent. The fact, nevertheless, leaves us uninformed as to the actual marriage law of the Fijians. But from the structure of their local tribes and the occurrence of the form of capture among them, we may infer with something like confidence that they were anciently exogamous. We have seen that the local tribes or nations are composed 1 Compare with Vasuing the position of the Tammahas in the Tonga Islands, see p. 252. 230 STUDIES IN ANCIENT HISTOR Y CHAP of clans, and that the clans had their several totems, or toten gods ; we have seen also that the same totem or god wa worshipped in more districts than one. The hawk, we are told, was the " shrine " of a god in Lakemba and also in Yanua Levu, the shark in various districts ; while a god of Mbau (wh< must have been a totem god) is mentioned who was worshippec under another name at Somosomo, and under a third name at several other places. It appears casually, too, that the people of Na Lava, in Vanua Levu, and the people of Ono were worshippers of the same god ; and so also the people of Vanu- amba-lavu and the people of Oneata ; and that accordingly they could treat each other like relations, and that different peoples who had this connection between them were said to be taivou to each other which suggests that the connection of which we hear in a few cases only was no uncommon one. These facts give us an interfusion of totem clans in every district, and the appearance of the same totem clans in various districts. Consistently with this we find that nearly every town or village had one or more temples, and that some had many (p. 221) each clan of the town having its own, no doubt. We find too that in time of war " relatives within a garrison are often bribed to befriend the besiegers by burning the town or opening the gates" (p. 52) a further trace of interfusion, which suggests plainly that the defenders of a besieged place often included some clan which was found also among the besiegers. Of interfusion like this among any people, the joint operation of exogamy and the system of kinship through females only is the sole adequate explanation ; and we may infer from it in this case that when the Fijian totem clans were continued through women, they were exogamous also. Again, marriage commonly followed upon betrothal, marriages of very young girls to elderly men being very frequent (pp. 167, 168). According to Mr. Williams, wives were not got by purchase. And yet women were "treated as a sort of property in which a regular exchange is carried on" (p. 168). On the large island, however, he tells us, "is often found the custom prevalent among savage tribes of seizing upon the woman by apparent or actual force in order xi FIJIAN GROUP 231 to make her a wife" (p. 174) which shows that there was on the large islands some observance of the form of capture, as well as some practice of forcible abduction. A practice of capture by the people of a certain district (combined with cannibalism) is also mentioned in Erskine's work. 1 Men get wives by capture only from other clans than their own, and of the association between the idea of capture and marriage which appears in the form of capture, there has never been offered an adequate explanation except that which derives it from a general practice or system of capture in times when friendly contracts could not be made, and exogamy had become the marriage law. V. BLOOD-FEUD The reader may consult under this head Williams, pp. 109, 110, 126, 129, and 186. At pp. 31, 32 he wiU find an account of the institution of Soro (which for all the world answers to the Irish eric) a contrivance, no doubt, for mitigating the effects of revenge among a people by nature terribly revengeful and bloody. At p. 30 he will find distinct indication of the institution in operation in New Zealand under the name of Mum the " taking up of the sticks " of the Cherokees. " Some offences are punished by stripping the house of the culprit. In slight cases much humour is displayed by the spoilers : the sang-froid of the sufferer is an enigma to the Englishman." No doubt the sufferer felt, as the New Zealander did, that the spoilers were only doing what was proper. At p. 32 it is said that the Soro worked badly, many crimes being committed after a calculation balancing the fruit of the crime against the Soro. This reminds one of the Irish chief who, on being told that a sheriff was to be sent to his district, remarked, " Well, you must tell us his eric, that we may know what we shall have to pay when we kill him." 1 I.e. p. 425, Jackson's narrative. It is stated that the people of Male " commit great depredations on the mainland, taking the women as prisoners and killing the men for food." CHAPTER XII THE SAMOA OR NAVIGATORS' ISLANDS THIS group consists of eight principal islands, and is situated between the parallels of 13 and 15 degrees of south latitude, and 168 and 173 degrees west longitude. The inhabitants are of the same race with the New Zealanders, the Kanakas of the Sandwich Islands, the Tongans, and the inhabitants of the Hervey Islands. The largest of the islands are Savaii, Upolu, and Tutuila. These are about 150, 120, and 80 miles respectively in circumference; but Upolu, though second in point of size, is the first in point of importance, being more fertile and more populous than Savaii. The small island of Manono, which lies so close to Upolu as to be virtually a part of it, was one of the most important districts in the group. Scarcely anything was known of Samoa (as the whole group was called by these islanders) until after 1830, in which year the Eev. John Williams of the London Missionary Society paid his first visit to the islands. Mr. Williams has given in his well-known work l a most interesting account of the Samoans. The islands were surveyed in 1839 by the United States Exploring Expedition, and are described at some length by Captain Wilkes in the narrative of that expedition ; ten years later they were visited by Captain Erskine, and some account of them is given in his Journal published in 1853. The work of the Eev. George Turner, published in 186 1, 2 is for the most 1 A Narrative of Missionary Enterprises in the South Sea Islands. By John Williams. The references which follow are to the edition of 1838. 2 Nineteen Years in Polynesia. London, 1861. CHAP, xii THE SAMOA OR NAVIGATORS' ISLANDS 233 part devoted to Samoa, where from 1843 onwards the author was settled as a missionary of the London Missionary Society. Mr. Williams tells us (p. 542) that the religious system of the Samoans differed essentially from that which obtained at the other Polynesian Islands then known to us. They had "neither maraes, nor temples, nor altars, nor offerings, and consequently none of the barbarous and sanguinary rites, observed at the other groups " " no altars stained with human blood, no maraes strewed with the skulls and bones of numerous victims, no sacred groves devoted to rites of which brutality and sensuality were the most obvious features." l They were accordingly considered an impious race, and their impiety became proverbial with the people of Earotonga; for when upbraiding a person who neglected the worship of the gods they would call him "a godless Samoan." The objects worshipped by the Polynesians, he proceeds, were of three kinds : their deified ancestors, their idols, and their etus. Many of the deified ancestors, as he explains, were not real ancestors, but fabulous persons who were believed to have conferred benefits on mankind, one being worshipped because (p. 544) he, by a most absurd process, "created the sun, moon, and stars " ; and another, the elevator of the heavens, because with the help of a myriad of dragon-flies he severed the heavens from the earth and raised them up ; while " besides this class they had the god of the fishermen, of the husband- men, of the voyager, of the thief, and of the warrior," all of whom were " said to have been men who were deified on account of their eminence in such avocations." 2 The idols, 1 Turner, however, speaks of temples and priests in Samoa, but it appears from his account that the temples were nothing but the family house " where they were all assembled," and that the person officiating as high priest was the father of the family. The god often spoke through the mouth of the father, or some other member of the family, telling them what to do (p. 239). 2 "Men had but one pair of primitive ancestors," viz. Rangi and Papa, or heaven and earth ; and by these were begotten the heads (each called god and father) of the different families of living things. Thus Tangaroa (Samoan Tangaloa) was the god and father of fish and reptiles. 234 STUDIES IN ANCIENT HISTOR Y CHAP. again, were not so much gods as god-boxes for making the deities portable ; they were different in almost every island and district ; Mr. Williams had seen no two precisely similar representations of the same deity, except those placed on the fishing canoes (p. 546). The etu "was some species of bird, fish, or reptile in which the spirit of the god was supposed to reside" (p. 436). In so far as it was worshipped there was no need of any idol. And it was through their worship of the etu that the religion of the Samoans differed so materially from that of the other islanders known to Williams. This form of idolatry prevailed much more at the Samoas than at any other islands. Here innumerable objects were regarded as etus, and some of them were exceedingly mean. It was by no means uncommon to see an intelligent chief muttering some prayer to a fly, an ant, or a lizard which happened to alight or crawl in his presence (p. 547). Mr. Williams says that, "in addition to these objects of adoration, the Samoans, like the islanders generally, had a vague idea of a supreme being whom they regarded as the creator of all things, and the author of their mercies." They called him Tangaloa. Mr. Williams was informed that at their great feasts, prior to the distribution of the food, an orator arose and, after enumerating each article, exclaimed, " Thank you, great Tangaloa, for this." But he does not seem to have known whether Tangaloa was otherwise worshipped. That the etu was truly a totem god appears clearly from incidents narrated by him. In general, on the natives of any island renouncing their paganism, they gave proof of the sincerity of their conversion by surrendering their idols to be burnt, or otherwise destroyed, or to be conveyed to England. In Samoa, " as the people generally have no idols to destroy," some other proof had t( be given, and what was the proper substitute appeared as sooi as a single convert had been made. It consisted in the convert eating his etu that is, one of the species to which the etu belonged an act regarded as one of absolute dese- See Grey's Polynesian Mythology , pp. 1-15. "Tangaroa signifies fish every kind " (p. 1 2). xii THE SAMOA OR NAVIGATORS^ ISLANDS 235 cration. A chief, who had resolved to become a Christian, first returned to his district and held a consultation with his people; and then invited the teachers to be present at his renunciation of his former worship. " On their arrival they found a large concourse of people, and after the usual saluta- tions the chief inquired if they had brought with them a fish-spear. They asked why he wanted that, when he replied that his etu was an eel, and that he wished one to be caught that he might eat it, in order to convince all of his sincerity. An eel was therefore caught, and being cooked was eaten by many who had formerly regarded it as their etu" (p. 437). The eel, therefore, was the etu of many in the district besides the chief. This was repeated when, Malietoa, a distinguished chief and titular king of Savaii, having been converted, his son resolved to follow him, and " not only his relatives, but nearly all his people abandoned their heathen worship " ; and it was thenceforth the sign of conversion throughout Samoa the liberty of eating the etu, indeed, becoming (as Mr. Turner tells us) a strong motive for becoming Christian as soon as the people found that the desecration of the etu was followed by no bad consequences. " The etu of Malietoa's sons was a fish called Anae ; and on the day appointed a large number of friends and relatives were invited to partake of the feast. A number of Anae having been dressed and laid upon newly- plucked leaves, the party seated themselves around them, while one of the teachers implored a blessing. A portion of the etu was then placed before each individual, and with trembling hearts they proceeded to devour the sacred morsel. The superstitious fears of the young men were so powerfully excited lest the etu should gnaw their vitals and cause death, that they immediately retired from the feast and drank a large dose of cocoa-nut oil and salt water." The onlookers expected "that the daring innovators would have swollen, or fallen down dead suddenly" (p. 438). It appears that the etu so eaten was not the god of an individual, but of a family, and that it was the god of many outside relatives of the family as well; also that the consequences expected to follow upon those who had been worshippers eating it were those which 236 STUDIES IN ANCIENT HISTORY CHAP. would have been expected to follow upon the eating of the totem by one of the totem kindred. Mr. Williams goes on to tell how Papo, a god of war, represented by " a piece of rotten old matting," which was regarded with great veneration, was soon after sentenced to drowning, and thereafter handed over to the teachers. But he says nothing further about Papo. He was aware that, just before his arrival in Samoa, the chief called Tainafaingo was believed to have the spirit of a god residing in him, was actually worshipped as if he were a god, and had in con- sequence become all-powerful throughout the islands; and that, as he grossly abused his powers, the people of Aana killed him. But he does not dwell upon the peculiarity involved in the case of Tamafaingo. From Captain Wilkes we learn that the god who was believed to ^reside in this chief was a war-god of the same name, and that there had been two successive Tamafaingos (United States Exploring Expedition, vol. ii. p. 107). Mr. Williams mentions by the way that vampire bats abound in Samoa, while they are unknown at the islands to the eastward of Samoa, except Mangaia, where also they are numerous ; and that the Samoans venerated them as etus (and therefore did not eat them), while the Savage islanders not far off consider them a great delicacy (p. 499). Turtle, he says, were far more numerous at the Samoas than at Tahiti or the Hervey group ; but the Samoans used them, while " the turtle was considered most sacred by the Rarotongans and the Tahitians" (p. 502). That the etu was a totem also appears from the little Captain Wilkes has to tell us about it. A chief of Tutuila told him that his etu had been fresh-water eels, and that these he used constantly to feed in former days in the brook near his village; and that then if any one had touched, disturbed, or attempted to catch one, he would have killed him immediately. Having turned Christian he had himself taken to eating them, and all of them had been caught and destroyed (I.e. p. 77). They all, this chief said, had had their etus, and had felt themselves obliged to do everything the etu commanded (of which by and by). Captain Wilkes mentions xii THE SAMOA OR NAVIGATORS' ISLANDS 237 a lake in Savaii which was believed to be inhabited by spirits in the form of eels, which were worshipped, and a cave sacred to the god " Moso " (who, from what is stated, may have been a swallow). He found that the pigeon was considered sacred and was not eaten, and that in one district (apparently Aana) to kill a king pigeon was thought as great a crime as taking the life of a man (I.e. p. 122). The people of Aana justified their regard for the pigeon by affirming that when they were driven from their homes eight years before the pigeons also abandoned the district, returning afterwards when they had returned. Consistently with this, he tells us that there were many gods of the etu kind, who watched over particular districts. He mentions besides three war-gods (one of them Tamafaingo, another a goddess), and also, besides Tangaloa (who was less worshipped, he says, than the war-gods), a god of earthquakes, a god who supports the earth, and gods of lightning, rain, whirlwind, etc., who were supposed to live in an island to the westward, the quarter from which their bad weather came. Like Williams, he thinks that Tangaloa was a supreme god, though not much worshipped; but in the account given him of the creation he found that Tangaloa worked under the guidance of his daughter Tuli, a snipe, through whose instrumentality it was that worms became men and women (p. 182). Turner confirms the evidence of Williams, saying that the dog, some birds, and fishes were sacred to particular deities, and might not be eaten. Thus a man would not eat a fish which was supposed to be under the care and protection of his household god, but would eat without scruple fish sacred to the gods of other families (p. 196). Besides his family god, every Samoan individually was supposed to be taken at birth under the protection of some tutelary god. Every village had its god, which was not the same as the household god, but had some particular incarnation, just as the household god had, appearing as a bat or a worm or an owl (p. 238, etc.) The people were made Christians by Mr. Williams and others fifty years or so ago, but no one will be surprised to find most of the ancient superstitions and customs in force. 238 STUDIES IN ANCIENT HISTORY CHAP. A correspondent of Mr. A. Lang furnished in 1874 some curious information concerning the Samoans. He had diffi- culty in finding out the real nature of their ways and beliefs, partly through imperfect acquaintance with the language, and partly through the variety of their dialects. He says he was much as a very deaf and rather dumb stranger in England, collecting information, would be. What he learned would be subject to numerous qualifications which he would miss, and he would lose all the finer touches. Mr. Lang's correspondent had long searched in vain for any trace of totemism in the Samoans. He says, " I had given up the idea as almost impossible In an island where so few animals exist ; however, one day while talking with two natives, Charlie and Jericho, a half-caste boy brought me a lizard cut in two with a spade, to put in spirits of wine; to my surprise, ^Charlie ran hastily up, took the creature, and in a reproachful tone asked the boy why he had killed it. 'Do you not know that he is my father ? ' ! After this Mr. Lang's friend found many traces of totemism, including the belief that if they were to eat the totem they would surely break out into sores, and the fact that the commands of the totem are given to them in dream. " Their belief in dreams is extraordinary ; dream is as real to them as waking life." Marriage. Williams mentions it as a remarkable fact, distinguishing these islanders from all the others, that with them marriage is by purchase (I.e. pp. 91 and 140). "One young woman was introduced to me for whom her husband gave the amazing price of upwards of 200 pigs, besides a quantity of siapo, or native cloth." Turner (p. 185) tells us that the consent of the father or brother had to be asked, the woman's consent being secondary, but that an elopement might take place if there was any probability of a refusal on the part of the lady's relations (p. 188). Marriage by capture also occurs, but the notion of marriage is developed beyond the point where possession by capture makes marriage. Williams (p. 114) cites the case of the daughter of a chief being captured by another chief in war, and declining to become his wife ; where- upon he slew her to prevent her becoming the wife of any one xii THE SAMOA OR NAVIGATORS* ISLANDS 239 else in his tribe. The lady was the favourite daughter of Malietoa. Her captor, Tangaloa, "Williams says, wished her, the captive, to become his wife, but to this she would not consent ; and it was also opposed by his own people, who said it was a base thing in him to take by force the daughter of so great a chieftain. This, of course, led to a terrible revenge ; the clansmen of the murderer even protesting against the act of their chief, that capture was not the legitimate mode of obtaining a wife of such distinction which intimates an approval of capture as a mode of wiving for the poorer sort. In dividing the spoil of a conquered people, the women were not killed, but taken as wives (Turner, p. 320). Both forms of marriage, that by purchase and that by capture, point to the parties belonging to different kindreds, if not clans. Mr. Williams (p. 91) says that when a sufficient price had been paid to the relatives of the lady, she seldom refused to marry the man; and (p. 92) that the law was, that if the new wife could run away home, her lord would have to repurchase her. Stronger evidence is forthcoming from Turner (p. 185), who says that care was taken to prevent the marriage of near relatives. As to kinship, it seems latterly to have been well settled that there was succession of sons to fathers at least in chieftainries. Anciently, however, the succession must have been universally according to the system of kinship through females only ; for we learn from Dr. Litton Forbes l that the inheritance to land is according to the system of kinship through females only. Polygamy occurs, modified by a rule, which is said to work well, of giving each wife supremacy for three days in turn. The brother of a deceased husband took his widow (Turner, p. 189). Village Communities. Turner says that in Samoa there are village communities of from 200 to 500 people, which consider themselves perfectly distinct from each other, and act as they please on their own ground and in their own affairs. Eight or ten of these unite by consent and form a district with some particular village for capital; of old, the 1 Paper on the Navigator Islands. Proceedings of the Royal Geo- graphical Society i 6th March 1877. 240 STUDIES IN ANCIENT HISTOR Y CHAP, xn head of that village had the title of king, and a parliament "of the heads of families" was held there (Turner, p. 287). In case of war these districts combine in twos or threes (Turner, p. 290). Each village has its chief (Wilkes, p. 274) ; and from ten to twenty titled heads of families. Chiefs trace pedigree to the ancient head of some clan, and there may be twenty persons besides the chief tracing their origin to the same stock, and called chiefs also, any one of whom may succeed to the title (Turner, p. 280). The land belongs to families, and is managed by the head of the family. Members of a family can depose their head, and the heads of families can take the title from the chief, and give it to his brother or uncle, or some other member of the chief family (Turner, p. 284). A proof of the strict unity of the clan is that it is a disgrace to a man to have it said that he paid his carpenter shabbily, and it is not only a disgrace to him, but to the whole family or clan with which he is connected. The entire tribe or clan is his bank ; being connected with it by birth or marriage, he has an interest in its property (Turner, p. 262). Blood-Feud. For proofs of the irresistible force of the law of blood-feud in these islands, see Williams, I.e. pp. 34, 64, 87, 138. He found that not even Christianity could dimmish the force of the obligation of the Ono, or systematic revenge, which prevailed through the whole of the Pacific Islands, a legacy descending from father to son for generations. Wilkes (I.e. vol. ii. p. 150) bears witness to the same fact, the right of retaliation vested in the friends and relatives of the slain, and extended against the relatives and friends of the guilty. Wilkes further mentions that the eric had been introduced to qualify vengeance, and that among the Samoans the tombs of chiefs were regarded as places of refuge. CHAPTER XIII THE TONGAN OR FRIENDLY ISLANDS THESE islands consist of the island of Tonga or Tonga- taboo, which gives its name to the group, the cluster called the Hapai, and the island of Vavaoo. We have scattered notices of the inhabitants in various works, but the principal authority respecting them, before they came under the influence of the missionaries, is a Mr. William Mariner, who was one of the crew of the Port- au-Prince that was captured by the natives, the bulk of the crew having been treacherously murdered. Mr. Mariner spent four years in the island as the adopted son of the king Finow, and his account of the natives we have as compiled and arranged for him by Dr. John Martin. 1 Mr. Martin's book, which I shall hereafter cite as Mariner, is most orderly and interesting, and, as far as it goes, apparently trustworthy. It leaves us, however, without any information on many essential points. Society in the Tongans is represented as divided into 1 An Account of the Natives of the Tonga Islands, etc. London, 1817. Two vols. R 242 STUDIES IN ANCIENT HISTORY CHAP. several distinct classes, which may be called respectively the families of Tooitonga and Veachi; the families of the How, or king, and the Egi, or nobles ; Matabooles, Mooas, and Tooas. The name Tooitonga means chief of Tonga, and the name of the family in which Tooitonga was hereditary was Fatafehi, a name of unknown meaning. The family claimed descent from a great god who formerly visited the islands of Tonga. So does the family of the Veachi, but whether their original mothers were goddesses or merely natives of Tonga, is a question which they do not pretend to decide. It is one, however, which we shall have to consider (Mariner, vol. ii. p. 82). The rank of Tooitonga, and also of the Veachi, seems to have been wholly of a religious nature. It was superior to that of the king himself, as was the rank of several families related to Tooitonga and Veachi. Many ceremonies show the veneration for Tooitonga. Once a year offerings were made to him of the first fruits of the year. There were special ceremonies con- nected with his marriage and his burial ; also with mourning for him when dead. He was the only Tongan not circumcised or tattooed, and a peculiar language was used in addressing him and in speaking of him. He was the greatest of the Egi, but his power was limited chiefly to his own family and attendanl The next to him, as already explained, was Veachi. whose name has no meaning that Mr. Mariner was abl< to discover. After Veachi ranked the priests, but at times when they were possessed by some god wer< xin THE TONGAN OR FRIENDLY ISLANDS 243 in short, god-boxes. At other times they took only the rank to which their standing as chiefs (usually they were of the lower order of chiefs), or as Matabooles, en- titled them. The king himself, however, has been known to become a god-box, the god being the chief of the gods. When a god was in a priest, the priest ranked according to the rank of the god. The How, or kingly family, came next. Mr. Mariner seems to think that the office of king was hereditary, and descended from father to son but he has produced only one example of this, and in that there might have been a contest for the succession, but for good management on the part of the dead king's son, and the adherence to his cause of his father's brother. Finow, his father, got the throne by murder and usurpation. The Hows before him were relations of Tooitonga, and it was not recorded what, during the four or five generations during which they reigned, the law of succession was. It may be doubted, however, w r hether unless the king married his own sister, or other female relation representing the principal line of female descent in the royal house, his son could succeed to him. " In every family," says Manner, speaking of the whole noble class without exception, "nobility descends by the female line, for where the mother is not a noble, the children are not nobles ; but supposing the father and mother to be nearly equal by birth, the following is the order in which the individuals of the family are to be ranked, viz. the father, the mother, the eldest son, the eldest daughter, the second son, the 244 STUDIES IN ANCIENT HISTORY CHAP. second daughter, etc. ; or if there be no children, the next brother to the man, then the sister, the second brother, the second sister, etc. ; but if the woman is more noble than the man, then her relations in like order take precedence in rank. But they do not inherit his property, as will be seen in another place. All the children of a female noble are, without exception, nobles" (vol. ii. pp. 89, 90). Elsewhere he states that " the right of succession to property in the islands is regulated by the order of relationship as given under the head of nobles (p. 89), and so in like manner is the right of succession to the throne." pise where he says that " children acquire their rank by inheritance from the mother's side ; if she be not a noble they are not, and vice versa. If a man, however high his rank, have .a child by a woman who is only a Tooa, no matter whether they are married or not, that child would not be a noble, though it were known that his father was a noble. . . . On the contrary, if a woman who is a noble were to have a child by a Tooa, the child would be noble" (p. 101). Hence we see that in no case could the father trans- mit his standing or quality. The god-fatherhood, therefore, of the families of Tooitonga and Veachi, stated apart from any knowledge or belief as to the original mothers of the families, appears in contradiction to the established law of succession. The pedigree must either have been invented at a time when descem was through fathers a hypothesis which may be pul aside as untenable or subsequent to a strong movemenl xni THE TONGAN OR FRIENDLY ISLANDS 245 having set in to alter the succession law, and enable fathers to transmit their dignities and titles to their sons. 1 Next to the Egi were the Matabooles, a sort of honourable attendants upon chiefs, their companions, counsellors, and ministers. They managed all cere- monies. The rank was inherited, but no man could assume the title of Mataboole till his father died. The heir to a Matabooleship was then the eldest son or, failing a son, the next brother. All the sons and brothers of Matabooles were Mooas. As the sons and brothers of Matabooles were only Mooas, the sons and brothers of Mooas were only Tooas. There was a further division of Tongan society into classes, according to trades or professions. Some trades Mariner represents as hereditary, others not. Mr. Mariner neither uses the term tribe nor clan, nor does he use the word family in any very extended sense, yet I think we can spell out the fact that the people were divided into clans, and that each clan had its own god or totem. For example, after the men of Chichia vanquished the people of Pau, the chief of Chichia resolved to give a great feast before the de- parture of one of his principal allies in the war, Cow Mooala. The day opened with a grand war-like dance, followed by the drinking of cava ; after which the cooks brought forward the feast. ( ' Immediately they advanced, two and two, each couple bearing on their shoulders a 1 I have been unable to find the " other place " referred to at p. 90 showing who would inherit a man's property in the case of the woman being more noble than himself. 246 STUDIES IN ANCIENT HISTORY CHAP. basket in which was the body of a man barbecued like a hog. The bodies were placed before the chief, who was seated at the head of his company on a large green. When all these victims were placed on the ground, hogs were brought in in like manner, after that baskets of yams, on each of which was a baked fowl. These being deposited in like manner, the number of dishes was counted and announced aloud to the chief, when there appeared to be 200 human bodies, 200 hogs, 200 baskets of yams, and a like number of fowls. The provisions were then divided into various portions, and each de- clared to be the portion of such a god, after which they were given to the care of as many principal chiefs, who shared them out to all their dependants, so that every man and woman in the island had a share of each of these articles (vol. i. p. 345). I think there can be no missing the meaning of this. Each chief and the members of his clan have for their share the portion of their god as the same had been publicly assigned to him. When we come to Mr. Mariner's account of the religion of the Tongans, and take it in connection with scattered indications of the real meaning of the system elsewhere found, we find that fancy and metaphysic had been at work in the Tongans to explain the origins of things and of good and evil in the world, and had pro- duced a vague jumble of gods, with names but almost devoid of attributes ; had pictured a heaven in the imagin- ary island of Bolotoo abounding with all good things, and creatures all immortal, in which the Tongan Egi [ xni THE TONGAN OR FRIENDLY ISLANDS 247 with their Matabooles took high rank after death, the lower class of people being shut out. The system of god-boxes we have seen, and how a man might for a time become important by feigning to be possessed by one of the primitive gods speaking through him, or even the spirit of a dead noble. But these were not the only god-boxes. Mr. Mariner says (vol. ii. p. 106) that the primitive gods also came into the living bodies of lizards, porpoises, and a species of water-snake, whence these animals were much respected; and (p. Ill) that these primitive or original gods were supposed to be rather numerous, " but the names of very few of them were known to Mr. Mariner." He says that only those whose attributes particularly concerned this world should be much talked of, and for the most part " the others are merely tutelar gods to particular private families." We may believe that the gods of the families were the gods of the clans also to which the families belonged, and that whatever confusion had arisen as to these clan gods in the island of Vavaoo, they had at least retained their pre-eminence in the island of Chichia. Among the indications of totemism may be taken the following : All the turtles in the world had their origin in a turtle that sprung from the head of the goddess-daughter of the sky-god Langi, which the god had in a rage severed from her body and thrown into the sea. " This story obtains almost universal credit at the Tongan Islands, and in consequence turtles are considered as almost a prohibited food, and there are many that will not eat turtle on any account " (vol. ii. 248 STUDIES IN ANCIENT HISTORY CHAP pp. 130-133). Elsewhere (vol. ii. p. 233) lie says, " Certain kinds of food, as turtle and a certain species of fish, from something in their nature, are said to be taboo, and must not be eaten until a small portion be first given to the gods." Again, as showing the superior antiquity of the animal gods, we have the following (vol. ii. p. 139): "It has already been stated that the gods are believed sometimes to enter into the bodies of lizards, porpoises, and water- snakes, but this power belongs only to the original gods, not to the souls of chiefs." The spirit of an Egi may make a god-box of any priest in the Tongas, but it is beyond him to enter appearance in one of the ancient totems of the country. It will be remembered that the Tongan tribe which we found settled in the Hervey group had for their god the lizard. It is a fact not a little remarkable that we should owe to the myths and songs of the Hervey group, which show the lizard-god with his temples and priests, the fact that in some of the Tongan islands, before metaphysic had elaborated its theogony of shadowy semi-human divinities, animal-gods must have been the principal deities of the country. 1 We have a confirmation of Mariner's statement as to the water- snake from the Eev. J. Williams in a story which shows that the snake was not a god-box, but the god himself. " While walking on one occasion," he 1 At the island of Tofooa, which was sacred to Tooitonga, shark worship seems to have prevailed (Mariner, vol. i. p. 251), and a suggestion that the shark was religiously regarded may be had at vol. i. p. 252. In the United States Exploring Expedition (vol. iii. p. 1 1 ) there is a mention of pigeons as sacred at Tonga. xiii THE TONGAN OR FRIENDL Y ISLANDS 249 says, " across a small uninhabited island in the vicinity of Tonga-taboo, I happened to tread upon a nest of sea- snakes. At first I was startled at the circumstance, but being assured that they were perfectly harmless, I desired a native to kill the largest of them as a specimen. We then sailed to another island where a number of heathen fishermen were preparing their nets. Taking my seat upon a stone under a Tou tree, I desired my people to bring the reptile and dry it on the rocks ; but as soon as the fishermen saw it they raised a most terrific yell and, seizing their clubs, rushed upon the Christian natives, shouting, ' You have killed our god ! You have killed our god ! ' With some difficulty I stayed their violence. . . . This incident shows not only that they worship these things, but that they regard them with the most superstitious veneration" (Williams, I.e. p. 547). In this case, again, we may see that the god belonged to the people of a small island constituting most probably a tribe of descent ; and, on the whole, though the evidence is purely circumstantial, we may infer that the Tongans were divided into clans on the totemic principle ; while from the force still preserved by the system of female kinship, we may infer that anciently they had kinship through women only. We have a note of this in a fact furnished by the Eev. J. Williams. He says that at the island of Lefuga in the Hapai group, the principal chief, who indeed exercised authority all over the group, was Taufaahau. This chief, having embraced Christianity, destroyed the gods and the Maraes first at his own 250 STUDIES IN ANCIENT HISTORY CHAP. island, and next as far as he could throughout the group. The fact which Mr. Williams brings out with emphasis is, that the deposed gods were not gods at all, but goddesses, of whom he observed, in one place, five hanging by the neck, one of which he obtained and sent to England (Williams, I.e. p. 84). I can find no evidence whatever bearing on the question of the jus connubii among the Tongans, nor any grounds for an inference, however faint, as to what it was. There appears to have been erected a religious obstacle to communication of any sort between brother and sister, 1 and from that we may infer that the marriage of brother and sister was impossible. There are none of the usual hmts even as to whether they permitted or forbade marriages between near relations. It is possible that the knowledge may exist, but the search I have made has not enabled me to find any record of it. We may believe that, where notes of totemic clanship are present, the law of exogamy, which, in so many quarters, among races of men so diverse, we have always found connected with them, cannot have been wanting. One circumstance that makes in favour of the position that they were not given to practices which we should call incestuous, is to be found in the fact that in the elaborate enumeration of the crimes which were with them habitual and not regarded as crimes, there is not a single hint of incest. Cannibalism, murder, theft, revenge, systematic rapes, were not held to be crimes. Everything that could be charged against a people has been set down 1 Mariner p , vol. ii. p. 155. xiii THE TONGAN OR FRIENDLY ISLANDS 251 without extenuation against them by Mr. Martin in his account of Mr. Mariner's experiences, but not one word is said to suggest that they ever contravened what is usually regarded as one of the most sacred laws of nature and morality, the law of incest. (See Mariner, vol. ii. chap, xix.) Marriage appears to have been usually monogamous, but polygamy prevailed to some extent. The women appear to have shown great conjugal fidelity. On the other hand marriage, which might be constituted without any ceremony, seems to have been dissolvable at the pleasure of the man. A woman often had in succession many husbands. When unmarried, or during periods of divorcement, custom set her perfectly free to have her lover or lovers. Most women seem to have been betrothed by their parents to some chief Mataboole or Mooa, that is, some elderly man. Mr. Mariner estimated that about one-third of the married women had been so betrothed. The marriages mentioned in his work, so far as the details are given, were between persons of different clans ; but that goes for nothing, as only the marriages of chiefs are mentioned. "It was thought shameful for a woman frequently to change her lover " (vol. ii. p. 174). The women are tender, kind mothers, he says (p. 179), and take care of the children, for "even in a case of divorce the children of any age requiring parental care go with the mother." Domestic quarrels are seldom known; no woman entertains the idea of rebelling against her husband's authority, and if she should, even her own relations would not take her part, unless the 252 STUDIES IN ANCIENT HISTORY CHAP. conduct of her husband was undoubtedly cruel. (See vol. ii. chap, xix.) The Maui myth is found in this group, but the goc who fished up the dry land was not Maui among the Tongans, but Tangaloa (vol. ii. p. 116). From Cook we may gather that Vasuing, or some- thing very like it, was known in the Tonga Islands We have seen that the king was not of the highest rank and Cook observed with astonishment that Latoohbooloo and three ladies were of higher. They are called Tammaha, i.e. chiefs. " We were told that the late king, Poulaho's father, has a sister of equal rank and older than himself; that she [by a Fijian] had a son and two daughters ; and that these persons, as well as their mother, rank above the king. We endeavoured in vain to trace the reason for this singular pre-eminence of the Tammahas ; for we could learn nothing besides this of their pedigree." The king could not eat in their presence. Latoohbooloo had the privilege of taking anything from the people even if it belonged to the king. (Cook's Voyages to the Pacific Ocean. Voyages in years 1776-1780. London, 1784. Vol. i. p. 413.) As to blood-feud, Mariner (vol. i. p. 69) says, "It was the Tonga custom not only to kill an enemy, but also all his friends and relations if possible." Vol i. pp. 140-147 : " It will be recollected that Toobo Neuha was the chief that assassinated Toogoo Ahoo ; ever since which period Toobo Toa's desire of revenge was most implacable, and he had made a vow never to drink the milk of the cocoa-nut out of the shell till he had accom- xiii THE TONGAN OR FRIENDLY ISLANDS 253 plished it. He did actually slay him, taking an oppor- tunity of lying in wait for him when he was unarmed, and knocking him on the head. At this moment a young warrior, whose name was Latoo Ila, and whose father had formerly been killed under strong suspicions of con- spiracy by Toobo Neuha, came up to the spot, possessed by a spirit of implacable revenge. He struck the body of the dead chief several times, and exclaimed, ' The time of vengeance is come ! . . . thou murderer of my father ! I would have declared my sentiments long ago . . . but the vengeance of my chief, Toobo Toa, was first to be satisfied. ... It was a duty I owed to the spirit of my father to preserve my life as long as possible, that I might have the satisfaction to see thee thus lie stinking (dead) ! ' A spirit of vengeance singled out in the subsequent war those who had been engaged in kill- ing Toobo Neuha." " It is not at all extraordinary that most of those who had assisted in the assassination of Toobo Neuha should fall victims in this battle to the vengeance of the enemy" (p. 194). In the account of the war between Hapai and Vavaoo under the leadership of Finow on the one side and his aunt on the other, we find that Finow was advised by the gods that he should first proceed to Vavaoo with three canoes only and with such men as had few or no relations at Vavaoo (Mariner, vol. i. p. 173). Afterwards, before the siege commenced of the queen's fortress at Vavaoo, an armistice was made that " each party might take leave of what friends and relations they might have among their opponents" (p. 188). On the armistice 254 STUDIES IN ANCIENT HISTORY CHAP. XIII being granted, many of the garrison came out to take farewell of their relatives. " Many tears were shed on both sides, and many a last embrace exchanged. The affecting scene had lasted about two hours when it was casually terminated ; how long it might have lasted but for the incident which made an instant challenge to fight there is no knowing." What this discloses is a singular interfusion of relations on the two sides. Was it due to exogamy and cross marriages, or is Mariner's explana- tion sufficient, that an old custom obliged every man in honour to fight for the chief on whose island he happened to be when war was declared ? CHAPTER XIV NEW ZEALAND THERE are several valuable works on New Zealand. The Eev. Eichard Taylor's Te-Ika-A-Maui, or, New Zealand and its Inhabitants, was published in 1855, long prior to the speculations which have given the facts he dis- closes their principal importance. His testimony there- fore may be accepted as beyond suspicion. Sir George Grey's Polynesian Mythology was published in the same year, and is accordingly subject to the same observation. Unfortunately, Sir George has not exhibited the same care in the use of terms in this work that he had shown in his earlier work on North- Western Australia. More- over, his versions of the myths are confessedly made up from portions got in different places from different per- sons, so that the versions are patchworks, and do not necessarily correspond with any forms of the myths that ever had acceptance among New Zealand tribes. Not- withstanding this, it is possible, I think, to make some use of this work as furnishing evidence of a supplementary sort. Lastly, there appeared in 1876 Old New Zealand, by a Pakeha Maori, with an introduction by Lord Pern- 256 STUDIES IN ANCIENT HISTORY CHAP. broke. No one can read this most amusing and instruc- tive book and suspect the Pakeha of being hampered by any speculations other than such as he fully discloses, and he certainly has none as to the origin or history of human society. No evidence could be more trustworthy, and for myself, I rely on it the more that it is mainly indirect and incidental, and of a sort that no ingenuity could fabricate, even could we imagine a sufficient motive in the case for the fabrication of false evidence. There are other works on the subject which I have examined, but with distrust, as influenced by specula- tions, and shall accordingly take no account of. What I propose here to do is (1) to give an idea of the structure of New Zealand native society as described by the Pakeha in Old New Zealand, and to do this by an analysis of the passages in his work in which the lead- ing terms relating to social structure are employed, such as tribe, clan, family, etc. ; (2) to corroborate, by the evi- dence furnished by Taylor and Grey, the inferences thus reached as to the state of society in New Zealand, partly by way of notes and partly in a separate exposition ; (3) to sum up all the inferences derivable from the evidence. The New Zealanders being obviously of one race with the bulk of the inhabitants of the Polynesian islands, any facts well made out as to them have a first-rate value outside the New Zealand group, as helping us to inter- pret aright the facts respecting kindred populations when they seem to be obscure. The Pakeha is often loose in the use of words. With xiv NEW ZEALAND 257 him " tribe " and " family " are sometimes interchange- able terms (e.g. pp. 79, 124, 200); sometimes "tribe" and "clan" are interchangeable; compare pp. 194 and 211, where the Ngapuhi are first spoken of as a hapu, or clan, and again as an iwi, or local tribe. But on the whole, and taking all his statements together, the mean- ings of his terms work themselves out clear as follows : 1. Iwi = local tribe or nation (e.g. p. 217). The natives are represented as believing that the English sailors " were quite a different hapu, though belonging to the iwi of England, and in no way ' related ' to the soldiers." The inference that being of a different hapu they were therefore no way related must not be drawn, although we may conclude that being of the same hapu implied relationship, e.g. in Heke's war, when the rockets were fired at Heke's pa, the natives taking part with England held their breath, and were in great fear. Why ? Because Heke and his men were, " although against us, all Ngapuhi, the same iwi as ourselves, and many of them our near relations." This, no doubt, refers to the cross connections by marriage between the several hapus, or clans, within the iwi. Again (pp. 229, 230), on a victory over Heke's men and capture of killed and wounded, " some of our young men, being hot with the fight, cried out to eat them raw at once ; but this was a foolish proposal, for although we were fighting Heke, we were all Ngapuhi together, and more or less related," and to eat a relation was a deadly sin. 1 Again (p. 221), 1 In the legend of Maui, as given by Grey, it appears that when Maui's S 258 STUDIES IN ANCIENT HISTORY CHAP. one native calling another slave, because his great grand- father had once been made a prisoner of war, the other softened the accusation by saying that even if his an- cestor had been made prisoner, it was by a section (i.e. hapu, or clan) of his own tribe (iwi), and consequently by his own relations he had been defeated." These references show that a sort of relationship, fainter than that of clanship, was supposed to bind together all of the same iwi, or local tribe. As already suggested, it was due to the cross-connections of all the clans by marriage. 2. Hapu clan, Iwi = & number of clans. When old Lizard-skin came to die, there was a concourse of people from far and wide to his obsequies. " Though this old rangatira was not the head of his tribe, 1 he had been for about half a century the recognised war- chief of almost all the sections, or hapu, of a very numerous and warlike iwi, or tribe, who had now assembled from all their distant villages and pas to see him die." The iwi, then, was in sections, and these were hapu; but hapu is clan (see pp. 154, 162, et ubique). Now were the hapu interfused as among the Iro- ogre ancestress was on the point of devouring him, she suddenly per ceived that he was a relation, " and her stomach, which was quite large and distended, immediately began to shrink and contract itself again (Grey, I.e. p. 34). The catastrophe, again, in The Dissensions a Hawaiki, turns on the eating of relations. "Then, indeed, a grea crime was committed by Hou and his family and his warriors in eating the bodies of these men, for they were near relations, being descendec from Tamatea-kai-ariki." Cowardice and fear fell on the tribe of Hou and they became fit for nothing in consequence of this sin (ibid. p. 131) 1 It is not clear what is here meant by tribe. xiv NEW* ZEALAND 259 quois, or separate and localised apart as among some Algonquins ? The first impression is that they were separate. " Native tradition affirms that each of these hills was the stronghold of a separate hapu, or clan, bearing its distinctive name," i.e. the fortress bore the clan's name. 1 The clan on this view was a village com- munity ; but we have yet to see its interior structure. Was it homogeneous ? I incline to think that the word hapu is used by the Pakeha in a double sense, to mean sometimes a group of kindred, forming a clan, con- sidered apart from the places of residence of its mem- bers, sometimes to mean a village community. In both senses hapu would be a section of an iwi, but the senses are quite different. (1) Take the case of Lizard-skin, p. 145 : "This old gentleman was not head of his tribe. He was a man of good family related to several high chiefs. He was the head of a strong family, or hapu, which mustered a considerable number of fighting men, all his near rela- tions." Here hapu = family = clan in the sense of a group of kindred claiming to have a common descent. And it included all who were of the same descent. But- (2) All relations, nay even all near relations, did not live in the same village (p. 133). Further, in the same village there would appear to be distinct families (Ngnati) (p. 134), for the one had concealments from 1 Compare Grey, I.e. pp. 115, 122, 221, and 73. Every New Zealand group is with Grey a tribe. There can be little doubt that the groups in the passages referred to were either clans or families. 260 STUDIES IN ANCIENT HISTORY CHAP. the other inconsistent with the idea of consanguinity. Again, in the account of " the calling up " of the spirit of a chief who had been killed in battle, we find (p. 122) " the priest (Tohunga) was to come to the village of the relations, and the interview was to take place in a large house common to all the population," implying the presence in the village of others besides the relations. Of course often the members of a hapu or the greater part of them might inhabit a village by them- selves, or a pa, or even a single large house. Such a case appears, indeed (p. 161), "a hapu, in number just forty persons." (3) But even when the whole of the hapu resided in one village, and no other whole hapu inhabited it, yet all in the village were not of one and the same hapu, as may be seen in the working of the institution of Muru, which shows that in some districts at least a man's children were held to belong not to his hapu, but to their mother's. " A man's child fell in the fire, and was almost burnt to death. The father was im- mediately plundered to an extent that almost left him without the means of existence : fishing nets, canoes, pigs, provisions, all went." How and why? Here is the Pakeha's explanation (pp. 85 if.): "The tract of country inhabited by a single tribe (iwi) might be, say, from forty to a hundred miles square, and the different villages of the different sections [hapu] of the tribe would be scattered over this area at different distances from each other. We will, by way of illustrating the XIV NE W ZEALAND 261 working of the muru system, take the case of the burnt child. Soon after the accident it would be heard of in the neighbouring villages; the family of the mother are probably the inhabitants of one of them ; they have, according to the law of muru, the first and greatest right to clean out the afflicted father a child being considered to belong to the family of the mother more than to that of the father in fact it is their child, whom the father has the rearing of. The child was, moreover, a promising lump of a boy, the makings of a future warrior, and consequently very valuable to the whole tribe (iwi) in general, but to the mother's family [? hapu] in particular. 1 ' A pretty thing to let him get spoiled.' Then he is a boy of good family, a rangatira by birth, and it would never do to let such a thing pass without making a noise about it. That would be an insult to the dignity of the families [lhapu] of both father and mother. Decidedly, besides being robbed, the father must be assaulted with the spear. True, he is a famous spearsman, and for his own credit must 1 hurt ' some one or another if attacked. But this is of no consequence, a flesh wound more or less deep is to be counted on ; but then think of the plunder ! It is against the law of muru that any one should be killed, and first blood ends the duel." We cannot cite in full the delicious narrative, and must abridge it. The taua muru war-party seeking compensation is probably 1 This occurs at p. 85. It is at p. 154 that the Pakeha uses the term hapu = clan for the first time obviously under a sense, developed as he went on, of the need of greater precision in the use of terms. 262 STUDIES IN ANCIENT HISTORY CHAP. headed by ike, brother of the mother of the child. The father hears of the taua coming. When it appears, " then the whole fighting men of the section [hapu\ of the tribe [iwi], of which he is an important member, collect at his back, all armed with spear and club." The taua comes on and dances the war- dance; the father's party dance the war-dance also. Then follows a fight with spears, apparently desperate, between the father and mother's brother, ending in both being slightly wounded. Then on the cry of murua, the father's place is completely sacked, the sack ending in a feast prepared ungrudgingly, since all the provisions were to disappear in any case for the robber relations. Strictly speaking, then, no whole hapu could live together. Members of it were in as many families in other villages as it furnished to these villages daughters for wives. In short, the clans were in some districts of New Zealand at least interfused as they were among the Iroquois, and that, too, owing to the same cause, viz. women marrying out of their own clans, and children being held to belong to the clans of their mothers. (4) Lastly, under the head of hapu, I notice that a strong force of one and the same hapu might exist in different villages within the territory of the iwi. This appears from the Pakeha's statements as to the voluntary slaughter of one of the same iwi as the murderer. He says (p. 91) the murderer usually went unpunished, by fleeing from the village or scene of the offence to "some other section of the tribe where he XIV NEW ZEALAND 263 had relations, who, as he fled to them for protection, were bound to give it, and always ready to do so ; or otherwise, he would stand his ground and defy all comers by means of the strength of his own family or section \hapu], who would defend him and protect him, as a mere matter of course." 3. Family. We have seen the word used as = tribe ; also as = clan. In a more restricted sense, it seems to mean the members of a man's household, including con- nections by marriage living with him, and his slaves. (1) P. 195. "So Heke went with his own family and people, and those of his elder relation Kawiti and the Kapatoe and some others, altogether about 400 men." (2) P. 200. " Then many men came to join Heke, but no whole hapu came, for most of the Ngapuhi chiefs said," etc. The whole hapu could go only under the chief, but heads of families were free to go without the chief with their own people. (3) P. 200. " Thomas Walker called together his family and all his friends ( = relations), and said he would fight against Heke." Ibid. " Te Tao Nui . . . brought with him all his family and relations, many fighting men ; only one man of the family [probably a marriage connection] did not come that man went to help Heke." Ibid. " The tribe [here again the meaning of tribe is not clear] of Ngnati Pou came to help Walker." . . . They brought forty men. 264 STUDIES IN ANCIENT HISTOR Y CHAP. (4) P. 201. " I, your friend, went also with my two younger brothers, my four sons and my daughter's husband, 1 and nine cousins and three slaves twenty men of us. ... I went [like all the others] because when the ancestors of Heke fought against mine, the ancestors of Walker came to help my forefathers, because they were related to each other ; so I and Walker are relations ; but I don't know exactly what the relation- ship is, for eleven generations have passed since that ancient war, but Walker and I are aware that we are related, and always come to each other's help in war." On which the Pakeha observes that a war-party is always composed of relations, and that "to be a man of many cousins " is to be a great chief. 4. Marriage. On this subject the Pakeha fails us entirely, if we except what he says as to wives hanging themselves, almost as a matter of course, when their husbands died, and husbands " very commonly " com- mitting suicide when their wives died (p. 135), which shows that marriage was a serious matter, whatever else it was. 1 That the son-in-law a Beena husband who went to live with his wife in her father's house, was bound to fight for his father-in-law, is stated by Taylor, I.e. p. 164, who says that a Beena husband in New Zealand was " looked upon as one of the tribe or hapu to which his wife belonged, and in case of war . . . was often obliged to fight against his own relations." He states, also, that this sort of marriage was common. The children would of course be of the mother's Jiapu, and by fiction of the father's also. We may be sure we have here a step in the transition to male kinship and descent through fathers. In the legend of Tawhaki (Grey, I.e. p. 59) it appears that that god went on his marriage to live with his wife's family. xiv NEW ZEALAND 265 I need not point, however, to the significance of the case of the burnt child, employed by the Pakeha as typical, to illustrate the working of the system. It is for all the world " the taking up of the sticks," as we find it among the Cherokees, Creeks, and other of the southern nations of America, and full of suggestions of totemism and exogamy, as it is demonstrative of female kinship. I find the following citation from Thomson's New Zealand, in Spencer's Descriptive Sociology : " Marriage among the New Zealanders was both endogamous and exogamous ; in the latter case the permission of both nations was required, and the neglect of this often led to war." I have not seen Mr. Thomson's book, and do not mean to consult it. Supposing the citation correctly made, it shows that he wrote since speculations on such subjects ^began, and has failed to comprehend them. An exogamous marriage, as the reader knows, is a marriage between persons of different clans or kinships, not entered into fortuitously, but because of law declar- ing it to be incest for a man to marry a woman of his own clan. If we find a case of marriage between persons of the same clan, we may safely conclude that there exogamy was not the law ; but we could not infer, from any number of marriages between persons of different clans, that exogamy was the law. It is needless to say that exogamy has no relation whatever to nations (local tribes, i.e. in New Zealand itvi), and that a universal practice of marrying within the nation would be con- 266 STUDIES IN ANCIENT HISTOR Y CHAP. sistent with the law being exogamy, provided the nation was composed of several clans, and the marriages were between persons of different clans, though of the same nation. Similarly, we should not be able to infer the existence anywhere within the nation of an endogamous group, with endogamy for law, because of an entire absence of international marriages. The only proof of the presence of endogamy would be evidence showing that some clan prohibited as an offence marriages except between members of the clan. All this of course has been misunderstood by Mr. Thomson, w r hose evidence accordingly goes for nothing. Have we, then, any evidence on the subject? Not much ; but what there is points to exogamy as having anciently been Maori law, and as still being uniformly observed in practice, so far as my search has led me. (1) All the marriages mentioned in Grey's Poly- nesian Mythology are between persons of different clans, but this does not count for much, for they are the marriages almost entirely of chiefs or heroes. The chieftain-class, forming a sort of caste, and seeking to unite their clans by marriage connections, might, on mere policy, wive among themselves the marriages always being between persons of different clans, but the reasons for them unrelated to exogamy. There is one cousin -marriage in the Mythology (p. 254) ; such a marriage, however, would be permissible, supposing exo- gamy the law. It is a point scored, then, that there is an absence of mention of any marriage between persons of the same clan. It is open to us to believe that exogamy xiv NEW ZEALAND 267 'may have been the law. The only suggestion of a doubt on this point is derived from the Legend of Tuwhakaro. This hero's sister was married to the son of the chief of the Ati-Hapai tribe [? = iwi, or hapu : hapu, I think]. He went to visit his sister at her new home, when " the young sister of his brother-in-law, whose name was Maurea, took a great fancy to him and showed that she liked him, although at the very time she was carrying on a courtship with another young man of the Ati-Hapai tribe." Were we to take this as a serious element of the story, we might doubt whether the "courtship" did not imply that the parties were marriageable. But we have reason for believing that, in Australia, court- ship, though carried as far as intrigue, would not imply the jus connubii between the parties. Also, the " young man of the tribe " might truly, though counted in it and of it in a way, be legally a member of another hapu, viz. that of his mother, and it would be by that that his capacity for marriage would be regulated. On the whole, I see nothing in this story to detract from the value of the negative point that we have no mention of marriage between persons of the same clan. 1 (2) The Mythology has some suggestions of capture of women for wives, and of elopements, but beyond this it is silent on the subject of marriage, except that it states one point in Maori marriage law which is interest- ing. It occurs in the story of Hine-Moa, who stole away from her people, and swam across a strait to join 1 See Grey, I.e. pp. 54, 59, 67, 99, 100, 163, 200, 206, 244, 247, 260, 262 ff, 296, and 310. 268 STUDIES IN ANCIENT HIS TOR Y CHAP. the man she loved. " They proceeded to his [the lover's] house, and reposed there ; and thenceforth, according to the ancient laws of the Maori, they were man and wife " l good old law, not only of the Maori, but of all peoples who had much experience of marriage by capture. I have elsewhere taken account of the case of Thibetan polyandry, recorded in the Legend of Kupe. 2 (3) Had the Maories a system of actual capture of women for wives, and had they in any of their districts the form of capture ? If they had, then we shall have an inference that they were anciently exogamous, as strong as the reader may suppose to be the connection I have elsewhere established between Exogamy and the Form of Capture. Not to appraise the strength of this con- nection overmuch, it will, I think, be likely to satisfy many minds that, all the other social circumstances being such including totemism, as w T e shall see as usually are accompanied by exogamy as marriage law, exogamy was anciently the law of the Maories. But I am able to show that they had both the system of actual capture and the Form of Capture, and this on evidence of a date ten years earlier than the first attempt to explain the origin of the Form of Capture the first occasion indeed of attention being called to capture at all The Rev. Mr. Taylor says in his valuable work (with which I was unacquainted when I wrote Primitive Marriage) : " The ancient and most general way of obtaining a 1 Grey, I.e. p. 243. 2 Ibid. p. 81 [Studies, first series, p. 98]. xiv NEW ZEALAND 269 wife [in New Zealand] was for the gentleman to summon his friends and make a regular tana, 1 or fight to carry off the lady by force, and oftentimes with great violence. Even when a girl was bestowed in marriage by her parents, frequently some distant relative 2 would feel aggrieved, and fancy they had a greater right to her as a wife for one of their tribe. 3 Or if the girl had eloped with some one on whom she had placed her affection, then her father or brothers would refuse their consent, and in either case would carry a tana against the husband and his friends to regain possession of the girl, either by persuasion or force. If confined in a house, they would pull it down, and if they gained access then a fearful contest would ensue. The unfortunate female, thus placed between the two contending parties, would soon be divested of every rag of clothing, and then would be seized by her head, hair, or limbs ; and as those who contended for her became tired with the struggle, fresh combatants would supply their places from the rear, climbing over the shoulders of their friends, and so edge themselves into the mass immediately round the woman, whose cries and shrieks would be unheeded by her savage friends. In this way the poor creature was often nearly torn to pieces. These savage contests sometimes ended in the strongest party bearing off in triumph the naked 1 Taua = war-party, composed, as we have seen, of " relations," directed necessarily against an unrelated hapu. 2 " Relative," used loosely, may mean connection by marriage ; may also mean a relation of the same hapu with a son for whom he wants a wife of a different hapu, namely, his mother's. 3 " Tribe " = hapu. 270 STUDIES IN ANCIENT HISTORY CHAP. person of the bride. In some cases, after a long season of suffering, she recovered, to be given to a person for whom she had no affection, in others to die within a few hours or days from the injuries she had received. But it was not uncommon for the weaker party, when they found they could not prevail, for one of them to put an end to the contest by suddenly plunging his spear into the woman's bosom, to hinder her from becoming the property of another. Even in the case where all were agreeable, it was still customary for the bridegroom to go with a party and appear to take her away by force, her friends yielding her up after a feigned struggle. A few days afterwards the parents of the lady, with all her relatives, came to the bridegroom for his pretended abduction. After much speaking and apparent anger, the bridegroom generally made a handsome present of fine mats, etc., giving the party an abundant feast." 1 Here we have a practice of capture so described that the picture might be supposed to be drawn from such natives of Australia as present us with capture in marriage in the highest perfection. There is alongside of this the Form of Capture, indicating that anciently capture was the chief incident associated with marriage, and, of course, that the wife was always obtained from enemies, i.e. not relations. I infer from such a persistence of capture in practice as to establish the form, that the captures were rendered, at some stage, a necessity through the operation of the law of exogamy. 1 Te-Ika-A-Maui, p. 163. xiv NEW ZEALAND 271 As already stated, Mr. Taylor explains that beena marriages were common, the man becoming a member of the wife's clan. He says (p. 163) " so common is this custom of the bridegroom going to live with his wife's family, that it frequently occurs, when he refuses to do so, his wife will leave him and go back to her relatives. Several instances have come under my own observation where young men have tried to break through this custom and have so lost their wives." On this evidence the proposition that the Maories were anciently exogamous, and till recently continued uniformly to observe the law of exogamy in practice, must meantime be allowed to rest. I may mention that I have not in my reading found a single thing to suggest that any Maori group ever had the law of endogamy. 5. Totems. We have found among the Maories the local tribes divided into clans, kinship and the clan taken from the mother, and strong suggestions, if not proof, of exogamy as the marriage law. Had the clans their totems? I shall content myself with adducing the evidence. (1) A girl being carried off by a ship-captain, her relations applied to the Tohunga to get the atua to recover her. The Pakeha was witness to the proceedings of the Tohunga. " About midnight I heard the spirit (atua) saluting the guests [i.e. the girl's relations]. I also noticed they hailed him as 'relation,' and then gravely preferred the request that he would drive back the ship that had stolen his cousin. This establishes a 272 STUDIES IN ANCIENT HISTORY CHAP. kinship between them and the Atua or god, believed in at least by them a manifest note of totemism. 1 Atua, aitu, etu we have seen what these in Polynesia usually are. The Pakeha says that the Tohunga called the Atua, who spoke through him, " the boy," and in that case the Atua may have become a man-god ; but the term of address may have meant no more than the phrase "old one" among ourselves. (2) We have another fact from the Pakeha, quite casually recorded. It occurs in his humorous account of his purchase of a piece of land, and of the numerous claims put in for participation in the price. " One man said his ancestors had killed off the first owners ; another declared his ancestors had driven off the second party ; another man, who seemed to be listened to with more respect than ordinary, declared that his ancestor had been the first possessor of all, and had never been ousted, and that this ancestor was a huge lizard that lived in a cave on the land many ages ago" (I.e. p. 61). Here we have a man who, and of course his family, if not his whole hapu, deduced their descent from a lizard. That of course is totemism. From Mr. Taylor we learn that lizards and sharks were among New Zealand gods ; and of one lizard god we get from him the name Moko-titi (Taylor, I.e. p. 33). But of course it is impossible to connect this particular lizard god with the lizard ancestor of the claimant on the Pakeha. (3) This inability is of less consequence, as we have a distinct statement on the subject from the Kev. K. 1 Old New Zealand, p. 120. xiv NEW ZEALAND 273 Taylor : " There were ancestors who became deified by their respective tribes \liapu\, and thus each tribe \hapu\ had its peculiar gods." After mentioning the gods of war, Tu and Maru, he mentions Kongomai, " the chief god of Tanpo ; ' (who, he tells us, p. 41, "was a god in the shape of a whale "), and proceeds to say, " There were also gods who had human forms, and others who had those of reptiles. . . . Whilst Tawaiki was of the human form, his brethren were lizards and sharks ; and there were likewise mixed marriages amongst them. These ancestral gods still hold their places in the genea- logical tables of the different tribes \]iapu] " (Taylor, I.e. pp. 32, 33). 1 We might here close the proof as conclusive, on evidence above all suspicion, that the ancient Maori were in clans organised on the totemic principle with female kinship and exogamy as marriage law ; but one or two further facts furnished by the authorities are worth noting, as they will help us hereafter in our inquiry into the structure of societies of which we have but partial records. The Kev. E. Taylor seems to have been impressed with the fact that the reptile gods of New Zealand were older than the men gods. " There were gods," he says, 1 See further as to clan-gods, "the curse of Manaia" (Grey, I.e. pp. 163, 164), where is also a proof of kinship being through the mother only. I may here cite, as a further note of totemism, the fact that the Maori had a pigeon-god (Rupe) ; and the Rev. R. Taylor mentions that some Maori dare not eat the pigeon (pp. 110, 168). Taylor mentions another of their gods Tuparaunui, "a large fly and an ancient god" (p. 121). It is impossible to doubt that we have here a demonstration of the totem stage. T 274 STUDIES IN ANCIENT HISTORY CHAP. " who had human forms, and others who had those oi reptiles. As in geology there is a reptile age, so th< was one in the mythology of New Zealand." As me1 physic worked on the basis of the mythology, it triec to invert this order of genesis, and assigned to the mei gods the earlier place. In the genealogy of Tawaiki which starts from a woman, we find one of his ancestoi (p. 37) was Kaiaia, a sparrow-hawk; while we have marriage, in the genealogy of descent from this commo] mother, between Punga [anchor of canoe] and Karil [net sinker], of which the offspring were the lizard, the shark, and the dog-fish, cousins of the man-god Tawaiki. 1 It is very obvious that the promoters "of faith in Tawaiki would not have put him into this connection but for the necessity of finding a place for him among acknowledged gods. The root of the genealogy in a common mother is another note of the primitive state of female kinship, the value of which cannot be too much insisted upon. Its occurrence in New Zealand mythology is most frequent. The Taninhas, great lizards, who play in New Zealand the part of dragons in European mytho- logy, trace their genealogy to a female progenitor. Tane, whom we found in the Hervey group, appears in New Zealand, according to some, as the mother, not the father, of Tui, and of birds in general, and of trees. Accord- ing to this account Tane was the wife of Parani, and Parani was the father of Tui. But Tane was a tree first, and a woman after. Among the first offspring of Eangi and Papa were first the Kumara (sweet potato), next 1 lc. pp. 35-37. xiv NEW ZEALAND 275 the fern-root, next Tane. 1 Next after Tane, Tiki was born, "from whom man proceeded," 2 but here again there is a dispute. Some traditions say Tiki is a woman (p. 23), so that we see the female idea in conflict in the traditions with the male idea. It was finally- overridden, as it was bound to be among people who had so extensive a practice of marriage by capture ; and the female progenitors were displaced on the whole by a long list of creative fathers, i.e. fathers supposed to have made nature in her different departments. 6. Blood-Feud. For instances of the operation in New Zealand of the blood-feud, the reader may consult Grey's Mythology (pp. 61, 98, 108); but, indeed, the whole mythology is full of it. The Rev. E. Taylor mentions the law of blood, in the appendix to his work, as one of the points of correspondence between the Maori and the Hebrews. As we have already seen in the account of the Muru, the law of retaliation, as its working is described by the Pakeha Maori, was so constantly in operation as in effect, in connection with minor offences of one member of a liapu against another, to keep movable estate continually shifting from one hand to another. The Pakeha says : " As the enforcers of this law were also the parties who received the damages as well as the judges of the amount, which in many cases, such as that of the burnt child, would be everything they 1 That Tane was a tree is proved by the legend given by Taylor, p. 21. "He [Tawhirimatea] placed his mouth to that of Tane-mahuta, and the wind shook his branches and uprooted him." 2 Taylor, I.e. p. 18. 276 STUDIES IN ANCIENT HISTORY CHAP. could by any means lay hands on, it is easy to perceive that under such a system personal property was an evanescent thing altogether" (p. 84). These executions were never resisted. In the more serious cases of homi- cide, the operation of the law of Muru had full swing only in the case of the homicide being accidental. In that case the penalty was the wholesale plunder of the criminal and his family. Where the slaughter was not accidental, the Pakeha observes that in nineteen cases out of twenty it would either be a meritorious action, or an action of no consequence, and the law of Muru was inapplicable. The explanation of this is, that where a man killed another by accident, the slain man would be in most instances, as a matter of course, one of his own tribe. Where the slaughter was intentional, the victim would usually be either a slave or a man of another tribe. A slave's death was nobody's affair. As to the man of another tribe, the slayer had but to declare that he killed the man in revenge for some aggression, either recent or traditional, when his whole tribe would support and defend him to the last extremity. The Pakeha says that murder, i.e. of a man of the same tribe \iwi\ y was comparatively rare, and went in most cases unpunished. "The murderer in general managed to escape to some other section of a tribe where he had relations, who, as he fled to them for protection, were bound to give it, and always ready to do so. Or otherwise, he would stand his ground and defy all comers by means of the strength of his own family or section, who all would defend him and protect him as a mere matter of course" (p. 91). In XIV NEW ZEALAND 277 this passage, for section of the tribe, we must read clan, as we have seen ; and then no statement of the obliga- tions of blood could be more complete. The case is not put of a man murdering a member of his own clan ; but in that case he would no doubt among the Maories, as elsewhere, suffer swift death at the hands of his own kinsmen, or be obliged to wander far from his relatives, a homeless outlaw. CHAPTEE XV AUSTRALIA THE Australians have commonly been represented as of a uniform type of savagery; but Australia is in magnitude about equal to all Europe, and it is not surprising that con- siderable diversities are found to occur in its population. To ascertain these, we must localise our information from point to point, to see to what tribes and districts it applies. Grey's account l is briefly as follows : " One of the most remarkable facts connected with the natives [of Australia] is that they are divided into certain great families, all the members of which bear the same name, as a family or second name. . . . These family names [they occasionally assume local form, and seven principal totem names and eight local variants are mentioned] are common over a great portion of the continent ; for instance, on the western coast, in a tract of country extending between four and five hundred miles in latitude, members of all these families are found. In South Australia I met a man who said he belonged to one of them, and Captain Flinders mentions Yungaree [one of the variants] as the name of a native in the Gulf of Carpentaria [i.e. on the north-east of the continent]. These family names are per- petuated and spread through the country by the operation of two remarkable laws : (1) that the children of either sex 1 Journals of Two Expeditions of Discovery in North-West and Western Australia, during the years 1837 and 1839. By George Grey, Esq. [Now Sir G. Grey.] Vol. ii. pp. 225 et seq. London, 1841. CHAP, xv A USTRALIA 279 always take the family name of the mother; (2) that a man cannot marry a woman of his own family name." Every family has "as its crest or sign" (kobong = totem) some animal or vegetable, between which and the family there exists a certain mysterious connection. A most remarkable law again " is that which obliges families connected by blood upon the female side to join for the purpose of defence and avenging crimes ; and as the father marries several wives, and very often all of different families, his children are repeatedly all divided among themselves ; * no common bond of union exists between them." As their laws are principally made up of sets of obligations, due from members of the same great family towards one another which obligations of family names are much stronger than those of blood it is evident, Sir George continues, " that a vast influence upon the manners and state of this people must be brought about by this arrangement into classes." Instead of classes he should have said families or clans. The description is that of a population of local tribes made up of exogamous totem kindreds the kindreds necessarily inter- fused in such a case being greatly interfused owing to the prevalence of polygamy, and being the same in all the local tribes throughout a great area. The clans or kindreds were named after the totem ; those of the same totem were bound together by the blood-feud and debarred from intermarriage ; 1 We may infer the operation of the blood-feud in connection with female kinship in New South Wales from the following passage in the report of the United States Exploring Expedition (vol. ii. p. 187) : " The natives [of New South Wales] have not, properly speaking, any distribution into tribes. In other conflicts, those speaking the same language, and who have fought side by side, are frequently drawn up in battle array against each other, and a short time after may be seen acting together." The bond of blood is stronger than the connection with the local tribe. There could be no better illustration than this of the kind of error into which observers are liable to fall in regard to primitive societies ; somewhat similar observations have not unfrequently been made in regard to tribes on the north-west Pacific shores of America by untrained observers. 28o STUDIES IN ANCIENT HISTORY CHAP. and they alone were bound together by all the obligations and restrictions of blood-relationship. Marriage within the totem kindred was held in the greatest abhorrence as being incestuous. And no individual of the species of the kobong or totem was ever killed without reluctance or without being allowed a chance of escape. The old men, it appears, managed to keep the women pretty much to themselves by means of a system of betrothing and exchanging daughters, which drove the younger men into stealing women. Sir George concludes by intimating that he had devoted much study to the " obliga- tions of family names," but was unable to bring out the results more fully when his book was published. His observations, he takes care to notice, 1 " can be only considered to apply, as yet, to the natives of that portion of Western Australia lying between the 30th and 35th parallels of south latitude, though," he adds, " there is strong reason to suppose that they will in general be found to obtain throughout the continent." Mr. Gideon S. Lang, who lived long in Australia in con- stant contact with the native population, declares that Sir George Grey's account of the natives confirms his own observa- tions in almost every particular. " One of the most remark- able facts connected with the aborigines," he says, 2 " is that over the entire continent, from Swan Eiver to Sydney, from Melbourne to the Gulf of Carpentaria, they are so exactly similar in appearance, complexion, customs, and language, in their weapons, and manner of fighting and hunting, that it would almost appear as if they were all descended from one canoe-load of people, and these again from the same tribe." " Of course there are points of difference between them, but they are only such as might be expected from difference of circumstance and locality. . . . The native inhabitants of the whole continent form, in fact, one people, all governed by the same laws and customs." Could we believe this, and have perfect confidence in Grey's statements, there would be an end of inquiry as regards Australia. We shall soon see, however, that Mr. Lang's 1 Vol. ii. p. 231. 2 The Aborigines of Australia, pp. 1 et seq. Melbourne, 1865. xv AUSTRALIA 281 statement must receive some qualifications. More important than his general statement is the evidence he gives on in- formation. "The Moreton Bay blacks," he says (p. 10), "led me to understand that they were divided into four clans. All the children take after the clan of their mother, and no man can marry a woman of the same clan, although the parties be born of parents in no way related according to our ideas." This is a fact as to some tribes on the east coast of Australia, adjoining the Kamilaroi, of whom hereafter, and it is in perfect agreement with Grey's observation of the tribes on the west coast. How the clans are named Mr. Lang does not say, but he must be understood as acquiescing in Grey's statement as to their being named after animals and plants. A casual observation at p. 21 discloses the fact that some natives in the Moreton Bay district name the con- stellations after animals, which may be taken as a hint of totemism. Another fact corroborative of Grey's general statement is obtained from the Kev. E. Fuller, missionary for some years in Frazer's Island, off the east coast of Australia, to the north of Moreton Bay. 1 This island, Mr. Fuller tells us, is eighty- five miles long by twelve broad. Since contact with Euro- peans the population has rapidly decreased in numbers, and in 1872 he estimated that there were not more than three hundred blacks surviving. The ancient framework of the society, however, still survived, for this " handful " of people were comprised in no fewer than nineteen distinct tribes. As to the marriage law, Mr. Fuller is most explicit. " A man cannot," he says, "marry a woman belonging to his own tribe, and the children are supposed to belong to the mother's tribe." Of totemism in the island he gives no hint. The islanders were naked cannibals. Mr. James Bonwick in his little work, The Wild White Man and the Blacks of Victoria (Melbourne, 1863, p. 58), recognises as prevailing among the blacks to some extent a 1 Letter dated 15th August 1872, to a friend of the missionary in Ipswich, printed in the Brisbane Queenslander, 7th September 1872. 282 STUDIES IN ANCIENT HISTORY CHAP. taboo on the use of the kobong or totem such as, according to Grey, existed on the west coast, but much cannot be made of his evidence on this point. "There is in some places," he says, " a sort of taboo of things selected by a tribe or family as a kind of badge, called a kobong. Governor Grey declared that the kobong had elements in common with the taboo of other people. Men cannot marry women of the same kobong." It is clear that we have not here a mere echo of the state- ment of Grey. But, at any rate, Bonwick does not seem to have been in a position to declare Grey's statement inapplic- able to the blacks of Victoria. Mr. Taplin's account of the " Narrinyeri " is more satis- factory, and it shows at once that Grey's statements must be taken as subject to qualifications and exceptions, while con- firming them in the essential points. 1 The name of these people means men, which they take pride in calling them- selves, as distinguished from other aborigines, whom they call wild black fellows. They occupy a tract of country which would be included within lines drawn from Cape Jervis to a point about thirty miles above the place where the river Murray discharges itself into Lake Alexandrina, and from thence to Lacepede Bay. " They are divided into eighteen tribes, and each is regarded by them as a family, every member of which is a blood relation, and therefore between individuals of the same tribe no marriage can take place. Every tribe has its Ngaitye, or tutelary genius, or tribal symbol, in the shape of some bird, beast, fish, reptile, insect, or substance." Elsewhere he says of the Ngaitye " that the tribe regard it as a sort of good genius who takes an interest in their welfare something like the North American Indian totem." Mr. Taplin gives the Ngaitye, or totems, of the eighteen tribes. They are as follows, viz. : 1. Black duck, and black- snake with red belly ; 2. Black swan, teal, and black-snake with grey belly ; 3. Black duck ; 4. Leeches, cat-fish (native 1 The Narrinyeri; an Account of the Tribes of South Australian Abori- gines, inhabiting the Country around the Lakes Alexandrina, Albert, and Coorong, and the Lower part of the River Murray. By the Eev. George Taplin. Adelaide, 1874. xv AUSTRALIA 283 pomerey) ; 5. Whip-snake ; 6. Wild dog, light colour ; 7. Wild dog, dark colour; 8. Mountain duck (chocolate shel- drake); 9. Mullet, called Kanmeri; 10. Kangaroo - rat ; 11. Butter-fish (native kungulde) ; 12. a kind of coot called Turi ; 13. Tern, a small kind of gull; 14. Bull-ant, and a kind of water- weed called by the natives Pinggi; 15. Whale (native kondarli) ; 16. Pelican; 17. Wattle-gum; 18. Musk-duck. Of the eighteen " tribes," all except the first, second, fourth, and fourteenth would appear to be pure families or tribes of descent; it is disclosed that three of them are directly named from the totem. The Whales (15) are Kondolin- yeri, from Kondarli = Whale, and inyeri, belonging to; the Coots (12) are Turarorn, from Turi, Coot ; and the Mullets (9) are Kannierarorn, from Kanmeri, Mullet. Nos. 1, 2, 4, 14, on the other hand, would seem to be local tribes in which two (and in one case three) distinct families or tribes of descent are united. No. 3, it will be seen, has for totem the black duck, which is one of the totems of No. 1, and no other. Both Nos. 1 and 3 reside on the river Murray. 1 Mr. Taplin has not mentioned whether marriage was permitted as freely between the Snakes and Ducks within a compound local tribe like No. 1 as between the same Snakes and the Ducks in No. 3. He shows, however, that the prohibition of marriage between those of the same tribe was not the sole limitation on the right of marriage. " The aversion of the natives to even second cousin marriages," he says, " is very great. They are extremely strict in this matter. The first inquiry with regard to a proposed marriage is whether there is any tie of kindred between the parties, and if there be, it prevents the match, and if the couple should cohabit afterwards they will 1 No restriction on marriage is spoken of beyond the prohibition of marriage within the tribe which in fourteen cases out of eighteen meant only prohibition of marriage between persons of the same totem except that (as is immediately hereafter stated) there was a device for preventing marriage between persons who, though not of the same totem, were con- sidered to be near blood relations. That an expedient was needed for preventing marriage in such cases shows, of course, that originally there was prohibition of marriage within the totem kindred only. 284 STUDIES IN ANCIENT HISTORY CHAP. be always looked on with dishonour." Elsewhere (p. 26) he explains an artifice by which persons of different tribes, but too nearly related to be allowed to marry, are prevented from ever marrying, viz. by " Ngia-ngiampe " being established between them, whereby they are prohibited from ever approach- ing or touching each other. A man's children (p. 10) belong to his tribe and not to their mother's. Property always descends (p. 38) from father to son, and from a man who dies childless, to his brother's children. We have here the same kind of totemic families and the same marriage law (speaking broadly) on the south of Australia which Grey found on the west, with two differ- ences : (1) kinship is no longer counted through the mother, but through the father. (2) As might in consequence be ex- pected, the tribes of descent, excepting in four cases, are no longer interfused ; all the men of the same "totem have drawn together, and groups have been formed which are homogeneous, which are at once local tribes and tribes of descent. The locality which each of the eighteen local tribes inhabits is mentioned by Mr. Taplin, and it appears that, except in the case of the Black Ducks, all of the same stock are together in the same locality ; while it would be reasonable to expect that the Black Ducks of No. 1 and No. 3, alike dwellers on the Murray River, would, if time were given, amalgamate and form a homogeneous local tribe, leaving tribe No. 1 composed of the people of Black-snake with red belly only. There would then remain three exceptional local tribes No. 2 with three totems, No. 4 with two totems, and No. 14 with two totems. The interfusion of kindreds in these tribes goes to show that the Narrinyeri formerly counted kinship through females only, as most of the Australian tribes continued to do. With kinship through females only a man being of a different totem from his wife or wives, and children being of the totem of the mother there would necessarily be such interfusion. The care taken to prevent marriage with relations on the side of the mother and grandmother relations whom civilised people may marry is another indication of the former pre- valence of this system of kinship. If it may be taken that all XV A USTRALIA 285 the Narrinyeri totems were formerly interfused, more or less, in the local tribes, it is easy to understand how the actual consti- tution of the local tribes was arrived at. With kinship counted through the father, all the men of a family would be of the same totem, and would be united for the purposes of the blood-feud, which would then be a far more serious thing than it could have been while (with the totem taken from the mother) it might array members of the same household on different sides. The more serious it was, the more need would there be of cohesion between those whom it united, and the men of the same totem would draw together until they formed a true clan within the local tribe or rather a portion of a clan, the other portions being in other local tribes. Here things might have rested, but among the Narrinyeri, in most cases, there has been a further drawing together of the men of the same totem, until in fourteen cases out of eighteen they have all coalesced in separate local tribes. In one case more (that of the Black Ducks) the segregation of the totem kindred had not been completed. In the remaining three cases the people of the same totem had all come to be together, but did not form separate local tribes. If, as seems clearly to have been the case (see note, p. 317), there was originally prohibition of marriage only between per- sons of the same totem among the Narrinyeri/the clans united in these three tribes would have originally been, and, with male kinship they might have continued to be, free to inter- marry with one another. But association in the same local tribe, with male kinship, and with such association become uncommon (and perhaps a frequency of the relationships they now acknowledged outside the totem), may have led to the prohibition being extended from the totem to the local tribe (the statements of Taplin and Meyer, however, do not make it clear that this was the fact). In this case two of the Nar- rinyeri clans would have been cut off from marriage with one clan besides its own, and one of them from marriage with two clans besides its own. Each of the Narrinyeri tribes had a chief (called Eupulle, which means landowner), elected by the heads of families, 286 STUDIES IN ANCIENT HISTORY CHAP. apparently from the family of the preceding chief. Wives were commonly got among them by exchange, and the brother, according to Mr. Meyer, gave a woman away more frequently than the father. It was considered disgraceful for a girl to elope. Polygamy was practised. The Narrinyeri did not kill the Ngaitye (Taplin, p. 48) unless it was an animal good for food, and when they did kill and eat it, they were careful to destroy the remains, " lest an enemy should get them, and by sorcery cause the Ngaitye to grow in the inside of the eater and cause his death." Mr. Taplin (p. 8) quotes the Eev. H. E. A. Meyer, who resided among the Narrinyeri before they had much inter- course with Europeans, as corroborating his own account at various points, and inter alia as to their exogamy. The evidence as to these tribes having exogamy in operation in conjunction with male kinship may, therefore, be accepted as being above suspicion. As to the law of blood-feud and the solemn obligation laid on the avenger of blood among these tribes to avenge it, by slaying the supposed murderer, or one of his relatives, see Taplin, pp. 16, 17, 22, and 89. At p. 89 Mr. Taplin pictures a small group of natives prowling about the bush at night in order to surprise some enemy or his relatives, "acting upon the native principle, if you cannot hurt your enemy, hurt his nearest relatives." Of the Adelaide district we have some information from Mr. Eyre, but it is rather negative than positive. All he says as to limitations on marriage is that "relatives nearer than cousins are not allowed to marry, and this alliance does not generally take place." 1 Elsewhere he quotes from Grey the passage in which it is said that children take the family name of their mother, and that persons of the same family name cannot marry. He does not expressly say so, but it is plainly implied that he had never heard of anything like this limitation on marriage, or of any limitation on marriage, other than that already quoted from him. And in a Latin note he says, 1 Journals of Expeditions of Discovery into Central Australia in the years 1840-41. By Edward John Eyre, vol. ii. p. 319. London, 1845. xv A USTRALIA 287 " Apud plurimas tribus, juventutem utriusque sexus sine discrimine concumbere in usu est " ; and that when the " Mooyum-karr " a machine in the shape of an egg, with mystic sculpturing had been produced and rattled, " liber- tatem coeundi juventuti esse turn concessam omnibus indicat " ; from which it seems that he thought there was no restriction upon the intercourse of the sexes of the kind declared by Grey, or of any other kind. 1 Again he says nothing of the 1 Mr. G. S. Lang (I.e. p. 38) has a few remarks in Latin to much the same effect as the Latin note of Eyre. He says, " Ut antea animadverti, senioribus pertinent mulieres permultae, junioribus vel perpaucse, vel nullae. Connubium profecto valde est liberum. Conjuges, puellse, puellulee cum adolescentibus venantur. Pretiuin corporis paene nullius est. Vendunt se vel columbae vel canis vel piscis pretio. Inter Anglos et Aborigines nihil distat." Yet Lang tells us that among the tribes under his observation, marriage was forbidden between persons of the same totem. Are we to surmise that this could be so, and intercourse without marriage free of restriction. 1 ? are we to take it that such inter- course as is vouched for by Eyre and Lang was subject to restriction, though a traveller might think it promiscuous ? From the nature of the case it is almost impossible to get evidence on the point raised in the preceding paragraph. In Kamilaroi and Kurnai (by the Eev. Lorimer Fison and A. W. Howitt, Melbourne, 1880) it is repeatedly asserted that intercourse never took place between natives who were not free to marry ; but for proof of this the authors appear to rely a good deal on the fact that, in lending a wife to a guest, a woman for- bidden to the guest would not be given, which is almost a matter of course, and should go for nothing. Of the statements from correspondents which they publish on this matter, only two or three appear to be in any degree trustworthy, and these are of course limited to the practice of particular tribes. And these statements are not all of the same tenor. The Kev. W. Julius Kuhn, writing of the Turra tribe (York's Peninsula, South Australia), says distinctly that at grand corrobborees the restric- tions observed in marrying were put aside, and, though less distinctly, he conveys that there might be intercourse between persons of the same totem (Kamilaroi and Kurnai, p. 286). "The men," he says, "were not confined to any particular totem ; " and as they were not confined to any particular totem at any time, he must have meant that the totem put no limit upon intercourse. 288 STUDIES IN ANCIENT HISTORY CHAP. relationships of the natives, as if he presumed them to be the same as those of white people. He has some observations, however, which bear upon kinship. The South Australian native, he tells us, adopts " some object in creation as his crest or tiende " ; and he thought it apparent there was very little difference between this and the kobong described by Grey, though he did not observe the unwillingness to kill the animal of which Grey had spoken. The same tiende seemed to de- scend from father to son, but he had heard occasionally of instances where this was not the case. And so far as his observations and inquiries had enabled him to ascertain, among the numerous tribes frequenting the Murray Eiver, the people had not family names or surnames " perpetuated through successive generations on the mother's side." Mr. Moorhouse, too, assured him that he had been equally unable to detect anything of the kind among the tribes frequenting the district of Adelaide. Evidence of this sort is not worth much. As regards certain tribes on the river Murray, however, we have already seen, on the authority of Taplin and Meyer, that names were not perpetuated on the mother's side, but on the father's. On the same authority we have seen, what Mr. Eyre failed to observe, totemism and exogamy, as described by Grey, in full force among these tribes, though in connection with male instead of female kinship. What we are told by Mr. Wilhelmi of the tribes about It will be found that Mr. Wilhelmi, who is quoted on the next page, says of the Port Lincoln tribes that intercourse between persons who were forbidden to marry was not considered incestuous even on ordinary occasions. There seems to be no further evidence to this effect. But Mr. Wilhelmi had much experience of those tribes ; and the fact he mentions was more easily ascertainable than the practice at corrobborees. As to the practice at corrobborees, if we trust the statement that the totem restriction was always disregarded in the Turra tribe, it will carry us far towards a conclusion as to ancient Australian practice. Nor would evidence that the totem restriction was in recent times generally observed by unmarried people in Australian tribes (were it forthcoming) make any obstacle. xv A USTRALIA 289 Port Lincoln, in South Australia, is also, as far as it goes, in full accordance with the statements of Sir George Grey. These tribes, Mr. Wilhelmi says, 1 "are divided into two separate classes, viz. the Matteri and the Karraru. This division seems to have been introduced since time immemorial and with a view to regulate their marriages, as no one is allowed to intermarry in his own caste but only into the other one. . . . This distinction is kept up by the arrangement that the children belong to the caste of the mother. There are no instances of two Karrarus or two Matteris having been married together; and yet connections of a less virtuous character which take place between members of the same caste do not appear to be considered incestuous." For caste read "family," gens, or tribe of descent, and we have here Grey's system realised except that there is no mention of totems, and subject to the Latin notes of Eyre and Lang. 2 "We now come to the Kamilaroi-speaking tribes of Eastern Australia, the first accounts of which represented them as having a marriage system entirely different from that with which we have been hitherto meeting a system, moreover, to the last degree strange and perplexing. The history of the evidence relating to those tribes will have to be considered at some length. The first statement as to the social organisation and the marriage law of the Kamilaroi (so to call them) was made by the Eev. W. Eidley in a lecture delivered by him in Sydney in November 1853. Mr. Eidley stated that the Kamilaroi were divided into castes, the castes being distinguished by peculiar names. " There are four names," he said, " of men Ippai, Murri, Kubbi, and Kumbo ; and four of women Ippata, Mata, Kapota, Buta. Every black has one of these names by birth." 1 Quoted in the Aborigines of Victoria, by K. Brough Smith, vol. i. p. 37. 2 Mr. Wilhelmi adds that there are besides " certain degrees of relation- ship within which intermarrying is prohibited." But these, unfortunately, he had not been able to trace. U 290 STUDIES IN ANCIENT HISTORY CHAP. " In one family all the sons are called Ippai, the daughters Ippata ; so that if you find a black man's name is Ippai, you may be sure all his brothers are Ippai, and his sisters Ippata. In another family, in like manner, all the sons were Murri and all the daughters Mata ; in a third the sons all Kubbi and daughters all Kapota ; in a fourth the sons all Kumbo and th( daughters all Buta." These being the family or caste names, " the following rules of intermarriage," said Mr. Eidley, " are most strictly observed : 1. Ippai may marry an Ippata (of another family) or any Kapota. 2. Murri may marry only Buta. 3. Kubbi mai marry only Ippata. 4. Kumbo may marry only Mata." Any attempt to infringe these rules, he said, would be re- sisted even to bloodshed. Polygamy was allowed and was common. The rules governing descents were given by Mr. Eidley as follows : " 1. The children of Ippai by Ippata. are all Kumbo and Buta. 2. The children of Ippai by Kapota are all Murri and Mata. 3. The children of Murri are all Ippai and Ippata. | 4. The children of Kubbi are all Kumbo and Buta. 5. The children of Kumbo are all Kubbi and Kapota. By tracing out the effect of these rules you may perceive," he said, " that descendants of every family come, in the course of a few generations, into the privileged class of Ippai, while the sons of these aristocrats inherit not their father's rank, but belong to the Kumb or Murri caste." Mr. Eidley further mentioned that the Kamilaroi had " commonly distinctive names added to their family titles " ; and that these distinctive names were often taken from animals. 1 According to this extraordinary statement the tribe among the Kamilaroi was, structurally, unlike any other tribe that had ever been described. Taking its divisions, as determined by the names of its members, to be truly castes, it was uncertain whether one should count four or eight castes within the tribe. The Ippais, however, were represented as being the aristocratic or highest caste, and except in their case the caste of children was represented as depending on that of their father. There 1 SQePritchaTd's Natural History of Man, vol. ii. pp. AQOetseq. Norris's edition. London, 1855. xv AUSTRALIA 291 was no suggestion of the tribe being divided into elans or gentes each with its animal totem, or of the marriage law being related to exogamy. The law of intermarriage was, according to the statement, dependent on the caste names solely ; and the caste name was inherited being directly determined by that of the father, indirectly by that of the mother whom he was confined to marry children, nevertheless, being never of the caste either of their father or of their mother. Murri, for example, because he was named Murri, could marry only a Buta ; if he took ever so many wives, as he might do, they all had to be Butas ; and his children were neither Murri nor Buta, but Ippai and Ippata. In the Journal of a tour made by Mr. Eidley among the aborigines of the western interior of Queensland in the year 1855, which is printed in the Appendix to Dr. Lang's work on Queensland, the foregoing statements are repeated in general terms ; and it is suggested that the peculiar classification of the people, and the marriage law connected therewith, were the invention of sagacious and comparatively civilised men among the remote forefathers of the Kamilaroi. The fact of there being similar " family or clan names " among the tribes at Moreton Bay is also here disclosed. 1 At some time before May 1871 Mr. Eidley made a fresh statement on the Kamilaroi marriage law to the Kev. Lorimer Fison, then a missionary in Fiji. This also agreed with that made in 1853, except at one point. Ippai was now repre- sented as by law free to marry Kapota and no other ; while the privilege of marrying Ippata (of another family), which had been assigned to him by the first statement, was now declared to be " an infringement of rule, allowed in favour of some powerful Ippai, and so continued." Apart from this little bit of theorising, the extraordinary character of the original state- ment remained unqualified in the early part of 1871. Before August in that year, however, Mr. Eidley had fresh occasion to visit the interior of Australia, and he now obtained new light on this subject, and thereupon made a further com- munication to Mr. Fison. He had now discovered that the Australia, by J. D. Lang, D.D. London, 1861. 292 STUDIES IN ANCIENT HISTOR Y CHAP. Ippais and Ippatas were in three s^&divisions, namely, Emus, Blacksnakes, and Bandicoots ; and the Kubbis and Kapotas likewise in three m&divisions, namely, Opossums, Paddy-melons (a sort of kangaroos), and Iguanas. Among the Kumbos and Butas, however, he found only two subdivisions, namely, Emus and Blacksnakes ; and also only two among the Murris and Matas, namely, Iguanas and Paddy-melons. By this date the term "caste " had disappeared from Mr. Bidley's exposition. It was replaced by a term even more inapplicable, namely, " class," used in the sense which that term has in connection with the classificatory system. And instead of the father's " caste " determining that of the children, Mr. Eidley was now able to state that both the " class " name and the subdivisional name (which the reader must have recognised as that of the totem, clan, or gens) were taken through the mother. Ippata's children, for example, were Kumbo and Buta, whether their father was a Kubbi or whether he was an Ippai ; and they were of the totem or clan of their mother, and not of the totem of their father. Moreover, Mr. Bidley had now found that the privilege of marrying within the " class " (a woman of the corresponding female name but of a different family, that is, totem) which he had supposed to be peculiar to the Ippais, and to have been obtained by them through an infraction of rule, was possessed by the other " classes " as well as by the Ippais, so that this was not a privilege but a common right. And a man's sons were not necessarily, as had been stated, all of the same " class." As none of the Kamilaroi were restricted to marrying in one " class," and polygamy was common, and the " class " of children was determined by that of the mother, when a man had wives of different " classes," his children by them were of different classes. And, indeed, all the four " castes " or " classes " might be represented in a single house- hold. This statement, which upset so much of what had been previously announced, and disclosed so much that was new, was yet itself obviously imperfect. There being three sub- divisions of the Ippais and Ippatas, it was impossible that there should not have been three of the Kumbos and Butas, the xv AUSTRALIA 293 children of the Ippatas. There being Bandicoots among the former, it was impossible there should not have been Bandicoots among the latter. The children of Ippata Bandicoot must have been Kumbo and Buta Bandicoot, according to the rules by which the gens and class names were said to be transmitted ; and Buta Bandicoot only could have been, according to those rules, the mother of Ippata Bandicoot. Similarly there being three subdivisions of the Kapotas, it was impossible that there should not have been three of the Murri and Mata ; as there were Opossums among the former, there must have been Opossums among the latter. And further research has since disclosed that the representation that the Kumbos and Butas, the Murris and Matas had each two subdivisions only resulted from an error of observation. In 1872 a further communication (giving the results of the same journey) from the pen of Mr. Eidley appeared in the Proceedings of the Anthropological Institute. 1 The errors of observation above noticed appeared in this also ; and it was, moreover, vitiated by the incorporation of inferences made by Mr. Fison from the statement of 1853. It gave, however, some further information as to the clan divisions. On the river Marron the Murri and Mata were said to be subdivided into Iguanas, Paddy-melons, and Opossums, while among the Kubbis and Kapotas there were Iguanas and Paddy- melons only. The Kumbos and Butas, again, were said to be subdivided into Emus, Blacksnakes, and Bandicoots, while among the Ippais and Ippatas there were Emus and Black- snakes only. The errors or inconsistencies already noticed were thus repeated, but with the " classes " reversed. Among the Wailwun tribes (below the junction of the Namoi and the Barwan rivers) the Murris and the Matas, and also the Kubbis and Kapotas, were found to have a fourth totem, namely, the Guru, a species of bandicoot. The Ippais and Ippatas, again, also the Kumbos and Butas, were found to have three totems, but only two of them the same as those previously noticed, viz. the Emu and Blacksnake, the Bundar (kangaroo) taking the place of the bandicoot. 1 Journal of the Anthropological Institute, vol. ii. p. 257. 294 STUDIES IN ANCIENT HISTORY CHAP. In other parts of the country about the Balonne, the Kumbos and Butas were stated to be Emus and Wombats, while the Ippais and Ippatas, instead of being Emus and Wombats (as they ought to have been if they were the children of Buta Emu and Buta Wombat), were Kangaroos and Black- snakes. The Murris and Matas, again, were here Opossums and Paddy-melons, while the Kubbis and Kapotas were Opos- sums, Paddy-melons, Iguanas, and Gulu, a species of bandicoot. On the Macintyre, among some tribes, the Ippais were either Emus, Blacksnakes, or Yuluma, which is untranslated. While the clans into which the so-called castes or classes are subdivided thus varied from district to district, the " class " names themselves are found to vary from district to district. Mr. Eidley, in one of his communications to Mr. Eison, had already stated that among tribes speaking the Balomer lan- guage, all the men bore one of the following four names, viz. : 1. Urgilla ; 2. Wunggo ; 3. Obur ; 4. Unburri ; and all the women one of the following four names, viz. : 1. Urgillagun ; 2. Wunggooun ; 3. Oburugun ; and 4. Unburrigun ; he had also previously mentioned that four names, different from any of the preceding, were in use as applicable to the males among the Moreton Bay aborigines, and four corresponding names formed from the male names by adding gun or un, as applicable to the females. Among the Wide Bay aborigines, he found five names different from the pre- ceding in use for the males, and five corresponding names for the females. It was suggested, though not expressly stated, that laws of marriage and descent similar to those of the Kamilaroi occurred wherever these male and female names were stated to occur. Let us here pause to note some results obtainable from the foregoing statements. Whatever may be the meaning or importance of the so-called class names, this much is certain, that they are not exceptional among the Kamilaroi, but occur among tribes ranging over a wide area of south-eastern Australia ; and that the names of the women among the Kamilaroi may be assumed to be in all cases feminine forms of the men's names XV AUSTRALIA 295 in each class. Further, we see that the names form themselves in two distinct sets as follows : Female. Buta. Ippata. FIRST SET. SECOND SET. Male. Female. Male. 1 1. Murri . . Mata. 3. Kumbo 2. Kubbi . . Kapota. 4. Ippai . TOTEMS. TOTEMS. Iguana. Paddy-melon (a sort of kangaroo). Opossum. Guru, or Gulu (a species of bandi- coot). Emu. Blacksnake. Bandicoot. Bundar = kangaroo. Wombat. Yuluma. It will be seen that each set is, so to speak, self-sustaining. Mata's children are Kubbi and Kapota ; Kapota's children are Murri and Mata. Whatever husband any one of these women takes, the children belong to her set and totem. Similarly of the second set. Next, let us observe, the totems in the two sets are absolutely different. Not one that appears in the one set appears in the other unless Guru, or Gulu, means the same thing as Bilba (bandicoot), which is not suggested, and Bundar (kangaroo) the same thing as Paddy -melon, which is not suggested. With reference to this, it must be remembered that the minutest difference in the type of animal makes an absolute difference in a totem. Two gentes might each have a dog for a totem ; but if the one had a terrier and the other a collie, the gentes would be thereby marked as absolutely distinct. We now come to Mr. Eidley's last communication on this subject, dated July 1874, which appeared in Nature on 29th October 1874. It is mainly directed to controverting Dr. Long's statement, 1 that among all the tribes, exogamy, pure and simple, is the marriage law, i.e. that no man can marry a woman of the same gens, clan, or totem with himself. " It is true," says Mr. Eidley, " that no man may marry a woman of 1 This statement is not accessible to me. the writer referred to.] [Mr. Gideon Lang is clearly 296 STUDIES IN ANCIENT HISTORY CHAP. the same names as his sisters. But it is by no means true, as Dr. Long stated, on imperfect information, that no one can marry a woman ' of the same clan/ taking the word clan in the common sense of the term as equivalent to gens. The rule that restricts marriage is founded on an exact law of pedigree and class names. It is as follows among the aborigines of the Namoi ; and other tribes have rules similar in the main, though the names differ widely. " The men are all divided into four classes [classes named]. The Murri are regarded as the most important ; the Kubbi are the lowest in esteem. [The women are similarly divided.] There is also another classification marked by totems, in which a second name is given to every one according to birth. Thus there are the Bundar [kangaroo], Meite [opossum], Duli [iguana], Nurai [blacksnake], Dinoun [emu], and others. 1 On these classifications are based the laws of marriage and descent. A Murri may marry Buta of the same totem, and of any other totem he may take a Mata though she bears the name of his own sisters, who are all Mata. So Ippai Dinoun [emu] may marry Ippata Nurai [blacksnake] but not Ippata Dinoun. But Ippai Dinoun may marry Kapota Dinoun." " Children always bear the second name (or totem) of their mother, and the first name of the child depends on the mother's." The effect of this is that (1) Between persons of the same " class " name the totem makes the only bar to marriage, and the "class" name does not bar marriage. A man may marry a woman of his own "class" provided she be not of his totem or clan. And, within the " class," the law is simple exogamy. 1 Mr. Ridley originally said that the Ippai were the aristocratic caste. Afterwards (Journal of the Anthropological Institute, ut supra) lie stated that, in some places, people affirmed that the Ippai were the highest class ; that, in other places, the first place was given to Kumbo, but that the most trustworthy witnesses said that the Murri stood first, the Kumbo next, the Ippai third, and the Kubbi lowest of all. Did his interlocutors and he understand each other ? There appears no reason why one class should be esteemed higher than another. xv AUSTRALIA 297 (2) Between the persons of different " class " name who are free to intermarry the totem makes no bar to marriage ; such persons may marry though they are of the same totem or clan. And, in this case, the law is not exogamy. To set forth the second point was, as appears above, Mr. Kidley's object in this communication. He has conveyed it by means of illustrative instances. Murri, he says, may marry Buta of the same totem. Ippai Emu may marry Kapota Emu. If, however, the table on p. 295 be looked at, it will be seen that there is no Buta of the same totem with any Murri, and that Kapota Emu does not exist. And, unless Mr. Eidley's previous statments were wrong altogether, the marriage between persons of different " classes " which he says are permitted, are all marriages which exogamy would permit. For there is no Murri of the same totem with any Buta, no Ippai of the same totem with any Kapota, no Kumbo of the same totem with any Mata, no Kubbi of the same totem with any Ippata. The totem can make no bar to marriage between these pairs, because the man and woman in each are never of the same totem. And there is nothing to make one doubt (for Mr. Eidley's statement to the contrary is evidently the result of an oversight) that in this case, as in the case of marriage within the " class," the law is simple exogamy. Mr. Eidley, no doubt, had in his mind the fact that there was no restriction on the marriage of Murri and Buta, of Ippai and Kapota, and so on. His first impression, formed in ignorance of the Kamilaroi totems, that the " caste " or " class " names had been devised for the regulation of marriage remained with him. And the totem, in fact, made no bar to marriage in these cases. Then he happened not to think of the distribution of the totems. And accordingly he concluded that they were in these cases overridden by the " class " that " class " was everything and totem nothing in the marriages of two " classes " which were free to intermarry. But it is quite clear that he came to this conclusion through inadvertence, and that all the marriages which are certainly permitted among the Kamilaroi are marriages between persons of different totems. Mr. Eidley's statements still leave each " class " cut off 298 STUDIES IN ANCIENT HISTORY CHAP. from marriage with two other "classes." Ippai may marry Ippata, and he may marry Kapota; but he may not marry Buta nor Mata. Murri may marry Mata, and he may marry Buta, but not Kapota or Ippata. And so on in the other cases, every man being cut off from the " classes " to which (according to Mr. Eidley) his wife's mother or his daughter may belong. In a case where the facts have so slowly been disclosed, so far as yet disclosed, in their true light a case in which they were at first so thoroughly misunderstood in essential features one must use discretion as to accepting the form in which for the moment they are cast. And that the facts in regard to the restrictions just spoken of are now fairly before us may most reasonably be doubted. Tor there is nothing to show that Mr. Eidley has ever made such inquiry about them as (circumstances having prompted him thereto) he made in regard to the totems and to marriage within the " class " with the result of shattering his own earliest statement. For anything that has appeared he has, as regards those restrictions, been content with his first impression. And it may well be that, on further inquiry, they will dwindle into rules or arrangements for restraining marriage between persons of different totems but of close family connection father and daughter, uncle and niece, son- in-law and mother-in-law, and perhaps also brother and sister. It is even highly probable the whole history of the evidence considered that something like this will yet be found to be the basis for Mr. Eidley's statement ; that, subject to some restraint of the kind suggested, Murri (to take one example) may marry Kapota, if of a different totem, as freely as he may marry Mata, if of a different totem ; and, by con- sequence, that the " class " names have nothing to do with the marriage law, and that that law is simply exogamy qualified by the restraint aforesaid. It is no reflection on Mr. Eidley to think he has been misled by appearances at this point, as he was at other points as to the " class " name being determined by the father instead of by the mother ; as to its being in virtue of an exclusive privilege, contrary to xv A USTRALIA 299 law, that some Ippais could marry some Ippatas, whereas we know now that any man in any " class " may marry a woman of his " class " if of a different totem ; as to the Ippais being the privileged and most important " class," which he now says the Murris are; as to the importance of the totems, over- looked by him altogether when he first formed his view about the " classes " and Kamilaroi marriage. And, however that may be, unquestionably every student, on the contrary, must feel grateful to Mr. Eidley for his perseverance in investiga- tion, and for his frankness and promptitude in rectifying errors of observation. The wonder is, looking to the circumstances including limited knowledge of the language under which his inquiries were conducted, as we see them disclosed in his journals, that he has been able to ascertain so much. 1 1 It was a Mr. Lance who first told Mr. Ridley of the Kamilaroi marriage law as he originally explained it. Mr. Lance, however, was aware (as appears from a subsequent communication made by him to Mr. Fison) that he did not thoroughly comprehend the matter. He had found (as appears in this communication) that the marriage system, as he conceived of it, was " crossed and complicated " by arrangements which he did not understand. Every black fellow was called after some animal, " implying some incomprehensible relationship " Kumbo being as a rule an emu, Ippai a blacksnake, Kubbi an iguana. And he had sometimes come across a man and wife whose names were not suitable (according to the impressions he had got) for the connection ; and, on inquiring how this could be, they would reply, " This Ippai is not a blacksnake, as most Ippais are, but an opossum ; that explains it." Something of this (very little, however) appeared in Mr. Ridley's first account of the Kamilaroi. Mr. Ridley thought he had come upon a division into castes, and that it was evidence of " an extinct and long- forgotten civilisation" which had existed in Australia. But had Mr. Lance or he been acquainted with the facts of totemism, should we ever have heard of Kamilaroi castes or classes ? That marriage was permitted between persons of the same " class " name who were not of the same totem was unmistakenly suggested by what Mr. Lance had noticed. That at least. He had at first magnified the fact that certain Ippais could not marry certain Ippatas, certain Murri certain Matas, into the much greater fact that no Ippai could marry any Ippata, no Murri any Mata, and so on. What is suggested in the text is 300 STUDIES IN ANCIENT HISTORY CHAP. Since the above was written, a further statement on the subject just dealt with has appeared in the February number of the Journal of the Anthropological Institute for 1878. This is a communication from Mr. Eidley to the Colonial Secretary for New South Wales, dated in July 1873, and it contains, inter alia, Mr. Eidley's account of information supplied to him by a Mr. Honery, respecting the Wailwun tribes who live near the junction of the Namoi river with the Barwan. It is much to be regretted that what Mr. Honery wrote has not been given as he wrote it, so that we might judge at once of the man and of his statements. These Wailwun tribes have the same family names (male and female) as the Kamilaroi. But the persons having these names are all divided, according to Mr. Honery, into " Kang- aroos, Emus, Brown-snakes, and Opossums." There are there- fore four classes of Ippai, namely [dropping the native names], Ippai Kangaroo, Ippai Emu, Ippai Brown-snake, and Ippai Opossum ; and so of the others, making sixteen classes of men and sixteen of women. that he made the same error in other cases, and that Mr. Ridley's inquiries never extended to these cases. (He inquired as to the totems, and as to marriage within the "class," at the instance of Mr. Fison.) Passing that by, however, had it been seen that marriage was permitted, subject to the totem restriction, between all people of the same name, the idea of such people being a caste must forthwith have been rejected. And the idea of their being a "class" appointed to intermarry with another "class" could never have been entertained. That, with exogamy for marriage law, they in other respects married one another freely, would have been enough to exclude this ; marriage among savage or bar- barous peoples is scarcely ever found freer. And to this by and by would have had to be added the fact that the people with whom they were known to intermarry without any restriction the supposed "class" with which they intermarried were all people whom exogamy allowed them to marry all people not of their totems. The theory of the Kamilaroi being in " intermarrying classes," which has been maintained by Mr. Fison and Mr. Morgan, has been founded on Mr. Lance's (no doubt very pardonable) misapprehension. It has appeared above that Mr. Lance's knowledge of the Kamilaroi was very limited, and that he was wrong in his facts at every point at which they have been tested. xv A USTRALIA 301 " When tribes go to war," this statement proceeds, " each carries its own representative animal stuffed, as a standard." " According to Mr. Honery, the only rules observed as to marriage and descent are these two : that a man cannot take a wife of the names corresponding with His own, and that parents may not give their children their own names. Thus Murri Opossum may not marry Mata Opossum, but he may marry Mata Brown-snake, or Ippata Opossum, or any woman except Mata Opossum. Ippai Brown-snake may marry any woman but an Ippata Brown-snake. The children of the Opossum and a Brown -snake must be either Kangaroo or Emu. It is likely enough that in some families the rules are more or less relaxed. The two rules above given are carried out in the more complete system which has been described in former reports. Mr. Honery also states that brothers and sisters have different animal names. Thus all the brothers of Ippai Brown-snake are also Ippai Brown-snake ; but his sisters are not Brown-snake, though they are all Ippata. Sometimes the brothers are Ippai Brown-snake and the sisters Ippata Opossum. "When Ippai Brown -snake marries Kapota Kangaroo, their children are Murri Opossum and Mata Emu ; when Kumbo Emu marries Mata Opossum, their children are Kubbi Brown-snake and Kapota Kangaroo." It is needless to say that we have here a fresh addition of perplexity to this already over -perplexed subject. This statement (1) represents the totems of the Kubbis and Kapotas, Murris and Matas, as being the same as those of the Ippais and Ippatas, Kumbos and Butas. At least it is silent as to their being different. (2) It represents a Murri as being as free to marry a Kapota of a different totem, as he is to marry a Buta or Mata of a different totem. (3) It repre- sents the children as taking the totem neither of their father nor of their mother, but, in accordance with some rule not stated, the one or the other, according to sex, of the other two totems known in the group. It is scarcely necessary to say I attach no value to this statement, not even where it supports one of my own surmises. It will be seen that Mr. Ridley himself has attached no 302 STUDIES IN ANCIENT HISTORY CHAP. importance whatsoever to it. His letter in Nature, dated a year after he had sent in this statement to the Colonial Office, contradicted it in every leading point. The conclusion to be arrived at on the evidence is that the Kamilaroi tribes and the tribes related to them, like the Australian tribes that had previously been noticed, consist of totem families or clans ; that female kinship prevails, so that children are of the clan or totem of their mother ; and that the marriage law is exogamy qualified if at all qualified in some way not yet ascertained by limitations on the marriage of persons nearly connected by family or blood, though not of the same totem or generally acknowledged blood connection. A brief account of the Brisbane district to the south of Moreton Bay (Queensland) has been given by Dr. Lang, 1 and it contains one point which must not be passed over. This throws light on some observations casually made by Eyre on a practice which he found among some tribes in the Adelaide district. Many of the Adelaide tribes, Eyre tells us, have no tattooing, but they are marked on the breast by singular- looking scars, occupying a space of six or eight inches each way upon the chest. These are called Renditch, and are made by fire. They are confined to particular tribes ; i.e. all tribes have not got them. Eyre had no opportunity of inquiring into their origin, but he expresses his opinion that they could not have been accidental. If we can trust Dr. Lang, these scars are the emblems or totem marks of the clans to which those having them belong. Tattooing, he explains, is unknown to the black race of the Western Pacific, to whom it would be of no service as an ornamant, from the colour of their skins. " But in lieu of this process, they make those singular scars which, although un- known among the lighter race, are universal among the aborigines of Australia." " The aborigines of Australia," he says at another place, " never mention the name of a deceased native, and they seem distressed when any European happens to do so ; but at Moreton Bay they usually carve the emblem 1 Queensland, Australia, by J. D. Lang, D.D. London, 1861. See pp. 316, 367, and 337. xv A USTRALIA 303 or coat of arms of the tribe to which he belonged on the bark of a tree close to the spot where he died. The first of these affecting memorials of aboriginal mortality which I happened to see, was pointed out to me near Breakfast Creek by Mr. Wade, on our return to Brisbane from the Pine Eiver. The rain was pouring down in torrents at the time, but I imme- diately reined up my horse to the tree, and remained fixed to the spot for a few minutes, till I fancied I could identify the rude carving on the bark with the raised figures on the breasts of the aboriginal tribe of the Brisbane district." This, if it can be trusted, is very interesting, and Eed American all over. Nothing could prove more clearly the structural importance of the totem. Dr. Lang tells us nothing further, however, of Australian totemism, marriage law, or kin- ship, except that marriage is generally contracted " with the consent of the relatives of the parties and the sanction of the tribe, and is never contracted between near relatives." To resume. We have now found in every district of Australia of which we have distinct accounts that the local tribes are composed of clans on the totemic principle. The accounts cover the whole west coast, the whole south coast, the east coast as far north as the Bay of Carpentaria, and the central portions of the continent, so far as explored sufficiently to be reported upon. In some cases, no doubt, the evidence goes no further than to suggest the totemic composition of the local tribes. Again, in the great majority of cases we have found the marriage law to be exogamy pure and simple, accompanied by the system of kinship through women only. In one case we have seen exogamy and totemism along with male kinship, with the usual result of the clans being drawn away from one another and localised more or less completely. Lastly, in all cases in regard to which we have information, we have seen that people of the same totem were bound together for common action for the redress of injuries; in fact, that the obligations of blood followed the totemic tie which thus was stronger than either the tie of family or the obligation which a man owed to his local tribe. 304 STUDIES IN ANCIENT HISTORY CHAP. THE KAMILAROI NAMES AND THE CLASSIFICATORY SYSTEM I have already quoted Mr. Darwin's remark that a false argument, however seemingly strong and convincing, does no harm as compared with a false fact. The systems of doctrine which men build up will surely be sifted and examined by other men, and, if false, will perish under criticism. But the false fact, put forward on what seems excellent authority, defies criticism. Because it cannot be included in a system with genuine facts, it arrests speculation ; or it may happen that it is itself made the special subject of speculation, which, having no basis in reality, cannot but be futile and pernicious. Now, in Mr. Eidley's original statement about the Kami- laroi, there were two leading false facts. The one was as to the marriage law. It represented Murri as bound to marry Buta and no other ; Kubbi as bound to marry Ippata and no other, and so on. The other was as to the relation of the father to the children. It was Mr. Eidley's first impression that the "caste" of the children depended upon that of the father. That would indeed have followed from the marriage law if that law had been as he conceived it. If Murri could marry only Buta, his children would always be Ippai and Ippata, and so on. But now we know for certain that he may also marry Mata, and that when he does so his children will be, not Ippai and Ippata, but Kubbi and Kapota. And if, as seems most probable, he can marry both Kapota and Ippata when of different totems from himself, it will be seen that his children may be of any "caste" or "class." In fact, we now know that the " class " and the clan are both taken from the mother. Upon Mr. Eidley's original statement, thus unwittingly false, the Eev. Lorimer Fison has founded a proof of the prevalence among the Kamilaroi of the Tamilian form of what Mr. Morgan has called the classificatory system of relation- ships, and his conclusions have been endorsed by Mr. Morgan, who has further discovered in the statement a proof of ancient xv A USTRALIA 305 wholesale communism of men and women ; in short, of the reality of his " communal family." x A single example of the process of discovering the Tamilian system among the Kamilaroi will suffice. Here it is : " Tamilian Characteristic. I, being male, the children of my brothers are my sons and daughters, while the children of my sisters are my nephews and nieces; but the grand- children of my sisters, as well as those of my brothers, are my grandchildren. " Take any male Kubbi. " (a) I, being male, am Kubbi. My brother is Kubbi. His son is Kumbo ; but Kumbo is my son ; therefore my brother's son is my son. " So it may be shown that his daughter is my daughter. " (b) My sister is Kapota. Her son is Murri. But Murri is not my son, for my son is Kumbo ; therefore my sister's son is my nephew. " So it may be shown that my sister's daughter is my niece. " (c) My grandsons are Kubbi, son of my son Kumbo, and Ippai, son of my daughter Buta. My sister's (Kapota's) grandsons are also Kubbi and Ippai, sons of her son Murri and daughter Mata respectively. But Ippai and Kubbi, as already shown, are my grandsons. In like manner it may be shown that her granddaughters are my granddaughters. Therefore my sister's grandchildren are my grandchildren." By similar reasoning all the Tamilian characteristics are discovered in the Kamilaroi " classes." It is almost certain that the classificatory system in some form, if not in several forms, exists in Australia, but Mr. Fison's proof of it is based upon misapprehensions, and therefore goes for nothing. The extraordinary thing is that this argument should have been published after it was well known that Ippai's right to marry Ippata was not an exceptional privilege, and that his son was not necessarily Murri, nor Kubbi's son, Kumbo ; after it was well known, in fact, that 1 This is named in Morgan's last work, Ancient Society (after having been expanded to the dimensions of a tribe or body of kinsfolk herding together), the Consanguine Family. X 306 STUDIES IN ANCIENT HISTORY CHAP. the whole basis of the argument was erroneous. It will be found that to the facts for which Mr. Eidley was primarily responsible others are added, by assumption, for which Mr. Fison and Mr. Morgan are alone responsible. It is assumed that all Kubbis are brothers, all Kumbos brothers, and so on ; that every Kapota is sister of every Kubbi, and every Buta sister of every Kumbo, and so on. But Mr. Eidley's state- ments, made after his visits to the Kamilaroi, justified no such assumption ; and we know from him that Kubbis, Kumbos, and the others are found in several distinct totem clans. Mr. Eidley had overstated the fact looking to the existence of polygamy in saying that all the sons in the same family are, say, Kubbis, and all the daughters Kapotas. But of any brotherhood or sisterhood between all the Kubbis or Kapotas comprised in all the different families in the districts inhabited by the Kamilaroi,'' he gave not a hint. No thought of this seems even to have entered his mind till he had seen Mr. Fison's argument. But that unauthorised assumption is as vital to Mr. Fison's argument as the fact that Kubbi must marry Ippata and no other, or the fact that his son must be Kumbo and no other. 1 1 See Australian Kinship, Pro. Amer. Acad. Arts, etc., vol. viiL p. 412. Fison and Morgan. In Kamilaroi and Kurnai (by the Rev. L. Fison and A. W. Howitt, Melbourne, 1880) Mr. Fison has not repeated the attempt to deduce the Tamilian terms from facts taken from the Kamilaroi. In this work he has substituted for the Kamilaroi a tribe the constitution of which, and its system of naming, are different from those of the Kamilaroi (and about which less is known all that has ever been heard of it being stateable in a sentence or two) a tribe which is said to be in two divisions, each comprising several totem clans which do not intermarry with each other, while intermarrying freely with the clans in the other division; and in which children take a divisional name (of which there is a form for males and a form for females), as well as the totem name, from the mother. He applies the term class in this work, primarily, to the divisions of such a tribe. In this sense, it is prima facie altogether inapplicable to the people of the same name the Murris, Ippais, etc. among the Kamilaroi, and Mr. Fison recognises this; but he endeavours, by means of a theory, to show that it is applicable. His theory, however, proceeds upon the same basis of error as his argument xv AUSTRALIA 307 Whatever the Kamilaroi names mean, they are obviously unrelated to the classificatory system. All the members of these tribes, as Mr. Eidley has told us, have three names one, which is individual and of no importance, for example, Eed feather ; one, which denotes the clan or totem, about which there is nothing unusual ; and a third, that which has been called the " class name," which belongs to a man or woman in respect of the so-called " class " of his mother. Now this " class " name is not of the nature of a term of address such as is applied to persons of the same class in the classificatory system. It is not relative, that is, but absolute ; and attaches to the individual for his life as an essential part of his designa- tion. A man is not Murri to one person, and Kubbi or Kunibo to another. He is once for all, and to all comers, Murri by name, as if he were by name John or Peter. But in the classificatory system every appellation is in its nature relative, not absolute is not a name at all, in short, but a term expressive of relation. The person who in the Hawaian system is Makua Kana to me his Keiki, is Kupuna to my son, who to me is Keiki, and to whom in turn I am Makua noticed above. He again treats the mistaken impression which Mr. Eidley received from Mr. Lance as if it were the truth about Kamilaroi marriage. For the purpose of deducing the Tamilian, or Turanian, terms, Mr. Fison now subdivides his class or division into groups or classes consisting of all the men and of all the women of the same generation ; and he assumes that all the men and women of the same generation in each division are brothers and sisters. He ignores the descent through the mother in existence of totem groups, and the few other facts stated about his tribe, except that of its being in two divisions which may intermarry with each other, and within which marriage is forbidden. He assumes also that marriage is (or was in the hypothetical state upon which he reasons) confined to the tribe, which is by no means the case now. And, within the tribe, the group or class consisting of all the men of the same generation in one division marry the group or class consisting of all the women of their generation in the other division, and are debarred from all women younger or older. What he tries to show is that the Tamilian or Turanian terms would embody the marriage relations which he has supposed to exist in his hypothetical tribe. 308 STUDIES IN ANCIENT HISTOR Y CHAP. Kana ; while in relation to one whom I call father, my son is not Keiki, but Moopuna. In fact, the Kamilaroi names are true personal names, like Tom, Dick, or Harry, while the classificatory names belong to classes only as relative to other classes above or below them, so that to an elderly person with a relative surviving in an earlier generation, every class name in the whole gamut of names must in turn be applied by persons of the different classes in addressing him. But, further than that, the names Murri, Mata, etc., do not belong to persons who fall into classes, properly so called, as names do in the classificatory system, which appropriates a common term to all of one generation in relation to all of some other generation. The Kamilaroi names alternate in successive generations in houses not specially connected with each other by blood or marriage, so that the same name belongs to persons related in every conceivable way or even unrelated, except by their being of one or other of several clans scattered over a wide area. Let us consider what the state of a single household may be. Murri Opossum marries Mata Iguana and Mata Paddy- melon. The children are Kubbi Iguana and Kapota Iguana, Kubbi Paddy -melon and Kapota Paddy -melon. At the same time, Mata Opossum, sister of Murri Opossum, marries it does not matter whom and has children, Kubbi Opossum and Kapota Opossum. There are now possible six legitimate marriages (or sets of marriages), Kubbi marrying Kapota, and Kapota marrying it matters not whom, which would yield a series of Murris and Matas of all the three totems which have been mentioned; and these Murris and Matas again may similarly, through their marriages, be followed by a series of Kubbis and Kapotas of the same three totems. Girls marrying, as they do among the Australians, before the age of puberty, there may be alive of one household at one time persons of five generations, all in each generation being alternately named Murri and Mata, and Kubbi and Kapota. Let us see which of those persons will be Murri and Mata by name. They will be : (1) The children last born in the xv A USTRALIA 309 house ; (2) the grandparents, and grand uncles and aunts of these children ; (3) the great-great-grandparents, and great- great-grand uncles and aunts of these children. The Kubbis and Kapotas again will be (1) The parents and uncles and aunts of the children last born in the house ; (2) the great-grand- parents, and great-grand uncles and aunts of these children. That is, those named Murri and Mata, instead of being a class, in the sense of the classificatory system, comprise persons related as grandparents to grandchildren, and in a variety of other ways, besides persons related as brothers and sisters, as first cousins, and so on ; and those named Kubbis and Kapotas also comprise persons necessarily related in a great variety of ways. It is, however, absurd to argue the matter further. The persons having those names not only do not form classes, but it is most extraordinary that they should have ever been regarded as being in classes. Still more extraordinary is it that they should have been regarded as forming castes. Mr. Eidley's misapprehensions as to the Kamilaroi may in some sort be illustrated by a supposition as to the errors into which an investigator might have fallen who visited the High- lands of Scotland two hundred years ago, and reported on the family system there prevalent on information which he had gleaned from persons who knew Gaelic only very imperfectly. He might have gone from glen to glen and found in every household (that had a sufficiency of members) such names for males as, say, Donald, Sandy, Malcolm, and Kory, and no other; and such names for females as Kursty, Kate, Maggie, and Mary, and no other. 1 He might have found those names in every home over large districts ; and, if fanciful, might suppose that this implied some systematic division of the people. Going back to clear his ideas, he might at a second visit have ascertained that the persons holding those names were subdivided in a peculiar manner. He would certainly be 1 He might have been long at home in a glen without knowing that any man or woman had any other name than one of those mentioned, unless it might be a name taken from a colour, e.g. Donald Roy (Red Donald), Sandy Bain (White Alick). 3 io STUDIES IN ANCIENT HISTORY CHAP, xv told that the Donalds and Sandys were not all of one clan, and might gather that they were subdivided into Macdonalds, Grants, and Erasers ; while the Eories and Malcolms were, in most districts, subdivided into Macleods, Mackenzies, and MacLennans. Every one will see how absurd that sort of misapprehension would be. Yet truly apart from the in- formation which misled him into thinking the Kamilaroi names connected with the marriage law Mr. Eidley's misapprehensions were of the same order. Eemembering that Kubbi, Murri, etc., are truly personal names, like Tom or Jack, or Tomson or Jackson, it is hardly possible to maintain one's gravity on its being discovered that the Murris are subdivided into three clans, with totems X, Y, Z, and the Kumbo also into three, with the totems P, Q, E. Of course the Kamilaroi naming system as described is most peculiar, and such as might from its nature prepare one for surprises. Looking to either set of clans, 1 we see that there are two male names for the set, and two female, the latter being manifestly feminine forms of the male names. In this per se there is nothing surprising : what is surprising is that the names appear in all the families of the set. Short of this we could match the system by the Scandinavian system of naming, as Mr. Elton has pointed out. Thus Per, son of Ole, is Per Olesen. His sister Serena is not Olesen by name, but is Serena Olesdatter. In turn Per's son Ole is Ole Persen, and his sister Serena is Serena Persdatter, and Ole and Per being always in use in the family, the successive generations are alternately all Olesens, Olesdatters, Persens, and Persdatters. But how come supposing any connection to exist between these cases all the lines to be covered by Oles and Pers ? I can throw no light whatever on this subject. It may merit investigation, however, and the first step, if it can be taken, should be to ascertain the etymologies of the different names. 1 See p 295. SECTION II AMERICA PLAN OF THE EXPOSITION FOR convenience in handling the evidence as to the organic structure of the native tribes on the Continent of America, I shall divide the continent into sections, and deal with them separately. The sections may be regarded as arbitrary, though they have in some cases been suggested by affinities between the tribes inhabit- ing them ; and the division of the whole area to be examined will thus be seen to resemble that which served Mr. Bancroft so well in his compilation of facts relating to the tribes of the Pacific States of North America. Could the tribes have been dealt with in natural groups of sufficient magnitude, it would have been better ; but such groups are not to be found. It will be seen, however, that within the several areas the orderly presentment of the facts has been facilitated by giving them as relative to the tribes in natural groups where they can be found. The sections I have made of the continent are as follows : I. NORTH AMERICA, EAST AND SOUTH. This section includes the whole of North America lying east of a line which is roughly given on the map 314 STUDIES IN ANCIENT HISTOR Y by the course of the Mississippi from the Gulf of Mexico northwards to its source ; thence northwards by the course of the Eed Kiver to Lake Winnipeg ; thence northwards in the line of the lake till it cuts the course of the Missinippi (or Churchill) Kiver, and thence along the course of that river to its mouth in Hudson's Bay ; but the Esquimaux portion of Labrador is excepted from the section. II. NORTH AMERICA, WEST. This area comprises so much of British America and United States territory as lies between the Pacific, west of the rough northernly line of the Mississippi as above described, and south of the parallel of N. latitude (about 57), which passes through the point where the Missinippi enters Hudson's Bay. This section includes, of course, British Columbia and a great portion of Kupert's Land. III. NORTH AMERICA, NORTH. This area comprises the whole of North America north of the parallel of latitude just mentioned, and not included in Section No. I. It includes the Esquimaux portion of Labrador. IV. MEXICO AND CENTRAL AMERICA. V. PERU. VI. SOUTH AMERICA, EXCLUDING PERU. In dealing with Sections IV. and V., I shall endeavour to combine modern and ancient accounts, so far as I know them. As to South America, other than tne Peruvian section, the information I possess, relating chiefly to the tribes on the Orinoco, Amazon, and those inhabiting the peninsula south of the Bio Negro, can quite conveniently be presented in block. CHAPTER XVI NORTH AMERICA, EAST AND SOUTH I. THE SOUTHERN NATIONS ALL the tribes within this area may be considered as having been comprised in four main groups the Muscogee, the Cherokee, the Iroquois, and Algonquin. These, again, may be taken together in sets of two, as follows : 1. The Muscogees and Cherokees may together be considered as the Southern nations. 2. The Iroquois and Algonquin may together be considered as the Northern nations. Tribes or bands of tribes foreign to these were found within this area about A.D. 1600, but they have either become extinct without any account of their customs being preserved ; or they may for our purposes be described in connection with one or other of the groups above mentioned ; or they represent tribes the main bodies of which will hereafter have to be described as belonging to some other area. Numerous local tribes of Florida, for example, seem within the period of our 316 STUDIES IN ANCIENT HISTORY CHAP. knowledge of the peninsula to have been exterminated by the Seminoles. We know nothing of their structure. The Natches, again, and other small nations of which we have accounts, though speaking languages radically different from the Muscogee, may yet be dealt with along with the Muscogees with whom they were in confederacy. Lastly, various local tribes or bands of the Dacotah stock, found east of the Mississippi, will be more conveniently examined along with the Dacotahs as a whole. And they on the whole lay west of the Mississippi. The Cherokees, and the MuscogeesJ and tribes united with the latter by blood, or in the Creek Confederacy, constitute then together The Southern Nations of this part of North America. These nations had as terri- tory the whole of Florida, Georgia, Alabama, Missis- sippi, Tennessee, and parts of North and South Carolina. The Cherokees held the northern and north- eastern portions of this territory ; the rest of the area was occupied by tribes which, so far as we are concerned, were related to, in league with, or one with the Muscogees. Beyond this, geographical detail would, I think, rather embarrass the mind than clear ideas. The accounts we have of the organic structure oJ the local tribes or nations of this area, as it has been usual to call them, apply, broadly speaking, to all oJ them. We shall take them, therefore, together. " The Cherokees," says Schoolcraft, " do not appear to have put forth any branches, and have come down xvi NORTH AMERICA, EAST AND SOUTH 317 to our times as a distinct people." It was not so with the Muscogees. In confederacy with them, or related to them by blood and language, we have the following tribes : 1. Hitchittees 4. Coosadas 2. Uchees 5. Seminoles 3. Alibamons 6. Natches Of these the Seminoles were pure Muscogee, while the Hitchittees spoke a dialect of Muscogee. Of the same race were the Yamasses and Catawbas, who are now extinct, and respecting whom but little information has been preserved. We learn, again, from Dr. Gallatin that the Chickasas and Choctaws were but two nations of one stock ; and from Mr. Schoolcraft, that the Choctaw and Muscogee were radically one and the same language, while he confirms what Gallatin says of the sameness of the Chickasas and Choctaws. " The Chickasas," he says, " are a scion of the Choctaws, as the Seminoles are of the Muscogees." On the whole, then, we seem entitled to add to the list of tribes in the Muscogee connection, now so recently existing, the following local tribes : 7. The Choctaws 8. The Chickasas. Of the numbers in all the southern nations, about 1835, we have an estimate made by the United States War Department, as follows : 318 STUDIES IN ANCIENT HISTORY CHAP. The Cherokees 15,000 The Choctaws, 18,500 } 24 Q00 The Chickasas, 5,500 J The Muscogees, Seminoles, and Hitchittees . 26,000 The Uchees, Alibamons, Coosadas, and Natches 2,000 Dr. Gallatin has inferred, from an examination of De Soto's marches, that these nations were located in De Soto's time much where they were when he wrote, and, on all the evidence, that till lately their habits and customs were much the same as they were at that earlier time. 1 Now we have from Dr. Gallatin "a full and most interesting account of the structure of society through- out the whole of this series of local tribes or nations. He says that, independently of political or geographical divisions, they were all divided " from time immemorial " into families or clans. " At present, or till very lately, every nation was divided into a number of clans, varying in the several nations from three to eight or ten, the members of which respectively were dispersed indis- criminately throughout the whole nation. It has been fully ascertained that the inviolable regulations by which those clans were perpetuated amongst the southern nations were, first, That no man could marry in his own dan; secondly, That every child belongs to his or her mother's clan. Among the Choctaws there are two great divisions, each of which is subdivided into four clans ; and no man can marry in any of the four clans 1 Arch. Amer., Cambridge, U.S. 1836, vol. ii. pp. 94, 98, and 101 ; Algic Researches, vol. i. pp. 13 and 14. New York, 1839. xvi NORTH AMERICA, EAST AND SOUTH 319 belonging to his division. The restriction among the Cherokees, the Creeks ( = Muscogees), and the Natches does not extend beyond the clan to which the man belongs." 1 This distinct statement by so careful an investigator is conclusive as to exogamy and female kinship prevail- ing universally among the southern nations ; but, oddly, just as Schoolcraft constantly mentions even hints for totems, and never mentions marriage law except in a footnote once, so it is in a footnote, and quite casually, that we learn from Gallatin that any of these southern clans had totems for their emblems. In explaining the operation of certain checks put, as he fancied, on the operation of the law of blood-feud, it appears inci- dentally that some Muscogee clans were named after the Wolf, Tiger, Bear, etc., and in a note (I.e. p. Ill) that the seven clans into which the Cherokees were divided were " the Deer, the Wolf, etc." From Mr. Schoolcraft's work, however, we can complete the list, at least for the Muscogees. Their totems were the Tiger, Wind, Bear, Wolf, Bird, Fox, Boot, Alligator, and Deer. The Choctaw totems I find nowhere mentioned, but we may believe one of them was the Deer, if we may trust Bernard Romans (cited by Gallatin), " that one-half of the Choctaws have never killed a deer during their lives." Of the totems of the Natches, again we have information which is at once partial and casual. All we know is, that they were divided into four clans, and that the totem of the chief clan was the Sun, and a 1 Arch. Amer., vol. ii. p. 109. 3 2o STUDIES IN ANCIENT HISTORY CHAP. note of totemism never to be lost sight of that the members of this clan claimed to be descended from the Sun, like other clansmen from their totem, whatever it may be. They were Suns, and their chief was the Great Sun. " The principle of clanship or totemism" says Mr. Francis Parkman, " existed in full force among the Natches, combined with their religious ideas, and developed into forms of which no other example equally distinct is to be found." The same, he says, was the case with another people (now extinct), the Taensas. With both the chiefs of the Sun clan had the attri- butes of demigods. "As descent was through the female," he continues, " the chiefs son never succeeded him, but the son of one of his sisters ; and as she, by the usual totemic law, was forced to marry in another clan that is, to marry a common mortal her husband, though the destined father of a demigod, was treated by her as little better than a slave. She might kill him if he proved unfaithful ; but he was forced to submit to her infidelities in silence." Beyond this we have no direct information as to the totems of these southern nations. On much scattered and indirect evidence we might conclude that every clan had its totem, and even in some cases determine what the totems were. But the exposition is not advanced enough to justify such an attempt at present. 1 1 Gallatin, in Arch. Amer., vol. ii. pp. 109-111 ; Hawkins's Sketch of the Creek Confederacy, coll. Georgia Hist. Soc., vol. iii. part 1, p. 69, Savannah, 1848, quoted in Antiquities of the Southern Indians, C. C. Jones, New York, 1873, pp. 14, 67 ; Discovery of the Great West, Francis Parkman, Boston, 1869, p. 279 ; Schoolcraft's Tlie Indian Tribes of North America, xvi NORTH AMERICA, EAST AND SOUTH 321 The general prevalence among the southern nations of clans organised on the totemic principle, with exo- gamy as marriage law, and kinship traced through women only, may be taken as well established on the evidence above adduced. It remains to point out, that all who belonged to the same clan were in these nations bound together by the bond of blood, for common action to avenge injuries or repair wrongs sustained or in- flicted by any of their number. However scattered they might be, this bond united them as if they had never ceased to be the members of one family contained in one household. 1 Dr. Gallatin, indeed, thinks that the object of "the unknown legislator," who arranged the southern nations on the totemic principle, was to prevent or soften the effects of private revenge by transferring the power and duties of revenge from " the blood relatives," whoever they may have been anciently, to a more impartial body, the clan. The notion of society being so strangely constituted for a purpose by a legislative act need not at present be examined. But it illustrates the import- ance of the law of blood-feud among these tribes, that so able a writer should imagine that the very bases of Philadelphia, 1851, vol. i. pp. 275 and 282 ff. The information collected by Schoolcraft is, so far as it goes, confirmatory of that supplied by Gallatin, but adds little to it except as stated in the text. 1 It is not stated whether the clan, for the purposes of revenge and reparation, was confined to all of the same totem within the same local tribe (or nation), or comprised as well all of the same totem in whatso- ever local tribes they were, provided the local tribes were federated or friendly. Most probably the clan for these purposes consisted of all of the same totem and local tribe. Y 322 STUDIES IN ANCIENT HISTORY CHAP. society should have been reconstituted with the view of affecting its operation. 1 " According to the ancient custom," says Gallatin, speaking of the Cherokees and Creeks, "if an offence was committed by one on another member of the same clan, the compensation to be made on account of the injury was regulated in an amicable way by the other members of the clan. Murder was rarely expiated in any other way than by the death of the murderer ; the nearest male relative of the deceased was the execu- tioner ; but, this being done as under the authority of the clan, there was no further retaliation. If the injury was committed by some one of another clan, it was not the injured party, but the clan to which he belonged, that asked for reparation. This was rarely refused by the clan of the offender ; but, in case of refusal, the injured clan had a right to do itself justice, either by killing the offender in case of murder, or inflicting some other punishment for lesser offences. This species of private war was by the Creeks called " to take up the sticks," because the punishment generally consisted in beating the offender. At the time of the annual corn- feast, the sticks were laid down, and could not again be taken up for the same offence." But this rule had exceptions. The Wind clan could take them up four times, and the Bear clan twice for the same offence. A further proof of the working in these southern nations of the law of blood-feud, is to be found in what are unquestionably arrangements for checking the spirit 1 Gallatin, I.e. xvi NORTH AMERICA, EAST AND SOUTH 323 of revenge, I mean the institution of cities of refuge which existed as well with the Cherokees as with the Hebrews, and the clothing of certain persons with such a degree of sacro-sanctity that blood could not be shed in their presence. Temporary asylums were thus found for those who had become subject to the vendetta, and these asylums, by suspending vengeance, no doubt would have in time introduced compensations in money or goods in lieu of blood for blood, which primitive law everywhere demanded. It is not part of Gallatin's statement, that the punishment for an offence could if the offender evaded it be legitimately made to fall on any of the clan to which he belonged. But " the taking up of the sticks " implied this in the case of a refusal of redress. II. THE NORTHERN NATIONS The Iroquois and Algonquin nations are for our purposes the northern nations of the area we are examining. It is not very material to know what, within this area, were the districts they respectively occupied. Their wars were constant, and the borders of their domains constantly shifting. According to an immigration hypothesis framed, so far as I know, by Mr. Schoolcraft, the Algonquins were the first Indian occupiers of the soil east of Mississippi. They crossed that river at a point in the south-west which he thinks can be fixed, and spread their settle- ments far and wide, eastwardly towards the Atlantic, 324 STUDIES IN ANCIENT HISTOR Y and north-easterly to the land between the southern spurs of Hudson's Bay and the Gulf of St. Lawrence. 1 Some time after they were followed by the fierce and predatory Iroquois, by whom many of their tribes were not only conquered but exterminated. Some Iroquois tribes advancing up the valley of the Ohio, which they occupied and named, took up a commanding and central position in western New York, and cut off all communi- cation between the northern and southern Algonquins ; other Iroquois tribes, turning in a south-easterly direc- tion from the Ohio, settled in the Carolinas, to the south of the most southern Algonquin settlements. What we know is that, at the time of our first certain knowledge of them, the more northern Iroquois were surrounded on all sides by Algonquin tribes, and were separated from their more southern congeners by several Algon- quin Lenape tribes that are now extinct. 2 THE IROQUOIS NATIONS The more northern Iroquois nations were the Mohawks, Oneidas, Onondagas, Cayugas, and Senecas, who were known as " The Five Nations " ; the Hurons ; the Neutral Nations ; the Erigas or Eries ; and the Andastes. Of the last three sets we know little except that they were Iroquois, and suffered terribly at the hands of the Five Nations. Of the Five Nations and 1 They found the country already inhabited, if their traditions may be believed, and exterminated and dispersed the earlier inhabitants to make way for themselves (Schoolcraft, vol i. p. 307). 2 Algic Researches, Introduction; Gallatin, ut supra, vol. ii. p. 22. xvi NORTH AMERICA, EAST AND SOUTH 325 of the Hurons, however, we have full and trustworthy accounts. The more southern Iroquois nations were the Tuscaroras, the Toteloes, Nanticokes, Conoys, and Nottoways. In the years 1714-15 the mass of the Tuscaroras, after suffering terribly in war, were admitted as a sixth nation into the League of the Five Nations, and we have a full account of them accordingly. The Nanticokes and Conoys (making one nation), and the Toteloes, seem for a time to have been admitted into the same League, but we have no special accounts of them. Their organic structure may, how- ever, very safely be assumed to have closely resembled, if not to have been precisely the same, as that of the Tuscaroras and the other nations in the League. The Tuscaroras, it may be mentioned, were far the most powerful and important of the southern Iroquois nations. What, then, was the structure of an Iroquois local tribe or nation ? This question is answered most fully for the Mohawks, Oneidas, Onondagas, Cayugas, Senecas, and Tuscaroras, by Lafitau, Golden, and Morgan. I shall hereafter have occasion to refer somewhat minutely to their evidence, but meantime the results are briefly as follows : 1. The local tribes were all divided into clans or families on the totemic principle. 2. No man could marry a woman of his own clan or totem. 3. Every child belonged to the clan of his or her mother, and all the inheritances were determined by the system which traced kinship through women only. As to the totems of the different tribes, the Mohawks and Oneidas had but three, the Turtle, Wolf, and Bear 326 STUDIES IN ANCIENT HISTOR Y CHAP. and in these clans the Sachemships, assigned by the League to these nations, were hereditary. The Senecas, Cayugas 3 and Onondagas had each eight totems, and the Tuscaroras had seven, according to Mr. Morgan. The eight totems were the Wolf, Bear, Beaver, Turtle, Deer, Snipe, Heron, and Hawk, but the existence of all these in any one local tribe may be doubted. The list of Sachemships furnished by Mr. Morgan show that, so far as known, those of the Onondagas were hereditary in the clans of the Bear, Wolf, Turtle, Snipe, and Deer; those of the Cayugas in the clans of the Bear, Turtle, Heron, Deer, and Snipe ; s and those of the Senecas in the clans of the Bear, Wolf, Turtle, Snipe, and Hawk. Thus the Sachemships furnish a proof that seven totems at least existed within the League, while they do not show that more than five totems were to be found in any one nation. It is consistent with this evidence, of course, that there may have been eight totems in some of the nations. If the Beaver clan held, as Mr. Morgan states, one of the Sachemships of the Onondaga, we should have a proof of there having been eight distinct totems altogether within the confederacy, and six in at least one nation. 1 We shall hereafter see reason for believing in the accuracy of the earlier statements we have, to the effect that the three original totems of the Iroquois were the Wolf, Bear, and Turtle. These were, at any rate, the only totems of the Hurons (Wyandots), who early 1 Schoolcraft, Indian Tribes, etc., vol. ill pp. 184 ff.; vol. v. p. 73. Morgan's League of the Iroquois, book i. chaps, iii. and iv. Kochester, 1854. xvi NORTH AMERICA, EAST AND SOUTH 327 separated from the main body of the Iroquois, as they were the only totems of the Mohawks and Oneidas ; and every Iroquois nation, so far as we know, had these three, whatsoever other totems they had in addition. The Hurons, being pure Iroquois, and with clans on the totemic principle, we should expect them to have exogamy as marriage law, and the system of kinship through females only, and this is what we find. 1 Of the Hurons, Parkman says (Jesuits, p. 52) : " The Huron nation was a confederacy of four distinct contiguous nations, afterwards increased to five . . . ; it was divided into clans ; it was governed by chiefs whose office was hereditary through the female," etc. See Jesuit Relations prior to 1650, and Champlain, Sagard, and Bressani. Lafitau and Charlevoix knew the Huron institutions only through others (Jesuits, p. 111). The totem seems to have established among the Iroquois, as among the Cherokees and Creeks, a bond of brotherhood between all of the clan. They were bound to one another alike for obtaining and giving repara- tion for injuries. In cases of murder, the murderer was given up to the private vengeance of the kindred of the slain. They could slay him wherever and when- ever they found him, without being taken to account. It was customary for the kindred of the murderer, and even the " tribe " (? local tribe) to which they belonged, to interfere with efforts to appease the kindred of the 1 Charlevoix, cited by Gallatin, Arch. Amer., vol. ii. p. 189 ; Johnston's Indian Tribes inhabiting Ohio (Arch. Amer., voL i. pp. 271 ff.), and see specially p. 284. Johnston's statements apply to the Wyandots and other Iroquois, and some Algonquins. 328 STUDIES IN ANCIENT HISTORY CHAP. slain. " In a doubtful case," says Johnston, speaking of the Hurons, Senecas, and other tribes inhabiting Ohio in 1819, "or an old claim for satisfaction, the family [? clan] consult the tribe [local tribe], and when they have resolved on having redress they take the guilty, if he is to be found, and if he flies, they take the nearest of kin. In some cases the family who have done the injury promise reparation, and in that case they are allowed a reasonable time to fulfil their promise ; and they are generally quite earnest of themselves in their endeavours to put the guilty to death in order to save an innocent person. This right of judging and taking satisfaction being vested in the family or tribe, is the sole cause why their treaty stipulations never have been executed. A prisoner taken in war is the property of the captor, to kill or save at the time of capture, and this right must be purchased" (Arch. Amer., vol. i. p. 28 2). * Mr. Morgan gives some detail as to the action of the tribes to which the parties belonged among the Five Nations, to appease "the family" of the murdered person, and to induce them to accept a pre- sent of white wampum, not as compensation for the crime, but as a regretful confession of it and prayer for forgiveness. If, he says, the wampum came too late, or was refused, " the family then either took upon them- selves jointly the obligation of taking what they deemed a just retribution, or they appointed an avenger, who resolved never to rest until life had answered for life." There is the same ambiguity in the term " family " here 1 Cp. Lafitau, vol. i. pp. 490 ff. I xvi NORTH AMERICA, EAST AND SOUTH 329 1 as in the previously cited statement of Mr. Johnston as ( to the tribes of Ohio. I think it can mean nothing but j the clan the brotherhood of the totem of the slain. | But may it have meant merely the immediate relatives through the mother of the slain his brothers uterine, mother's brothers, etc. ? We shall see that it would be I contrary to all primitive law thus to interpret a term, used with obvious thoughtlessness as to its precise meaning being important. " In the eyes of an Iroquois," says Mr. Morgan, " every member of his own tribe, in whatever nation, was as much his brother or his sister as if children of the same mother." It would be hard to reconcile with this statement any interpretation of the term " family " in regard to vengeance that fell short of including at least all of the same tribe i.e. totem included within the same nation? THE ALGONQUIN NATIONS These have been arranged by Gallatin under four heads, the Northern, North-eastern, Eastern or Atlantic, and Western. (1) The Northern division of the Algonquins com- 1 See Morgan, Iroquois, pp. 81 and 331 ff. As illustrating the reasoning in the text, let me cite Mr. Warren's account of the Chippeway warrior, Ba-be-se-gun-dib-a. " His totem was a Crane, one of the oldest families in the tribe now residing mostly at Lake Superior." Here " family " clearly means " clan," and tribe means local tribe, band, or nation. Schoolcraft, Indians, etc., vol. ii. p. 162. Another illustration in point occurs in Schoolcraft's Algic Researches, vol. ii. p. 146. " All the individuals of a particular family," says Schoolcraft, speaking of the Algonquins generally, " such as tha Deer, Crane, Beaver, etc., when called 330 STUDIES IN ANCIENT HISTORY CHAP. prised the Crees, the Algonquins, Chippeways or Ojib- ways, Ottawas, Potowotamies, and Mississagues. It is uncertain whether "Algonquin" was not a generic term. (2) The North-eastern division comprised the Algon- quins of Labrador, the Micmacs, the Etchemins, and the Abenakis. (3) The Eastern or Atlantic division comprised the New England Indians, the Long Island Indians, the Delawares and Minsi of Pennsylvania and New Jersey, the Nanticokes of the eastern shore of Maryland, the Susquehannocks, the Powhattans of Virginia, and the Pamlicos of North Carolina. (4) The Western division comprised the Menomonies, the Miami and Illinois tribes, the Sauks, Foxes, Kicka- poos, and, finally, the Shawnoes. I have set forth this extensive array of tribal names that the extent of my ignorance of their social structure may be frankly confessed. The only "nations" of which the works accessible to me have yielded information, after a careful search, are the following : (a) In the Northern division, the Chippeways or Ojibways, and the Ottawas. (6) In the North-eastern division, none. (c) In the Eastern division, the Delawares and Minsi. (d) In the Western division, the Sauks, Foxes, and upon for the signatures, affix their respective family marks without regard to specific names." The family marks here are distinctly stated to "be the totems, and so family = clan. xvi NORTH AMERICA, EAST AND SOUTH 331 Shawnoes. There is, indeed, abundance of information respecting the others, the boundaries of their territories, their wars, their villages and population, and in some cases their alliances, and what may be called their political system. But of the form of the. family, the gentile bond, the blood ties and obligations, I find nothing but general statements or surmises. As an example, take the following as to the Illinois : " The Illinois were an aggregation of distinct though kindred tribes [it is not said whether tribe here means clan or local tribe]. Their general character and habits were those of other Indian tribes," and so on, the account disclosing some of their improper habits, telling us in what direction they traded and sold their captives, and beyond that nothing. 1 In some cases the evidence comes provokingly near to touching the subjects of our inquiry, and suddenly leaves them unexplained. Not- withstanding, it is possible that we have a sufficient number of examples, casually presented, to give us an assurance that all the " nations " in the group were of a type, without pressing inference to the limit ex uno disce omnes. Of the Miamis, we know that they were divided into clans, from a casual mention of their cannibalism, for there was, it seems, a clan of the Miamis whose heredi- tary duty and privilege it was to devour the bodies of prisoners burned to death. 2 Totemism, the division of clans on the totemic prin- ciple, we should expect as common, if not universal, in 1 Parkman, I.e. p. 206. 2 Id. p. xl. 332 STUDIES IN ANCIENT HISTOR Y CHAP. this group. The very name Totem = Dodaim = town- mark, is Algonquin, furnished by the common language of the group, as has been already explained. It may be remarked that it has not been, perhaps, a fortunate circumstance that this term, having such an origin, should have come into common use to denote the emblem of a clan, apart from the common residence of the clansmen in the same village or town. It is like any other term, however, if we allow it to mean no more than it actually does in the run of cases. Up to this point we have applied it to denote the clan, though the members of the clan have been dispersed in a variety of villages and even in a variety of nations. It occurs then, at once, that in coming to the birth-group of the term, we have now reached a case where totemism has some distinctive accompani- ment. What this was is not difficult to guess. We have reached tribes at last where kinship is traced through males, and not through females, and where in consequence, the gentile bond remaining as before, and being no longer counteracted by cross ties of family between the different gentes, the gentes tend to separate and each to assume for itself its own home, village, or town, or one or more villages or towns in vicinity to each other. To comprehend the cohesive force of clanship operating with exogamy as marriage law, and the system of female kinship, let us consider the case of a local tribe like the Mohawks with three totems the Wolf, Bear, and Turtle. A Wolf man marries, say, a xvi NORTH AMERICA, EAST AND SOUTH 333 Bear woman ; his son and daughter are Bears. Suppose, now, his son to marry a Turtle and his daughter a Wolf, in a household in two generations there will be family ties binding together persons of the three clans, and tending to counteract any tendency of the clans to sever through a conflict of their duties and interests such as, we saw, the blood - feud might on occasions bring on. As every household would in its composition resemble every other, would be, in short, a true epitome of the nation, we may see how the family affections, giving stability to the family group, directly resisted the disintegration of the nation. The clans were doubly and triply jointed into one another, and bound together not only by marriage ties, but by the parental, filial, and fraternal affections. However weak we may suppose these to have been, compared with the feelings of gentile kinship, we see their whole force would go to soften clan conflicts when they arose, and support clan cohesion. In fact, when we think of all a separation of clans thus welded together implies, we see it is almost impossible that it should ever take place, or be more than temporary for the accomplish- ment of some definite clan purpose. Mr. Parkman has aptly likened the bonds that united the clans to cords of indiarubber. " They would stretch, and the parts would be seemingly disjointed, only to return to their old union with the recoil. Such was the elastic strength of those relations of clanship, which were the life of the League" (Jesuits, p. 337). Suppose, now, with clanship and exogamy remain- 334 STUDIES IN ANCIENT HISTORY CHAP. ing, that the system of kinship undergoes a change. A Wolf man marries say a Bear woman : his son and daughter now are Wolves and not Bears. When his son marries, his children shall be Wolves, whether he marries a Bear or a Turtle. His daughter must marry a Turtle or Bear. Whichever she marries, her children will be Turtles or Bears, as the case may be, in another household. Every household will now become homo- geneous comprising Wolves alone, Bears alone, or Turtles alone, excepting the mothers, who through the change of kinship have lost importance and go for nothing. When in a course of generations the house- holds of a nation have been transformed, have become homogeneous, should conflicts arise between the clans, there will be nothing to assuage them in the constitu- tion of the family. In the constant shiftings of village settlements, incident to the wandering life of the people, nothing would be more natural than that all the Wolves, Bears, and Turtles, should establish their homes near one another respectively. Whether this separation took the form of wards or quarters in the same town or village, or of separate villages, it is apparent that a condition of things would now have arisen favourable to a final separation of the clans. In " the taking up of the sticks " for redress which was refused, the enemies would find themselves confronted with one another in an antagonism from which there would be no force to withdraw them and resettle them in the nation. A clan beaten in a conflict might take up a residence permanently apart from, though perhaps near to, that xvi NORTH AMERICA, EAST AND SOUTH 335 of the other clans; or even in disgust and enmity separate itself from them for ever, and wander away in search of a new territory. 1 We may see how separation in another form might take place, through bands from the different clans setting out on some expedition under a popular leader, and remaining independent on experiencing success under his leadership. Such conjunct bands would resemble, however, the parent group in their composition, and their settlement in one village or town never could introduce for them a common totem. At least it would so appear prima facie. Such separations must have been frequent among the Algonquins, and the chief cause of their weakness compared with their immediate enemies of the Iroquois stock. They were far more numerous than the Iroquois, but such were their subdivisions that they could no- where oppose a sufficiently solid front to their enemies. That the subdivisions were due to such separations as we have been explaining, appears from the case of the Delawares. They were in three clans the Turtle, Wolf, and Turkey. The Wolf clan separated itself from the Delawares so completely that the Wolves, though maintaining an intimate connection with the other Delaware clans, spoke a different dialect. The Turkey and Turtle clans remained united. From this case we should be bound to infer that the Delawares had male kinship, though the fact were not recorded. 1 The Algonquins have numerous traditions of the separations of clans and bands from their various nations, e.g. Schoolcraft, vol. ii. p. 139. 336 STUDIES IN ANCIENT HISTORY CHAP (a) The Chippeways or Ojibways and Ottawas. Dr. James represents the Chippeways as divided int< numerous clans on the totemic principle, and a list o eighteen of their totems which he had made fell far shor of being a complete list. It is not accessible to me, bu a pictorial Petition presented at Washington in 184 by a delegation of Chippeways from Lake Superior tc the President of the United States, which has beer analysed by Schoolcraft, discloses the following as the totems of the chiefs who formed that delegation namely : 1, Crane ; 2, Marten ; 3, Land -Tortoise ; 4 Black Bear; 5, Man-fish; 6, Catfish; 7, Brant; 8 Subterranean Bear ; 9, Sturgeon; 10, Springduck; 11 Eagle; 12, Loon; 13, Elk. 1 Elsewhere we seem have evidence of other four, viz. : 14, Turtle ; 15, Swan 16, Crow; 17, Woodpecker, and yet want one to com plete 2 such a list as Dr. James had formed. It may b seen in Schoolcraft, casually mentioned, that the Chippe- ways had the Loon the totem of the royal family the Bear, of the family of the war -chief; the Crane and the Marten the latter imported by adoption from the Munduas, whom the Chippeways had nearly exter- minated. 3 That the Ottawas were divided into clans on the totemic principle, appears incidentally in the course oi an Ottawa myth, given by Schoolcraft in his Algic Researches. The old spirit being in need of tobacco, managed to have captured and brought to him a mori 1 Indians, etc., vol. i. p. 419. 2 Ibid. vol. iv. p. 491. 3 Ibid. vol. ii. pp. 138, 139, 140, 141. xvi NORTH AMERICA, EAST AND SOUTH 337 named Wassamo, to whom he gave his daughter in marriage, on the condition of his procuring a proper supply of the weed. Wassamo being allowed to return to his people, disclosed the need of the god, whereupon tobacco for him was immediately forthcoming to such an extent that it took two sacks " of dressed moose skin to hold it." " On the outside of these skins the different totems of the Indians who had given the tobacco were painted, and also those of all persons who had made any request." On this Schoolcraft, in a note, takes occasion to discourse on the meaning of totems, carefully avoiding of course, as usual, all reference to their connection with marriage law, and mentioning the Deer, Crane, Beaver, etc., as totems ; but he does not say that they were totems of the Ottawas, and it does not follow that they were, though mentioned in an illustration of an Ottawa myth. 1 If we may trust Algonquin traditions, given by Schoolcraft, the Chippeways, Ottawas, and Potowotamies formed originally one nation, and so we may hold as true of the Potowotamies whatever we have ascertained as true of the Chippeways and Ottawas. 2 Dr. James is Gallatin's authority for exogamy as marriage law among the Chippeways. " Dr. James informs us," he says, " that no man is allowed to change his totem, that it descends to all the children a man may have, and that the restraint upon intermarriage which it imposes is scrupulously regarded." This is 1 Algic Researches, vol. ii. p. 146. 2 Indians, etc., vol. i. p. 308 ; and vol. ii. p. 139. CHAP. 338 STUDIES IN ANCIENT HISTOR Y confirmed by Tanner (Narrative, p. 313) : "They prefer to consider it highly criminal for a man to marry woman whose totem is the same as his own ; and the; relate instances where young men, for a violation oJ this rule, have been put to death by their own nearest relatives." Johnston's statement regarding the Ottawas is les direct. He says, " A man seldom or never marries his own tribe." 1 As this statement was thought bro; enough to cover exogamy in the case of the Huron* and Seneca Iroquois, whose case it also covers i] Johnston's account, of which we have direct and in- dependent evidence, we must hold it to imply exogam; in the case of the Ottawas. The traditions that these Algonquin tribes had been anciently one peopL point to the same conclusion. As to kinship, Dr. James's statement points out thai children took the totems of their father among th< Chippeways ; and several of their traditions point t< succession of son to father as the rule. It may not, however, have become the rule among all the tribes oJ Algonquin stock. I find no express statement 01 the subject as to the Delawares, but I have already indicated that I infer the Delawares had male kin- ship like the Chippeways. What Johnston says however, of the Ohio tribes, would make one doubt this inference. The tribes he directly deals with are the Delaware!- Wyandots (Hurons), Shawanoese, Senecas, and Otta^ 1 Arch. Amer., vol. i. p. 284. xvi NORTH AMERICA, EAST AND SOUTH 339 of which the Hurons and Senecas are Iroquois, having the system of female kinship, and the rest Algonquin. Now the statement he makes as applicable to all of these is, that in their marriages the brothers and uncles of the woman on the maternal side are consulted as to a proposed match, " and sometimes the father, but this is only a compliment, as his approbation or opposition is of no avail " ; and elsewhere he states that marriage gives no right to the husband over the property of his wife, "and when they part, she keeps the children and the property belonging to them and to her. Not un- frequently they take away everything the husband owns, his hunting equipage only excepted." This is a state of things consistent with kinship being traced through females only, and children being accordingly counted to belong to the tribe of their mother taking of course her totem ; but it is incon- sistent with male descent, and children taking their father's totem. Mr. Johnston was, when he wrote, agent for Indian affairs at Piqua, and his account is official. It comprises a population table (I.e. p. 270), from which we have the Indian population as follows : Men, Women, and Children. Wyandots or Hurons . . . . 542 Senecas . . . . . .551 Mohawks 57 Total of Iroquois stock . . . 1150 340 STUDIES IN ANCIENT HISTORY CHAP. Men, "Women, and Children. Shawanoese . Ottawas (settled and migratory) . 377 Delawares . . . . 8 Total of Algonquin stock . .1257 We have already seen the correctness of his state- ments respecting this population, so far as regards the Iroquois portion of it. There is no doubt as to that. But it must be assumed that the same state- ments were applicable on the whole to the Algon- quin, or major part of the population, and this fact is interesting as showing that some Algonquins who according to tradition, had anciently been united with the Chippeways, continued in Ohio to have female kinship after kinship had shifted among the Chippe- ways to the male side. It yields of course the infer- ence that the Chippeways also had originally th< system of female kinship. But while it is inconceiv able that his account did not apply on the whole to the Algonquin population, it may well have been inapplicable to the "mere handful," as he calls them of Delawares living at Upper Sandusky, of whom there cannot have been more than say sixteen families. There were among them in all but twenty-one women. I do not hold his account, therefore, as conclusive against the inference that the Delawares had male kinship But further inquiry may clear up the doubt. (c) The Delawares and Minsi. We have already seen that the Delawares at first consisted of three clans i NORTH AMERICA, EAST AND SOUTH 341 or tribes, with the totems of the Turtle, the Wolf, and the Turkey respectively, and that they say that the Wolf clan, or Minsi, who were originally mixed up with them in the "nation," separated from the others and became, though remaining intimately connected with them, a separate local tribe or nation the Minsi, or Wolves ; the composite local tribe of Turtles and Turkeys retaining the title of Delawares proper. We might be sure, were there only indirect evidence, that the Minsi intermarried with the Delawares, and that the law of the whole " nation " was exogamy. The fact is, however, not left to inference. It rests on the most direct and distinct evidence. Loskiel is quoted by Gallatin as saying : " The Delawares and Iroquois never marry near relations. According to their own account the Indian nations were divided into tribes for no other purpose than that no one might ever, either through temptation or mistake, marry a near relation, which at present is scarcely possible, for whoever intends to marry must take a person of a different tribe." I have already examined the evidence as to kinship among the Delawares. Does it make for my view or against it, that they were subdivided into numerous small tribes, distinguished by local names ? 2 (d) Of the Shawnoes in the Western division we have an account by Mr. Johnston, who calls them Shawanoese. It is singularly interesting, as introducing us to a man-totem. 1 History of the Moravian Missions, Part I. p. 56. 2 Gallatin, ut supra, p. 46. 342 STUDIES IN ANCIENT HISTORY CHAP. The Shawnoes were in four tribes or clans : 1. "The Piqua Tribe, which they say originated as follows : In ancient times they had a large fire, which being burnt down, a great puffing and blowing were heard in the ashes ; they looked, and behold a man stood up from the ashes ! hence the name Piqua a man coming out of the ashes, or made of ashes." 2. " The Mequachake Tribe, which signifies a fat man filled a man made perfect, so that nothing is wanting. This tribe has the priesthood. They perform the sacrifices, and all the religious ceremonies of the nation. None but certain persons of this tribe are permitted even to touch the sacrifices." 3. The Kiskapocoke Tribe. The signification of the term is unknown, at least it is not given. 4. "The Chillicothe Tribe. Chillicothe has no definite meaning. It is a place of residence." The statements of Mr. Johnston as to the other tribes in Ohio, in 1819, apply to the Shawnoes, and ascribe to them exogamy as marriage law, and the system of kinship through females only. 1 Blood-Feud. Johnston's statement may be taken as giving the law for the Algonquin and Iroquois stocks. It is, as might be expected, the same for both. ? "'If murder be committed, the family [ = clan or 1 Arch. Amer.j vol. i. pp. 275 and 284. 2 In regard to his statement, note his use of the terms nation, tribe, and family, e.g. (p. 273) of the Shawanoese. ..." The people of this nation [ = local tribe] have a tradition that their ancestors crossed the xvi NORTH AMERICA, EAST AND SOUTH 343 tribe of descent] of the deceased only have the right of taking satisfaction ; they collect, consult, and decree. The rulers of a town, or of the nation, have nothing to do or say in the business. The relations of the deceased person consult first among themselves, and if the case is clear, and their family [ clan] not likely to suffer by the division, they determine on the case definitively. When the tribe [ ? nation] may be affected by it, or in a doubtful case, the family [ = clan] consult the tribe [ = nation], and when they have resolved on having redress they take the guilty, if he is to be found, and if he flies they take the nearest of kin. In some cases the family [ = clan] who have done the injury promise re- paration ; and in that case they are allowed a reasonable time to fulfil their promise, and they are generally quite earnest of themselves in their endeavours to put the guilty to death, in order to save an innocent person. This right of judging and taking satisfaction being vested in the family or tribe, is the sole cause why their treaty stipulations never have been executed " (p. 281). sea. They are the only tribe [ = nation = local tribe] with which I am acquainted, who admit of a foreign origin." Again (p. 275) : "All the Indian nations [ = local tribes] are divided into tribes [ = clans, families, or tribes of descent] after the manner of the Jews." Again (p. 280) : "War is always determined on by the head warrior of the town which feels itself to have been injured. ... If ... he goes out he is followed by all who are for war. It is seldom a town is unani- mous : the nation [ = all the clans] never is ; and within the memory of the oldest men among them, it is not recollected that more than one half of the nation have been for war at the same time." 344 STUDIES IN ANCIENT HISTORY CHAP. TOTEMS OF THE IROQUOIS THE IROQUOIS LEAGUE Lafitau, in his work on the Ked Indians, published in 1724, gives us the following tradition of the Iroquois, as to the origin of the earth, and as to their own origin : "This is how the Iroquois relate the origin of the earth and their own origin. In the beginning there were, they say, six men (the people of Peru and Brazil have also the like number). Whence came these men, then ? This they did not know. There was as yet no earth, they wandered about as the wind wafted them. They had also no wives, and they felt that their race was about to perish with them. At last they learned, I know not where, that there was a woman in heaven. Having held a council they resolved that one of them, named Hogouaho or the Wolf, should transport himself thither. The enterprise appeared impossible, but the birds of the air agreed together and raised him up, making a seat for him with their bodies supporting one another. When he arrived, he waited at the foot of a tree till this woman should go out as usual to draw water at a spring near the spot where he had halted. The woman did not fail to come according to her custom. The man who was waiting for her entered into con- versation with her, and made her a present of bear's grease, of which he gave her some to eat. A curious woman who likes chatting and who receives presents is soon overcome. This woman was weak even in heaven ; she allowed herself to be seduced. The master of heaven xvi NORTH AMERICA, EAST AND SOUTH 345 perceived it, and in his anger he turned her out and precipitated her, but in her fall the Tortoise received her on its back, on which the otter and the fish, drawing some clay from the depths of the waters, formed a little island, which increased in size ; by degrees it extended itself into the form in which we see the earth at the present day. This woman had two children who fought together. They had unequal arms, of which they did not know the strength ; those of the one were offensive and those of the other were incapable of doing harm, so that the latter was killed without difficulty. " From this woman are descended all the other men through a long series of generations ; and it is such a singular event which served, they say, as a foundation for the distinction of the three Iroquois and Huronese families, the Wolf, the Bear, and the Tortoise, the names of which are like a living tradition, which keeps before their eyes (recalls to them) their history of the earliest time" (vol. i. p. 93). Here we have a distinct and comparatively early attestation, in Iroquois tradition, to the existence from the first times they knew of, of the tribes of the Wolf, Bear, and Tortoise (or Turtle, as it is more commonly called) among them and the Hurons, and the suggestion that, whatever other tribes were included from the first in the Iroquois nation, these three were the most pro- minent. Indeed, it may be believed, on the analogy of like cases, that if any other totem, other than that of a tribe in a position of thorough insignificance, had existed in the nation at the time this legend took shape, a place 346 STUDIES IN ANCIENT HISTOR Y CHAP. and part would have been found for it to play in the story. Mr. Cadwalader Golden, who, we are informed by Mr. Schoolcraft (vol. iii. p. 195), had often been a com- missioner to the Iroquois during the reign of George II., and was familiar with their history and customs, gave, in 1747, an account of the Iroquois in perfect agreement with that just cited from Lafitau. Enumerating the nations in the League Mohawks, Oneidas, Onondagas, Cayugas, and Senecas he says (Schoolcraft, vol. iii. p. 184) : " Each of these nations is divided into three tribes or families who distinguish themselves by three different arms or ensigns the Tortoise, the Bear, and the Wolf ; and the Sachems, or old men of these families, put this ensign or mark of their family to every public paper when they sign it." Such a statement as this, from an officer who must have had with the Iroquois numerous public transactions, should be conclusive. It is confirmed by what Schoolcraft says of Joseph Brant, the celebrated Mohawk leader ("who," says Morgan (p. 74), "from his conspicuous position and the high confidence reposed in him, had conceded to him by some writers the title of Military Chieftain of the League;" "but this," says Morgan, " is entirely a mistake, or rather a false asser- tion "), that he signed with a triune badge of a bear, turtle, and wolf. Also, by the separate testimony of Charlevoix, writing in 1744 (Histoire de la Nouvelle France, Paris, 1744, vol. iii. p. 266; cited by Dr. Gallatin, Arch. Amer., vol. ii. p. 109), who says, 4 'Among the Hurons the first tribe is that of the Bear ; xvi NORTH AMERICA, EAST AND SOUTH 347 the two others of the Wolf and the Turtle. The Iroquois nation has the same divisions, only the Turtle family is divided into two, the Great and the Little." It is further confirmed by the fact that certain of the Iroquois nations have not now any other than these three totems, and assert that they never had more. " The descendants of the ancient Oneidas and Mohawks," says Mr. Lewis Morgan, 1 "affirm that their ancestors never had but three tribes, the Wolf, Bear, and Turtle, and on old treaties with these nations, now in the state department, these titles appear as their only social divisions/' Charlevoix's authority may be a mere repetition of Lafitau, but the authority of Golden must be accepted as independent, and the evidence of the public documents in the state department agreeing with the other evidence would appear conclusive. More recent researches by Mr. Schoolcraft, representing the United States Govern- ment, confirm what precedes. He says (vol. iv. p. 666, pub. 1856): "The Iroquois originally appear to have had but three totems, the Turtle, the Bear, and the Wolf." And elsewhere (vol. v. p. 73), " There appear to have been originally three totems that received the highest honour and respect, i.e. the Turtle, Bear, and Wolf. These were the great totems of the Iroquois. The other totems appear of subordinate, secondary, and apparently newer origin." If further proof of this were wanted, we have it in the fact that all the Sachemships in the Mohawk and Oneida nations were held by the Turtle, Wolf, and Bear tribes, while of the forty-seven 1 League of the Iroquois, p. 81. 348 STUDIES IN ANCIENT HISTORY CHAl Sachemships in the League, of which the line of descent is alleged to be known, thirty-four were hereditary in the families of the Turtle, Bear, and Wolf, against thirteen of the families of the Snipe (six), Deer (four), Heron (two), and Hawk (one). Moreover, while the Sachemships vested in the three first -named tribes included all the most dignified titles, there is no evi- dence to support the proposition that the thirteen Sachemships belonging to the minor tribes were con- ferred upon them at the time of the formation of the League, and not subsequently, as these tribes were developed within the nation and became of importance. Moreover, the Turtle, Wolf, and Bear are the only tribes that appear in all the nations ; while from the fact that no tribe of any totem but one of these three ever formed itself within the nations of the Mohawks or Oneidas, conjunctly with the law which deduced descents through females, and counted children to be of the tribe of the mother, we are constrained to believe that the three primary tribes of the Mohawks and Oneidas, at least, must have intermarried among themselves agreeably to the law of exogamy, which we know to be now their marriage law, and to have declined marriage with women of the inferior totems. If they had married with women of such totems, it must have infallibly happened that gentes of the Snipe, Deer, Heron, and Hawk totems must have appeared as subdivisions in the nations of the Mohawks and Oneidas. Before producing Mr. Morgan's account of the nature and origin of the League of the Iroquois, it is well to xvi NORTH AMERICA, EAST AND SOUTH 349 note the sources from which he has drawn it. It is mainly founded on the tradition that the League, with every feature of it, was fully formed by an act of legisla- tion, at a single congress of the chiefs of the different nations entering into it, and thereafter continued with- out the smallest change of importance down till near the end of last century ; so that an account of the League as it was at any time in its history would be true for the moment of its origin. As to when it originated, Mr. Morgan says (p. 8), that some circumstances tend to show that it was a century old at the date of the Dutch discovery, 1609. That would refer the origin to about the year 1500. On the other hand, Mr. Morgan says that the principal Iroquois traditions indicate a period far more remote. If we assume this not to exceed a hundred years, we may have the origin given by tradition about the year 1400. We are asked to believe that the constitution of the Iroquois League, matured at all points from the first by a single legislative act, remained unchanged for four hundred years, so that we can entirely trust an account given of it by Mr. Morgan in 1851, founded on state- ments made to him by Seneca Indians when he was a young man. It may be remarked that the fact here alleged, if true, would be absolutely without parallel in the history of mankind. Fancy a constitutional history of England in the fifteenth century written in the nine- teenth, on the authority of a Welsh peasant ! Let us now proceed to consider what this marvellously un- changing constitution was, and how it was founded. 350 STUDIES IN ANCIENT HISTORY CHAP. Mr. Morgan says (p. 7) that the original Iroquois group having been disrupted into the five nations, " the severance was followed by a gradual alienation, finally resulting in a state of open warfare, which continued for an unknown period." Up to the moment when this League was formed, then, the nations who formed it were open and inveterate enemies. The project of the League, it is said, originated with a wise man of the Onondaga nation. This man must have been a far- seeing statesman, and the most thorough draftsman that the world ever knew. He propounded the League to the nations as a scheme already matured. The confederation, says Morgan (p. 60), if we may trust Iroquois testimony, was not of gradual construction under the suggestions of necessity. The plan of the League, which the wise Onondaga had projected for the union of the hostile nations, was considered at a place on the north shore of Lake Onondaga, " where the first council- fire was kindled, around which the chiefs and wise men of the several nations were gathered, and where, after a debate of many days, its establishment (i.e. that of the League) was effected." The plan of the wise Onondaga was of the most extraordinary sort, to propose to nations at war with one another. In the first place he proposed to divide each of them into eight tribes, as a device for creating new relationships by which to bind them all more firmly together (p. 91), and he proposed with the same view, and to make these relationships operative in the best manner, that it should be a law of the League that descent should be counted in the female line, and xvi NORTH AMERICA, EAST AND SOUTH 351 the children accordingly be. " in perpetuity " of the tribe of their mother. Then with a view to making the division of the people in each nation into eight tribes more effectual the most perfect union, Mr. Morgan says, of separate nations which was ever devised by the wit of man (p. 81) the wise man of the Onondagas planned the division of each of the eight tribes into five portions, and planned the distribution of these portions throughout the several nations. " In effect the Wolf tribe was divided into five parts, and one-fifth of it placed in each of the five nations. The remaining tribes were subjected to the same division and distribution." And it must be assumed, as part of the plan of the wise man who invented these tribes and subdivisions of tribes, that, by a law of the League, all of the same tribe should count themselves as brothers and sisters, bound to each other by the ties of consanguinity. It is easy to see, assuming it possible to create by enactment sentiments of brotherhood, and to effect such a marvellous series of transplantations of sections of newly-created divisions of hitherto hostile nations, how thoroughly well welded together the nations would become in respect of the newly-created cross relationships. " If either of the five nations," says Mr. Morgan, "had wished to cast off the alliance, it must also have broken this bond of brotherhood. Had the nations fallen into collision, it would have turned hawk tribe against hawk tribe, heron against heron, brother against brother. The history of the Iroquois exhibits the wisdom of these organic pro- visions, etc." 352 STUDIES IN ANCIENT HISTORY CHAP. The reasoning by which our admiration is claimed for these organic provisions would force us, even if we were not expressly told, to think of them as novelties and devices of the wise man of the Onondagas, since the nations are represented as having been continually at war with one another before the League. If we think of them as being before the League divided into tribes of the same name, we see that we should have, in case of war, hawk against hawk, heron against heron, etc., which it is suggested could not be, and certainly could not be if a common totem implied common blood. The tribal distribution, then, must have been, as Mr. Morgan says, an organic device; and at p. 91 it is expressly stated that "the Iroquois claim to have originated the idea of a division of the people into tribes, as a means of creating new relationships by which to bind the people more firmly together. The fact that this division of the people of the same nation into tribes does not prevail generally among our Indian races, favours the assertion of the Iroquois." The plan of the wise man of the Onondagas proposed not only an entire reconstruction of the whole of the five nations socially. It proposed, further, a scheme of political government in perpetuity for each of them ; a scheme of government for the whole of them in their union as a League ; the creation and distribution of a variety of noble offices, and solemn methods of "raising up" from time to time successors to the various national and League functionaries. No wonder that it was only after a debate of many days that its establishment was xvi NORTH AMERICA, EAST AND SOUTH 353 effected. The Iroquois traditions, which have guided us so far, say, indeed, that the wise man of the Onondagas met the chief opposition to his scheme from a Sachem of his own nation. Tododa'ho (p. 67) was an Onondaga ruler who had become very powerful in war. "Tradition says that he had conquered the Cayugas and the Senecas. It represents his head as covered with tangled serpents, and his look when angry as so terrible that whoever looked upon him fell dead." This terrible person, of the reality of whose existence, as harmonising with the rest of the narrative, we can have no doubt, " was reluctant to consent to the new order of things, as he would thereby be shorn of his absolute power and be placed among a number of equals." He was bribed into acquiescence by naming the first Sachemship after him, and dignifying it above the others by special marks of honour. Two facts, true to this day, support the tradition respecting Tododa/ho. His name is to the present day among the Iroquois the personification of heroism, forecast, and dignity of character; and the Mohawk Sachem who, when the League was formed, combed the snakes out of Todo- da/ho's hair, is still called Ha-yo-went'-ha, the man who combs (p. 68). The Iroquois traditions would not appear to have been very consistent with one another, or Mr. Morgan has sifted them, inclining sometimes to one form of them and sometimes to another. For instance, we saw that they claim to have originated the division into tribes as a means of welding the people together, and that 2 A 354 STUDIES IN ANCIENT HISTORY CHA1 Mr. Morgan inclined, though not without some hesita- tion, to favour this view (p. 91). Elsewhere the trib* divisions are assumed as existing before the Leagui "The founders of the Iroquois Confederacy," say* Mr. Morgan, "did not seek to suspend the tribj divisions of the people, to introduce a different soci; organisation ; but, on the contrary, they rested th< League itself upon the tribes, and through them soughl to interweave the race into one political family" (p. 79). "In each nation," he says, "there were eight tribes which were arranged in two divisions and named follows: Wolf, Bear, Beaver, Turtle; Deer, Sni] Heron, Hawk " (p. 79). " Tradition declares that the Bear and the D< were the original tribes, and that the residue wer< subdivisions" (p. 80). " All the institutions of the Iroquois have regard to the divisions of the people into tribes, originally with reference to marriage. The Wolf, Bear, Beaver, and Turtle tribes being brothers to each other, were no1 allowed to intermarry. The four opposite tribes being also brothers to each other, were likewise pi hibited from intermarrying. . . . Whoever violate< these laws of marriage incurred the deepest detesta- tion and disgrace. In process of time, however, th< rigour of the system was relaxed, until finally th< prohibition was confined to the tribe of the individual, which among the residue of the Iroquois is sti] religiously observed. They can now marry into tribe but their own. Under the original as well xvi NORTH AMERICA, EAST AND SOUTH 355 modern regulation, the husband and wife were of different tribes. The children always followed the tribe of the mother" (p. 83). It will be seen that the ultimate facts are totemism, exogamy, and female kinship. Assuming that there is anything in the tradition that anciently there were among the Iroquois eight tribes in each nation, and that these were arranged in two divisions of four tribes each, within which marriage was interdicted, we see in these divisions nothing inconsistent with simple exogamy ; for the one of them contains the Bear and the other the Deer, and Mr. Morgan gives the tradition that the others were mere subdivisions of these stocks. If the Wolf, Beaver, and Turtle were truly of the Bear stock, marriage among them would by that fact be interdicted according to the law of exogamy. So, if the Snipe, Heron, and Hawk were truly of the Deer stock, would marriage be interdicted between them by the same law ; and the phenomena of such divisions touch not the question of the origin of exogamy. They exhibit a phenomenon posterior to the establishment of exogamy, the formation of bands on the totemic model, mindful for a time of their being of one stock with one another, as well as with the group from which they sprung, and for long obeying the law of exogamy on that footing ; but in the lapse of time, the memory of their origin growing feeble, or it may be falling under doubt, or being overridden by new notions of origin, treating one another as if they were of different stocks. That a tribe which for many generations had passed for 356 STUDIES IN ANCIENT HISTORY CHAP. XVI Wolves, albeit they were derived from Bears, should in time come to regard themselves as being really Wolves and not Bears, is what might be expected ; and they would apply the marriage law accordingly. Mr. Morgan's account, then, is of importance in connection with the origin of exogamy only so far as it ascribes it to a legislative device ; and we have already seen how absolutely incredible it is that it had such an origin. CHAPTER XVII NORTH AMERICA, WEST THE MOQUIS GOVERNMENT among the Moquis is hereditary, but does not necessarily descend to the sons of the chief, if the people prefer any other blood -relation. They give the following account of their origin : " Many years ago their great mother brought from her home in the West nine races of men, in the following forms. 1. The Deer race. 2. The Sand race. 3. The Water race. 4. The Bear race. 5. The Hare race. 6. The Prairie-wolf race. 7. The Eattlesnake race. 8. The Tobacco-plant race. 9. The Eeed-grass race. Having placed them in the spot where their villages now stand, she trans- formed them into men who built the present Pueblos, and the distinction of races is still kept up. One told me he was of the Sand race, another of the Deer, etc. They are firm believers in metempsychosis, and say that when they die they will dissolve into their original forms, and become bears, deer, etc., again. The Chief Governor is of the Deer race." (Re- ported by Dr. Ten Broeck, Assistant-Surgeon United States Army.) They say that the great mother gave them all the domestic animals they have. They keep up a sacred fire; the women propose marriage to the men ; polygamy is unknown among them. 1 [A different list of totems among the Moquis is given by 1 Schoolcraft, voL iv. pp. 85, 86. 358 STUDIES IN ANCIENT HISTORY CHAP Bourke; besides animals and plants, it includes Water, Fire- wood, Sun, and Cloud. " My informant said he himsel: belonged to the Butterfly gens, that his wife and children were of the Eagle, his father a Deer." This proves female kinship among them, and suggests exogamy. Both are distinctly stated further on : " When a man marries he goes to the house owned by his wife ; the act of marriage does not sever the connection with his own clan, but his children follow their mother's clan. Women exercise the right of choosing their own husbands. Property owned by the wife descends, upon her death, to her daughters. Marriage must be exogarnous as concerns the gens, i.e. a man and woman belonging to the same gens or clan cannot enter into the marital relation. A Badger man cannot marry a Badger woma'n. . . . The women have the management, control, and ownership of the houses. A man will never sell anything in the house unless his wife consent, and then she pockets the money. This feature of domestic life is noticeable among the Pueblos generally." l ] THE NAVAJOS These are nomadic tribes; they have no chiefs, but are organised on a somewhat patriarchal type, each wealthy man owning a band composed of retainers and servants, called his family, and resembling a Scottish Highland clan. 2 They tell this story of their origin : Many years ago they all lived under ground, along with the Pueblos and other tribes. Among the Navajos were two dumb men who played the Indian flute. One of them accidentally touching one day the top of the cave, there was heard a hollow sound, and immediately the old man conceived the idea of boring through to see what was inside. Then follows a story in which the Racoon, Moth- worm, and four great White Swans figure. The four Swans, on a Moth-worm boring a hole through the roof, attacked him, the first piercing him with an arrow ; then 1 The Snake-Dance of the Moquis of Arizona, pp. 116, 135, 248. London, 1884. 2 Schoolcraft, vol. iv. pp. 89, 211. xvn NORTH AMERICA, WEST 359 saying, " He is of my race," and retiring. After a variety of adventures, the men and animals came up ; the Navajos first , after them the Pueblos and other Indians. The old men of the Navajos made the sun ; the old men of the other tribes made the moon, heaven, and stars. The scattering of the stars about anyhow was the work of the Prairie-wolf. The few constellations show how the Navajos would have em- broidered them had the Prairie-wolf let them alone. Corn was brought to them by a Turkey-hen from the morning star ; also, at a later time, white corn and wheat; in fact, all the seeds they possess were brought to them by this benevolent bird. This all savours of totemism, and further proof is that they "never eat the flesh of the grey squirrel, nor could I (Major Backus) induce them to give any reason for declining it; yet they eat the Prairie-dog, which is in no respect pre- possessing." 1 Another account of their origin runs thus : " At the first twelve Navajos, six men and six women, came out of the earth in the middle of the lake which is in the valley of Montezuma. They were preceded in their ascent by the Locust and Badger, the Locust being the foremost, and boring the hole for the others; but, as he was not very successful, the Badger made the hole larger, so as to enable the Navajos to come out. On arriving at the surface of the earth, the Navajos were provided with fire in the following manner. The animals now found on the earth were already in existence. The Coyote, Bat, and Squirrel were the special friends of the Navajos, and agreed to aid in procuring fire for them. The animals, neither Deer nor Moose being yet created, were engaged in playing mocassin or shoe-game, having a fire to play by. The Coyote stole a light and ran off with it. The Bat in time relieved him, and when tired made the fire over to the Squirrel, who conveyed it to the Navajos." They think the Americans may be descended from twelve other Navajos, who came up after the first twelve. After a time there were but three Navajos, an old man and woman, and young woman. 1 Schoolcraft, vol. iv. pp. 89, 91, 211, 212. STUDIES IN ANCIENT HISTORY CHAP. The young woman had a son by the Sun, and after that they increased in numbers. 1 Marriage. Girls are considered the property of their parents till they marry ; the husband makes a contract with the father, the usual price being five or six horses. Navajo wives are " very independent of menial duties, and leave their husbands upon the slightest protest of dislike." *' Kinship among them appears, therefore, now to be male, but to have been once female, to judge from the myth quoted above, wherein they trace their descent to a legendary ancestress and the Sun. THE DACOTA GROUP Under this head are here included the Omahas, Otoes, Winnebagoes, lowas, Dacotas proper or Sioux, Quappas. The tribal organisation appears to be totemic. The Iowa and Sac tribes have the notion that they are descended from animals or birds, and bear the title of the particular animal 01 bird from which they sprung. " They have eight leading families, though some of them are now extinct. These are the Eagle, Pigeon, Wolf, Bear, Elk, Beaver, Buffalo, and Snake. These families are known severally in the tribe by the parti- cular manner in which their hair is cut. The other families with their peculiar badges are lost." It is only the male children who have their hair cut in this manner, and it is done once a year. " They pay a kind of religious adoration to som< animals, reptiles, and birds. There is a species of the hawk which they never kill, except to obtain some portions of it body to put with their sacred medicine." 3 The Osagees attribute their origin to a snail and a beaver. 4 The Hidatsas or Minnitarees, a branch of the Dacotas, a their divisions ddki ; " In the Hidatsa daki," says Matthews "we have apparently a modification of the totem system." 1 Schoolcraft, vol. iv. pp. 218, 219. 2 Id. vol. iv. pp. 214, 217. 3 Id. voL iii. pp. 261, 268, 273. 4 Id. vol. iv. App. H. 5 Ethnography and Philology of the Hidatsa Indians, p. 139. "Wash- ington, 1877. xvii NORTH AMERICA, WEST 361 " Meats forbidden are strictly observed by the Indians (Dacotas), but all differ in the different kinds of meat for- bidden," x a clear sign of totemism. But the Dacotas possess a somewhat elaborate hierarchy of gods, of which there are several classes. One class is called Onkteri ; they resemble the ox in form, but are very large, and " subordinate to the Onkteri are the serpent, lizard, frog, leech, owl, eagle, fish, spirits of the dead, etc." Another class is called the Wakinyan or " fliers," which are in bird form, and are the Dacota war-gods. Thirdly, there is a god who is " invisible and ubiquitous. He resides in the consecrated spear and tomahawk, in boulders (which are hence universally venerated by the Dacotas), and in the ' Four Winds.' " Subordinate to him, again, are the " buzzard, raven, fox, wolf, and some other animals of a similar nature." There are many other kinds of gods besides. 2 The Winnebagoes have a cosmogony full of suggestions of totemism. Their Great Spirit dreamt, waked, took a piece of his body and of earth, made a man, then other three, talked to them, and made a woman, the earth, grandmother of Indians. The four men are the four winds. The earth was unsteady, and so he made four beasts and four snakes to steady it. From a piece of his heart he made a man. He gave to man tobacco, and to woman grains. The Great Spirit invented fighting 118 years after, to prevent men dying of old age. He tried to make an Indian and made a negro ; he then tried to make a black bear and made a grisly bear. Another version runs : " The Great Spirit then made a man from a he-bear, and made a woman from a she-bear. After these men were created, they held a council, and it was agreed that the second man that came down from heaven should be a war-chief, and that the man made from a bear should be his second in com- mand." A distinct statement of totemism follows. " This tribe was anciently divided into clans or primary families, known by the names of Bird, Bear, and Fish families. The clans have not at the present day any badges designating their order or rank." 3 1 Schoolcraft, vol. iii. p. 243. 2 Id. vol. iv. pp. 643-645. 3 Id. vol. iv. pp. 236, 239. 362 STUDIES IN ANCIENT HISTORY CHAP. The Omahas are divided into two principal sections, which are again divided into eight and five clans respectively. Each of these derived its name from some animal, plant, or other natural substance, and was not allowed to eat it. 1 MAERIAGE Very little evidence under this head is forthcoming. It is, however, definitely stated that among the Omahas " even a very remote degree of consanguinity is an insuperable barrier to the marriage union," 2 and as we have found them to possess the totemic organisation, there can be no doubt that this means exogamy. " In the Omaha nation numbers of the females are betrothed from their infancy ; and as polygamy is extremely common, the individual who weds the eldest daughter espouses all the sisters successively, and receives them into his house when they arrive at a proper age." After marriage the young wife remains some time with her parents, occasionally visited by her husband, before she goes to reside in his lodge. 3 Polygamy is the rule also among the Dacotas. 4 The Omahas have the levirate. " If the deceased has left a brother, he takes the widow to his lodge after a proper interval, and considers her as his wife without any preparatory formality. If the deceased has not left a brother, the relations of his squaw take her to their lodges." 5 Though polygamy now prevails among the Hidatsas, a man often marrying several sisters, yet there is some evidence of former polyandry of the Thibetan type in a legend " said to belong originally to this tribe, and to have been known to it from time immemorial." Two demigods, Long -Tail and Spotted Body, lived together in a lodge ; " a woman lived with them, who took care of their lodge, and who was their wife and sister; and these three were at first the only beings of their kind in the world." 1 James, Expedition from Pittsburgh to the Rocky Mountains, vol. ii. p. 47. 1823. 2 Ibid. vol. i. p. 213. 3 ftfa vo i i p . 209. 4 Schoolcraft, vol. iii. p. 240. 5 James, vol. i. p. 222. NORTH AMERICA, WEST 363 Marriage by purchase and marriage by elopement are both recognised among them, but the latter is looked on as undigni- fied, and they have different terms for the two kinds. Divorce is easy, but rare among the better class of people ; the women are as a rule faithful. "A man usually takes to wife the widow of a brother, unless she expresses an unwillingness to the arrangement, and he may adopt the orphans as his own children." A man may not talk with his mother-in-law, a custom which holds with other western tribes ; " but this custom seems to be falling into disuse." Hence it may be inferred that it was once lawful to marry a mother and daughter, which is still common in parts of California; the prohibition even to speak to the mother is the result of an excessive rebound in the opposite direction. They have the classificatory system of relationship, according to the common American type. 1 Blood-Feud. The death of a relative is commonly avenged, sometimes two or three being killed for one, but compromises are frequently made by the offender giving large presents. He often flees from justice, and may be killed years after the offence. They have no cities of refuge. 2 Succession. The Dacotas seem to know nothing of heirship in property, or if they do, they have no chance of leaving it to their children. Most of it goes to medicine men, in death-bed expenses, and what remains is taken by the other Indians. 3 COLUMBIANS AND CALIFORNIANS The Columbians dwell between the 43rd and 55th parallels of latitude on the north-west American coast. The " Nootka Columbian" tribes are: 1. Haidahs, etc., Queen Charlotte's Islanders ; 2. Nootkas, Vancouver's Islanders ; 3. Sound Indians, Puget Sound area; 4. Chinooks, on the banks of the Columbia and along the coast both north and south. The Cascade Mountains run north and south ; on the coast side lies 1 Matthews, pp. 63, 52, 53. 2 Schoolcraft, vol. ii. p. 184. 3 Id. vol. ii. p. 194. 364 STUDIES IN ANCIENT HISTORY CHAP. a strip 1000 miles long full of noble forest game, etc. On the inland sides are plains or table -lands. The chief nations of the Haidahs are the Massets, Skiddegats, and Cumshawas (all Queen Charlotte's Islanders) ; Kaiganies (Prince of Wales' Archipelago) ; Chimsyans, Nass, and Skeenas, Sebassas, Hailt- zas, Bellacoolas. These are further subdivided " into numerous indefinite tribes." ] The Calif ornians. Between 43 and 32 30' and back irregularly to Eocky Mountains. Divisions Northern, Central, and Southern, and 4th, the Shoshones inland beyond the Sierra Nevada. For toternism among these tribes see the article on totemism, infra, pp. 368 ff. BLOOD-FEUD Among the Nootkas, Northern, Central, and Southern Californians, blood-feuds are mentioned as continually raging ; thus among the Nootkas "private, family, and tribal feuds continue from generation to generation." Among the Northern Californians the eric is sometimes substituted for vengeance. The temple was a refuge for the murderer in South California, but vengeance was only deferred, and was exacted later on by the descendants of the murdered man either from the murderer himself or his kindred. Among the Acagchemem nations every vanguecli or temple of Chinigchinich was a city of refuge. "Not only was every criminal safe there, whatever his crime, but the crime was as it were blotted out from that moment, and the offender was at liberty to leave the sanctuary and walk about as before ; it was not lawful even to mention his crime ; all that the avenger could do was to point at him and deride him, saying, 'Lo, a coward, who has been forced to flee to Chinigchinich ! ' This flight was rendered so much a meaner thing in that it only turned the punishment from the head of him that fled upon that of some of his relatives ; life went for life, eye for eye, and tooth for tooth, even to the third and fourth generation, for justice' sake." 2 1 Bancroft, vol. i. p. 155. 2 Id. vol. i pp. 91, 134, 207, 348, 386, 409 ; vol. iii. p. 167. xvii NORTH AMERICA, WEST 365 CAPTURE Capture is continually going on throughout the whole district west of the Eocky Mountains. " Some of the smaller tribes at the north of the island (Vancouver) are practically regarded as slave-breeding tribes, and are attacked periodically by stronger tribes." ] " Although by no means a bloodthirsty race, the Chinook tribes were frequently involved in quarrels, resulting, it is said, from the abduction of women more frequently than from other causes." '* A similar statement is made of the Northern, Central, and Southern Californians alike. 3 The Eogue Kiver Indians " kill all their male prisoners, but spare the women and children." In the middle of this century the Yosemite Indians were nearly exterminated by the Monos, and all their women and children carried off. The Southern Californians, who are always at war, spare no male prisoners, and sell the females or retain them as slaves. At Clear Lake " rape exists among them in an authorised form, and it is the custom for a party of young men to surprise and ravish a young girl, who becomes the wife of one of them." 4 MARRIAGE Exogamy we have seen already to be the rule among the Nootkas. Indirect evidence may be found in the statement that in Southern California, " if a man ill-used his wife, her relations took her away, after paying back the value of her wedding presents." " The Spokane husband joins his wife's tribe " (Columbia). 5 Apart from these two cases, there appears to be no evidence for exogamy. Polygamy is the rule among the chiefs of the Nootkas and throughout California. Among the Chinooks only the very rich have several wives. "Eich old men almost absorb the female youth and beauty of the tribe " in Northern California, 1 Sproat in Bancroft, vol. i. p. 195 note. 2 Bancroft, vol. i. p. 235. 3 Id. vol. i. pp. 343, 380, 412. 4 Id. vol. i. pp. 344, 401, 407, 389 note. 5 Id. vol. i. pp. 277, 413. 366 STUDIES IN ANCIENT HISTORY CHAP, xvn " while the younger and poorer men must content themselves with old and ugly wives." Yet " polygamy is almost universal, the number of wives depending only on the limit of a man's wealth." The common people of the Nootkas and South Californians have only one wife apiece, and among the Cahrocs there is no polygamy. In Central California a man often marries a mother and all her daughters. Polygamy is every- where practised by the Haidahs. 1 Polyandry is said to be forbidden among some of the Central Californians. 2 Marriage is generally by purchase among all the above nations, despite the capture which exists beside it. Incest is very common. Among some Central Californians " parentage and other relations of consanguinity are no obstacles to matrimony." Of the Haidahs it is said that the women " cohabit almost promiscuously with their own tribe, though rarely with other tribes." 3 1 Bancroft, vol. i. pp. 168, 196, 197, 241, 349, 351, 388, 410, 437. 2 Id. vol. i. p. 388 note. 3 Id. vol. i. pp. 168 note, 389 note. CHAP TEE XVIII NORTH AMERICA, NORTH THE tribes of this district are grouped in five principal divisions: 1. Esquimos ; 2. Koniagas, or Southern Esquimos; 3. Aleuts; 4. Thlinkeets ; 5. Tinneh. Their time is spent in a continual struggle for life, their daily food depending on daily battles with beasts, birds, and fishes. The Koniagas are divided again as follows : 1. Koniagas proper on the Archipelago; 2. Chugatshes; 3. Aglegmutes ; 4. Keyataigmutes ; 5. Agulmutes; 6. Kuskoquigmutes ; 7. Magemutes; 8. Kwichpagmutes ; 9. Chagmutes; 10. Anlyg- mutes; 11. Kaviaks; 12. Malemutes. Of the Aleuts there are two tribes, the Unalaskans and the Atkhas. The Thlinkeets live along the coast and in islands from Mount St. Elias to the river Nass, or to Columbia river, accord- ing to Holmberg. They comprise : 1. Yakutats; 2. Ugalenzes; 3. Chilkats ; 4. Hoodsinoos ; 5. Takoos ; 6. Auks ; 7. Kakas ; 8. Sitkas; 9. Stikines ; and 10. Tungas. The Sitkas on Bara- noff Island are dominant. They are warlike, and settle tribal quarrels by trial by combat. La Perouse says they are more advanced than the South Sea Islanders, except as to agri- culture. The Tinneh consist of four great families of nations: 1. Chepewyans ; 2. Tacullies ; 3. Kutchins ; 4. Kenai. 368 STUDIES IN ANCIENT HISTOR Y CHAP. TOTEMS The notices of totems in Bancroft's compilation are casual and fragmentary; but probably they contain all the information the authorities have to furnish. It is certain that the liabilities to errors of observation are enormous in regard to this class of facts. A traveller meets a Thlinkeet, for example. The Thlin- keet is of the Bear clan, and declares his horror at the idea of either killing a bear, or eating his flesh. The record is in general terms, " The Thlinkeets will not kill bears or eat bear's flesh." This is of course misleading ; but, making due allow- ance for errors of this class, it will be often possible to spell out from the records, more particularly when there are several of them, the true state of the facts. An example may be seen in what is said of the tribes grouped in the compilation as " Southern Calif ornians." It is noticed by Schoolcraft l that superstition forbids them to eat " the flesh of large game." They believe " that in the bodies of all large animals the souls of certain generations, long since passed, have entered." But the statement in the text of Bancroft's compilation is, " Bear meat the majority refuse to eat from superstitious motives," 2 from which we infer that some will eat bear meat. Others, we are directly told in the text, will eat deer. So that we have the general statement instantly qualified, and are free to infer that it had no other foundation than that some of these natives abstained from one species of large game, others from another, and that the abstention in each case was as regards the sacred animal or totem of the clan to which they respectively belonged. The cases of totemism that are directly given are few indeed. They are as follows : (1) The Thlinkeets, who are represented as being divided into totem clans in two main divisions, Wolf and Eaven, with exogamy as marriage law. 3 (2) The Kutchins, a great nation included under the more general term Tinneh, who are said to be in three " castes " 1 Schoolcraft, vol. v. pp. 215, 216. 2 Bancroft, vol. i. p. 405. 3 Idm volt i p . 109< xvm NORTH AMERICA, NORTH 369 clans obviously being intended ; and the marriage law, again, being exogamy. " Two persons of the same caste cannot marry, but a man of one caste must marry a woman of another." The mother, it is stated, gives " caste " to the children, i.e. descent is through females. 1 (3) The Nootkas in Vancouver's Island and on the adjoining mainland. The family (clan) is indicated by some animal adopted as the family crest (totem), and exogamy is the marriage law. Persons of the same crest must not marry. The child takes its crest from the mother, and, as a rule, descent is traced from the mother and not the father. 2 (4) The people of Cueba counted among the Isthmian Indians. It is directly stated that they have totems, but their case appears as one of transition from descent in the female to descent in the male line, and a dis- tinction is taken between the "ancestral" totem and an adopted totem, which indicates a break-up of the ancient sacredness of the sign and a shifting of kinships from the female to the male side. 3 (5) Certain Utah Indians. The text of the compilation says nothing of totemism among them, but in the description of their boundaries there is a casual disclosure of its existence in a citation which has no bearing on the matter of boundaries. It appears that the different " bands " another term often used for clan of the Pah Utes are named after " some article of food not common in other localities " an obvious guess of the traveller. Thus Ocki, the name of one band, means " Trout." 4 While the direct information as to the prevalence of totemism is thus sparse, the indirect is overwhelming, and will be best appreciated if we recur to the cases above specified and consider some at least of them in detail, and the kind of legends and beliefs connected with totemism which they present. The Thlinkeet nation, inhabiting the coast and islands from Mount St. Elias near the Copper Eiver on the north to the river Nass on the south over about five degrees of 1 Bancroft, vol. i. p. 132. 2 Id. vol. i. p. 197 note. 3 Id. vol. i. p. 753. 4 Id, vol. i. p. 466. 2 B 370 STUDIES IN ANCIENT HISTORY CHAP. latitude is represented as being in two great divisions, which are named respectively after the Wolf and the Eaven. The clans comprised in the Wolf division are named after the Bear, Eagle, Dolphin, Shark, and Alca respectively the Wolf clan, or Coquotans, being dominant in the division. The clans comprised in the Raven division, again, are named after the Frog, Goose, Sea -lion, Owl, and Salmon respectively the Eaven clan being dominant in the division. Between members of the same clan war is interdicted and so is marriage. On the other hand, all of the same clan are bound to stand by one another in quarrels ; and the young Wolf who to-day marries a Eaven may be called upon to-morrow to fight his father-in- law over some hereditary feud. 1 Some casual entries in portions of the Bancroft compilation illustrate though feebly the totemism of the Thlinkeets. Tor example, we are told that the Thlinkeet will kill a Bear only in case of great necessity, " for the Bear is supposed to be a man that has taken the shape of an animal " a statement true, no doubt, of the Thlinkeets of the Bear clan. 2 Again, there are the conflicting cosmogonies of the Thlinkeets. Those of them with whom the Eaven is in the first place as the totem of the dominant clan believe Yehl, the Crow or Eaven, 1 See Bancroft, vol. i. p. 109; and the footnote. There is no con- formity between the authorities cited in the footnote and the statement in the text. When Lisiansky (Voy., pp. 238, 242) says the Thlinkeets are divided into " tribes," lie means clans, for he proceeds to enumerate them, viz. " the tribe of the Bear, of the Eagle," etc. His " tribes " are then genuine clans, or gentes, and not local divisions. When he says, " The tribe of the Wolf has many privileges over the other tribes," we must read, " The Wolf clan or gens has many privileges," etc. ; whence it appears that the Wolves form a clan like the Bears and Eagles, and are not, as stated in Bancroft's text, a "trunk" or division of the nation subdivided into Bears, Eagles, etc. In the text again the word " tribe " is used without thought. Thus, " tribes of the same clan may not war on each other," which makes nonsense. I interpret it to mean that the bond of a common blood and religion restrain men of the same clan from conflicts with one another. 2 Id. vol. iii. p. 129. xvni NORTH AMERICA, NORTH 371 to be the Creator of all things, and in particular to have furnished the world with light, fire, and fresh water. The adherents of the Wolf, on the other hand, claim these honours for the Wolf, Khanukh, the progenitor and totem of the Wolf clan and they assert not only his superiority to Yehl, but his priority. A story, invented no doubt by the priests of the Wolf division, represents Yehl himself as having been con- strained to admit this inferiority. It runs that Khanukh had all the fresh water, and the Kavens and their allies had none. Yehl having determined to procure water for his people, pro- ceeded to Selka, an island in which the water was stored, and was there confronted by Khanukh, when the following con- versation took place between them. "How long hast thou been in the world ?" asked Khanukh. Yehl proudly answered, "Before the world stood in its place, I was there." "But how long hast thou been in the world ?" asked he in turn of Khanukh. " Ever since the time that the liver came out from below." This allusion, to us incomprehensible, was clear to Yehl, who instantly confessed, " Then art' thou older than I." 1 The story which has a parallel in the Lin'ga Purana proceeds to show how Khanukh then terribly frightened Yehl, notwith- standing which the latter managed to steal some of the fresh water, and flying back to the mainland with it, scattered it as he flew in all directions, " and wherever small drops fell there are now springs and creeks." The rivalry of the gods which this story reveals could not but be unfriendly to national unity ; and the tendency every- where as a condition of peace and quiet within a nation is, in the long run, towards a harmonising of the claims of rival divinities through co-ordinating them somehow in a system. Equally certain is the universal tendency to anthropo- morphise the original animal gods, and more often than not, to establish amities among them by ties of kinship. Thus when we come on a new version of the story of Yehl and Khanukh, which represents them as men and brothers sons of one mother and accepted as the two leading gods of the Thlinkeets with separate domains Khanukh being god of war 1 Bancroft, vol. iii. pp. 98-102. 372 STUDIES IN ANCIENT HISTORY CHAP. it is only what we should have looked for. On this step being taken, Yehl is a man-god named after his " favourite " bird the Eaven ; and Khanukh a man-god named after his " favourite " animal the Wolf. At first neither had any trace of humanity. 1 To the north and west of the Thlinkeets are the tribes grouped together in Bancroft's work as the Koniaga nations, and extending from Behring Straits to the strait of Alaska. In the direct account of these tribes no mention is made of totemism among them. Their mythology however, as else- where disclosed, enables us to infer it one portion of them having the Eaven and another the Dog in the place of honour in Olympus. They will not eat pork, moreover, and have a festival to an owl with a man's head, a sea-gull, two partridges, and bladders of animals. 2 The Koniagas proper, who inhabit the island of Kadiak, claim descent from the Dog, i.e. those, as we may believe, that belong to the Dog clan do so ; but the form of the descent is not disclosed. We are told, however, of the form of this descent believed in by their neighbours the Aleuts. 3 Some Aleuts trace their origin to a dog and bitch as first parents, and are thus Dogs on both sides of the house ; others have it that they are sprung from a bitch and a man father ; while a third account represents their first father as having fallen from heaven in the shape of a Dog the maternity in this case being undisclosed. Other neighbours of the Koniagas, again, on their north-east the Tinneh furnish a fresh variant of the myth. The Dog clan of this great nation refer their origin to a woman and a male Dog, who yet was a man, and from whom indirectly they derive not only all human beings, but the fishes, birds, and animals. This completes all the possible variations that could be played on the canine and human elements in their ancestry. One clan of the Tinneh, it may be noted, have a bird presumably the Eaven as a totem. But the clans would appear not to be interfused, if we can credit the statement that the bird has no place in the 1 Bancroft, vol. iii. p. 149 note. 2 Id. vol. i. pp. 75, 85 ; vol. iii. p. 104. 3 Id. vol. iii. p. 104. xvin NORTH AMERICA, NORTH 373 northern Indian branch of the Tinneh. The statement is however most probably incorrect. Among the Tinneh we have a further note of totemism. They will not eat the Dog, and abominate all that do so. This is most probably true of all the Dog clans among these northern nations, but we have the fact for some Tinneh only. 1 Of clan distinctions, or totemism, among the tribes of California not a word is said. We are told, however, that the Californians describe themselves, in most cases, as originating from the Coyote (i.e. the Prairie-wolf), which would indicate a Coyote clan ; and there are indications that they have Hawk, Crow, Duck, and Bear clans at least as well. One tribe, the Potoyantes, give a metaphysical account of the transformation of the first Coyotes into men. " There was an age in which no men existed nothing but Coyotes," and the account is made to explain the origin of cremation. Another Californian myth tells of a great flood, when only a Coyote survived he and a single feather tossed on the water. As he looked at it, this feather became an Eagle and joined the Coyote on " Keed Peak," and after a time, the two, feeling lonely, created men. 2 Among the Acagchemem, inhabiting the valley and neigh- bourhood of San Juan Capistrano, California, the Creator is Chinigchinich, whose image is the skin of a Coyote, or that of a Mountain Cat, stuffed with feathers, so as to look like the live animal. The same people have the great buzzard in veneration, and a special ritual connected with it. They kill it with religious ceremonies and bury it, mourning over it, and, strange to say, they believe " that the birds killed in one same yearly feast in many separate villages were one and the same bird." But, strangest of all, this bird, which they called the Panes, was not the Creator but a metamorphosis of him. It was Chinigchinich and at bone a Coyote ! 3 1 Bancroft, vol. iii. p. 105. 2 Id. vol. iii. pp. 87, 88. 3 Id. vol. iii. pp. 166, 168. In Oregon we have the Coyote anthropo- morphised as Italapas, the Coyote and creator of the human race and first instructor. It was the Coyote who stocked the rivers with salmon for the Cahrocs (vol. iii. pp. 137, 155). Boscana, in Kobinson's Life in 374 STUDIES IN ANCIENT HISTOR Y CHAP. Among the Cahrocs in North California, we again meet the Coyote as the benefactor of men, in conjunction with the Cougar, the Bear, the Squirrel, and the Frog, while a special legend explains the superior cunning of the Coyote and his friendship for human beings. 1 But in North California, in the neighbourhood of Mount Shasta, it is the Bear that is in the first place, the master of all animals, and a terror even to his Creator, the Great Spirit. This Great Spirit had his wigwam on Mount Shasta, and lived there with his little daughter. A great storm blew the daughter away to the land of the grisly bears, who at that time walked and talked like men. She was brought up in a family of the bears, and married the eldest son. " Their progeny w T as neither grisly bear nor Great Spirit, but man." When- the Great Spirit heard of this he was enraged, and doomed the race of grislies to perpetual dumbness. " No Indian, tracing his descent from the spirit mother and the grisly, will kill a grisly bear ; and if, by any evil chance, a grisly kill a man in any place, that spot becomes memorable, and every one who passes casts a stone there till a great pile is thrown up." 2 Before leaving the Californians, I notice that the Spaniards of Vizcaino's expedition in 1602 found the Indians of Santa Catalina Island venerating two great black crows, which, according to Seiior Galan, were probably a species of bird known in Mexico as king of the Turkey-buzzards ; he adding that these birds are still the objects of respect and devotion among most Californians. 3 Among the Ahts we meet a variety of totems birds, animals, and fishes. Among these it is the Cuttle-fish that holds the place of honour. He alone of all creatures possessed fire. The other animals in vain tried to steal fire from the Cuttle-fish. The Deer at last succeeded in stealing some, and brought it into general use. Not all animals were made at once the Loon and Crow were metamorphosed men. In California, p. 169, describes certain other Californians as worshipping their chief god in the form of a stuffed Coyote. 1 Bancroft, vol. iii. pp. 90, 115. 2 Id. vol. iii. pp. 90-93. 3 /^ vo i t ^ p 134 NORTH AMERICA, NORTH 375 Queen Charlotte's Island, however, men are deduced from crows not crows from men. The Haidahs "gravely affirmed and steadfastly maintained " their descent from crows. " Certain owls and squirrels are regarded with reverence and used as charms." * Shift the district, shift the gods. On the Palouse Eiver certain Cayuses, ISTez Perces, Walla Wallas, etc., refer their origin to a miraculous Beaver. The Cayuses sprung from the heart of this Beaver, and for this reason they are more energetic, daring, and successful than their neighbours. In British Columbia, again, among the Tacullies, it is the Musk- rat that has the place of honour. Some Navajos, again, unite the Coyote, Bat, and Squirrel as benefactors. Others, again, have traditions of origin which mix up the Moth-worm, Swan, Eacoon, and the Prairie-wolf the Coyote. Among all the tribes visited by Mr. Low, from the Eraser Eiver on the west to the St. Lawrence on the east, the Owl " was portentously sacred." 2 The Greenlanders ascribe their origin to the transformation of dogs into men. Torngarsuk, their good spirit, is described by some of them as a bear. 3 Female kinship is evidenced among them by the fact that a man's goods go to his sister's children. 4 In case of divorce children always follow their mother, and even after her death will not help their father in his old age. If a husband divorces his wife she goes back to her relations. 5 This certainly points to exogamy, and it is clear, taking all these details together, that the Greenlanders have come through the stage of totemism and its usual accompaniments, although there is no evidence that they still keep up the clan system. The Kamschatkans, who are allowed on all hands to belong to the same race of mankind, have animal dances, and regard the Walrus, Orca, Bear, and Wolf with reverence. 6 They speak to 1 Bancroft, vol. i. p. 171 ; vol. iii. p. 96. 2 Id. vol. iii, pp. 81, 95, 98, 117, 128. 3 Cidmtz' s Greenland, voli.p. 188. Eng. Tr. London, 1820. Klemm, Allgemeine Cultur-Geschichte der Menschheit, vol. ii. p. 3 1 6. Leipzig, 1843-52. 4 Crantz, vol. i. p. 176. 5 Id. vol. i. p. 148. 6 Klemm, vol. ii. p. 329. 376 STUDIES IN ANCIENT HISTORY CHAP. the animals with respectful forms and never name their names, nor will they kill and eat any land or sea beast without first excusing themselves to it, and asking it not to take it ill. Also if a man kills an otter, he will eat it, but it is a sin. 1 MAREIAGE The Esquimos are devoid of jealousy and lend wives com- monly ; polygamy is practised among them, but polyandry also if women are scarce. The Koniagas have no idea of morality ; "the Kaviaks practise polygamy and incest; the Kadiaks cohabit promiscuously, brothers and sisters, parents and children." The Malemutes have only one wife at a time, but no marriage ceremony, and divorce at pleasure. " Two husbands are also allowed to one woman ; one the chief or principal husband, and the other a deputy, who acts as husband and master of the house during the absence of the true lord; and who, upon the latter's return, not only yields to him his place, but becomes in the meantime his servant." The Aleuts also have no marriage ceremony, but are poly- gamous. "Wives are exchanged by the men, and rich women are permitted to indulge in two husbands." 2 Other authorities give more interesting details. A very peculiar form of marriage is found among the Esquimos, according to Boss. A man has never more than two wives, so far as he observed, but the more able and useful of these two maintains a second husband ; despite this mixture of polygyny and polyandry, the greatest harmony reigns in the establishment. 3 As a rule among the Polar races the wife is brought to live with her husband's family, and the form of capture is observed. 4 Among the Aleutians, owing to wife- lending and similar habits, the husbands cannot claim the children of their wives with any certainty ; the mother alone has full power over them, and even the uncle on the mother's *. 1 Klemm, vol. ii. pp. 328, 329. 2 Bancroft, vol. i. pp. 65 ft, 81, 82, 92. 3 Apud Klemm, vol. ii. p. 204. 4 Id. vol. ii. p. 205. xvm NORTH AMERICA, NORTH 377 side has more importance with them than the father. Children of different women married to the same husband are not accounted to be brothers and sisters, and are allowed to marry, whereas those of the same mother are not allowed to do so even if the fathers be known to be different. 1 Here female kinship and exogamy are still more clearly denoted than they were above among the Greenlanders. 1 Sarytschew, Reise, erster Abschnitt, pp. 166, 167. CHAPTEE XIX MEXICO AND CENTRAL AMERICA THESE nations are divided into the Aztecs to the north ; Mayas to the south; Chichimecs to the north of the Aztecs. The Aztec empire was a confederation of the Acolhuas, Aztecs, and Tepanecs, whose respective capitals, all near each other, were Tezcuco, Mexico, and Tlacopan or Tacuba. The federation was made a hundred years before the advent of the Spaniards. The Aztecs were at first a small tribe, but gradually acquired the predominance. The annals of the confederation go back to the sixth century, and are divided into three periods the Toltec, the Chichimec, and the Aztec. Toltec was for five centuries a confederacy like the later Aztec. It was ruined by civil wars, which were also religious. The kings were overthrown in the middle of the eleventh century, and probably went south with the nobles, priests, etc., and played a part in the history of the Maya Quiche nations. The Chichimec empire lasted till early in the fifteenth century ; there were two main divisions of them, a line of kings at Tezcuco, and wild tribes to the north. After the Chichimecs the valley was entered by the Matlaltzincas, Tepanecs, Acolhuas (the Chichimecs par excellence), Teo- Chichimecs (Tlascaltecs), Malinalcas, Cholultecs, Xochimilcas, Chalcas, etc., and Aztecs. The third empire, that of the Aztecs, lasted till the Spaniards came. The legends of the pre-Toltec period declare that Votan, a serpent, divided the land, taught the people, built the great city Nachan, i.e. city of the serpents, and finally wrote a book CHAP, xix MEXICO AND CENTRAL AMERICA 379 with proof of his claims to be considered one of the Chanes or serpents (Bancroft, vol. v. pp. 159 ff.) Quetzalcoatl found food for men and gods (vol. v. p. 193). " At that time Azcatl, the ant, going to Tonacatepetl (mount of our subsistence) for maize, was met by Quetzalcoatl, who said, ' Where hast thou been to obtain that thing ? Tell me.' At first the Ant would not tell, but the Plumed Serpent insisted, and repeated, ' Whither shall I go ? ' Then they went there together, Quetzalcoatl metamorphosing himself into a black Ant." The yellow Ant accompanied him respectfully, and they got maize. Then the gods got maize and gave some to men. According to the Popol Yuh, he got maize by aid of "the Coyote"; and according to the same authority men were changed into monkeys after the Flood. In the Popol Vuh is a Bat legend of origin of Fire. Zotzil = a bat. " But one family stole the fire, the family of Zotzil of the Cakchiquels, whose god was Chamalcan, and whose symbol was the bat." 1 According to the Codex Chimalpoca men were changed into " dogs," i.e. Chichimecs, after the Flood. Bancroft says (vol. ii. p. 103) that the Chichimecs were called " dogs " by their more polished neighbours. These legends suggest totemism strongly, and there is l '' plenty of more direct evidence. The etymology of some of the civilised nations of North- West America, so far as known, is thus given by Bancroft (vol. ii. p. 125). Acolhuas one meaning, from Colhua". . . chose courbee. Dela le nom de la cite* de ColJvuacan, qu'on traduit indirTerem- ment, ville de la courbe, de choses recourbe'es (des serpents) et aussi des ai'eux, de coltzin, aieul." Aztecs from Az, an ant, according to Brasseur de Bourbourg; but Buschmann rejects this view. Chalcas jade, according to Brasseur. Cheles a species of bird, according to Brasseur. Chichimecs Chichi = dog ; Mecatl = race race or line of the dogs. Cocomes Cocom " est le pluriel de cohuatl = serpent." Huexotzincas, from Huexotla, willow forest. Itzas " suivant Ordonez, le mot itza est compose de itz, doux, et de M, eau." Mayas " Mai," a divinity or ancient person after whom the country 1 P. 548. 380 STUDIES IN ANCIENT HISTOR Y CHAP. was called Maya ; also said to mean earth, or earth without water. Mizquicas an acacia. Miztecs Mixtli = clouds. Olmecs probably from Olli, people of the gum. Pokomams Pokom, white tufa. Quiches tree or forest. Tlahuicas cinna- bar. Toltecs people of Toltan and others, from Tolin = willow, reed, according to Buschmann. Tutul-Xius Totol, a bird. "Each nation [of the Aztecs] had its own standard, on which were painted or embroidered the armorial bearings of the state. That of the Mexican empire bore an eagle in the act of seizing a tiger or jaguar. That of the republic of Tlascala a bird with its wings spread as in the act of flying, which some authors call an eagle, others a white bird or crane. Each of the four lordships of the republic had also its ensign ; Tiratlan had a crane upon a rock, Tepeticpac a wolf with a bunch of arrows in his paws, Ocotelulco a green bird upon a rock, and Quiahuiztlan a parasol made of green feathers. Each com- pany had also a distinct standard, the colours of which corre- sponded to that of the armour and plumes of the chief" (vol. ii. p. 411). Among the Zapotecs "it was customary to assign some animal or bird to a child as its nagual or tutelary genius." The days of the Aztec month are named as follows : 1. Sword- fish ; 2. Wind ; 3. House ; 4. Lizard ; 5. Snake ; 6. Death ; 7. Deer; 8. Babbit ; 9. Water; 10. Dog; 11. Monkey; 12. Brushwood; 13. Cane; 14. Tiger; 15. Eagle; 16. Vulture; 17. Movement; 18. Flint; 19. Bain; 20. Flower. Like the signs of the Zodiac, these names suggest totemism. According to a legend of the founding of Mexico Tenochtitlan, it was built where an eagle with a serpent in its beak and a nopal (a kind of tree) were found together. Diaz noticed in the temple at Mexico idols " half human, half monstrous in form, and found the rooms blood-stained," i.e. they sacrificed to these idol sphinxes. " Among other divinities (of the Zapotecs) a species of parroquet with flaming plumage, called the Ara, was worshipped in some districts" (Bancroft, vol. ii. pp. 211, 277, 511, 560, 583). Idols, etc., are found in Guatemala, as also in Copan. " In one the human figure has a head-dress, of which an animal's xix MEXICO AND CENTRAL AMERICA 381 head forms a prominent part, while in another the head is half human and half animal" (vol. iv. p. 113). " Statues in stone representing human beings, but in some cases animals and monsters, also have been found (in the Chontal Province), and described to the number of about sixty." The island of Zapatero yielded some seventeen idols of great size. Ten are figured : 1. Human and a cross. 2. Huge tiger. 3. Human. 4. Human. 5. Head of a monster surrounding head of a seated human form (said to be " a common device in the fashioning of Mcaraguan gods"). 6. Human. 7. Crouching human figure, on whose back is a tiger or other wild beast grasping the head in its jaws. 8. Human. 9. Human with animal mask. 10. Something like human (vol. iv. pp. 40, etc.). From the island of Pensacola come three idols of stone. 1. Human face surmounted by a monster head, and by its side the open mouth and fangs of a serpent appear. 2. Animal clinging to the back of a human being (alligator, I think). 1 3. Hideous monster with eyes of owl or bird of prey (vol. iv. pp. 48-51). " At the Indian Pueblo of Subtiava, near Leon, many idols were dug up. ... The natives have always been in the habit of making offerings secretly to these gods of stone, and only a few months before Mr. Squiers' visit a stone bull had been broken up by the priests " (vol. iv. p. 54). Observe how, as by an accident, Bancroft notices totems among the Cuebas. The chief on getting the title chose a certain device "which became that of his house." If his son adopted " the ancestral totem," he could not afterwards change it on becoming chief. " A son who did not adopt his father's totem was always hateful to him during his lifetime " (vol. i. p. 753). Among the Zunis " the Frog, the Turtle, and Eattlesnake are minor gods" (vol. iii. p. 132). The ancient monuments of this region are full of indica- tions of totemism. In the department of Guanacaste, near the Gulf of Mcoya, was found a little frog on grey stone. A hole 1 An alligator's head from the ruins of Copan is figured at p. 101. 382 STUDIES IN ANCIENT HISTORY CHAP. near the fore-feet would seem to indicate its use as an orna- ment or charm. At Tollan was a temple to the goddess of water or the Frog-goddess. Mendieta says that the Mexicans painted the earth-goddess as a frog with a bloody mouth in every joint of her body (Bancroft, vol. iii. p. 351 note ; vol. iv. p. 24; vol. v. p. 265). In Nicaragua are found carvings representing the monkey and the coiled feathered serpent, the Aztec Quetzalcoatl or Quiche Gucumatz, and other animals (vol. iv. pp. 37, etc.). In Yucatan a row of turtles is sculptured on an ancient building. From Uxmal comes a double-headed stone animal. The serpent is everywhere in Yucatan; serpent -heads are found in decorations. There was a serpent temple. "Two serpents, each with a monster's head, between the open jaws of which a human face appears, and the tail of a rattlesnake, almost entirely surround the front above the lower cornice " of a temple at Uxmal. Serpents are found again in a " stone ring" at Chichen (vol. iv. pp. 166, 168, 183, 185, 186, 231). A coiled serpent from Mexico is figured at vol. iv. p. 498. Eelics found in the valley of Mexico are (inter alia) a crouching monster of stone, a lizard of stone, a Maltese cross, a flat-fish, a coiled serpent in red porphyry, a rabbit in low relief on stone. At Tepeaca have been found a serpent and a monster eagle. At Cuernavaca (this word means " place of the eagle ") an eagle is carved on an isolated rock (pp. 467, 482, 497). "In the city of Mexico I continually saw serpents carved in stone in the various collections of antiquities . . . two feathered serpents among them. ... On the benches [of the museum] around the walls and scattered over the floor are numberless figures of dogs, monkeys, lizards, birds, serpents" (Bancroft, vol. iv. p. 5 54, quoted from Mayer's Mexico as it was). "On the slope of the hill four leagues north-west of Santiago, at the foot of Lake San Juan, was found a crocodile of natural size, carved from stone, together with several dogs or sphinxes, and some idols, which the author [Senor Eetes] deems similar to those of the Egyptians " (Ban- croft, vol. iv. p. 575). In the volcanic region south-west of the city of Durango was found among other things " a very small stone turtle, not over half an inch in diameter, very perfectly xix MEXICO AND CENTRAL AMERICA 383 carved from a hard material." A series of boulders are found in Arizona and New Mexico, covered with the usual totem marks and writing familiar among the Eed men. Eock-paintings and rock-inscriptions of precisely the same character are found in Utah (vol. iv. pp. 600, 620, 641, 716, 717). Suggestions of totemism are found also among the modern tribes in this region. The Apaches and Navajos will not eat or kill bear or rattlesnake. "The Papagos stand in great dread of the coyote, and the Pimas never touch an ant, snake, scorpion, or spider." The wild tribes of Central America usually give a child the name of some animal, " which becomes its guardian spirit for life." In Guatemala "the Itzas hold deer sacred," and again, " certain natives of Guatemala, in the province of Acalan," kept tame deer, holding them sacred (Ban- croft, vol. i. pp. 491 note, 553, 703, 707 ; vol. iii. pp. 131, 132). In the more civilised nations of this region we find a jumble of tribal gods and metaphysical refinements. Thus J. G. Muller wonders that Acosta did not know of Teotl, the highest invisible god, " he through whom we live," and " he who is all things through himself" (Bancroft, vol. iii. p. 183). This pantheistic idea is manifestly a late development; no prayers were made, no offerings or feasts given in honour of this god ; in fact, he was not a god of the common people at all. As a general rule, the Mexican gods are only emerging from the totem stage. Some of them " were shaped like men, some were like women ; . . . some were like wild beasts, as lions, tigers, dogs, deer, and such other animals as frequented the mountains and plains ; . . . some like snakes of many fashions, large and coiling; . . .of the owl and other night- birds ; l and of others, as the kite, and of every large bird, or beautiful, or fierce, or preciously feathered they had an idol. 2 But the principal of all was the sun. Likewise they had idols of the moon and stars, and of the great fishes, and of the water lizards, and of toads and frogs, and of other fishes ; and these, 1 In the Aztec mythology figures an evil genius called the " Owl-man." 2 The "Menagerie" at the palace was then a real Olympus an assembly of the living gods. 384 STUDIES IN ANCIENT HISTORY CHAP. they said, were the gods of the fishes. . . . They had for gods fire, water, and earth, and of all these they had painted figures. ... Of many other things they had figures and idols, carved or painted, even of butterflies, fleas, and locusts " (Motolinia, in Bancroft, vol. iii. p. 196). The image of Quetzalcoatl had the body of a man but the head of a bird, the sparrow with a red bill. But he was also a snake-god. 1 He was adored in Yucatan as " Cuculcan," that is, "snake covered with god-like feathers." The entrance to his temple at Mexico represented the jaws and fangs of a tre- mendous snake. And he was also worshipped as a stone. At Achiuhtla, " in a cave the interior of which was filled with idols, set up in niches upon stones dyed with human blood and smoke of incense, was a large transparent chalchiute, entwined by a snake whose head pointed toward a little bird perched on the apex. This relic, worshipped since time im- memorial under the name of the ' heart of the people,' has all the chief attributes of Quetzalcoatl, the stone . . . the snake, and the bird " (Bancroft, vol. iii. p. 449). It was said to be Votan as well as Quetzalcoatl. If Mliller is right, Votan was a pure snake-god to whom the bird attribute was given at a later time. The god Huitzilopochtli was the anthropomorphic god of the Aztecs, as Quetzalcoatl was of the Toltecs. His name signifies " on the left side a humming-bird," and his image had feathers of this bird on the left foot. This was not the only decoration. The god had also a green bunch of plumage upon his head shaped like the bill of a small bird. The shield in his left hand was decorated with white feathers, and the whole image was at times covered with a mantle of feathers. He had a spear or a bow in his right hand, and in his left some- times a bundle of arrows, sometimes a round white shield. As the sparrow head in Quetzalcoatl's case, so the hum- ming-bird in Huitzilopochtli's points him out as an original animal god. " The general mythological rule that such animal attributes refer to an ancient worship of the god in question, 1 Schoolcraft (vol.v. p. 105) says that Quetzalcoatl seems to mean " great serpent." Humboldt says, " serpent clothed with green feathers." xix MEXICO AND CENTRAL AMERICA 385 and in the form of an animal, points out this in his case," l and the myth of Huitziton confirms this. This Huitziton (literally "small humming-bird") led the Aztecs from Aztlan as the Picus led the Sabines to Picenum. He, according to the fable, was a man, " and heard the voice of a bird which cried ' Tihui,' that is, ' Let us go/ " Therefore, Huitziton and Huitzilopochtli were originally one, which is the conclusion arrived at by the learned Italian Boturini. " Previous to the transformation of this god by anthropomorphism, he was merely a small humming-bird, huitziton ; by anthropomorphism the bird became, however, merely the attribute, emblem or symbol, and name of the god" (Bancroft, vol. iii. p. 305). But Huitzilopochtli is also a snake-god, and is connected by myth and image with numerous snakes (vol. iii. p. 321). This attribute, added to the original humming-bird attribute, gave rise to a compound animal god such as is familiar in many religions. SUCCESSION Among the Pimas it is said that a man's property is divided up among the tribe at his death. In the provinces of Cueba, Comayagua, and other parts of Darien, the eldest son succeeded to the government on his father's death. It is said generally of the tribes of Guatemala, Salvador, and Nicaragua, that sons inherit " equally " their father's property. " Goods and lands are equally divided among the sons." The sons were heirs among the Southern Mexicans (Bancroft, vol. i. pp. 508, 545 note, 664, 700, 769). Eight of Succession to the Throne. The succession in Mexico is said to have been collateral and elective. " Zurita states that in Tezcuco and Tlacopan and their dependent provinces, 1 Le droit de succession le plus ordinaire e"tait celui du sang en ligne directe de pere en fils ; mais tous les fils n'he'ritaient point, il n'y avait que le fils aine" de 1'epouse principale que le souverain avait choisie dans cette intention. Elle jouissait d'une plus grande consideration que les autres, et les sujets la respectaient davantage. Lorsque le souverain prenait une de 1 Bancroft, vol. iii. p. 304/ 2 c 386 STUDIES IN ANCIENT HIS TOR Y CHAP. ses femmes dans la famille de Mexico, elle occupait le premier rang, et son fils succedait, s'il etait capable.' " He goes on to say that in default of direct heirs the succession became collateral. If there was no heir the successor was elected by the nobles. " In a previous paragraph he writes : ' L'ordre de succession variait suivant les provinces; les memes usages, a peu de difference pres, etaient regus a Mexico, a Tezcuco et a Tacuba.' Afterwards we read: 'Dans quelques provinces, comme par exemple a Mexico, les freres Etaient admis a la succession, quoiqu'il y eut des fils, et ils gouvernaient suc- cessivement.' " " Prescott affirms that the sovereign was selected from the brothers of the deceased sovereign, or in default of them, from his nephews." According to Clavigero (Storia Ant. del Messico, torn. ii. p. 112), s they established by law, that one of the brothers of the dead king should succeed him, and, failing brothers, one of his nephews, and if ever there were none of these, one of his cousins. The electors had only choice between the brothers, etc., in their order. That the eldest son could put forward no claim to the crown by right of primogeniture is evident from statements made by Las Casas (Bancroft, vol. ii. p. 135 note). The following is the order of actual succession of nine Mexican kings from Codex Mendoza : No. 1. Son. Acamapichtli, succeeded by his son, Hui- cilyhuitl. No. 2. Son. Huicilyhuitl, succeeded by his son, Chimal- pupuca. No. 3. Father's brother. Chimalpupuca, succeeded by Yzcoaci, son of No. 1, i.e. uncle, father's brother. No. 4. Brother. Yzcoaci, by his own brother, Huehue- moteccuma. No. 5. Grandson. Huehuemoteccuma, succeeded by Axayacaci, grandson of No. 4. No. 6. Son. Axayacaci, succeeded by his son. No. 7. Succeeded by brother. No. 8. Succeeded by brother. This list gives four cases of collateral, and four of direct. It would agree with succession to " eldest male of the family." xix MEXICO AND CENTRAL AMERICA 387 Bancroft states the law thus : " Eldest surviving brother was generally elected, and when there were no more brothers, the eldest son of the first brother that had died." " The father of Montezuma II. had 150 children, of whom Montezuma killed all his brothers." The tenant or feuar of crown lands was succeeded therein by his eldest son ; so the lands of the nobility were transferred by inheritance from father to son. Herrera, however (Hist. Gen., dec. ii. lib. vi. cap. xvii.), says that brothers inherited estates, and not sons. " Throughout Zapo- tecapan and Miztecapan landed property was transmitted from male to male, females being excluded." This means from father to son. In the province of Panuco the eldest son was the sole inheritor of the land (Bancroft, vol. ii. pp. 135, 183, 224, 228, 230). TJie Maya Nations. The order of succession to the Quiche* throne, all are agreed, preferred the brother to the son. " Padre Ximenez implies, perhaps, that the crown descended from brother to brother, and from the youngest brother to a nephew who was a son of the oldest brother." This would seem the sense of the authorities as cited by Bancroft. At Mayapan, according to Brasseur de Bourbourg, the king's brother suc- ceeded to the throne. " Ce n'etaient pas ses fils qui suc- ce'daient au gouvernement, mais bien Faine' de ses freres." But Bancroft says there is no text for this, and that the eldest son succeeded, the brother ruling only during minority. "At a man's death his property, in Yucatan, was divided between his sons equally." As a general rule the daughters got nothing ; a son who had helped his father might get more than a share (Bancroft, vol. ii. pp. 634, 639, 653). MARRIAGE The New Mexicans are polygynous; they acquire their wives by purchase ; in case of separation the children follow the mother. The Pueblos also get wives by purchase, but are monogamous ; " if dissatisfied," they divorce them, and in that case, " if there are children, they are taken care of ly {heir grandparents, and both parties are free to marry again." 388 STUDIES IN ANCIENT HISTORY CHAP. But it is not stated which grandparents are meant. Among the Papagos " polygamy is not permitted," but " they often separate and marry again at pleasure." The Lower Californians, described as houseless wanderers, have no marriage ceremony, and no word to express marriage. " Like birds or beasts, they pair off according to fancy." Among the Pericuis, a tribe now extinct, a man married as many women as he pleased ; among the Guaicuris and Cochimis polygyny is less common than it was among the Pericuis, because they have more men than women (Bancroft, vol. i. pp. 512, 548, 549, 565, 570). The Tahus buy their wives ; the chief or high priest has jus primae noctis. The Southern Mexicans (Zapotecs, Miztecs, Mayas, etc.) are far advanced in laws, government, manufactures, and agriculture. They have monogamy and male kinship, and practise " marriage within the rancho or village." This is not a clear case of endogamy, however, for we find it said that they are " all relatives since reduction by cholera," so that previously they were not all relatives. Brothers and sisters could not marry (Bancroft, vol. i. pp. 584, 657 ff., 663). The wild tribes of Central America were all polygamous at the time of the Conquest. They marry early, and " usually within the tribe," but " the tribe " is not defined, and seems to mean nothing here but a local tribe. Chiefs consult as to the expediency of any marriage between different tribes. " Several tribes in Guatemala are strictly opposed to marriage outside of the tribe, and destroy the progeny left by a stranger." Does this imply endogamy ? They " remain under the parents' roof until married, and frequently after, several generations often living together in one house under the rule of the eldest" (Bancroft, vol. i. pp. 633, 702, 703, 704). In Nicaragua it is stated that " no one might marry within the first degree of relationship, but beyond that there was no restriction." The Caribs have a separate house for each wife ; widows belong to the husband's relatives (vol. i. p. 731 ; vol. ii. p. 666). Among the Aztecs the men married from twenty to twenty- two, girls from eleven to eighteen. " Marriages between blood relations, or those descended from a common ancestor, were not xix MEXICO AND CENTRAL AMERICA 389 allowed." This is a clear note of exogamy and male kinship. " A brother could, and was enjoined to, marry his deceased brother's wife, but this was only considered a duty if the widow had off- spring by the first marriage, in order that the children might not be fatherless." This seems extremely improbable, for elsewhere the levirate is instituted to meet the case of a widow who has no children. Bancroft, moreover, contradicts himself on the point by elsewhere quoting Las Casas as an authority to the effect that " it was customary for a man to raise up seed to his deceased brother by marrying his widow" (vol. ii. pp. 251, 466). Concubinage was permitted throughout the Mexican empire. Prostitution was tolerated, the law taking cognisance of the women (vol. ii. pp. 2 64, 2 6 6). It is stated that " the Toltec kings could only marry one woman, and in case of her death could not marry again, or live in concubinage with any woman." Among the Chichimecs " marriage with near relatives was never permitted, and polygamy strictly prohibited," but we are not told what "near" means. Although marriage was early contracted, there could be no legitimate intercourse till the wife was aged forty (vol. ii. pp. 2 6 5, 2 6 1, 2 6 2). The Guate- malans recognised no relationship on the mother's side only, and did not hesitate to marry their own sister provided she was by another father. Thus if a noble lady married an inferior or slave, the children belonged to the father's order. Torquemada says they sometimes married their sisters-in-law and step-mothers. " Among the Pipiles of Salvador an ancestral tree, with seven main branches, denoting degrees of kindred, was painted upon cloth, and within these seven branches or degrees none were allowed to marry," except as a reward for services. " Within four degrees of consanguinity none on any pretext might marry." "In Yucatan there was a peculiar prejudice against a man marrying a woman who bore the same name as his own, and so far was this fancy carried, that he who did this was looked upon as a renegade and an outcast. Here also a man could not marry the sister of his deceased wife, his step-mother, or his mother's sister, but with all other relatives on the maternal side, no matter how close, marriage was perfectly legitimate." Here we find exogamy and agnation. 390 STUDIES IN ANCIENT HISTORY CHAP. " In Nicaragua no one might marry within the first degree of relationship, but beyond that there was no restriction " [?]. Among the Maya nations the widow was regarded as the pro- perty of the family of her deceased husband, to whose brother she was invariably married, even though he might have a wife of his own at the time. If there w T as no brother, then the nearest relative on the husband's side married her. Yet it is also said that monogamy seems to have been the rule among the Maya nations, and many say polygamy did not exist. Prostitution was " tolerated, if not encouraged " among these nations. There were public houses, and parents were in the habit of sending daughters on town to earn marriage portions (vol. ii. pp. 664-676). " According to the system of relationship in use amongst the Indians of Yera Paz, it frequently resulted that brothers must marry sisters, of which the reason was as follows. They were accustomed not to make marriages between the men of one village and the women of the same village, and they sought the women of the other villages, because they did not count as belonging to their own family sons born in a foreign tribe or lineage, even though the mother might have issued from their own lineage ; and the reason of this was because relationship was counted through the men alone. So that if any chief gave his daughter to another of another village, even if the chief had no heirs except his grandchildren (sons of his daughter), he did not recognise them as grand- children, or as relations, so as to make them his heirs, on account of their being sons of another chief of another village ; and so there was sought for such [? the latter] chief, a wife who belonged to another village, and not to his own. And on account of their (the kindred) being in another village, thus it came about that they did not hold the sons of these women as related to the kindred of their mother. And this must be understood as applying to marriages with them (women of other villages), which they considered lawful, although, in other respects, they recognised the relationship [literally, they recog- nised one another]. And because the reckoning of relationship was through the men alone, and not through the women, on xix MEXICO AND CENTRAL AMERICA 391 this account they did not consider that there was any impedi- ment to marriages with such relations ; and so they married in all degrees of consanguinity, because they considered any woman of their own lineage (although the relationship might be very remote, and though they had no recognition of the degree of relationship) more as a sister than they did the daughter of their own mother who might have had a different father ; and through this mistake they married their sisters through the mother, but not through the father " (Torquemada, Monarq. Ind., vol. ii. p. 419 ; apud Bancroft, vol. ii. p. 664).^ From this confused statement it appears that in Vera Paz the Indians lived in clans in distinct villages (Pueblos) with male kinship and exogamy. Marriage with a half-sister on the mother's side would therefore be lawful, and naturally must have happened often enough. CHAPTER XX PERU THE authority used for the empire of the Yncas was : " The First Part of the Royal Commentaries of the Yncas, by the Ynca Garcilasso de la Vega. Translated and edited by C. E. Markham. London, printed for the Hakluyt Society, 1869. Two volumes. TOTEMS Stripped of his explanation of the origin of the different cults, the statement of Garcilasso is as follows : Before the times of the Yncas each province, each nation, each house (gens ?), had its gods different one from another. They (i.e. some of them) worshipped herbs, plants, trees, high hills, great rocks, chinks in rocks, caves, pebbles, and small coloured stones, such as jasper. Some adored the emerald, particularly in the province now called Puerto Viejo. They (i.e. some of them) worshipped different animals, such as the tiger, lion, and bear ; if they met them they went down on the ground to worship them instead of flying for their lives. They (i.e. as ante) adored foxes and monkeys, the dog, the cat, the bird called the cuntur, and some of them the eagle, " be- cause they thought they were descended from it as well as the cuntur." Others worshipped falcons, others the owl, others the bat, others serpents, as among the Antis ; others snakes, lizards, toads, and frogs. " In fine, there was not an animal that they did not look upon as a god . . . merely differing CHAP, xx PERU 393 one from the other as to their gods." On the coasts some worshipped the whale ; others sardines, others the dogfish ; others the golden fish, the crawfish, and the crabs. If their gods were birds or beasts, they offered to them in sacrifice what they usually saw them eat, and what appeared to be most agreeable to their tastes (vol. i. pp. 47, 48, 49, 53). The Huancas, a warlike people, " pretend descent from one man and one woman, who came out of a fountain." They live in small fortified villages, and often dispute over their boundaries ; yet are said to be of one race. " Before they were conquered by the Yncas these people worshipped the figure of a dog, and had it in their temples as an idol, and they considered the flesh of a dog to be most savoury meat." " They also had an idol in the form of a man, from which the devil spoke." They were allowed by the Yncas to retain this, but the dog idol was destroyed. Head-dresses of different colours were instituted by the Ynca in their three provinces. In the province of Chucurpu they worshipped the tiger. In that of Huamachucu they worshipped stones such as jasper ; kept them in houses, and offered human sacrifices to them (vol. ii. pp. 128, 129, 131, 137). The Chunchus went about almost naked, but wore great plumes on their heads composed of the feathers of macaws and parrots. The Chachapuyas worshipped serpents, and had the bird Cuntur as their principal god ; near them the Huacra- chucu worshipped serpents, " and had them painted as idols in their temples and houses" (vol. ii. pp. 264, 322). In the large and populous province of Huancapampa, " each tribe worshipped many gods, each household having its own." These gods were animals, plants, hills, fountains, etc. The inhabitants of the province of Canaris worshipped trees and stones, e.g. jasper. In Quitu the gods were "deer and great trees." In Caranque various people worship tigers, lions, goats, serpents, " offering human hearts and blood to them as sacrifices." The victims were captives, and were eaten. On the shore near Tumpiz they worshipped lions and tigers, and sacrificed human hearts and blood. The Manta nation " worshipped the sea and fishes ; also tigers and lions, great 394 STUDIES IN ANCIENT HISTORY CHAP. serpents, and other reptiles ... a great emerald as large as an ostrich egg." They sacrificed to it from far and wide, and offered little emeralds to it (vol. ii. pp. 329, 330, 335, 344, 350, 425, 441). Chimu, chief of all the valleys from beyond La Barranca to the city now called Truxillo, was conquered by the Ynca, who forgave his resistance on the condition that he should adore the sun, and cast his idols to the ground, "being the figures of beasts and fish" (vol. ii. p. 195). As a general rule the Indians of the coast, over 500 leagues, from Truxillo to Tarapaca, which are at the north and south extremities of Peru, worshipped the sea in the shape of fish (in addition to the special idols which were peculiar to each province). Fish was used both as sfood and manure. " They also generally worshipped the whale . . . and besides this, some provinces adored one kind of fish, and others another." It appears that some Yuncas worshipped the fox, and had an image of it with other gods in a temple to Pacha- camac the invisible the only temple he had in Peru, where human sacrifices were sometimes offered to him (vol. ii. pp. 147, 186). The people of the province of Aymara sent messages to the Curacas (chiefs) of Uma-suyu. They were told by the Ynca " that their own idols, which they called the gods of their land, were merely the figures of vile and filthy animals!' The gods of tribes in the coast valleys of the province of Camana were fishes (vol. i. pp. 237, 245). The Chancas of the province of Antahuaylla were rich and warlike, and claimed descent from a lion. "Wherefore they adored the lion as a god." They used to dress as Hercules is painted, "covered with lions' skins, and their heads thrust into the skulls of lions." Under the name of Chancas were several tribes, who all boasted their descent from various fathers, such as a lake, a hill, or a fountain. One of these tribes is noted by Markham as occupying a wild part of the Cordillera (vol. i. p. 323). The Indians called Antis " usually worshipped the tigers as gods. Also the serpents that they called Amaru. . . . xx PERU 395 They also adored the herb called cuca or coca" (vol. i. p. 330). " The Collas consist of many different nations, and believe themselves descended from various things ; e.g. some from the lake Titicaca, whom they adored as a mother, offering sacrifices on its banks ; others drew their lineage from a fountain ; others from caves and recesses in rocks places in all cases sacred and sacrificed to ; others from a river, as their father the fish in it their brothers." "There was only one deity which all the Collas united in worshipping and holding as their principal god." This was a white sheep, "and they offered to it lambs and grease as sacrifices" (vol. i. p. 168). In the times before the Yncas, a beetle was worshipped in Chuquisaca, but this was prohibited by the Yncas (vol. i. p. 121). The Yncas established the worship of the sun everywhere, and tried to put down the totemistic religions. "After subjugating a province, the first thing the Ynca did was to take the principal idol as a hostage and send it to Cuzco, ordering it to be kept there in a temple until the chief and people . . . were taught the idolatry of the Yncas." Among other things there were offered to the sun images of all the animals in the provinces, each imitated from nature in gold and silver. Collections were made in and near Cuzco of the various sacred animals of the provinces, e.g. Garcilasso names the part of the city where the Amarus were kept, called Amaru cancha. Markham notes a huge stone lintel with two serpents in relief upon it, in a house still in that district (vol. ii. pp. 30, 35, 160). Though the Yncas called themselves children of the sun, and insisted on all their subjects becoming worshippers of the sun, it is clear enough that they had once belonged to another totem themselves. For the Ynca bore the device of two wing-feathers of a bird called Coraquenque, a bird looked on as sacred. And " in the time of the Yncas each Indian was known by his head-dress, which showed to what tribe and nation he belonged" (vol. ii. p. 334, and compare p. 179). It seems, therefore, that at some time the Yncas were known by this sacred bird, and the presumption is that, when they became 396 STUDIES IN ANCIENT HISTORY CHAP. the rulers of the empire, they took the sun for their totem instead as being more dignified. The Ynca Manco, who is said to have first established solar worship, was readily be- lieved by the Indians to be descended from the sun, because the Indians believed similar stories about themselves, " though they did not comprehend how to select ancestors as well as the Ynca, but adored animals and other low and earthy objects" (vol. i. p. 83). Manco also distinguished the tribes by ear- holes and ornaments " signs intended to prevent confusion between one tribe and another," and extended the name Ynca "to principal vassals." In the temple of the sun at Cuzco was a cloister with five halls, four of them respectively dedi- cated to the moon, stars, lightning, and rainbow, and the fifth set apart for the high priest. Thus a full-blown nature-religion reigned over the previous totemism. And the " philosophers " of Peru went still further, maintaining that Pachacamac (creator or sustainer of the world) was invisible and could only be adored mentally ; the sun was not identical with him, but put by him above all, to be adored by all as a god (vol. i. pp. 101, 106). SUCCESSION The name Ynca = person of blood royal, was applied to all who were descendants " in the male line, but not in the female line." Queen = Ccoya. Concubines of the king, being his relatives, and all other women of royal blood were called Palla. Succession to the throne was from father to son ; but note the marriage law. If the Ynca had no child by his first sister, he married the next, and so on till he had children. Failing sisters, he married the most nearly related cousin, or niece, or aunt in the royal family, and on the failure of male heirs, " she might have inherited the kingdom as in the laws of Spain." In defence of sister marriages they pleaded that the heir should be legitimate by both father and mother, " for otherwise they affirmed that the prince might be bastardised through his mother." Failing sons by the legitimate wife, the eldest legitimate relative of pure blood inherited. " It was on xx PERU 397 account of this law that Atahuallpe (not of pure hlood) destroyed the whole royal family" (vol i. pp. 95, 96, 159, 309, 310). Among the Curacas (chiefs), who were lords or vassals, various customs prevailed as to inheritance of estates. 1. In some provinces the eldest son inherited, and succession was from father to son. 2. In others a son succeeded, but there was election among the sons. 3. In others sons "inherited according to their respective ages." When the father died the eldest son succeeded him, then the second, and so on ; when the sons all died the succession went to the sons of the eldest, and afterwards to those of the others. Some Spaniards allege this (3) as the succession law in Peru, but Garcilasso stands to it as above (vol. i. p. 311). These customs were prior to the Yncas. As to Atahuallpe, King of Quitu, murdering all the blood royal, (1) it was necessary that the inheritor of the kingdom of Cuzco should be the son of a legitimate wife, who must be a sister of the king, " the inheritance of the kingdom being derived as much from the mother as from the father." (2) In default of such an heir the inheritor ought to be at least of the legitimate blood royal son of a Palla of untainted descent. He had neither qualification ; therefore he met the difficulty " by the cruel destruction of the whole royal blood," not only of possible heirs, but of possible claimants with such qualifica- tions as he had himself. " This is the remedy usually resorted to by all those kings who have usurped power by violence, for they have believed that if there is no legitimate heir to whom the vassals can turn, they will be secure in conscience and justice. Both ancient and modern history give full testimony to this. ... It will suffice to allude to the bad custom of the house of Othman, which is, that the successor to the empire shall bury all his brothers with their father, that he may be safe from them" (vol. ii. pp. 515, 516). MARRIAGE According to tradition, in the old times, men and women cohabited like beasts ; there was no law of incest, but men 398 STUDIES IN ANCIENT HISTORY CHAP. lived with their sisters, daughters, and mothers. Others excepted their mothers. " In other provinces the nearest relations of the bride and her most intimate friends had connection with her, and on this condition the marriage was agreed to, and she was thus received by her husband." Among the Collas license was allowed to women before marriage, and "the most dissolute were most sought in marriage." The Ynca Manco Ccapac taught the people to live in villages, to cultivate the land, build houses, construct channels for irriga- tion, etc., to respect the wives and daughters of others ; decreed death to adulterers, homicides, and thieves ; instituted mono- gamy, " and that marriages should take place between relations so as to prevent confusion in families " ( = endogamy). A native of one province might not marry the native of another, or remove from one province to another. " All were to inter- marry in their own village, and with members of their own families, in order that the lineages and tribes might not be confused and mixed" (vol. i. pp. 58, 81, 169, 308). " Besides the legitimate wife, these kings had many concu- bines, some of them being relations of and within the fourth degree, and others no relations." It seems that relation beyond the fourth degree was not taken into account at all. "All those of the blood royal married with their relations to the fourth degree, but they reserved the daughter, whose marriage to a brother was only permitted in the case of the king" (vol. i. pp. 310, 311). Though the assertion that there had been " no law of incest " previous to Manco Ccapac is no doubt a gross exaggeration, owing to the ignorance of tradition as to what the law really was, still it is clear that the highly advanced form of marriage we find in Peru had been developed out of a savage state of things. Endogamy is represented as a comparatively recent institution. If there is no clear relic of exogamy, the evidence for its accompaniment, totemism, is plain enough. That female kinship had once been the recognised form of kinship is shown by the sister-marriages of the Yncas, "that the heir might inherit the kingdom as much from his mother as from his father" (Garcilasso, vol. i. p. 93). xx PERU 399 CLASSIFICATION The general language of Peru had two names for a son. The father said Churl, and the mother Huahua. Both words meant a child, including those of both sexes and numbers. . . . The parents could not misuse the words without making a male female, and a female male. To distinguish the sexes they had the words which signify male or female. ... There were four words to express brothers and sisters. The male to the male said Hua/uque for brother. The female to the female Nana for sister. But if a brother said Nana to his sister he would be making a woman of himself. In like manner if a sister said Huauque to her brother, though it means brother, she would be making a man of herself. The brother said to his sister Panci, and the sister to her brother Tom. But a brother could not say to his sister Tora, nor a sister to her brother Pana. " Thus there are words of the same meaning appropriated some to the use of men, and others to the use of women" (vol. i. p. 314). A further note of the classificatory system is found at vol. ii. p. 345. Huayna Ccapac "never refused a request made to him by a woman, whatever might be her age, rank, or condi- tion, answering each one according to her age. To those who were older than himself he said, ' Mother, do that which you desire.' To those who were about his own age he said, ' Sister, let it be as you wish ' ; and to those younger than himself he said, ' Daughter, let it be as you would have it.' " The system seems to have fallen into disuse by the time of Garcilasso, who evidently does not understand the bearing of what he says. CHAPTER XXI GUIANA THE Arawaks of Guiana show the whole totemic system in a very perfect form, though Mr. Brett 1 thinks their organisation " was probably much more perfect in former times than it is at present. They are divided into families, each of which has a distinct name, as the Siwidi, Karuafudi, Onisidi, etc. Unlike our families, these all descend in the female line, and no indi- vidual of either sex is allowed to marry another of the same family name: Thus, a woman of the Siwidi family bears the same name as her mother, but neither her father nor her husband can be of that family. Her children and the children of her daughters will also be called Siwidi, but both her sons and daughters are prohibited from an alliance with any indi- vidual bearing the same name, though they may marry into the family of their father if they choose. These customs are strictly observed, and any breach of them would be considered as wicked." [How the families were distinguished otherwise than by their names, and what the names themselves signify, Mr. Brett does not tell us. But that they were genuine totem clans is shown plainly by another work (which Mr. M'Lennan never saw). Mr. Irn Thurn 2 gives a list of forty-six of these family names, with translations ; among these we find grass, rain, deer, ourali, two sorts of wild plantain-tree, black monkey, 1 Indian Tribes of Guiana, p. 98. London, 1868. 2 Among the Indians in Guiana, pp. 176-184. London, 1883. CHAP, xxi GUIANA 401 hyawa tree, red-breasted bird, dakama tree, tortoise, awara palm, rat, three kinds of bee, mocking-bird, wild thorn-tree or white ant (the interpreters differ on this point), a kind of parrot, armadillo, hawk, tree-palm, an insect called the razor-grinder, and a night- jar. They say generally that they are descended from these animals or plants " their fathers knew how, but they have forgotten."] Beena marriage is the form commonly practised among the Arawaks. " The wife's father expects the bridegroom to work for him in clearing the forest, and in other things, and the young couple often remain with him until an increasing family renders a separate establishment necessary." 1 It may be inferred from Brett's account that in such cases monogamy is the rule, but that after a man has set up for himself he often has several wives. It is among this people that the couvade is found along with female kinship as noticed above. 2 " On the birth of a child, the ancient Indian etiquette requires the father to take to his hammock for some days as if he were sick, and receive the congratulations and condolence of his friends." 3 This of course shows what might have been expected a priori, that the couvade was first instituted before male kinship was actually established, for after it had been established the couvade would have been unnecessary. The ordinary view that it was a means of giving to the father some right over his child is obviously in harmony with this. The Caribs may be considered to furnish the classical in- stance of the couvade, but, unlike the Arawaks, have progressed to male kinship, as is evident from the following story given by Brett (p. 354). "A high-spirited Caribi girl, indignant at being given in marriage to an elderly man, who had already other wives (one being her own sister), ran away from him, and bestowed her hand on one of the Essequibo Caribs, a younger man whom she liked better. After a while the old man visited that quarter ... to claim compensation for the loss of her services. It was willingly allowed ; and for a gun, a barrel of salt, or some article of like value, the woman was 1 Brett, p. 101. 2 Supra, p. 29. 3 Brett, loc. cit. 2 D 402 STUDIES IN ANCIENT HISTORY CHAP, xxi left with the man of her choice, who perhaps thought himself secure, and the business well ended. But the next year the old man, who well knew what he was doing, paid them another visit, still, as he said, in quest of compensation. On being reminded by the husband that he had already been paid for the woman, he replied, ' Yes, for the woman ; but she has since borne you a child, you must now pay me for that' The unwritten law of Caribi usage was decidedly in the old man's favour, and he received compensation for that child. For each succeeding birth he could, if he chose, reappear like an unquiet spirit, make a similar demand, and be supported therein by the custom of his nation." The children therefore belonged to the husband, not the wife. In harmony with this is another story of a young Carib who attempted to avenge his father's death. "Whether he considered himself as bound by their fearful custom to be the avenger of blood, we know not" (p. 195). Their marriage system was polygamous, and they princi- pally depended upon capture for their wives, insomuch that the women in a group spoke a different language from the men. On this point see Studies, first series, p. 33, with the references to Humboldt. Whether, however, a definite law of exogamy prevailed among them does not seem to be known. Polyandry is mentioned as coexisting with polygamy among the Waraus. An old Warau objected strongly to putting away either of his two wives ; " on being asked why a man should have two wives, and a woman not be allowed two husbands, he directly said that his tribe did not consider either practice to be bad ; and that he knew a Warau woman who had three " (p. 178). Blood-feud existed in all its force throughout Guiana ; Brett considers the worst feature in the character of the natives to be " their proneness to blood revenge, by which a succession of retaliatory murders may be kept up for a long time." Even if a man dies naturally a sorcerer is employed to point out the guilty cause of death, and " a near relative of the deceased is then charged with the work of vengeance." He is supposed to be possessed by a destroying spirit called Kanaima, and is so called himself. SECTION III AFRICA CHAPTER XXII GUINEA THE KINGDOM OF ASHANTEE OF Ashantee, its geography, history, religion, laws, and customs, we have an excellent account from the pen of Mr. T. Edward Bowdich, who conducted a mission from Cape Coast Castle to the capital of that kingdom in 1817. The motive of the embassy was chiefly political, but it was an exploring ex- pedition as well. Mr. Bowdich's report was published in London in 1819. 1 Little is known about the Ashantees of earlier date than the year 1700. They are mentioned by Bosnian and Barbot. Their first contact with the English seems to have occurred in 1807, when an Ashantee army reached the coast, waging war against the Fantees. Twice thereafter, in 1811 and 1816, they struck at the Fantees, and these wars were the immediate cause of the embassy. 2 1 Mission from Gape Coast Castle to Ashantee, by T. Edward Bowdich, Esq., Conductor. John Murray, London, 1819. 2 The Ashantee kingdom is supposed to have been established about the beginning of the eighteenth century. It was founded, and afterwards greatly extended, by conquest. The first king, Sai Tootoo, was one of the leaders of the expedition from a country nearer the coast by which it was established, and became king by election, or war-chief rather, for in civil matters the other leaders remained his peers. The kingly power, however, though still in form controlled in affairs of government by the representatives of these leaders, had become virtually despotic by 1819. The ordinary rule of succession (for which see p. 411 infra) being held 406 STUDIES IN ANCIENT HISTOR Y CHAP. TRIBES, CLANS, AND TOTEMS In Mr. Bowdich's account of Ashantee, the term nation is employed in the same sense in which it is used of the North American Indians. Mr. Bowdich, for example, speaks (p. 231) of the Braffoes as a Fantee nation, as we speak of the Mohawks as an Iroquois nation ; and at p. 146 we are told how Apokoo " had himself conquered five nations since the beginning of his reign," and how twenty -one nations paid tribute to Ashantee. He does not use the word tribe at all, except in the sense of family or tribe of descent, and in this sense he constantly uses it. The word clan he does not use, but it would be the term applicable to each of the branches of his families or tribes of descent which were contained in the several nations. As to the history of the kingdom of Ashantee, the scanty traditions that exist indicate that it arose from the consolidation of several local tribes or nations, under the authority of a nation more warlike and enterprising than the rest. The case of the Iroquois is recalled by the facts disclosed in Ashantee, and had the Onondagas, after having overmastered the Mohawks and the Oneidas, estab- lished themselves under the sovereignty of their war-leader, instead of entering into the Iroquois league, they would have presented us with the nucleus of such a kingdom as Ashantee became. Nor do the conditions under which the Ashantee kingdom formed itself appear to have differed much from those under which the Iroquois league attained its pre- eminence. The Ashantees, to give the term a general mean- good for the kingship, Sai Tootoo was succeeded in turn by three of his brothers. The last of these is said to have been succeeded by his grand- son his daughter's son so that either there was a failure of heirs in the female line (in which case a son would have been entitled to succeed, and a daughter's son was in the same position) or daughter and grandson are in this statement only terms of address denoting a sister's daughter and her son. A younger brother of this grandson or grandnephew was reign- ing at the time of Mr. Bowdich's visit. The family name of all the kings was Sai. xxii GUINEA THE KINGDOM OF ASHANTEE 407 ing, were of one stock with the Fantees and people of Akim, Warsaw, Assin, and Aquapim, as the Iroquois were one with the Hurons, the Eries, the neutral nations, and the Tuscarora, and they consolidated their power on the ruins of their kinsmen. Mr. Bowdich states that the Ashantee, Fantee, Akim, Warsaw, Assin, and Aquapim languages were indis- putably dialects of the same language, their identity being more striking than that of the dialects of the ancient Greek. And he adduces curious evidence of the former identity of these peoples and of part of the Ahanta nations in a tra- dition " that the whole of these people were originally com- prehended in twelve tribes or families, the Aquonna, Abrootoo, Abbradi, Essonna, Annona, Yokb, Intchwa, Abadie, Appiadie, Tchweedam, Agoona, and Doomina." In these, he adds, " they class themselves still without any regard to national distinction. For instance, Ashantees, Warsaws, Akims, Ahantas, or men of any of the nations before mentioned, will severally declare that they belong to the Annona family ; other individuals of the different countries that they are of the Tchweedam family ; and when this is announced on meeting they salute each other as brothers. The king of Ashantee is of the Annona family, so was our Accra and one of the Fantee linguists ; Aman- quatea is of the Essonna family." The Aquonna, Essonna, Intchwa, and Tchweedam, he then tells us, " are the four patriarchal families, and preside over the intermediate ones, which are considered as the younger branches. I have taken some pains," he says, "to acquire the etymology of these words [the names of the families], but with imperfect success ; it requires much labour and patience, both to make a native comprehend, and to be comprehended by him. Quonna is a buffalo, an animal forbidden to be eaten by that family. Abrootoo signifies a corn-stalk, and Abbradi a plantain. Annona is a parrot, but it is also said to be a characteristic of forbearance and patience. Esso is a bush -cat, forbidden food to that family. Yoko is the red earth used to paint the lower part of the houses in the interior. Intchwa is a dog, much relished by native epicures, and therefore a serious privation. Appiadie signifies a servant race. Etchwee is a 4 o8 STUDIES IN ANCIENT HISTORY CHAP. panther, frequently eaten in the interior, and therefore not unnecessarily forbidden. Agoona signifies a place where palm oil is collected. These are all the etymologies in which the natives agree" (pp. 229, 230). Not to trouble the reader with speculations of Mr. Bow- dich " as to the meaning of these families as primeval in- stitutions," special attention may be drawn to the fact that the Ashantee, Fantee, and other peoples mentioned still class themselves as belonging to these families without any regard to national distinction; i.e. the various families or tribes of descent are distributed throughout the nations in which these people are comprised, and form of course clans within these nations, just as the people of the bear, wolf, and turtle stocks were distributed throughout the Iroquois nations, and formed clans within them. Besides the totems above enumerated we seem to have indications of many others as represented in the population when Bowdich knew it. Some of these appear on the stan- dards of the chiefs, or on their umbrellas (which are much the same thing) ; for example, the leopard, the elephant, the common barn fowl, etc. (see pp. 34, 57, 2*76, and the picture of the yam custom). The fowls indeed appear to have been the totem of the royal family, for we are told (p. 266) that " some families never eat beef, others abstain from pork ; fowls and beef are the fetish [totem] of the king's family, 1 and con- sequently never eaten by it." 2 Incidentally (p. 265) we 1 The state umbrella, however, is surmounted by an elephant in the picture of the yam custom. 2 It is to be noted here that the puzzling fact of the king's family appearing to have two totems may have its explanation in the marriage law, since the king's children might be bound to abstain from eating his totem because it was his, while they would be bound also to abstain from that which descended to them through the mother. And probably the explanation is of this sort. But it would seem that the family meant was the Sai family to which the king belonged, or to which he had succeeded ; and if it was as a grandson (daughter's son) of a former Sai that he was in the line of succession, this fact may give the explana- tion. He was by his birth debarred from the totem of his mother, and xxii GUINEA THE KINGDOM OF ASHANTEE 409 have a light thrown on the totems of the tribes in between Coomassie and the sea. " The different states of the water- side," we are told, " revere different animals as fetish ; the hyena is esteemed so at Accra, the alligator at Dix Cove and Annamaboe, and vultures universally. ... A black man killing a hyena at Accra would incur a serious penalty. . . . In a freshwater pond at Dix Cove there is an alligator about twelve feet long, which always appears on the bank at the call of the fetish men, who then throw it a white fowl." The area of totemism is further extended by this work into the region of the Gaboon. Bowdich found among the negroes there the same limitations on eating as among the Ashantees: "Like other negroes," he says, " different families have different fetish : some will not eat a cock, nor others a hen." 1 The Ashantees calling all the slaves brought to the coast Dunkos, it had been supposed that there was a country and people of that name, but Bowdich found Dunko meant bar- .barian. He first suspected this from some Dunkos being cut in the face, and some not; and because their languages were various and unintelligible to each other. As to the cuts, some had three on each cheek-bone, and three below, with one horizontal under the eye ; others, three deep continued cuts ; others, three very deep and long, and one under the eye ; others were cut in the forehead ; others still all over the body in fine, small, and intricate patterns. In each case the tattoo is said to have indicated the people of a particular place (p. 183). A few words on the religion and superstitions of the Ashantees. They had had long contact with the Moors, and lie was also debarred from the totem of the former monarchs. We are told that lie belonged to a fowl family (tribe of descent) the Annona or parrot. While the king's fetishes or totems, which he could not eat, were fowls and beef, it appears (p. 319) that eggs and milk were forbidden to all Ashantees. 1 And cf. Winwood Reade's account of the initiation to manhood recounted to him by a Mpongwe at which time the injunction to refrain from particular food is laid upon the novice (p. 245), and Reade's description of the gorilla dance among the Mpongwes (p. 194). Savage Africa, by W. H. Reade. London : Smith, Elder, and Co., 1863. 410 STUDIES IN ANCIENT HISTORY CHAP. had heard, of course, of a God of the Universe. They may have got from them also other deities and superstitions. Bowdich says (p. 262) that some of their gods l are supposed to inhabit particular rivers, woods, and mountains. "The present favourite fetish of Ashantee is that of the river Tando. Cobee, a river in Dankara, and Odentee, on the Adirree, are two of the others." These gods were honoured in proportion to the success in predictions of their respective priests. The priesthood was in some cases hereditary in families. There were other priests who took rank, like the god-boxes of the Tongans, in respect of their power of feigning convulsions, and their adroitness in rendering oracles. KINSHIP Though the form of the family would have permitted of kinship being traced through fathers, in point of fact, in Ashantee, it was traced through women only, and all suc- cessions, honours, and estates were inherited according to the law peculiar to that form of kinship ; except that the king, a despot without scruple, had constituted himself heir to all 1 " Their fetishes or subordinate deities," i.e. their gods, as distin- guished from the God of the Universe, from whom they were " alienated," the alienation, however, causing them no despondency. After death the spirits of the common people were supposed to inhabit the houses of the fetish, those of superior wisdom being charged with a supervision over the people who acknowledged it. One object of killing men and women at the funerals of people of rank was, as usual in such cases, to provide them after death with attendants suitable to their rank. The hereditary priests dwelt with the fetish. The others lived among the people and were resorted to as fortune-tellers or conjurers, and the number of these are stated to have been " frequently augmented by those who declare that the fetish has suddenly seized or come upon them, and who, after inflicting great severities on themselves in the manner of the convulsionists, are ultimately acknowledged." There were fetish women who belonged to this class. At the coast, a common fetish day was observed (Tuesday) on which men neither fished nor worked in their plantations. In Ashantee different families kept different days of the week as their fetish day ; and every one similarly observed the day of the xxii GUINEA THE KINGDOM OF ASHANTEE 411 the gold left by his subjects, a law frequently evaded by means of gifts inter vivos. The first notice of the law of succession which occurs in Bowdich will be found in his account of the history of Ashantee (p. 234). " Sai Apokoo, brother of Sai Tootoo, was next placed on the stool (i.e. throne). Had there been no brother, the sister's son would have been the heir. This extraordinary rule of succession, excluding all children but those of a sister, is founded on the argument that if the wives of the sons are faithless the blood of the family is entirely lost in the offspring, but should the daughters deceive their hus- bands it is still preserved." The law thus stated as the law of succession to the throne, Mr. Bowdich declares in his chapter on the constitution and laws of Ashantee to have been universally binding. " The most original feature of their law," he says (p. 254), "that of succession, has been mentioned in the history with the argument on which it is founded. It is universally binding. The course is, the brother, the sister's son, the son, the chief vassal or slave to the stool. In the Fantee country the principal slave succeeds, to the exclusion of the son, who only inherits his mother's property, frequently considerable, and inherited from her family in- dependently of her husband." l The daughters seem to have got for their share only a small part of the " ornamental gold." The stool, it may be mentioned, exists in every house, even the humblest, so that the terms in which the law of succession week on which lie was born. Besides denoting the god whom a family acknowledged, the word fetish is used by Bowdich in the sense of a charm ; the things of this sort which were most esteemed by the Ashantees came to them from the Moors (pp. 261-271). At the fetish house or cemetery at Bantama, where the kings were buried, human sacrifices, we are told, " are frequent and ordinary, to water the graves of the kings " (p. 289) ; and a ghastly account is given by Mr. Hutchison (pp. 419-421) of the proceedings at an Adai custom when the king " washed the bones " of his mother and sisters to propitiate the fetish. Mr. Hutchison says, " But in such doings we see the germ of ancestor worship." 1 Throughout Equatorial Africa female kinship is of common occurrence. 412 STUDIES IN ANCIENT HISTOR Y is stated are of general application. There is an independent corroboration of this statement of the law in the diary of Mr. Hutchison, which is appended to Bowdich's report. In con- versation with him Odumata began to boast to Mr. Hutchison of the number of his wives and children as compared with those of Englishmen. " I told him," says Hutchison, " there was a possibility of an Englishman knowing his father, but no black man could tell his. They were all slaves, and rendered incapable of inheriting their father's property. None of his children need to thank him, he neither could give them any- thing while alive, nor leave them anything when dead " (p. 416). But this statement of the law of succession appears to be over-pressed. At the time when Bowdich and Hutchison wrote, as their accounts show, adultery was punishable with death, the harem was fenced in every "way jealousy could dictate, and women were sometimes put to death for what we should call slight indiscretions. The people were certainly not wanting in the feeling of jealousy, nor is it credible the family system being monandrous that a father should have been so completely cut off as Mr. Hutchison states, from his children. In fact gifts inter vivos from father to children are mentioned by Bowdich as a means of evading the king's heir- ship in gold. Moreover the wife seems usually to have been obtained either by purchase or gift from the king. Bowdich's statement, too, distinctly sets forth that, failing sisters' children, the son comes in as heir. In the chapter on the laws of Ashantee we have a statement which throws some light on the relation of a father to his children. " If a husband is not heard of by his wife for three years," it is there said (p. 260), " she may marry again, and if the first husband returns, the claim of the second is the better ; but all the children of the after marriage are considered the property of the first husband, and may be pawned by him." It is a fortiori that children by himself should have been his property : they might not be his acknowledged kith or kin, but they were his goods and chattels ; he was entitled to the possession of them ; and that being so, it is not in human nature but that he should have been interested in their welfare, and should seek to further xxn GUINEA THE KINGDOM OF ASHANTEE 413 their interests by gifts inter vivos, as fathers do, in like cases, in India and in many other quarters. 1 We may here note that in the Gaboon, where we have seen already that there were traces of totemism, in the succes- sion to the throne the brother came in before the son. " Kings are numerous in the Gaboon, and scarcely comparable even with the petty caboceers of Fan tee." But while a primitive succession law is thus seen lingering round the succession to chieftainries, it had ceased to be the law of succession in private families. "All children share the property of the father in equal portions, except the eldest son, who has about half as much again as any other" (p. 437). 2 EXOGAMY There is not one word in Bowdich's report on the right of 1 A transition in kinship had evidently set in among the Ashantees. Among the Fantees, who were the same people, a son never came in as heir. Among the Ashantees he regularly did, failing sisters' children, and his relationship to his father was to this degree acknowledged, though the older kinship through the female line (formerly, no doubt, the only kinship, as among the Fantees) continued in the first place. Bowdich mentions (p. 258) that no one is punished for killing his slave, but is punished for the murder of his wife or child, and that the death of an inferior is generally compensated by a fine to the family ; and at p. 260 he says that if a woman involves herself in a palaver she involves her family but not her husband. Thefts must be compensated by the family of the thief, who can only be punished by his own family. 2 In the Gaboon, failing brothers, the son succeeded his father in the kingship (and not the sister's son), and sons succeeded father to the exclusion of brothers in all other cases, the eldest son getting a special share. It is stated, too, that " the acknowledged heir to a property may bring a palaver against his father, or whoever may be possessor of it, for killing a slave unjustly, or otherwise injuring the property, and oblige him to make good the injury." There was therefore a thorough establish- ment of family property and of male kinship, with an acknowledgment of special claims for eldest sons, and with a preference of brothers to sons remaining in the succession to dignities all of which would be derivable from Thibetan polyandry. 414 STUDIES IN ANCIENT HISTOR Y CHAP. intermarriage, and the subject of incest is only mentioned once in the account of Gaboon where the prohibition against a man looking at or conversing with his mother-in-law on pain of a heavy, perhaps a ruinous fine, is given as " founded on the tradition of an incest " (p. 437). Whether prohibitions of this class, which are of frequent occurrence, can be referred to the necessity for rules to prevent marriage or connections between persons nearly related by family ties, who, being of different stocks, would nevertheless be free to intermarry, is a subject to be inquired into separately. That the marriage law of the Ashantees and the people allied to them was anciently exogamy may, however, be confidently inferred from the interfusion of the same clans, each forming a stock, throughout their various nations. No hypothesis adequate to explain that kind of interfusion has ever been framed other than that which refers it to the joint operation of exogamy and the system of female kinship. The tradition given us by Mr. Morgan of the Iroquois nations having of set purpose divided themselves into totem stocks on the motion of a wise Sachem of the Onondagas, and then placed a portion of each stock in each of their nations, cannot be taken seriously. And, indeed, no other explanation of such an interfusion has ever been proposed. And that the system of kinship is exogamous is shown also by the statement (at p. 437) that if a man kills one of his wives he pays a fine to her family, who, and not the husband, are involved in all her palavers. Again there are casual statements made by Bowdich which suggest that, even when he wrote, husband and wife always belonged to different kinships ; and which show that " the family " was firmly held together by joint interests and obliga- tions. If a man committed a theft, it was his family which suffered for it; the family was bound to compensate the accuser, and might "punish their relative or not" as they thought fit (p. 259). If a woman involved herself in a palaver (accusation) it was her family which had to bear the conse- quences ; and the case is mentioned of a famous beauty (p. 259) who, having driven one of her lovers to kill himself in despair, xxii GUINEA THE KINGDOM OF ASHANTEE 415 killed herself in turn to save her family from a ruinous palaver with the family of the deceased. And in Ahanta (which is said to have had some peculiar customs) a creditor could seize for debt not only the family of his debtor, but any of his towns- men (p. 257), which shows that all of the same town were liable for each other in Ahanta, and suggests that all of the same town were kinsfolk there ; that they formed a clan (necessarily with male kinship) and that all of the clan were liable for each other. In Ashantee also, however, when a man had brought a frivolous palaver against another, we find that he was bound to give an entertainment "to the family and friends of the acquitted" (p. 259); from which it appears that in Ashantee also the right to compensation, and therefore the liability for injury done, extended how far there are no means of judging to relatives outside the family in our sense of the word. It was her family (whatever its limits) that was answerable in a married woman's palavers ; her husband had nothing to do with them (p. 260) ; and in this it is implied that he could have no responsibility for her as a relative. Moreover, wives were got by purchase (where the wife was not a gift from the king), and it appears that a captain could sell his wife, her relatives, however, having the right to get her back from him on repay- ment of the marriage fee (p. 260); and there clearly could be no kinship between husband and wife where these things could happen. It will be seen immediately that wives received in gift from the king would, in most cases, probably have been foreign women. The law allowed the king 3333 wives " which number," we are told, " is carefully kept up to enable him to present women to those who distinguish themselves, but never exceeded, being in their eyes a mystical one" (p. 289). These women were carefully secluded. Some of them were married to the king while still infants at the breast. Others were selected from the slaves taken by the Ashantees in their frequent wars. Many of them the king has probably never seen. Speaking of the women of the upper classes apparently, Bowdich remarks that their beauty is not surprising " when we recollect that they are selected from or are the daughters of the handsomest 416 STUDIES IN ANCIENT HISTORY CHAP, xxii slaves or captives, or are expressly chosen by their interior neighbours to compose part of their tribute to the King of Ashantee, who retains but a small proportion" (p. 318); and there are other indications that the better sort of people were really to a considerable extent provided with wives by gift from the king. In general, however, wives were got by purchase, and there was a system of betrothals. " Infants are frequently married to infants for the connection of families, and infants are as frequently wedded by adults and elderly men" (p. 302). There was a graduated scale of price for a wife, for the rich and for the poor, also a graduated scale of damages for intrigues, for the rich and for the poor. Curiously enough, though the husband bought his wife, and the right to the children she might have, and even, as we have seen, to her children by another husband, he did not get her property, which remained independent of him. If a woman was ill- treated by or disliked her husband, and her family tendered back the marriage fee, the husband had to accept it, when the woman returned to her family, but was not allowed to marry again (p. 260). The king's sisters, through whom the royal line was continued, were " not only countenanced in intrigue with any handsome subject, but allowed to choose any eminently so, however inferior otherwise, as a husband." If the wife died first in the latter case, the husband was expected to kill himself, and so also if the only male child died. When a male child was born the father did it homage as vassal in the most abject manner (p. 291). Speaking of polygamy " tolerated to such an excess amongst the higher orders " Bowdich tells us that " most of the lower order of freemen have but one wife, and very few of the slaves any" (p. 317). He says also that the slaves form the greater portion of the military force, and he estimates this force as 204,000 men in a population of about a million. It would result from this that the bulk of the adult male popula- tion were without wives. CHAPTEK XXIII GUINEA THE GOLD COAST BOSMAN'S work, entitled A Neiv and Accurate Description of the Coast of Guinea, divided into the Gold, the Slave, and the Ivory Coast, was written in Dutch about 1700, and was translated into English, and published in London in 1785. 1 William Bosnian was the chief factor for the Dutch at the castle of St. George D'Elmina. He appears to have been a man of much acuteness and learning, and during his fourteen years' stay upon the coast of Guinea he had ample oppor- tunities of studying the habits and customs of the people. These he describes in a very interesting manner, always stating where he thought he knew the matter fully from experience, where he only partially comprehended it, and where, as in regard to some inland negroes, he only wrote on hearsay evidence. The condition of things which we have seen in Ashantee, and the country immediately surrounding it, seems to prevail, according to all the representations, throughout Guinea, except that kinship is not uniformly through the mother only. Bos- man represents the negroes on the Gold Coast generally as being in five ranks : kings or captains, which he says are synonymous words ; caboceros, or chief men ; rich men, counted noble ; the common people engaged in tillage, fishing, etc.; and, 1 His work is contained in Pinkerton's collection, and the references given are to Pinkerton, vol. xvi. (General Collection of Voyages and Travels in all Parts of the World, by John Pinkerton, 17 vols. 4to, 1808-14). 2 E 418 STUDIES IN ANCIENT HISTORY CHAP. lastly, the slaves " either sold by the relations, taken in war, or come so by poverty" (p. 392). The chiefs are elected from the elders, young men being seldom admitted to the office. Those represented as noble in respect of their wealth are so considered only by foreigners ; they were not truly noble, either by birth or creation. They have no poor, for the reason that at any pinch a man rents himself out. TOTEMISM In giving an account of the religion of the negroes, Bosnian declares that "it is so various there is no village or town, nay, I had almost said no private family, which doth not differ from another on this head" (p. 396). It is true, he says, they believe in one God, who created all things, but they have no just idea of Him ; they owe the notion of such a God to the Europeans. Two reasons convinced Bosnian of this : first, they neither make offerings to God, nor call on Him in time of need, "but in all their difficulties they apply them- selves to their fetish, and pray to him for success in their undertakings ; secondly, the different opinions of some of them concerning the creation" (p. 396). As to this, Bosnian men- tions that a great part of the negroes believe that man was made by Ananse, i.e. a great spider. Mr. Bosnian gives a description of this spider at p. 469, as follows : " Going to niy chamber at night in order to go to bed, I found a hideous great spider against the wall. We found his body long am his head sharp, broader in the fore than hind part, but nol round as most sort of spiders are ; his legs were as large a man's finger, ten in number, being hairy and the thickness of a little finger. The negroes call this spider Ananse, am believe that the first men were made by that creature, and, notwithstanding some of them by conversation with Europeans are better informed, there are yet a great number that remai] of that opinion, out of which folly they are not to be reasoned. 3 In his account of their religion he speaks habitually oi fetishes, but it will appear plainly that the fetish is simply the totem. Bosnian says that the negro word is Bossui xxm GUINEA THE GOLD COAST 419 which means God. The phrase " Let us make fetish " is the equivalent of " Let us perform our religious worship." Fetish made with a view to the injury of another is of the nature of an incantation. Obligatory swearing is in their phrase, " Confirmation by fetish." Oaths taken with the oath-draught, and an imprecation that the fetish may kill them if they do not keep it, are inviolable. Each priest has his peculiar idol. When a negro is to take an oath " the priest asks him the name of his idol, each having a particular one " ; the name being given, the priest calls the fetish by its name, 1 and the oath proceeds. The statement that each negro has his particular fetish is consistent with the notion of the fetish being truly hereditary in the clan. 2 Bosman mentions public worship of a whole nation or town as customary on a bad season occurring. The chief of the town or nation advises with the priest as to the best course for removing the calamity, and an order is issued in conformity with their determination. The kind of orders given is not disclosed, nor the nature of the "public general religious exercise " on such occasions. When the fishery is at a low ebb, they make offerings to the sea. The public worship is frequently performed in the sacred groves which exist in every village. At p. 400 Bosman says that the Bossum is the individual god, answering to the " private medicine " of the Eed Indian, and that the individual worships it on that day of the week on which he was born. Most of the negroes have besides this another weekly day sanctified to their fetishes. 3 These fetishes then are distinct from the individual Bossum; they are wor- 1 The fetish is supposed to come, on being called, into the god-box, to hear the oatn^ the god-box being a great wooden pipe, filled with earth, oil, blood, etc., making a confused heap in the pipe. 2 "They have a great number of false gods; each man, or at least each housekeeper, hath one " (p. 400). 3 As soon as the child is born, says Bosman (p. 423), if above the common rank, it has three names bestowed upon it ; the first is that of 420 STUDIES IN ANCIENT HISTORY CHAP. shipped by sacrifices of a cock or sheep, which is eaten by the sacrificer, " his friends and acquaintance " no doubt his rela- tions. We owe a disclosure of a leading note of totemism to a speculation of Bosnian's, that if the negroes could be con- verted, it would be to popery. " The Eomanists have their allotted times for eating particular sorts of food, or perhaps wholly abstaining from it, in which the negroes outdo them, for each person here is forbidden the eating of one sort of flesh or other. One eats no mutton, another no goat's flesh, beef, swine's flesh, wild fowl, cocks with white feathers, etc. This restraint is not laid upon them for a limited time, but for their whole lives, and if the Komanists brag of the antiquity of their ecclesiastical commands, so if you ask the negroes why they do this, they will readily tell you, ' Because our ancestors did so from the beginning of the world, and it hath been handed down from one age to another by tradition.' The son never eats what the father is restrained from, as the daughter herein follows the mother's example, and this rule is so strictly observed amongst them that it is impossible to persuade them to the contrary" (p. 40 O). 1 The value of this casual testimony cannot be over-estimated; the disclosure of totemism is complete, as is the disclosure that the fetish is the totem. The last sentence again implies that the father and mother were always of different totems, which' suggests, if it does not necessarily imply, exogamy ; while kin- ship appears in a transitional state, as we saw it in the Hervey group. As in marriages between Protestants and Eoman Catholics, the girls are of the religion of the mother, and the boys of that of the father. Bosnian says he could never learn how their gods are represented to them; he thought every head of a house had the day of the week on which it is born ; the next, if a son, is generally his grandfather's; if a girl, her grandmother's. And on subsequent occasions other names are added from time to time, so that sometimes a man has as many as twenty. 1 " Before they eat or drink," says Bosnian (p. 402), " they are accus- tomed to throw away some ... it is for their false god, or sometimes for their deceased friend." xxin GUINEA THE GOLD COAST 421 one, and these family gods rewarded good and punished evil. 1 They believed in a future life, but it would be useless to study their ideas on that subject. Apparently, they have devils, but whether these be the gods of confederate tribes or a European importation is not clear. The statement is express (p. 403) that " on the Gold Coast the natives are not in the least acquainted with image worship, but at Arbra there are thou- sands of idols." Bosnian has no definite idea attached to the word idol. Compare p. 398, where the idol is described as a wooden pipe filled with oil, blood, etc. obviously a mere god-box, into which the fetish is to be brought, and where, in the same passage, the idol is confused with the fetish and p. 499, where he speaks of idol serpents : the serpents not being idols at all, but the living creature which is worshipped, and the species of which he describes. In the same page snakes are spoken of as being reverenced and idolised. NATIONS, TRIBES, AND CLANS Bosnian makes no direct statement as to the constitution of numerous monarchies in Guinea, i.e. as to their being com- posed of local tribes or nations, as we saw was the case in Ashantee ; but of the monarchies situated near the castle of St. George d'Elmina he says, that when they were at war, " Though each of the two contending armies was composed of five or six several nations, they would not together make 25,000 men" (p. 412). That is, the fighting force of a nation, on the average, would be about 2000 men, so that the nations, 1 " Most of them believe that after the death of any person he goes to another world . . . where he makes use of all the offerings of his friends and relations made here after his death, and that he revisits his dwelling for several nights after his death" (Bosnian, pp. 401, 402). " Slaves are sacrificed at the grave of the deceased, with some of his wives, especially the Bossums. The funeral ceremonies are repeated a year after the death, and again some ten or twelve years later " (pp. 429, 430). 422 STUDIES IN ANCIENT HISTORY CHAP. we may see, were very small, and strictly speaking local tribes, j We must see if we cannot ascertain this more distinctly. In speaking of the kings, he says that some of them have not more land under their jurisdiction than the bailiff of a village, and that the word for king came by a misapprehension into use among the negroes, who by the term Olin, in use before contact with us, understood the commander of a country (? district), or town, or nation (p. 415). These kings led simple lives, and, except as representing their community abroad, had little authority, and were treated with little respect The king would appear, however, to have had a right of hirin< out the forces of his nation to his neighbours, to assist thei in their wars. Where, as was sometimes the case, the king was absolute despot, his subjects were bound to fulfil any bargaii he made. In other cases the subjects, or many of them, joint him in a war with a view to the pay, but chiefly with a vie 1 ' to booty. The pay indeed was not much. Bosnian mentions that a four years' war which he had conducted did not cost quite 6000, and yet the Dutch had had successively five nations in their pay. Where the authority of the king and his council was less absolute, it would appear that a free negn went to the war or not as he liked. The Manceros, a kind oi chiefs, had each a fort, but command only over his slaves, as it is stated ; no free negro owning authority, even to the king, unless under compulsion. " If their leader is disposed to march up first towards the enemy, he may, but will not be followed by many." The suggestion is, but it is by no means clear, that within a district or country there might be several nations, and within the nations several towns ; and there is the suggestion, from the law of debt, that within the nations there were clans. (1) " One of the leading men in one country hath money owing to him from a person in an adjacent country, which is not so speedily paid as he desires. He causes goods, freemen, or slaves to be seized in the country where his debtor lives, so as to pay himself." This seems to involve the principle of xxm GUINEA THE GOLD COAST 42.3 joint liability of the people of the debtor's country for the debts of their compatriot. Where the debt is clear, the goods, etc., which are distrained, are sure to be promptly redeemed by the payment of the debt, but where it is doubtful this sort of proceeding is apt to lead to war. (2) At pp. 421, 422 we find that negroes, that is the free negroes, were responsible for their slaves, of whom they often have many ; also for their sons, nephews (sisters' sons), and other relations (these, as we shall hereafter see, are the relations on the mother's side) ; though in this case the rela- tions help each other by a mutual contribution, each giving something towards it according to his circumstances. Here we have reproduced the obligation which lay on the Irish Geilfine tribe, and on the Welsh kindred. If we imagine these relations to have a common fetish or totem, being all forbidden to eat the same animal, we have a clan within the nation on the totemic principle, because, as we shall hereafter see, kinship in Guinea was counted through the mother only. 1 (3) In Bosnian's account of the inland negroes, which he says is collected from people he could trust, we have several notes of clanship, (a) A negro committing a fault might not only be ruined himself, but his relations suffered with him. (6) A negro, when injured, calls his relations to assist him, " who readily lend him their helping hand," each being sure to get something of the compensation, (c) Bosnian had heard of 1 " Murder," says Bosnian (p. 406), " is punished two several ways : one is by the death of the murderers, and the other by a pecuniary mulct. If anybody kill a freeborn negro of Axim, and the murder is to be remitted by a fine, the usual sum of 500 crowns is demanded of him, though the whole demand is seldom paid, the murderer generally getting some abatement, according as the relations of the murdered man stand affected, for it is in their choice to be contented with as little as they please, and them only he is obliged to agree with ; quite contrary to what a certain writer affirms, viz. that the fines accrue to the king, which is so false that he has no pretence to the least share, unless he hath been assistant in the getting of them, when indeed he is paid for his trouble only. But if a murderer cannot pay his fine he is executed in a cruel 424 STUDIES IN ANCIENT HISTOR Y CHAP. fines to the amount of 5000 being paid on account of adultery. He never knew a negro possessed of such a fortune not even a king. The implication is that either the inland negroes were very rich or the fine was raised by assessment of the clan (see p. 422). KINSHIP At one place (p. 392) Bosnian has stated that the dignity of -captain or king in most countries descended from father to son ; but when he makes in his chapter on marriage a full statement on the subject he has a sentence to the very opposite effect. " The children," he says, " follow the mother, and that in Guinea has passed into an unalterable rule." The following is a summary of his statement as to mar- riage law and succession : Marriage is made almost without ceremony or previous courtship. The wife goes to the hus- band's home, and most wives work for their husbands. Poly- gamy is permitted, the chief check on it being the cost of the marriage feast. In a rich polygamous household two wives at least are exempt from labour, the chief wife and " the second, who is consecrated to his god"; from which we infer that the chief wife is not. The second wife is a slave bought with a design to be consecrated to his god, and of her the husband is declared to be specially jealous. " As to his other wives, he doth not watch them so narrowly" (p. 420). The probability is that her children take the father's totem. Married people have no community of goods ; each has his or her own pro- perty. They bear jointly the charges of housekeeping, but the man clothes the family. 1 On the death of either the man or wife, the respective relations come and sweep away all the property, leaving nothing to the survivor; but this statement must be taken in connection with that as to the law of succession. He says that all along the Gold Coast the children which a man has by his wives never inherit their parents' property, except at Accra only. " The eldest son, supposing the father a king or 1 The children of rich people are often married in infancy (p. 424). xxiii GUINEA THE GOLD COAST 425 a captain of a town, succeeds him in his office only ; but be- sides his father's shield and sabre he has nothing more to pretend to, so that it is no advantage to be descended from rich parents ; unless, which seldom happens, paternal love obliges him to bestow something on his children in his life- time, which must be done very privately, otherwise the relations after the father's death will oblige the children to return it to the uttermost farthing." " As far as I could observe," he says, " the brothers and sisters' children are the right and lawful heirs in the following manner : They do not jointly inherit, but the eldest son of his mother is heir to his mother's brother or his son (?), as the eldest daughter is heiress of her mother's sister or her daughter." Neither the father himself nor his relations, as brothers, sisters, etc. (it is obviously meant, by the same father), have any claim to the goods of the defunct, for what reason they cannot tell. " But I am of opinion that this custom was introduced on account of the licentiousness of the women, herein following the custom of some East Indian kings, who, as authors say, educate their sister's son as their own, and appoint him to succeed to the throne, because they are more sure that their sister's son is of their blood than they can be of their own, for being obliged to trust a woman no way related to them, 1 if she commit adultery the child may be entirely estranged from their blood" (p. 421). Bosman here was struggling with the intricacies of the system of succession peculiar to kinship through women only. He is very frank about his difficulties. After stating that, failing the heirs mentioned, the brothers and sisters of the defunct were the heirs, " and failing them the nearest relation to the mother of the defunct," he proceeds to say : " But their account of this subject is so perplexed and obscure that hitherto no European has been able to obtain a clear description of it, as I am certain they never will, notwithstanding that the negroes are so accurately perfect in it, that they never commit any error on this head : not but that great disputes sometimes arise amongst them on this occasion, but these are never owing 1 Observe the suggestion of exogamy here. 426 STUDIES IN ANCIENT HISTORY CHAP, xxm to their ignorance who is the heir, but happen from the next heirs being too potent in men and arms, and therefore stretch- ing beyond the due bounds of inheritance" (p. 421). This is a statement, containing one or two obvious errors, of the law of succession peculiar to the system of kinship through women only : one important error being due to the principle of symmetry. It is quite intelligible that the eldest daughter should be heiress of her mother's sister or her daughter, though the particular appointment might be some- what curious ; but that the eldest son should be heir to the son of his mother's brother is contrary to the whole principle of the succession law of which Bosman was endeavouring to give an account. As to the reasons for this law, it is obvioi that, intrigue being punishable with death (p. 422), and the family system allowing certainty of fatherhood, the origin oi the law must be looked for in a different family system fro] that found on the Gold Coast. CHAPTER XXIV GUINEA : SLAVE COAST FIDA THERE are various "kingdoms" on the Slave Coast, of which Mr. Bosnian has given some information, but the kingdom of Fida (otherwise Whidah) is the only one of which he has given a full account. He found the inhabitants of Goto, one of these " kingdoms," to resemble those of the Gold Coast " in politics, religion, and economies," except in the matter of idols ; by which, however, he means living animals religiously regarded, not fabricated images of the gods. They depend on plunder and the slave trade, as did the inhabitants of Popo, an adjoin- ing kingdom. Popo and Goto were continually at war, but both drew their chief supplies of slaves from the inland countries. As to Popo, Bosnian states that his account of Fida applies to it especially as regards the government. What follows is mainly taken from his account of the kingdom of Fida. NATIONS, TRIBES, AND CLANS Fida is treated of as a single nation or kingdom. There is no mention of tribes or clans, and the families appear to be patriarchal groups, composed of a man and his numerous descendants and co-relatives living together. Kinship was through fathers, with, apparently, a right of primogeniture in the eldest born son (p. 480). I can find absolutely no material in Bosnian's account of Fida from which to infer the existence of local divisions within the population answering to 428 STUDIES IN ANCIENT HISTORY CHAP. local tribes composed of different clans, or of the interfusioi of such clans in the local tribes ; but, as I have explained, ii discussing the case of the Algonquins, and also of the Narrin- yeri in Australia, the almost immediate effect of shifting the kinship from the female to the male side is to undo the primitive interfusion of clans, and to render the population ol the local tribes, towns, and villages homogeneous. Fida was very populous. Bosman says that in one village alone, as the king's village, or any of his viceroy's villages, " there were as many people as in a common kingdom on the Gold Coast." These large villages were numerous, the smi ones innumerable, often within a gunshot of one another, " fc those who live out of the great villages or towns build am settle where they please, so that each family builds a si village, which increases as that multiplies" (p. 477). The greater men have many wives, and a proportional number of children, and he says he has known men who have above 200 children. The king told him on one occasion that one of his viceroys, assisted by his sons and grandsons, wit their slaves, had repulsed a powerful enemy which came against him, and that this viceroy, with his sons and grand- sons, could make out the number of two thousand, not reckon- ing daughters or any that were dead (p. 481). In such patriarchal group we have the germ of a homogeneous tribe descent, such as was believed to have been the primitive typ( of group. KINSHIP At p. 479, after having given an account of the Chim like industry of the Fidisians in agriculture, so that all lam not covered by villages, or occupied by footpaths, was unde constant cultivation ; of their excessive ceremonial in sock intercourse, which reminds one of the Tongans and Fijians and of the diligence of the people in the slave trade, whicl enabled them to deliver 1000 slaves per month, he proceec to say : " The remaining customs and manners of the native of Fida not affecting their religious worship are very like the of the Gold Coast, excepting only that these exceed the oth< xxiv GUINEA: SLAVE COAST FID A 429 in all particulars of living, for whereas the former content themselves with 1, 2, 3, and the most considerable men with 8, 10, 20 wives, they have here 40 or 50, and their chief captains 300 or 400 some 1000, and the king between 4000 and 5000." This excessive polygamy is, no doubt, made possible by the slave trade, and the lucrative nature of farming, the work of which is mostly done by the women. The more beautiful women do not, however, work in the fields ; they wait on their lord at home. " No rich negro will suffer any man to enter the houses where his wives reside " (p. 479). Un- like some of the negroes on the Gold Coast, " they are strangely jealous " of their wives, and on the least suspicion sold them to Europeans. The custom of Fida made an intrigue, especially with the wife of a wealthy man, involve the death of the guilty person, and the enslavement of his whole family. It may be presumed that this jealous guarding of the harem grew up with and introduced the new system of kinship through fathers. As Bosman imputed the system of female succession to the licentiousness of the women, so his narrative in the case of Fida connects the system of male kinship with the jealousy of husbands and the law against adultery. The law of succession of Fida is stated as follows : " Upon the father's death the eldest son inherits, not only all his goods and cattle, but his wives, which he immediately holds and enjoy eth as his own, excepting his own mother. . . . This custom obtains not only with the king and captains, but also among the commonalty." It will be observed that the custom which excluded the mother implied a law of incest. There is but one other reference in the account of Fida to that subject. The king married two of his own daughters, but they dying soon after, he imagined that the gods punished him that way for his crime (p. 480) an intimation of the connection between the law of marriage and religion. As elsewhere, disputed successions to the throne were frequent in Fida. The taking possession of the king's court and wives was looked upon by the common people as like a solemn making-up of a title to the throne ; " and succeeding happily in these particulars the claimant to the throne need 430 STUDIES IN ANCIENT HISTORY CHAP not doubt the remainder, for the commonalty will not easily consent that after that he shall be driven from the throne. This seems somewhat like Absalom's design on his fathe David. To accomplish this design, the younger brother's part are always careful enough that he is near at hand in order take possession of the court" (p. 492). TOTEMISM Bosnian gives a pretty full account of what he calls the public gods of Fida, i.e. the gods receiving worship from the whole people. He opens his statement as to their religion by representing them as having gods by tfie thousand. An account of the nature of the fetish that has been often quoted represents them, when resolved on any undertaking of import- ance, a3 choosing as their god for the occasion the first creature or thing they see dog, cat, or other animal, or a stone or a piece of wood and making offerings and promises to it, and rejecting it as a useless tool in case of failure. " We make and break our gods daily/' said his informant (p. 493), "and consequently are the masters and inventors of what we sacrifice to them." This account seems to have amused Bosnian, who says, " This divine service is not new in the world, nor were the first men strangers to it"; in saying which he appears as a theorist. What is certain is, that such instability in the relations of worshippers to gods is nowhere else disclosed in the world, and that the report of it here is incredible. In fact, it is instantly disclosed that his informant was making fun, had been educated among the Trench, whose language he spoke perfectly, and had acquired "such a just notion of the true God," and how to worship him, as he could learn from Eoman Catholicism. Bosnian proceeds to say that the people of Fida have a faint idea of the true God, but do not pray to Him or offer any sacrifices to Him, saying that He has committed the government of the world to their own gods, and this is their firm belief. The faint idea mentioned they could not but xxiv GUINEA: SLAVE COAST FIDA 431 acquire from the Dutch, French, and English who carried on among them the slave trade. Their principal gods, owned as such throughout the country, are of three sorts. (1) A certain sort of snakes, who possess the chief rank amongst their gods. The species of snake is one streaked with white, yellow, and brown. Bosnian says that the biggest he ever saw was about a fathom long, and the thickness of a man's arm. (2) Certain trees, the nature of which is not disclosed. (3) The sea. These are the public deities worshipped and prayed to throughout the whole country. Each of them has a special jurisdiction or province, but that of the snake includes those of the other two. The offerings made to the snake are very valuable, consisting of European and African commodities, all sorts of cattle, food, and drink : it is supposed the priests appropriate them as they do the offerings made to the trees. They little encourage offerings to the sea, as these are not recoverable ; yet, in bad weather, all sorts of goods are thrown into it, so that it may permit of the arrival of ships. The snake, who appears here in the first place among the gods, must have been the totem of the stock to which the kings of Eida belonged. The kings of Eida used to make annual pilgrimages to the snake-house near the king's village, where they offered magnificent presents to the grandfather of all the snakes, who was supposed to reside there, to be as thick as a man, and of immeasurable length. Of course all the snakes of the divine species were religiously regarded, as well as their grandfather. Some Englishmen having found a snake in their house and killed it, in ignorance of this, the negroes killed them all, and burnt their houses and goods. After this, the English for a time having withdrawn their trade, the negroes made a rule of explaining to Europeans that they must not hurt the snakes, because they were gods. Bosnian was of opinion that even when he wrote it would be death for a man even to kill a snake by accident, unless he could instantly procure the protection of the king; and he relates how a pig having devoured a snake that had bitten him, an order went forth for the instant slaughter of all the pigs in the country. The slaughter went on till the race was threatened 432 STUDIES IN ANCIENT HISTORY CHAP, xxn with extermination, and it was stopped by the authority of the king. The snake is chiefly invoked in bad seasons for weather, for the preservation of the cattle, and in all politic difficulties ; the trees seem to be prayed to chiefly for healtl The offices of public religion are celebrated by priests an< priestesses associated together. Their persons are sacred (pj 493-500). CHAPTER XXV LOANGO, ETC. History of Loango, KaJconga, and other Kingdoms in Africa, by the Abbe Proyart; Paris, 1776. Pinkerton, vol. xvi. OUR author was a missionary, and a leading purpose of his history was to make out the people of Loango, etc., to be purer and better in all respects than they had been represented as being in the Histoire G6n6rale des Voyages, which he never mentions except to execrate. The Abb fully states the difficulties of observation in such a country, and the liability to error through misunderstanding, and it is not the least remarkable feature of the case which we shall state upon his testimony that he clearly was not in the least aware that he was stating such a case. His work treats of a portion of the West Coast of Africa, extending from the equator about six degrees south, and containing several kingdoms Loango, Kakonga, called also Caconda, lomba, etc., some of the kingdoms having various names. NATIONS, TRIBES, CLANS In each of these kingdoms the government is represented as being purely despotic, the king being regarded as a demi-god and the people as his slaves ; but within each kingdom appeared to be various states having their own princes, whom the Abbe* regarded as the king's vassals, many of whom were not much 2 F 434 STUDIES IN ANCIENT HISTORY CHAP. inferior to the king in power, and asserted for themselves a high order of independence. Within the several states the people are divided into two classes, freemen and slaves : the latter, so far as they belonged to the king or the princes, were well treated to prevent their running away; the former were the victims of arbitrary exactions unfavourable to agriculture and other forms of industry. The people within the states were grouped in towns or villages, each of which had its governor or chief, said to be king's officers, though the statement seems altogether in- consistent with the alleged independence of the princes. It is needless to say that in the Abbess exposition not a single word is said of either tribes or clans. FAMILIES AND KINSHIP The first impression derived from a perusal of the Abbe Proyart's seventh chapter, on Societies, is that we are in th( presence of a perfect example of what has been called th( patriarchal system; but on reading further the illusion is absolutely destroyed, and we find ourselves in presence of a society of the Iroquois type. " The people of these countries, like ourselves, inhabit towns and villages, and they present] a most striking image of the origin of society. They are not drawn together so much by reciprocal wants as by ties of| blood, which hinder them from separating. The families dol not disperse as with us, so that in the same town, and even! in the same village, you discern an infinite number of little hamlets which are so many families, each having its patriarch for a president. A family which finds itself too crowded, and? does not wish to confound itself with the neighbouring one,j may go and settle on the first piece of land which is not already occupied, and there found a hamlet ; it is the affair of a single day in a country where the father of a family is able, with the help of his wife and children, to carry away at- one journey his house and all his furniture, goods, and chattels. The heads of families are the first judges of them. When any xxv LOANGO, ETC. 435 dispute has arisen among them they confront the parties, and after hearing the pleadings on both sides, they pronounce a sort of sentence in juridical form. This domestic tribunal is the model of the other superior tribunals. The laws do not allow a woman to appeal from the sentence of her husband, nor a son from the judgment of his father. Indeed, they never think of doing so. But in the sequel we shall see that from the tribunal of the chief of each village there is an appeal to the governor of the province, and, lastly, to the king " (p. 560). The towns are said to be only great villages. Now let us see the facts in the light of which this state- ment must be read. They are all of them casually disclosed. (1) If a man wanted a wife, he never applied to the girl's father, but to her mother only, and the presents he made, he made to the mother only : a note of kinship through women only (p. 569). (2) Women of the royal stock married whom they liked, and the only case in which a husband is said to have had authority to divorce his wife was where a princess required him to do so, in order that she might marry him. Having married him, the princess kept him under guard as in a harem, and scrupled not to have his head chopped off if he was detected casting eyes on any woman when promenading under escort. The licence of the princesses, recalling Ashantee, is a further note of female kinship (pp. 569, 570). (3) " The commonalty of goods between husbands and wives is not held in this country. ... As to successions, the children do not inherit from their father, but only from their mother. The goods of the father are reversible after his death to his eldest uterine brother, if he has one ; in default of brothers, to the eldest son of his eldest uterine sister ; or, lastly, to the eldest son of his nearest maternal relation" (p. 5*71). This law of succession, stated for rich and poor alike throughout these kingdoms, is a perfect statement of the law of succession peculiar to the system of kinship through women only, and shows that, as respects heirships at least, a man had no kindred except through his mother. The only limit assigned to the operation of this law is that the heir to lands 436 STUDIES IN ANCIENT HISTORY CHAP. and lordships succeeded only on petition, supported by presents to the government, but the same person was heir of a fief as succeeded to a chattel. (4) The position of fathers in relation to the education of their children is so stated (p. 571) as to recall the case of the Iroquois, in which it is said the father had no right either to their nurture or education. (5) "Nobility does not descend except by the females, so that all the children of a princess mother are princes or princesses, though begotten by a plebeian father. On the other hand, the children of a prince, or even of a king, are not nobles unless their father has married a princess, which scarcely ever happens" (p. 579). The reason for the absence of marriages between princes and princesses is said to be that the man, in order to enter into one of these, would have to renounce the right of poly- gamy. (6) In the statement of the powers of the father occurs the phrase, " The heads of families are the first judges of them. When any dispute has arisen among them, they (that is the heads of families) confront the parties." The parties here appear to be within the family, but who are they ? The statement is obviously made in reference to the case detailed at p. 570, viz. the case of a man with many wives. "The husband, in order not to excite jealousy among his wives, uses no familiarity with any of them. He always dwells alone 1 in his hut, and each of them in hers with her children. This j separation of dwellings does not prevent differences from arising I among them now and then, which the husband, according to I the usage of the country, has a right to terminate juridically. I On the complaint which has been referred to him he orders I the two rivals to appear before him ; each pleads her cause ] kneeling, while he sits on the ground with his feet crossed. ] Having heard them he pronounces sentence. They retire in silence, testifying the most entire submission to his judgment." What appeared, then, to be a broad statement of the powers of a father, similar to the patria potestas, resolves itself into a right of arbitrating between wives in a quarrel, and is not, xxv LOANGO, ETC. 437 either in the leading statement or elsewhere, extended to include a power of any sort over children. Moreover, even this case of arbitration must have been very rare. Proyart thought there were rather fewer women than men in the country he wrote about, " so that a grandee of the country cannot marry twenty women without placing at least nineteen of his fellow- citizens under the necessity of observing celibacy " ; besides, he thought the women preferred the monogamous marriage. Polygamy, in short, he declares to be a privilege used only by the rich ; adding that, as the class of rich persons is far from numerous, all the free men, and even most of the slaves, still find means to marry, i.e. monogamously (pp. 568, 569). It is very obvious after this statement that the general absence of community of goods between husband and wife in the country cannot possibly be referred, as it is at p. 571, to the practice of polygamy : no general law or custom can be referred to the practice of a few. We shall presently see clear reason for believing that the people were united in clans on the totemic principle ; and that being so, we may believe that the system of female kinship which gave the succession law was the only operative system, and that the clans were interfused in the different communities, and bound to a common action by the blood-bond. This being so, it would have been incredible that a father should have had judicial authority over his children without appeal. We may be sure their relations would have something to say to his decisions, by the way of " Muru " or otherwise. TOTEMS The way in which a disclosure of the totem system of the family is made by the Abbe is simply charming. He opens by saying that the people, in order not to expose their religion to contempt, are very reserved in speaking of it to Europeans. What he found out was that they acknowledged a supreme Being, a just and perfect God meaning his own, and also a god of wickedness, by which he understood their own. " They who know only the theology of the country, persuaded that the good God will be always sufficiently favourable, think 438 STUDIES IN ANCIENT HISTORY CHAP only of appeasing the god of wickedness. Some, to rende him propitious to them, never eat fowls or game, others ea only certain sorts of fish, fruits, or vegetables : not one amon them but makes profession of abstaining all his lifetime fron some sort of nourishment. The only way of making him offer ings is to let die under their feet in honour of him some shrub laden with their fruits. The banana-tree is that which the;; consecrate to him in preference" (p. 594) from which w learn that the " good God " was no god of theirs at all, am not worshipped by them, and that the god of wickedness wa in each case the totem, the benignant protector of the clan. From the rest of his statement it appears that the people or some of them, had idols; that they had amulets -anc charms, like other people all over the world ; and that the}/ had priests among them, called Ganga, whom they believec in seriously ; that there was a consecration of children on thei birth to the god, no doubt, of the mother's clan or family, fo it is stated that a result of the dedication was the impositior of some superstitious practice to which they were to be faith ful all their lives, and of which their mothers were bound tc remind them as they grew up. Lastly, we owe to the Abbe' sense of humour the disclosure that one totem, at least, hac obtained universal regard, as the serpent had at Fida, and th( spider on the Gold Coast. " The Ganga," he says, " who in other respects do no pique themselves on uniformity in their doctrine, unanimously teach everybody that there would be an extreme danger in eating partridges, and no one dare hazard the experiment' (p. 595). The example of the Europeans in killing and eating the partridges seems to have been successful in undermining this scruple, for some of the natives at least were found to kill them with a view to exchange them for ammunition. EXOGAMY There is not one word said by the Abbe on the law o intermarriage, nor are there any statements made by him from ! xxv LOANGO, ETC. 439 which it can be surely inferred what the law was ; but does not the law of succession imply that it was exogamy ? and may not the absence of marriage in the princely clans be imputable to exogamy ? The Abbe* says : " There is in each kingdom a family, or if you please a class of princes, for they are very numerous, and they know not the order of their genealogy so correctly as to know if they be of a common origin. It is sufficient to be a prince in order to have the right of pretending to the crown, and it must necessarily be so in order to possess certain noble fiefs, which are held more immediately on that tenure " (p. 579). This seems to imply in each kingdom a clan of princely stock whose memberslield the best of the lands, and one or other of whom succeeded to the throne, either as successor- designate of the king, which was permitted in certain kingdoms, or by establishing himself as king by force of arms after a civil war. We are told that marriage between man and woman of this princely stock scarcely ever happened. A king who had the right of nominating his successor might not seek a marriage which would naturally make his son his heir. Were the cases in which such marriages occurred cases where the king or prince had not that right of nomination, and, as despots will, defied the law ? Lastly, was this princely stock the same in all the kingdoms ? Was its totem the partridge, and is this the explanation of the universal religious regard for that bird throughout the kingdoms ? CHAPTER XXVI EQUATORIAL AFRICA IT is only casually that a disclosure of facts interesting to the student of history is made in the pages of M. du Chaillu, but though few and far between, they show a sharpness of observation, and care in ascertaining facts, rare among travellers. The result of all the observations that he made in Western Equatorial Africa on the structure of tribes is stated in effect as follows. He found tribes with different names considering themselves different nations, though speaking the same lan- guage ; and he found tribes speaking the same language fre- quently separated from one another by tribes speaking different languages. He found all these tribes or nations divided into numerous clans, and the clans independent, and often at war with one another. Sometimes there were several villages belonging to one clan within the same nation, and he found villages of the same clan sometimes at war. Each village had its chief, assisted in the government by elders, each elder being the head of a separate portion of the village. Each clan again had its acknowledged head, in the position of " father to the clan, and in this sense it is that he calls their system of government patriarchal. As to the origin of the clan system he writes as follows : " I have never been able to obtain from the natives a know- ledge concerning the splitting of their tribes into clans : they CHAP, xxvi EQUATORIAL AFRICA 441 seemed not to know how it happened, but the formation of new clans does not take place now among them." * We here have the marriage law in Africa for the first time expressly stated : Tribes and clans intermarry with each other, but people of the same clan cannot intermarry. It is considered an abomina- tion that there should be marriage between persons in the least connected by consanguinity. Nevertheless the nephew has no objection to take his uncle's wives, or the son his father's, except his own mother, as to which remark it may be observed that the heir in both cases supposed was of different blood from the wives he succeeded to. 2 This leads to the succession law and the marriage system. Du Chaillu found the wives of West African chiefs almost as independent as their lords, each with her own plantation and property. 3 In case of polygamy each wife had a separate house built for her. Beyond this the statement that the marriage law was exogamy, that women were on the whole very well treated among the Ashangos, and that formerly it was their custom a species of inverted suttee that when the woman died the man should die also, we learn nothing of their mar- riage customs except what may be inferred from the succession law (see pp. 171, 259, 331, and 394). As to heirship the statements are distinct. (1) The palm- trees in the quadrangles are the property of the chief man of each group of houses, and being valuable property pass on his death to his heir, the next brother or the nephew, i.e. sister's son, as in other tribes. (2) A case of succession is stated as follows : 0, the nephew and heir of M, died, leaving two wives, the one young, the other old. One of the widows was taken as a wife by M, the other by O's younger brother. Du Chaillu says: "My notion was that the younger brother ought to have had all the property and the pick of the wives, but I was told that the elder brother inherits the property of 1 A Journey to Ashango Land, etc., by Paul B. du Chaillu, p. 425. London: John Murray, 1867. 2 Id. p. 427. 3 Id. p. 171. 442 STUDIES IN ANCIENT HISTORY CHAP. the younger brother. If O's younger brother had died would have taken his wives and property, but O having no elder brother, M, his uncle, mother's brother, had the right of dividing the property as he thought fit." Here we have a system of succession law corresponding to kinship through women only. It was customary, however, and proper that some of the wives should go to the younger brother. (3) Du Chaillu says that the laws of inheritance were the same through- out Ashangoland except among the Bakalai. The reader will readily see that the main difference between the law of the Bakalai and that just stated is that, among the Bakalai, M would not be O's heir in preference to O's younger brother. " The next brother inherits the wealth of the elder, but if the youngest dies the elder inherits, and if there are no brothers the nephew. The headship of the clan or family is hereditary, following the same law as that of inheritance and property. In the case of all the brothers having died, the eldest son of the eldest sister inherits ; and it goes on thus until the branch is extinguished, for all clans are considered as descended from the female side." l Marriage within the clan is prohibited among the Bakalai and other tribes. 2 The Obongos are disowned by the Ashangos, who will not intermarry with them, and declare that the Obongos intermarry among themselves, sisters with brothers, doing this to keep the families together as much as they can. Du Chaillu says that the communities of these wretched creatures are small, and so isolated as possibly to necessitate close interbreeding, and that this may be the cause of their physical deterioration. 3 1 A Journey to Ashango Land, pp. 389, 390, 429. 2 Explorations and Adventures in Equatorial Africa, by the same author, p. 388. John Murray. London, 1861. 3 A Journey to Ashango Land, p. 320. An instance is given of what seems an endogamous tribe on the Niger by E. and J. Lander (Journal of an Expedition to explore the Niger, London, 1833) : "It appears, say they (vol. i. p. 21 3), that the Falatahs inhabiting Acba, though very numerous, are but one family, for we are told that their ancestor separated himself from his friends . . . and travelled hither. . . . The sons and daughters of his descendants intermarry only among their own kindred, and they are XXVI EQUATORIAL AFRICA 443 In his Wild, Life under the Equator, London, 1869, Du Chaillu makes of the African clans the general statement : " Every clan has some kind of animal they do not eat " (p. 128). An unmistakable note of totemism, which, being put on record by the traveller prior to the publication of any speculation about totemism, may be taken as satisfactory evidence that the equatorial African tribes, which we have seen to be composed of clans with exogamy as marriage law and female kinship, were altogether composed on the totemic principle, like the tribes of the Iroquois. It is merely as illustrating the curi- osities of the metaphysic connected with totemism that I notice the reasons in one or two cases assigned for the absti- nence from particular kinds of food. The Abonga cannot eat buffalo ; they believe that if they were to eat it disease would creep over them, and they would die. Stranger still, their women would give birth to buffaloes (p. 127), as tradition tells them they did once before. Are we to suppose that Abonga means buffalo, and that the clan were named from the animal ? Lastly an infallible sign of a totem we have the belief in what we may call the were-leopard. " Nothing is so terrible as the leopard that was once a man" (p. 254). Du Chaillu, writing of the tribes of Equatorial Africa, says : " The Shekiani tribe is divided into clans, and though these families grow very large sometimes, marriage between the members of the same clan is prohibited." 1 The Shekiani would appear to be a general name, like Iroquois or Bechuana, for we are told there are various tribes (nations) of them known by different names. They inhabit between the 9th and 10th parallels of N. latitude. Among the Bakalai it was customary for the heir to give up some of the wives to one or more of those who would have succeeded had he not been in the way. In the case mentioned in the reference the heir gave two of the seven wives to a betrothed to each other in infancy and childhood." These Falatahs are described by the Landers as far superior to their neighbours in manners and intelligence. 1 Explorations and, Adventures in Equatorial Africa, p. 163. 444 STUDIES IN ANCIENT HISTORY CHAP, xxvi younger brother and one to a cousin. 1 A village was so much distressed because the heir, being affronted, declined to take any of his brothers' widows, that they came in a body, and begged him to take at least two. 2 The town was agitated all day on this important question. As early as 8th January 1866 Du Chaillu, in giving an account of his travels to the Eoyal Geographical Society, says of the peoples he had visited : " The people are divided into clans very much like the Scotch, only a man belongs to his mother's clan." Du Chaillu says of the Fans that they buy the dead of other families, i.e. clans, in their own tribe, for food, 3 and else- where he makes the following general statement : " I found the tribes of Equatorial Africa greatly dispersed, and in general no bond of union between parts of the same tribe. A tribe is divided into numerous clans, and these again into numberless little villages, each of which possesses an independent chief. The villages are scattered, are often moved . . . and not unfre- quently are engaged in war with each other." 4 As to the frequent wars, he says : " Unlawful intercourse with the women of a neighbouring tribe or village is the cause of nearly all the palavers and wars and fights in Africa. If a tribe wants to fight they make this the cause, by getting one of their women to intrigue with a man of the other tribe or village, and if they do not want to fight even, they are often forced into it." ' He repeatedly says that in Equatorial Africa wives and slaves are the only property, there being no cattle and no property in land. Frequently women desert their husbands and run off to another village. It is a point of honour not to give them up. This is a fertile source of war. 6 1 Explorations and Adventures in Equatorial Africa, p. 239. 2 Id. p. 262. 3 Id. p. 88. 4 Id. p. 329. 5 Id. p. 51. 6 Id. p. 162. CHAPTEE XXVII SOUTH AND SOUTH-EAST AFRICA * THE MATABELE WE have accounts of this horde, the Gagas of South Africa, from both Mackenzie l and Livingstone ; they are in general agreement. Mackenzie spent five months among them, and says he gave some attention to their customs, so that we may follow his account with some confidence. The Matabele were Zulu by origin. But few of them, however, and these old men, were Zulus when Mackenzie knew them; the rest of the 10,000 or 12,000 soldiers of the Matabele belonged to every tribe through which the original Zulu band had forced its way north. It is said that the band had detached itself from another Zulu band, the chief of which was Tshaka, whose law was that his soldiers should not be free to marry. The Mata- bele, under their chief Moselekatse, had the same law, so that the horde was recruited by children captured in war. It is of no interest to us by what processes the captives are trained to become Matabele warriors. The captive girls had no choice but to be shall we call them wives ? to the warriors. " The Matabele soldier town," says Mackenzie, 2 " has nothing domestic about it : it is not a town, but barracks. 1 Ten Years North of the Orange River, from 1859 to 1869, by John Mackenzie, of the London Missionary Society. Edinburgh : Edmonston and Douglas, 1871. 2 Mackenzie, p. 329. 446 STUDIES IN ANCIENT HISTORY CHAP. The voice of the infant, the song of the mother, are almost unknown there ; only after some signal service does the chief bestow as a great reward to the soldier, a captive girl to be his wife, who has no choice in the matter, but is delivered over to her new owner, as an ox is given to another man, whose deeds have been less meritorious." Nothing could be more horrible or heartrending than the narrative of the wholesale massacres by which this robber horde, dispensing on the whole with the troubles of rearing children, secure grown-up boys and girls with whom to recruit the ranks. It would of course be absurd to expect in such a horde the germs of civil society, except so far as subdivisions of the horde present themselves, holding together in respect of their common derivation, being captured from the same tribe ; and it would "appear that so long as any of one blood remained in the tribe, they endeavoured to carry out the traditions of their ancestors, so far as they knew them, especially in the matter of funerals. Mackenzie remarks, however, that most of them were captured too young to permit of their native mores or religion being taught to them, and of course received no instruction from the Zulus, so that they were absolutely godless. Livingstone in his Zambesi makes mention of more than one Zulu horde of the Matabele type, as ravaging the country along the course of that river, especially the Mazitu. It is curious to learn, on the evidence of Mr. Mackenzie, that the chief of the Matabele, who was constantly authorising, directing, or leading in person expeditions which always had for their purpose the wholesale slaughter of men and women who had never offended him, the murder of infants at the breasts of their mothers, in fact of all infant and adult life, in whatever district he struck, was a man of gentle mien and kindly nature ; nay, even of much tenderness, so as to be even moved to tears at the sight of a poor motherless child who happened to be presented to him. He was no fiend, says Mackenzie in effect, but a very good fellow, who was the creature of circumstances, and whose existence, as well as his power and authority, were dependent upon the control of his savage soldiery. P. I XXVII SO UTH AND SO UTH-EAST AFRICA 447 THE MAKOLOLO As a tribe the Makololo have ceased to exist. The notices we have of them are scattered, confused, and confusing, and it would not be worth while to attempt to piece them together. It would be to attempt to make a picture of a whole that never existed, for evidently the ways of these people were various, and frequently changing under the varying conditions of their lives. Where they came from seems somewhat uncertain ; their progress, during the forty years of their career, seems to have been on the whole from south to north. Under their vigorous leader, Sebetuane, they conquered a great variety of tribes. Mackenzie l speaks of them as " the mighty people who spread dismay in the neighbourhood of Kuruman, who in their north- ward journey conquered the Bangwakatse, the Bakwena, and other tribes in that region, who drove the Bamangwato before them like antelopes before the lion, whose track can be marked by the usual signs of savage conquest, the wasted towns, the devastated country, the silent grief of the widowed and orphaned captives." It appears that it was their practice to incorporate in their horde the captive women and children. In this respect they resembled the Matabele, but the original element in the horde, which seems at one time to have been called the Basotas, never was reduced to so small proportions as the Zulu element among the Matabele. They are spoken of in Livingstone's Zambesi as comprising a great variety of tribes, and it appears that the headmen of the genuine Makololo had entrusted to them in vassalage the various conquered provinces, with the subject races as vassals cultivating the soil and attending to the cattle, all of which was held in property by the Makololo. Their destruction was a consequence primarily of the death of Se- betuane and of the incapacity of Sekeletu, his son, whom he made his heir, whether in conformity to custom or not does not appear. On this man's succession, many of the real 1 Mackenzie, p. 247. 448 STUDIES IN ANCIENT HISTORY CHAP. Makololo, on various pretences, went or were driven into exile, and when on his death a civil war among the chiefs took place with regard to the succession, their numbers were so reduced that the subject tribes, suddenly rising, were able in one night to put nearly the whole of them to death. Of the few who escaped the majority were subsequently treacherously slain by pretending friends with whom they had taken refuge ; and the Makololo ceased to exist. Similarly, we believe, must perish every organisation founded upon mere brute force : on the enfeebling of the iron will of the leader, or the withdrawal of his iron hand, the various elements held together by mere force must tend to stand apart, and the group to be dissolved. The Makololo women are described as being "vastly superior" to any seen by Livingstone in "Africa. The tribe practised polygamy, which Livingstone explains by saying, " The wealthy old men who have plenty of cattle marry all the pretty young girls." l Of course the young men, who had no cattle, had to get on without wives ; a state of affairs, says Livingstone, probably leading to a good deal of immorality. Wives were got by purchase ; the price, when desired, being increased to include the right to the children, which otherwise belonged to the woman's family. We may infer from the system of marriage that a man could not marry in his clan, and from it being necessary to purchase the right to the children, that anciently there was kinship through women only. They appear to have been anciently organised in clans, and in a general survey which we shall take of the evidence of totemism in the districts which they inhabited, we shall see reason to believe that the clans had their totems but in Sekeletu's time the clan system would appear to have been practically broken up. We are told 2 that families frequently leave their own headman, and flee to another village, and sometimes a whole village decamps by night, leaving the headman. Sekeletu rarely interfered with the liberty of the 1 Narrative of an Expedition to the Zambesi and its Tributaries 1858-64, by David and Charles Livingstone, p. 284. John Murray, London, 1865. 2 Id. pp. 291, 292. xxvir SOUTH AND SOUTH-EAST AFRICA 449 subject to choose his own headman : a statement which recalls the arrangements of the Matabele and other Zulu conquering tribes, who allowed the incorporated captives at their will to change their masters, 1 but seems inconsistent with the relations between clansmen and an hereditary clan or chief. The Makololo had not only pretty women but many children ; they were both pastoral and agricultural, and made their chief raids for the purpose of lifting cattle. We get some light casually thrown on their marriages. 2 Livingstone says but few good-looking women appear in the first Batoka villages (beyond the Makololo settlements), because the Makololo marry all the pretty girls. THE BECHUANAS The Bechuanas had a career of some duration as a con- quering tribe in South Africa. Since Livingstone knew them they were struck, and, in a sense, destroyed by the Makololo, and the destruction of the Makololo themselves was afterwards recorded by Mackenzie, as already mentioned. From the south-east, Caffraria and Natal, waves of conquest have within our knowledge of the South of Africa repeatedly swept eastwards and northwards, to the destruction of the inland "tribes." These waves have sometimes collided with waves of conquest from the north-west, and the victors of to-day have been the vanquished of to-morrow. While, on the one hand, as the result of this state of perpetual war, the land is strewn with the detritus, as it were, of ancient organisations, a small clan alone representing a once powerful tribe of descent, or a few scattered individuals representing a once powerful clan, on the other hand, we seem to see forming under our eyes new combinations of a tribal sort, absorbing into them- selves the elements of the destroyed communities. The Mata- bele on the Zambesi, and north of the Orange Kiver, are in all respects the Gagas of the West Coast. Originally a band of Zulu warriors, under the command of Moselekatse, being successful in a raid towards the north, they never returned to Zululand ; starting without women, they plundered all the 1 Zambesi, p. 385. 2 Id. p. 311. 2 G 450 STUDIES IN ANCIENT HISTOR Y CHAP tribes along the route of their young girls and boys, putting to death the old men, women, and children. The young women became their wives, and the young men in time entered their ranks as warriors. Their ranks swelling through constant incorporations of this sort, the forces of the Matabele at last numbered 12,000 warriors, and as they continued to grow on the same principles, it will be seen that the growth and continuec existence of this robber horde was dependent on perpetual massacres of the communities with which they came in contact The Eechuanas, or nation of equals, as their name implies constantly remind us by their domestic arrangements, no less than by their use of cow-dung and clarified butter, of the earl) Vedic nations. They were comprised in a great variety o: tribes, the word tribe being by all the authorities used vaguely sometimes to mean local tribe or nation, but more frequently as in the citation which follows from Livingstone, to mean clans. " The different Bechuana tribes 1 are named after certain animals, showing that probably they were, in former times addicted to animal worship, like the ancient Egyptians : the term Bakatla means ' they of the monkey ' ; Bakwena, ' they of the alligator ' ; Batlapi, ' they of the fish ' ; each tribe having a superstitious dread of the animal after which it is called They also use the word Bina, to dance, in reference to the custom of thus naming themselves, so that when you wish to ascertain what tribe they belong to, you say, What do you dance ? It would seem as if that had been a part of the worship of old. A tribe never eats the animal which is its namesake, using the term ila, hate or dread, in reference to killing it. We find traces of many ancient tribes in the country in individual members of those now extinct, as the Batau, ' they of the lion ' ; the Bauoga, ' they of the serpent ' though no such tribes now exist." Here we have every note of totemism coming together, and cannot doubt that in the passage just cited " tribe " means tribe of descent or group of kindred in the sense of " clan " 1 Missionary Travels and Researches in South Africa, by David Living- stone, p. 13. London : John Murray, 1857. xxvii SOUTH AND SOUTH-EAST AFRICA 451 among the Iroquois and " family " among the Australians : a fact to be kept in view when we come to inquire into the general social structure of the Bechuana nations. I will corroborate the preceding statement cited from Livingstone by casual proofs of totemism in South Africa to be found in Mackenzie. 1 The common goat is stated to be the sacred animal of some Bushmen ; the alligator, as stated by Livingstone, the sacred animal of the Bakwena. There is evidence of the lion and the baboon being religiously regarded, and there are indications of clans of the leopard and zebra. Elsewhere occurs the following statement : 2 " There are many things which occur in the daily life of a Bechuana man to cause him misfortune, according to the old belief. Each tribe has its sacred animal to which it is said to dance ; the ' Puti ' was the sacred animal of the Baniangwato " (or of their chief, vide p. 4 5 4 infra). There is a suggestion of a crocodile clan and pig clan ; 3 in fact all the animals mentioned by Livingstone in his general statement are casually seen in Mackenzie's book to be religiously regarded. We have already seen what trust to repose in a general statement made by a missionary that the system of govern- ment among a people is patriarchal. Livingstone and Mackenzie agree in saying that among the Bechuanas the patriarchal system prevails. After examining all the facts, I am of opinion that this at least is true, that the chiefs among the Bechuanas, unlike the chiefs in an Iroquois or Australian group, have despotic powers, and that kinship has shifted from the mother's side to the father's. "The government," says Livingstone, 4 " is patriarchal, each man being by virtue of paternity chief of his own children ; they build their huts round his, and the greater the number of children, the more his importance increases. Hence children are esteemed one of the greatest blessings, and they are always treated kindly. Near the centre of each circle of huts there is a spot called a 'Kotla,' with a fireplace. Here they work, eat, or sit and gossip over the news of the day. A poor man attaches himself 1 Mackenzie, pp. 135, 151, 256, 437. 2 Id. p. 391. 3 Id. pp. 390, 393. 4 S. A., p. 15. 452 STUDIES IN ANCIENT HISTORY CHAP. to the Kotla of a rich one, and is considered a child of th< latter ; an under-chief has a number of these circles round and the collection of Kotlas around the great one in the middle of the whole, that of the principal chief, constitutes the town. The circle of huts around the Kotla of the chief is compose( of the huts of his wives and those of his blood relations. He attaches the under-chiefs to himself and his government b] marrying their daughters, or inducing his brothers to do so. They are fond of the relationship to great families." 3 We learn that the chieftainship is inherited. Livingstone says that there are signs of tribes (i.e. groups like that in th( town described above) splitting up. 2 He speaks of an " ori- ginal tribe " breaking up into Bamangwato, Bangwaketse, anc Bakwains, the Bakwains retaining the hereditary chieftain- ship. 3 Again he says that the Basutos arrange the natio] in three divisions : (1) Matabele or Makonkobi ; (2) Bakoni 01 Basutos ; (3) Bakalahari or Bechuanas ; and he adds that th( Basutos include a great variety of tribes, as Batau, " they of the lion " ; Baputi, " they of the Puti," etc. 4 The people had a pro- found respect for their legitimate chief. When Sebituane's elder brother, becoming blind, gave up the government, Sebituane continued to call his elder brother Kosi, or chief, and the descendants of the elder brother paid no tribute to the throne It will complete Livingstone's account of the Bechuana towi society to mention that on a child being born boys bein^ always more welcome than girls the birth was duly reporte( to the chief, though whether of the town or Kotla is not stated. This is said to be the patriarchal system, but is it ? On th< patriarchal theory families tend to multiply families rounc them, the families of their descendants, as the banyan tends to surround itself with its own offshoots, and the whole collection of related families, all derived from the central one as re- 1 8. A., p. 45. 2 His words are : *' There are several vestiges besides of very ancient partitions and lordships of tribes." 3 S. A., p. 45. The Bakwains are the same as the Bakwena. 4 Id. p. 201. xxvii SO UTH AND SO UTH-EAST AFRICA 453 [presenting the stock, form the tribe, which may include some ;who are relatives of the rest only by a fiction, but on the I whole consists of blood relations derived from a single stem. I The town, the composition of which Livingstone has just ex- plained, is built up on a different principle. The chief at the centre does not represent a stem except within his own, the principal centre, within which are said to be contained his wives and their families, and the wives and families of his relations. It seems to be implied in this that within the circle all were related by marriage or blood, but that the tie of blood did not connect the several circles, which therefore, it is stated, it was the policy of the chief to connect with his own circle by ties of marriage. The town, therefore, while within the several circles it may be composed on the patriarchal principle, is not, on the whole, patriarchal, and the cohesion of the whole of the sub-groups to the centre is seen to depend on the sovereignty, which is despotic, of the principal chief, and the marriage connections dictated by policy. We must inquire further, therefore, before presuming to form a definite opinion on this so-called patriarchal group. Let us now turn to the account of the Bechuanas found in Mackenzie's Orange River. Mr. Mackenzie uses his terms vaguely, like the other writers we have had to do with. The Bechuanas are a tribe, they are a race also. Among the Bechuanas are very many tribes, and of these there are often many portions in subdivisions. The term clan is used, but it means tribe, and tribe has no meaning (pp. 128, 177, 209, 356, 485). Mr. Mackenzie tells us (p. 356) that the Bamangwato are a tribe of the Bechuanas. So are the Bahurutse, who claim a precedence as to rank among the tribes in 1ST. Bechuanaland. The Bangwaketse and Bakwena are Bechuana tribes, and he says they were originally one people with the Bamangwato. " Tradition contains a glimmer- ing of the circumstances of their separation. The Bakwena included the Bamangwato when they separated from the Bangwaketse, but afterwards a subdivision took place, the Bamangwato being the younger or minor party. We have already seen that afterwards the Bamangwato again divided, 454 STUDIES IN ANCIENT HIS TOR Y CHAP. the minor party being now the Batowama, at present residin at Lake Ngami." Taking all these as tribes of the Bechuan tribe means group in general, and anything said about it i general reveals nothing as to its internal structure and recurrin to the statement first quoted from Dr. Livingstone, that Bakwen means the Alligators, I ask : Did the names Bangwaketse an Bamangwato have animal meanings also ? If so, when th three were one people they were a local tribe of three clans o the totemic principle. Kinship shifting from the female to th male side, their separation would be facilitated, as we find in th case of the Algonquins in North America, and the Narrinyeri in South Australia ; but on that view the Bamangwato, for instance, would be, however they split up, of one stock and have a common totem, and we saw that Mackenzie declared that the whole tribe had but one sacred animal. The Puti (p. 391), he says, was the sacred animal of the Bamangwato. If we assume this as true, the patriarchal group included in various circles were all of one totem or stock, and the several circles must be regarded as comprehending " kindreds " that had grown up and solidified under their several heads, within the stock- tribe, corresponding to the kindreds of the Welsh and Irish. There are difficulties, however, in the way of this view. There is reason to think the Puti was not the totem of the Bamangwato, but of the royal house of Sekhome's clan, and possibly not of his father's but of his mother's clan. The Makololo, says Mackenzie, kill it and dress its skin, but may not wear it in the town. The town referred to is Shoshong, its king Sekhome, and at p. 392 is a proof that the Pu was the totem of Sekhome's clan, and therefore, we ma be sure, religiously regarded throughout the town, so tha Mackenzie might well believe it to be the sacred animal o: the Bamangwato in general instead of being the totem of on of their clans. The Bamangwato may have been, therefore, strictly local tribe, and the various sections of it clans o different descent. The inquiry may not seem worth pursui yet as I have made it, it may be worth while, having examine the facts so far as disclosed, to exhibit them. The succession of son to father, so far as disclosed in th xxvn SOUTH AND SOUTH-EAST AFRICA 455 case of the chieftainship, and it is not disclosed in any other case, did not give the office to the eldest son of the king, but to the son of his wife of highest rank, whenever born. Sekhome was not the heir, though he was the eldest son of Khari ; 1 he slew one brother who had a superior title to the throne, and tried to slay Macheng, the only remaining rival, who was then child, the undoubted heir. Macheng was saved by his mother, carried into exile to the Bakwena, no doubt to rela- tives of his mother, and ultimately returning to Shoshong, became king. Macheng's father was Khari, but he was a putative father only, for Macheng was not born till some years after his death. He was counted the child of Khari on the principle of being seed raised up to Khari, but this prin- ciple would not have applied unless Khari, when he took Macheng's mother to wife, had paid in cattle the price for her and her offspring. Her child then was his on the same principle on which, as we see elsewhere in Africa, the children born to a wife who had been duly purchased, in the absence of her first husband, whom she believed to be dead, were counted to be the first husband's, though the second marriage held good. Mackenzie says : " Having paid her price in cattle, she and her offspring are to be reckoned to Khari, though the children should be born a dozen years after his death." When Sekhome was vexed, Mackenzie says he used to declare that Macheng was the " child of cattle," meaning that the price paid for Macheng's mother at her marriage with Khari was her son's only title to the chieftainship, "but even Sekhome's warmest supporters could not question the goodness of the title." 2 The same was the rule with the Makololo : marriage was by purchase, but in two ways. The price might merely cover the right to the wife, or it might also be made to cover the right to detain in the husband's family any children she might have, otherwise the children would belong to the family of the wife's father. Even when this price has been paid there is not a complete separation of the children from the house of 1 " As his mother was not the first wife of Khari as to rank" (Mac- kenzie, p. 359). 2 Mackenzie, p. 364. 456 STUDIES IN ANCIENT HISTORY CHAP. the mother. On the wife's death the husband has again to pay a price to her family " to give her up," and cause an entire severance of her children from them, a charge which will remind the student of the Mundoo of the tribes round Munniepore. 1 These are strong notes of the ancient filiation of children to their mother's clan, but what appear to be stronger ar( behind. Sekhorne, for example, lived in the house of his mother, and had there all his private wealth. While even the king again could not transmit his rank, every woman could. Lastly, marriages among the Bamangwato being effected by purchase were in this sense intertribal, that they were between persons belonging to different circles. Within each circle the children whose birth was reported to the chief belonged to the community rather than to the parents ; and the community being, except in the matter of marriage, highly communistic, the very idea of marriage by purchase, as the only marriage, carries with it that the persons marrying were of different communities. 2 Lastly, the youth who were formed into regiments on their initiation into manhood lived, according to both Livingstone and Mackenzie, in different circles in the town, while yet their union under a chiefs son, not by any means always or necessarily a son of the king, must have been made on some principle ; and the only one that occurs is, that however scattered they were throughout the circles they were still, through their mothers, of his clan. Having now, after making every allowance for polygamy in the royal house, seen so many things at variance with the patriarchal system, accord- ing to the ideas of it that have been formulated, let us look more closely at the structure of a Bechuana town. Mr. Mackenzie has given some details as to the town of Shoshong. It contained a population of 30,000, and was in five divisions ; round it there were some small towns, apparently affiliated to Shoshong, all under one chief, and probably altogether ranking as a sixth division. 3 All Bechuana towns would appear to be built on one 1 Livingstone, Zambesi, p. 285. 2 Id. p. 126. 3 Pitsan (says Moffat, p. 388), the principal town of the Barolong tribe, xxvn SOUTH AND SOUTH-EAST AFRICA 457 general model, which is thus explained. 1 In laying out a Bechuana town the first thing is to ascertain where the chiefs courtyard, with the public cattle-pen, is to be placed. As soon as this is settled the remainder is simple. As after the tabernacle was placed in the wilderness each one of the twelve tribes knew on which side he had to take up his position, so in the case of a Bechuana town. As soon as the chiefs position is ascertained, one says, " My place is always next the chief on this side " ; another adds, " And mine is always next on that side " ; and so on till the whole town is laid out. The town is called by two words one meaning urbs and the other home, and only freemen are entitled to reside in it. Here the different circles are likened to the tribes of the children of Israel. Presently Mr. Mackenzie speaks of the several circles as towns by themselves, as, for example : 2 " Those headmen whose towns are on the east of the chief, have their cattle-posts and hunting stations on the east side of the country," etc. And again : 3 " The headmen have power over their own towns, and over their own vassals and property ; and the chief himself is distinguished from other headmen only by having a larger number of vassals and more live stock," i.e. he has no right over their vassals or property. Thus the town seems to be composed of several towns, each containing a distinct group, and the whole united by confederacy and not otherwise under an hereditary king, who is at once king, general, judge, and high priest or sorcerer to the whole. But the evidence leads us further ; the people in the different circles were distinguished included a numerous division of Bahurutse and another of the Bangwaketse. It was under the government of three chiefs (p. 389). "As in all other towns, there were sections composed of the inhabitants of other tribes who congregate under chiefs of their own, and retain the name and peculiarities which distinguish their nations." At p. 353 he gives an address of the king to the Bechuana Parliament, " I command you, ye chiefs of the Batlapis, Batlaros, Bamairis, Barolongs, and Bakotus, tliat you acquaint your tribes," etc. At p. 248 he says that "tribe includes a number of towns and villages " (Missionary Labours and Scenes in South Africa, by Robert Moffat. J. Snow, London, 1842). 1 Mackenzie, p. 367. 2 Id. p. 368. 3 Id. p. 373. 458 STUDIES IN ANCIENT HISTORY CHAP. from one another by names, which Mr. Mackenzie says were their old intertribal names. When a Bechuana tribe ad- vanced into a new country, the whole scheme of their social life was gradually redeveloped there. The town was settled in the manner previously described, so as to be a counterpart of the one last left, the divisions in the town keeping up their old intertribal names, which of course apply also to the possessions in the country ; thus one part of the town of Shoshong (or rather a division of its people living together) is called Maloshu; a certain belt of garden ground also goes by that name, as well as a cattle-post and a hunting station. 1 If we knew what Maloshu meant, should we have a totem name ? Any way the whole of Mackenzie's statements go to show that the town or local tribe of the Bechuanas was composite of independent groups, and it is most probable that these were clans of different stocks. 2 We conclude this view of the organisation of the Bechuanas by a brief account of their political organisations and of the initiation of their young men to manhood, both of which matters throw light upon the questions we have been considering. The chieftainship was hereditary, the king judge, general, and high priest; he decides nothing of importance, however, without its being brought before a public assembly of free- men, to which a herald summons the people from the several divisions of the town. In going to such an assembly, the men march under their 1 Mackenzie, p. 370. 2 In his general statement on the races of South Africa, Mackenzie has some remarks on the character and disposition of the different tribes. He says : " In a Hottentot or Koranna town the houses are in the middle, sur- rounded by the cattle enclosure. A Zulu town, on the other hand, is built round a cattle-pen" (p. 501). He remarks that the Basutos, Bechu- anas, Mashonas, and Makalakas have better arrangements. "Whilst there is a great cattle-pen close to the public courtyard of the town, each subdivision has its own headman, its own pen for cattle, and fold for sheep and goats " (p. 501). No one can read of the arrangements without recalling the gotras (cow-pens) of the Indians, and the law which made ^it incest for two persons belonging to the same cow-pen to marry. xxvn SOUTH AND SOUTH-EAST AFRICA 459 own headmen, and in cases of dispute,- the headmen range them- selves under the chiefs they prefer, and thus march to the council. 1 A prompt disclosure is thus made of the relative strength of parties. The chief has in his court as assessors the headmen of the division of the town, and all men of distinction in it. Offences are all capable of being atoned for by payment of an eric. The Bechuanas have what we have found nearly all the world over, a ceremony of initiation to manhood, of which circumcision is a feature. It takes place only once in five or six years, when there are a sufficient number of young people ready for it. On the initiation they are formed into what both Mackenzie and Livingstone call regiments, though surely some other term would be more applicable, as the women, who at the same time are initiated to womanhood in a somewhat similar manner, are similarly formed into regiments. Living- stone (S. A., p. 146), after mentioning that all the Bechuana and Kaffir tribes practise circumcision, adds that there is a second part of the ceremony called Sechu, practised only by three tribes. The statement of Livingstone is that all the boys of an age between ten and fourteen or fifteen are selected to be companions for life of one of the sons of the chief. They are taken out to some retired spot, where the old men teach them to dance, initiating them at the same time into all the mysteries of African politics and government. After various ceremonies, inclusive of severe beatings of the boys, they are formed into bands or regiments (which are distinguished by different names, as Malsalsi, the suns, Mabusa, the tribes), and though living in different parts of the town, they turn out at the call of, and act under, the chief's son as their commander. They recognise a sort of equality and partial communism ever after. 2 This statement gives the idea that all the youth ready for initiation at one time were formed into one regiment under a son of the chief, on which view the institution would seem to be mainly political and not ancient, because chiefs' sons are not ancient. Mackenzie's account differs little from the pre- ceding, except that it recalls the initiation into the clan as we have seen it elsewhere where there were no chiefs' sons ; 1 Mackenzie, p. 373. 2 Livingstone, 8. A. y p. 148. 460 STUDIES IN ANCIENT HISTORY CHAP. and suggests that the ceremony was religious, and of the nature of confirmation or solemn dedication otherwise in the religion of the clan. He says that no youth can marry till after this ceremony ; that it includes instruction in the tradi- tional wisdom of the tribe. He agrees that all initiated at one time are banded under a regimental name, so as to be a compact body for warlike purposes, while, as civilians, owing allegiance to different headmen. The initiation is given to the youth "at the hands of the elders in his own family," who, headed by the priests, march in procession to the camp of the novices. They join in a sacred dance, and afterwards the initiated select their own relatives among the boys for the purpose of severely scourging them. All this is what we should look for on an initiation into the clan. My opinion is that we have here merely that initiation plus a device for strengthening the hands of the chief of the state. When Sekhome's sons declined to undergo the initiation, the chiefs who stuck to the old customs blamed them for refusing to enrol themselves as subjects and to enlist as militia-men, and Mackenzie remarks that if Christianity made progress among them, they would no doubt have to invent a new military organisation. THE BANYAI The country of the Banyai, called Shidima, as exhibited on the map accompanying Livingstone's Zambesi, is repre- sented as lying south of that river, between the parallels of east longitude 32-34. He seems not to have come in con- tact with them in his Zambesi exploration, but he gives an in- teresting account of some of their customs in his South Africa. The government of the Banyai, he says, is rather peculiar, being a sort of feudal republicanism ; the chief is elected, and they choose a son of the deceased chief's sister in preference to his own offspring. When dissatisfied with one candidate they even go to a distant tribe for a successor, who is usually of the family of the late chief or a sister's son, but nev.er his own son or daughter (p. 317). We have here female kinship and a strong suggestion of exogamy. xxvn SOUTH AND SOUTH-EAST AFRICA 461 The newly elected chief on taking office acquires as property the wives, goods, and children of his predecessor. "The children of the chief of the village," says Living- stone, "have fewer privileges than common freemen. They may not be sold, but they are never elected to the headship." In the village, beneath the class of freemen, are the slaves. The sons of freemen, between twelve and fifteen, live under leaders outside the village, under strict rule, and cannot marry till a fresh batch of the village youths is ready to take their places. As we might expect from the form of kinship indicated, the marriage system was peculiar. Livingstone remarks on the great influence of their women, and says that the wives are masters. Marriages are what are called fieena in Ceylon (S. A., p. 622). " When a young man," says Livingstone, " takes a liking to a girl of another village, and the parents have no objection to the match, he must go and live at their village." He there has to serve his mother-in-law for his wife, as Jacob served Laban for Eachel and Leah, but under more disagreeable conditions, for he is allowed only one pre- scribed attitude, and that an uncomfortable one, in which to sit before his mother-in-law. If he wearies of this service, and would return to his own people, he must leave behind all the children : they belong to the wife. This, says Livingstone, is only a more stringent enforcement of the law which pre- vails so very extensively in Africa, known to Europeans as buying wives. Such, virtually, it is, but it does not appear quite in that light to the actors ; so many head of cattle or goats are given to the parents of the girl " to give her up," as it is termed, i.e. to forego all claim on their offspring, and allow an entire transference of her and her seed into another family. If nothing is given the family to which the wife belonged can claim the children even when the marriage is deega, and the payment is made to put an end to this right. Livingstone remarks that among the Banyai there was a preference for having daughters married without a payment, as thereby the village was increased. Of course the whole conception of marriage of this sort involves the idea that 462 STUDIES IN ANCIENT HISTORY CHAP. marriage was what we call intertribal, if not exogamous. As to the village communities, if the marriage was even in the great run of cases beena, they would be groups very much of the Nair type, with the difference that the family circle, as among the Kocch, included the husband. Of totemism among the Banyai, I have found as yet no information. The Banyai lie to the north-east of the land in which were found the tribes of the Bechuanas and the Makololo. TRIBES, CLANS, AND CLAN MARKS It is needless to say that the same want of precision in the use of terms that we have had elsewhere, meets us in the nar- ratives of exploration in Eastern Africa. s Tribe, clan, caste, order, family, are used indiscriminately as names of groups, without any disclosure being made as to the description of group they were severally designed to designate. I shall rapidly put in juxtaposition the main statements to see what will come of them. Livingstone could not find out whether the nudity of the " go-nakeds " was a badge of a par- ticular order among the Bawe. They only referred to custom. The Bawe are in villages. When at strife within the village they are careful not to kill one another. When two villages are at war they are not so particular (Zambesi, p. 226). The Batokas, again, are brave hunters of buffaloes and elephants. The Batokas are a tribe, and a body of this tribe was induced to settle near Tette (Zambesi, p. 230). The body in question bore the name of Basimillongwe. Were they a clan ? Eecurring to the " order " of the Bawe, called Baenda pezi (go- nakeds), the Livingstones are found conjecturing whether they might not be an order similar to Freemasons. The hypothesis is rejected, because secret societies cannot be found among native Africans, except perhaps in Angola, where there is a sort of brotherhood, with community of right to food in each other's huts. Had the go-nakeds as much as or more than this ? If so, they were probably a clan, and their nudity a badge. Eecur- ring to the Batokas, a strong clannish feeling exists among them, as among all the other tribes. "In travelling, those xxvn SOUTH AND SOUTH-EAST AFRICA 463 belonging to one tribe always keep by themselves, and help one another " (Zambesi, p. 311). The tribe of Batokas therefore contains many " tribes," and each of these " tribes " is animated by a strong clannish feeling of brotherhood. The Batokas, in fact, are in clans. A dense population was found on the shores of Lake Nyassa, in the south, an almost unbroken chain of villages. All the natives are tattooed from head to foot, the figures being characteristic of the tribes, and varying with them. The Balonda were tattooed with small cicatrices (p. 272), and those who came with Sheakondo filed their teeth to a point, as do the Chiboque (p. 342); the people on the Loajima had cicatrices of various forms, some in shape of stars (p. 451). The Matumboka raise up little knobs on the skin of their faces, after a fashion that makes them seem covered all over with warts or pimples. Tribes here must mean stock-tribes or clans, for we have too often now seen the true meaning of such skin -marks to doubt their significance in a new case. At p. 437 a people had a bad name even among their own tribe. Were they a clan within the tribe ? In the country of Chonga michi, about eighty miles up the river, were found decent people of the same tribe. Were they another clan? Near them were settled a body of Makoa from the south. They were known by a cicatrix in the forehead, shaped like a new moon, with the horns turned downwards (p. 438). 1 1 Livingstone (L. /., vol. i. p. 49) says that the marks on the forehead of the Matumbwe seem a sort of heraldic ornament, for they can tell by his tattoo to what tribe or portion of a tribe a man belongs ; and of these people he says that the son takes the father's tattoo. The Makoa have the half or nearly full moon ; others have the cross or wavy lines. One tattoo seems to be a tree. The tribes or portions of tribes seem also to have peculiar tooth-marks. The Banyamwezi, he says (vol. ii. p. 180), derive their name from the new moon, which also provides them with the ornament that they wear round their neck. The Makoa, he is told, have surnames Miraze, Melolo, Chimposolo. "All have the half-moon mark when in the south-east, but now they leave it off a good deal and adopt the Waizan marks." They show, he says, no indication of being named after beasts and birds. At p. 110 he says, " The lines of tattoo of the different tribes are resorted to most by women ; it is a sort of heraldry closely resembling 464 STUDIES IN ANCIENT HISTORY CHAP. Other tribes have as their distinctive mark something done to their teeth. The Batokas, for instance, knock out the uppei front teeth, giving as a reason that they wish to be like oxen, and that otherwise they would be like zebras. The Batimpe knock out both the upper and lower front teeth. Among the Bakaas a child who cut the upper front teeth before the under was, says Livingstone (S. A., p. 577), always put to death, as he believed to be the case also among the Bakwains. In the Last Journals, vol. i. p. 276, he records a practice among Casembe's people and among the Makololo of knocking out the upper front teeth, and says (vol. ii. p. 120) of a boy from the Lomaine, " The upper teeth extracted seemed to say that the tribe have cattle. The knocking out of the teeth is in imitation of the animals they almost worship." The Babisa file their teeth to points ; other Manganja notch each of the upper fore teeth ; the notch is in some cases angular, in others round. The latter style gives the edges of the upper front teeth a semi-lunar shape. Other tribes make an opening of a triangular shape between the upper front teeth. What tribes these may be is not dis- closed, but it appears that the Babisa are Manganja, and dis- tinguished from other Manganjas by the style of their teeth. It appears also at the same place that the Matumboka, who we saw make knobs on the skin of the face, were also Manganjas, and that the only difference between them and the rest of the Manganjas was in the mode of tattooing the face. They have same language, but though the people found higher up the rive call themselves Matumboka, their tattooing differed from that oi people of the same name lower down. " Their distinctive marl consists of four tattooed lines diverging from the point betweei the eyebrows." Have we here confusion within confusion ? Man- ganja, a term for a state, or let us say empire ? The Matumboks a local tribe within it, and that again in clans ? (Zambesi, p. 5 24). ] the Highland tartans " (Last Journals in Central Africa, etc., by Dr. D. Livingstone. J. Murray, London, 1874). 1 The Chipeta, a division of the Manganja, have special ear and toot marks as their tattoo (Livingstone, L. J. t vol. i. pp. 1 30, 1 40). The Balungi are marked by three or four little knobs on the temple, and by distension of the lobes of the ear (Id. p. 201). xxvii SOUTH AND SOUTH-EAST AFRICA 465 This may be doubted, for though the Babisa are in a sense Manganja, yet they speak a different dialect, though they can con- verse (Zambesi, p. 547), and their relations are not a little difficult to be understood. The travellers noticed a very curious circum- stance at one stage of their journey. " Wherever a Manganja village was placed, a Babisa one was sure to appear in the vicinity." The former are the owners of the soil, but the latter did not appear to be considered intruders, so we are in a puzzle. Are we helped by what is found at p. 592 ? Gangs from one village plunder neighbouring hamlets of their children. " Next comes the system of retaliation of one ham- let against another, to make reprisals, and the same thing on a larger scale between tribes. The portion of the tribe which flees becomes vagrant, attacks peaceful tribes, etc." Let us see whether any light can be thrown on the nature of these groups from the character of their chiefs. It were useless to consider a group like the Makololos. We have only a disclosure of their powers in a form to be of use in two cases. Chinsamba was a chief in Mosapo. He had a great deal to do in attending to the affairs of his people. They consulted him on all occasions. He was eloquent in advice, and showed an intimate knowledge of his district. He knew "every rood cultivated, every weir put in the river, every hunting net, loom, gorge, and every child of his tribe. Any addition made to the number of these latter is notified to him, and he sends thanks and compliments to the parents " (p. 5 5 8). Now tribe here must mean a compact little group, such as might live in a hamlet or village, with the chief in the position of protector or father, in short, in loco parentis to all the members of it. Speaking apart from the question of kinship, which we shall consider hereafter, we may safely believe tribe here to mean clan. It must be in reference to this sort of group as the constituent sub-group of African society that Livingstone says (Zambesi, p. 598) that the African form of government is patriarchal, and, according to the temperament of the chief, despotic or guided by the counsels of the tribe. " The Manganja generally live in villages, each of which has its own headman, and he may be ruler over several adjacent 2H 466 STUDIES IN ANCIENT HISTORY CHAP. villages." The people are regarded as Ms children (Zambesi, p. 108). Child here means no more than subject or vassal. Thus (Zambesi, p. 198) : "Pangola is the child or vassal of Mpend< Sandia and Mpende are the only independent chiefs froi Kebrabasa to Zumbo, and belong to the tribe Manganja." It is said the Assenga appear to be of the same family as the rest of the Manganja and Maravi. Formerly all the Manganjf were united in an empire extending from Lake Shirwa to th< river Loangwa. It fell to pieces on the death ol Undi, th( chief who had established it, the Banyai, on the south, absorb- ing a large portion of it. TOTEMISM In Livingstone's Zambesi we find various traces of totem- ism, which may as well be brought together as they occur, without attempting to connect them with particular tribes. 1 Many of the indications are no stronger than the prohibition to eat particular food. First, it is said of certain natives of the Zambesi that they have as great a horror of hippopotamus meat as the Mohammedans have of swine's flesh. Livingstone's pilot would not even cook his food in a pot which had con- tained the meat : he would go without food rather ; yet he at with relish the flesh of the foul-feeding marabou. At Tette, where a great number of tribes converge, many superstitions converge, but Livingstone does not enu- merate them. The serpent is worshipped ; Morungo, who live above the stars, is not worshipped ; the manes are all go( and they make some offerings to them. There is a wide-spi superstition, shared by the Makololo, that to plant a mango ti would be death the disclosure here referring to Tette. W( come upon both the elephant and the crocodile, with theii respective schools of medicine, whose priests the doctors- drove a trade resembling the Koman Catholic sale of indul- gences. The crocodile doctor, for instance, sold a charm whicl protected its owner from crocodiles. Not enough is disclosed to show the true nature of these cases, but the following is very 1 Zambesi, pp. 46, 47, 51, 67. SOUTH AND SOUTH-EAST AFRICA 467 suggestive : " Unwittingly we offended the crocodile school of medicine while at Tette by shooting one of these huge reptiles as it lay "basking in the sun on a sandbank. The doctors came to the Makololo in wrath, clamouring to know why the white man had shot their crocodile." Again, some Africans in these parts believe that souls at death pass into the bodies of apes. We have already seen the ape as a totem of the Bechuanas, and we find him again religiously regarded at Sinjere, on the Zambesi route. " The monkey is a sacred animal in this region, and is never molested or killed, because the people believe devoutly that the souls of their ancestors now occupy these degraded forms, and anticipate that they themselves must sooner or later be transformed in like manner." At a Kebrabasa village the party were saluted by a man who claimed to be able to change himself into a lion. The Makololo believed that he could, that he was a Pondoro, and could change his form at will. When he takes the lion form he is sometimes absent for a month ; at times the transformed man returning claims to have provided the village with game : " Go and get the game that I have killed for you " (Zambesi, pp. 159,160). They saw another Pondoro at another village, the Pondoro of the village ; and this office we must connect with the belief that lions are sacred, the souls of departed chiefs being in them. Livingstone tells an amusing story of one of his men, im- bued with the belief that the lion was a chief in disguise, scolding a hungry one that had come close to the camp, " You a chief, eh ? " etc., the questions being all of the nature of con- temptuous chaff. Another of the men, equally believing that he was speaking to a chief, expostulated with him on the im- propriety of such conduct to strangers who had never injured him (p. 160). Among some tribes (though whether local or clan is not apparent) the dead were given to the crocodiles (p. 231) ; and (p. 232) a man refused to sell a boat, seeing a large serpent on a tree overhead, and alleging that this was the spirit of his father coming to protest against the sale. A donkey having died, Livingstone's men were shocked at the idea of eating him ; " it would be like eating man himself " (p. 335). 468 STUDIES IN ANCIENT HISTORY CHAP In the whole book, if we except the case of the mango which is faint, and that of the fig tree (p. 188), as seeming to have heen held sacred, I find no traces of plant worship, yet (at p. 521) Livingstone makes the following statement: "The primitive African faith seems to be that there is one Almighty Maker of heaven and earth, that He has given the various plants of earth to man to be employed as mediators between Him and the spirit world," and so on. This is qualified by the statement that they do not consider themselves amenable to the Supreme God, from which it must be inferred that they are amenable to the plants. Then this religion of theirs is saic to be mild in its nature, " though in one small corner of the country, called Dahomey, it has degenerated into a bloody superstition. Human blood takes the place of the propitiatory plants which are used over nine-tenths of the Continent. The reckless disregard of human life mentioned by Speke anc Grant is quite exceptional." Now the narrative of the exploration of the Zambesi is, more than anything else, a narrative of massacres and spoliations, and when we recall th< names of the Makololo, Matabele, and Azimus, to say nothing of Zulus and the Christain and Mohammedan slave-traders, it is rather surprising to read this statement. Human sacrifices again, are of frequent occurrence all over Africa. Nor is Dahomey the only place on the West Coast where their extent is appalling. Again, although the opening statement inclinec one to regard the plants as spiritual powers, their being put in contrast to blood shows that they are thought of as things to sacrifice to a power, and not as powers. The general statements unfortunately, are not supported by any detail, except th( statement that the Africans pay no regard to Morimo, which is vouched for on every hand, as fully as it is declared on al hands that the Africans have no idea of who Morimo may be and apparently owe what they have to the missionaries. In his South Africa Livingstone says of the Makololo (p. 199^ that they had no feeling about killing men, their lives from infancy being passed in scenes of blood, while (at p. 159) h( cites the case of a Bushman boasting of having killed five other Bushmen : " Two," said he, counting on his fingers, " were xxvn SOUTH AND SOUTH-EAST AFRICA 469 females, one a male, and the other two calves." "What a villain you are to boast of killing women and children of your own nation ! What will God say when you appear before Him ? " " He will say," replied he, " that I was a very clever fellow." For him God meant the chief of the tribe. The notices of totemism in the works of Speke and Grant are few. In the mythology of the royal house of Karague* we find that on the death of Kurnanika's successor, the body, sewn in a cow's skin, was left in a boat on a lake for three days, when, decomposition having set in, maggots were en- gendered from the royal person, of which three were taken into the palace and put in charge of the heir-elect. These maggots seem to have symbolised the blood that flowed in his veins. One became a lion, another a leopard, and the third a stick. This miracle over, the dead king was duly put under ground, " with five maidens and fifty cows " (Speke's Journal, p. 221, Edinburgh, 1863). Speke's curiosity being roused, he obtained from Eumanika some genealogical disclosures. He related that his grandfather, Eohinda VI., having died, and his body been taken to the hill to be buried, a young lion emerged from the heart of the corpse and kept guard over the hill, from whom other lions came into existence, until the whole place became infested by them, which had since made Karague" a power and dread to all other nations, for these lions became subject to the will of Eohinda's son and heir, who, when attacked by the countries to the northward, instead of assembling an army of men, assembled his lion force, and so swept all before him. How much of this may be pure misunderstanding it would be useless to inquire. But one other trace of animal worship occurs in Speke. Eumanika, who yearly sacrificed a cow at his father's grave, also made periodical sacrifices to a large stone on the hillside, and mentioned that if, when leading an army to battle, he were to hear a fox bark, he would retire at once, knowing that that meant evil. Speke adds the general statement, which is of little use to any one, there are many other animals, and lucky and unlucky birds, which all believe in (p. 241). 470 STUDIES IN ANCIENT HISTOR Y CHAP. It illustrates the nature of the casualties on which infor- mation on such matters depends, that while Speke gives us so little bearing on the subject, Grant, by pure accident, gives us a good deal. At a village of Ukuni he found it was forbidden to eat a particular kind of antelope. Grant was proud of having shot it ; his follower pretended not to be able to carry it, but Grant insisted. When close to the village he began to learn from his attendant that there was something wrong. " M'weeko, i.e. it is not customary ;" it is a " Phongo, never eaten." Grant made an appeal to the Sultan to get his beast secured ; the Sultan flew into a rage. It was finally explained to Grant, " If you eat it, you will lose your fingers a"nd toes, get scab all over, etc." They would not admit even the skin inside the village. The Seedees, who were with Grant, said it was never eaten in their country. Grant says it was as nice an antelope as he ever saw. He sold it to a native caravan, who were glad to eat it. Another animal, he says, which it is not customary to eat, is the N'grooweh, also a species of antelope. 1 The next note of totemism in Grant occurs at p. 126. Of game-birds, he says, the most plentiful was the guinea-fowl, near the cultivations. The natives of Usui will not eat the fowl, but the Walinga, a class of people who work in iron, have no objection. On the next page we have some marvellous stories about the king of birds and the tippet monkey, with a bird of good omen to Grant's men, called Kong-ot'a. They were certain to have luck after seeing it. In Ugogo a lizard would appear to be sacred. There are many species of them, but only one, about twelve inches long, appeared to be religiously regarded (p. 36). At another place, among the Wagogo (p. 41) a lizard was also religiously regarded. Grant shot a lizard, and was told he had hurt their feelings, and must pay for his folly. Among the Watusi, " a curious and distinct race," in the province of Unyanzembe, Grant found what is common in India, and is an infallible sign of cow worship. These people, 1 A Walk across Africa, etc., by J. A. Grant, Capt. etc., pp. 90, 91. Edinburgh, 1864. xxvii SOUTH AND SOUTH-EAST AFRICA 471 previous to milking the cows in the morning, washed themselves, their teeth, and their wooden milk vessels or gourds, with the urine of the animal. Here the men are the milkers (p. 51). We have seen the lion in a religious connection with the family of Kurnanika, and that one of the animals sprung from the body of Eumanika's father ; and at p. 141 is something further of the relation of the leopard to that family. Eumanika hav- ing been consulted as to some strange animals that had come at night to the camp, recommended that should they come again they should be challenged three times, and fired upon if they did not answer. His belief was that they were enemies sent by his rebel brother. Should they prove to be leopards, however, they were not to be molested. For all leopards, says Grant, they have a great reverence, as the late sovereign is believed to be still protected by them, and, on an invading army coming from Uganda, this Sultan had the power to send leopards to disperse them. Leopard-skins were worn only by royalty or its followers. Speaking of the religion of Eumanika and his people, Grant says he could not trace any distinct form of it. A sacred signification was attached to a piece of copper, said to represent the horns of cattle. "A tree was considered the greatest object in the creation, not even excepting man." Lions protected the mausoleum of the former Sultan. "No kingdom was so powerful as this ; no one dared attack us. Lions guard us" (p. 145). Apparently an invisible god lived in the lake, and the reason why Speke could not kill any hippopotami there was that he had not conciliated that god by a present. The Sultan of Unyoro could divide the waters of the lake with a rod (p. 145) perhaps the mystic stick of the maggot. The curious thing is that though Eumanika believed that lions garrisoned the country, neither Grant nor Speke saw one in the country, dead or alive. As to the god of the lake, he was possibly the king of the otters, whom the people ^superstitiously regard, saying that the king of the otters is as white as an old man's beard. Eecurring to the subject of religion at Karague, Grant adds to his previous account that he found that there was a belief that certain 472 STUDIES IN ANCIENT HISTORY CHAP. animals were possessed of devils, but were in the power of the soothsayers (p. 186). In addition to the notes of totemism already mentioned, numerous others might be cited, as Oxen amongst the Bechuanas, Barolongs, Batokas (where the veneration for cattle is coupled with antipathy to the Zebra), and elsewhere ; 1 the Hare amongst the Hottentots ; 2 the Elephant amongst the Koossas, 3 though there is reason to suppose that the Elephant was only the totem of the chief clan, and that another totem of the Koossas was the Hysena ; 4 the Duck (perhaps) amongst the Maravi ; 5 Fish amongst the Banyamwezi, Matabele, and some of the Bechuanas ; 6 the Hippopotamus amongst the Matabele ; 7 the Hyaena amongst the Bangwaketse, the Koossas, and the Makoa ; 8 the Leopard also amongst the Makoa and Makonda ; the Wagtail amongst the Bechuanas and elsewhere ; 9 the Whydah Bird, 10 the Snake, 11 a Beetle, 12 the Lion, 13 the Alligator amongst the Bamangwato and Bakwains. 14 SUCCESSION AND KINSHIP There are two reports by officers of the Indian Service on some tribes within the dominion of Zanzibar. Lieutenant- 1 Moffat, pp. 278, 349, 580; Pinkerton, vol. xvi. p. 690; Living- stone, L. J., p. 128 ; Id., S. A., pp. 263, 339, 450, 462, 466, 532. 2 Narrative of an Explorer in Tropical South Africa, by Francis Galton, p. 188. Ward and Lock, London, 1852. 3 Walckenaer, vol. xviii. pp. 187, 199 (Collection des relations de voyages, etc., par C. A. "Walckenaer. Paris, 1842). 4 Id. pp. 187, 189, 199. 5 Livingstone, S. A., p. 577. 6 Matabeleland and the Victoria Falls, by F. Gates, p. 111. Kegan Paul, London, 1881. 7 A Hunter's Wanderings, etc., in Africa, by F. C. Selous, p. 135. Sampson Low, London, 1881 ; and vide Livingstone, Zambesi, p. 39. 8 Moffat, p. 401. 9 Livingstone, L. J., vol. i. pp. 190, 281. 10 Id. vol. i. p. 281. 11 Id. vol. i. p. 281 ; vol. ii. p. 344 ; and cf. Zulu regard for snake, Leslie. 12 Id. vol. ii. p. 27. 13 Id., S. A., pp. 282, 304. w Id. p. 255. xxvn SOUTH AND SOUTH-EAST AFRICA 473 Colonel Rigby's report (Bombay, 1861) relates chiefly to the slave trade and Arab customs. The manners and customs of the Mukhadim, the aborigines of Zanzibar, are not described. They are said to resemble those of the natives of the East Coast of Africa. Lieutenant- Colonel Play fair's official letter, rather than report (Bombay, 1864), relates to the Wanika, a tribe on the East Coast. They keep cattle, and have the phrase cattle-pride. They have ceremonies of initiation to manhood, like those we have everywhere found. They have hereditary chiefs of little or no authority, and the real govern- ment of a group is carried on by the elders. Among the Wanika, children belong not to the parents but to the mother's eldest brother, who has absolute power over them. Marriage is by purchase, the price paid to the bride's family. Divorce is frequent : if it is the wife who retires the price must be repaid, but not till she has got another husband. There is a system of betrothals, and marriage takes place early, but the bond is a slight one, and continency is hardly known. Of course this implies that the Wanika have exogamy and a system of female kinship. We find female kinship again on Livingstone's route, and in connection with the Kebrabasa, among whom we found a trace of totemism. A sister's son, says Livingstone of this people, has much more chance of succeeding to the chieftainship than the chief's own offspring, it being un- questionable that the sister's child has the family blood. Again, among some unnamed tribes near the Sinjere, where also we found unmistakable signs of totemism, the position of women is exceedingly high ; they traded in the camp, and seemed both sensible and modest. Perhaps in this particular the women here occupy the golden mean between the Man- ganja hill tribes and the Jaggas of the north. It is said that the Jagga women do all the trading, have regular markets, and will on no account allow a man to enter the market-place. The same sign of women being in the first place occurs in the Upper Shire Valley, which is under the government of a lady. In her dominions women rank high (Livingstone's Zambesi, pp. 108, 192, and 486). Among some Manganja, near the 474 STUDIES IN ANCIENT HISTORY CHAP. village next to Nkoma's, the women were found to assume a superiority over their husbands. " A man was selling a goat " to the explorers (p. 550) "when his wife came in and pulled him up : ' You appear as if you were unmarried, selling a goat without consulting your wife ; what sort of a man are you ? ' The bargain was fairly off. If this was a fair specimen of domestic life, the women here have the same influence that they have in Londa and farther west, and in many parts north of the Zambesi, where we have known a wife order her husband not to sell a fowl, merely, as we supposed, to show us strangers that she had the upper hand. We conjectured that deference was commonly shown to women here, because, as in the West, the exclamation most commonly used was, ' my mother ! ' We heard it frequently some thirty miles east of this, where the inhabitants took us for the Mazitu. South of the Zambesi the exclamation oftenest heard is 'My father!" South of the Zambesi is a wide term. The Bamgi are south of the Zambesi, and we know the position of women among them. Mackenzie notices that as far south as the neighbourhood of Shoshong the Makalaka swore by their mothers and not by their fathers, as it is said the other tribes did. 1 In considering society at Shoshong I omitted to point out, what Mackenzie repeatedly states, 2 that in questions of succession the division of the town to which the mother of a claimant belonged invariably supported the claimant. " mother ! " is the exclamation of the Wangwana, Zanzibar. 3 So it is amongst the Balonda 4 and the Makonde. 5 Notes of female kinship are in fact abundant, and it is unnecessary to multiply instances ; it is sufficient to refer the reader to various passages where they occur. 6 On the other hand, in some cases kinship is reckoned through males, as appears to be the practice amongst the 1 Mackenzie, p. 508. 2 /^ p. 410. 3 Through the Dark Continent, vol. ii. pp. 34, 264, by H. M. Stanley. Sampson Low, London, 1878. 4 Livingstone, S. A., p. 298. 5 xd., L. /., vol. i. p. 28. 6 Id. vol. i. pp. 77, 223, 289 ; vol. ii. pp. 69, 88 ; Id., S. A., pp. 268, 273, 281, 309, 502. xxvn SOUTH AND SOUTH-EAST AFRICA 475 Bechuanas, whose chief Sebituane installed his daughter as chief before his death ; " he wished to make her his successor, probably in imitation of some of the negro tribes with whom he had come into contact, but being of the Bechuana race, he could not look upon the husband except as the woman's lord," so he wanted her to have men as he had women, but not to have a husband. One man whom she chose was even called her wife, and her son the son of her wife ; but the arrangement did not please her, so, as soon as her father died, she made over the government to her brother Sekeletu. 1 This story is not incompatible with the tracing kin through females, but it seems to be fully proved that amongst the Bechuanas, Barolongs, and a good many other tribes the habit exists of paying a higher price for a wife for the purpose of securing her children, who would otherwise belong to the wife's family, showing at once the existence of kinship through women side by side with that through men, and indicating the manner in which the system of kinship may gradually change. Among the Bechuanas the son inherits his father's wives, and their children to him are called his brothers ; and when the elder brother dies the next brother takes his wives, as among the Jews, and the children that may be born of those women he calls his brothers also ; he thus " raises up seed to his departed relative." The chief has a head wife or queen, and her children inherit the chieftainship. Some of the wives are given to under-chiefs. 2 Among the Bangalas of the Cassange valley the chief is chosen from three families in rotation ; a chief's brother inherits in preference to his son. The sons of a sister belong to her brother, and he often sells his nephews. 3 This is a pretty wide scattering of notes of female kinship. Here and there we must believe kinship has shifted, as in Australia, to the male side, but enough evidence remains to show that anciently the marriage system was such that the children always belonged to the mother's family. We have seen in many cases property in the children acting as a means of destroying the ancient filiation to the mother's 1 Livingstone, 8. A., p. 179. 2 Id. p. 185. 3 Id. p. 434. 476 STUDIES IN ANCIENT HISTOR Y CHAP, xxvn family ; and, as in the case of the Wanika, we see that when one can get at the facts, the mother's family means her family through the mother, her mother's brother appearing as the head of the house. In Speke we seem to see an intermediate stage, which recalls an arrangement prevailing among the polyandrous Todas. Two men at Karagu^ having married one woman laid claim to her child, which, as it was a male one, belonged to the father. 1 That is to say, we find a case of polyandry, with an appropriation of the female children to the mother, and the male to the father, an arrangement which occurs elsewhere in Africa. Polyandry, especially of the Thibetan type, should not surprise us at Karague". Grant (p. 186) says that among the Wahuma kings it was lawful to co- habit with a brother's wife, or with an own sister. The latter statement tells directly on the succession law, and to the effect that the succession was according to the system of female kinship. The former statement is all that Mr. Grant has yet published in support of his statement to me personally, that Thibetan polyandry was the marriage system of the Wahuma. 1 Speke, p. 239. CHAPTER XXVIII KAFFIRS AND ZULUS THE term Kaffir, says Theal, is usually restricted to members of the Amaxosa tribe, 1 but he adds that all along the East Coast as far as the great Fish Eiver the country is thickly peopled with Kaffir tribes, each of which has its own title, usually the name of its first great chief. In the year 1683, he says, the country between the Cape and Natal was found to be occupied by four great tribes the Amanipondomsi, Amampondo, Abatembu, and Amaxosa, whose language was the same, though much mixed with that of the Hottentots, owing to the marriage of women after defeating and killing the men, and by the adoption of Kaffir refugees, as was done by the Guanaqua tribe of Hottentots. 2 Other writers use the term Kaffir in a much more extended sense, and sometimes apply it to all the allied races of the Bantu family, such as the Bechuanas and Zulus, who live on the East Coast. The Zulus, who believe that they sprang from a reed, 3 are a typical race of the Bantu family, and we are fortunate enough to possess a fairly complete account of their customs from the pen of David Leslie, who was well acquainted with them. We have a very important statement as to the marriage laws and relationships of the Kaffirs by Theal, who treats the Tembus, 1 Kaffir Folklore, by G. M. Theal, p. 1. Swan Sonnenschein and Co., London. 2 Id. p. 2. 3 Among the Zulus and Amatongas, by David Leslie, pp. 76, 207. Edinburgh, 1875. 47 8 STUDIES IN ANCIENT HISTORY CHAP. Pondos, Zulus, and other tribes as all belonging to the Kaffir race. 1 It appears that kinship is now generally traced through males, 2 but there are indications that female kinship formerly prevailed. Santos says that the king's brother at Quiteva was elected to the throne in preference to any of his sons, 3 and in the Amatonga country on the borders of Zululand, where the women are much better treated than amongst the Zulus, 4 the nephew always succeeds to the chieftainship. 5 Santos, who wrote in 1684, describes the Kaffirs generally as a nomadic race subject to terrible famines, and, at Tette, as feeding on human flesh. He states that polygamy was prac- tised, and that marriage was by purchase, and that the husband might send the woman back if displeased with her. He men- tions that the inhabitants of some of the islands off the coast of Melinda lived in harmony, but divided into war parties on the mention of religion, and that at the death of the king of Quiteva, who set up to be the only sorcerer in his dominions, his wives devoted themselves to death. 6 Polygamy is practised, but amongst the Zulus the king's regiments were not allowed to marry without permission. The result of granting such permission to a regiment was that all the marriageable girls were swept up at once. The purchase of wives appears to be an universal custom throughout South Africa, and amongst the Zulus, although the form of capture is no longer a part of the marriage rite, there seem to be unmistakable traces that such a custom formerly existed. Leslie, who had attended many Zulu marriages, tells us that the girl, who has been escorted to the kraal of the bridegroom, which may be, and is always feigned to be, at a great distance from 4 her home, always makes several attempts to escape, and if she succeeds in getting out of the gate of the kraal it is a great disgrace to the bridegroom, and the whole ceremony has to be gone through again. 7 Leslie adds that he had heard that formerly the bridegroom went to the bride's 1 Theal, p. 198. 2 Id. p. 198. * Pinkerton, vol. xvi. p. 683. 4 Leslie, p. 125. 5 Id. p. 126. 6 Pinkerton, vol. xvi. pp. 344 (inter- polated after p. 674), 696, 710, 723. 7 Id. pp. 115 et seq., 196. KAFFIRS AND ZULUS 479 kraal and took her away. If any cattle paid for the wife die within the year they must be replaced, and if the wife dies the cattle paid for her can be reclaimed, but generally a sister is sent " to raise up the house of her that is dead." l The Kaffirs, according to Theal, have a strictly exogamous marriage law. " Marriages," he says, " are absolutely pro- hibited between people of the same family title. ... A man may marry a woman of the same clan (clan must, of course, be used here in the sense of group or village) that he belongs to, provided she is not a blood relative, but he may not marry "a woman whose father's family title is the same as his own, even though no relationship can be traced between them, and the one may belong to the Xosa and the other to the Pondo tribe. . . . Among the Tembus, the Pondos, the Zulus, and many other tribes are people with the same family title. They cannot trace any relationship with each other, but wherever they are found they have ceremonies peculiar to themselves. Thus the customs observed at the birth of a child are exactly the same in every part of the country among people of the same family title, though they may never have heard of each other; while neighbours of the same clan (here again clan evidently means local tribe) but of different family titles have these customs altogether dissimilar. All the children take the family title of the father, and can thus marry people of the same family title as the mother, provided they are not closely related in blood." 2 In the above statement is a proof that some, if not all, of the South African races are strictly organised in clans upon the totemic principle. Theal mentions one of the family titles Amaywabe but does not give its meaning. Theal adds 3 that to the Kaffirs what we term cousins are brothers and sisters, and that this does not arise from poverty of language, for they have words to express shades of relation- ship where we have none. They use the same word for father 1 Leslie, p. 141. 2 Theal, p. 198. Theal's evidence is clear in this passage that kin- ship is traced through males, but some of the Bantu races seem to trace kinship through females. 3 Id pp. 210, 211. 480 STUDIES IN ANCIENT HISTORY CHAP. and paternal uncle and father's half-brother. In the same way their word for mother means any of their father's wives. That is to say, the Kaffirs distinguish very clearly between the family of the father and the family of the mother, and when a Kaffir woman marries her husband's parents give her a new name, by which she is known to his family, and on the birth of a child she is frequently described as the mother of the child. 1 Burchell's testimony is in harmony with that of Theal. He says that those who are descended from the same ancestor, however far off, are called brothers and sisters, and conse- quently the chiefs of the Amakoses usually take their principal wives from the Tamboukies, because all the families of rank among them are descended from the same stock. Yet it is not forbidden to marry a deceased wife's sister. Amongst the Zulus and some, if not all, of the other Kaffir races it is considered a disgrace to a woman to be seen by her son-in-law, or to mention the name of her father-in-law. This custom is called Hlonipa, and is observed so strictly amongst the Zulus that the worst oath applicable to women is " O'mka ninazala," which means that she does or will bear children to her father-in-law. For a minute account of this custom the reader is referred to Leslie. 2 Leslie says that there are three kinds of Hlonipa the family, the tribal, and, in the case of the Zulus, the national. The tribal Hlonipa is equivalent to " taboo," that is to say, a man will not use the name of his chief or progenitors in everyday conversation. National Hlonipa applies only to words sounding like the name of the king. By stating that national Hlonipa is confined to the Zulus. Leslie seems to imply that family Hlonipa is customary throughout the country. Theal also mentions the custom of family Hlonipa, whereby the wife is prohibited from sitting in the house where her father-in-law is seated, or pronouncing the names of any of their husband's male relations in the ascending line. 3 Indications of totem worship abound amongst all the Bantu i Theal, p. 211 ; Livingstone, S. A., p. 126. 2 Leslie, pp. 172 et seq. 3 Theal, pp. 10, 202. xxvm KAFFIRS AND ZULUS 481 races. The Koossas, says Lichtenstein, venerate the elephant, and regard the hysena as sacred. 1 Cattle are only killed for food on solemn occasions. 2 The snake is treated with great respect, 3 and the Zulus believe that the dead revisit their old kraals in that form. Every one is supposed to have two snakes, " ehloseY' a good one and a bad one. 4 The whole Zulu nation, says Leslie (in 1870), is broken up into little tribes, the remnants of those conquered by Chaka. Each tribe has its Esebongo or name of thanks ; for instance, one tribe is called Emtetwa or scolders, and so on. Each of these tribes has its peculiar habits and customs. " One, the Mat-e-enja (dog's spittle), will not eat goat flesh, because they always leave a goat on the grave of their dead." " If they eat any part of a goat unawares they are seized with epilepsy and die. Even the young children in the kraal, who are too young to know anything of this, when a piece of goat flesh is given to them, will not eat it, but carry it in their hands for a little and then throw it away ; and be it remem- bered that meat is their greatest dainty ! " 5 Theal gives the names of two dances, 6 and it is not uncommon to hear of different dances danced by different " tribes." Vengeance appears to rest with the family ; 7 and Barrow says 8 that the Kaffirs did not make war for conquest, but only in revenge for some direct insult or act of injustice towards the whole community or one of its members. Nearly all the wars are occasioned by cattle raiding. In other words, we have in South Africa an example of the struggle for existence by mankind in its primitive form. The Matatees, says Moffat, a pastoral people, robbed of their cattle by the Matabele, had nothing left but to die or rob ; and from being wild men became like wild beasts. 9 In South Africa, says Livingstone, most wars are about cattle. Several tribes have no cattle to avoid war. Livingstone had only heard of 1 Walckenaer, vol. xviii. pp. 187, 189, 199. 2 Id. vol. xviii. p. 198. 3 Theal, p. 22. 4 Leslie, pp. 47, 120. 5 Id. p. 146. 6 Theal, p. 197. 7 Leslie, p. 154. 8 Walckenaer, vol. xvii. p. 310. 9 Moffat, p. 372. 2 I 482 STUDIES IN ANCIENT HISTORY CHAP, xxvm one war on another account, namely, amongst the Barolongs, where three brothers fought about a woman. 1 How keen the struggle must be is apparent from the effects of a prolonged drought, or from the wholesale loss of cattle in war. The Matatees, a great and successful nation, were found to be in a state of starvation. With all their conquests they were dying of hunger, and for hundreds of miles their march might have been traced by human bones. 2 The usual con- comitants of such a state are found in the steps taken to economise food by the destruction of sickly and deformed children and weak and aged persons, and in some cases by the practice of cannibalism. 3 The habit of recruiting a tribe by the seizure of women and children from other communities seems to be universally prevalent. 4 And, as we see every- where, marriage is by purchase ; and in some cases, if not in all, the marriage rites indicate a form of capture as part of the ceremony. Examples recur in these pages, and it is unnecessary to dwell upon the point here. 5 1 Livingstone, A, pp. 10, 213. 2 Moffat, p. 369. 3 Livingstone, L. J., vol. ii. p. 189 ; Id., S. A., p. 577. Pinkerton, vol. xvi. pp. 694, 710, 717. Moffat, pp. 57, 132. 4 Stanley, Dark Cont., p. 253. Livingstone, L. /., vol. i. p. 205. 5 Pinkerton, vol. xvi. p. 696. Livingstone, L. J., vol. ii. p. 73. And see the Eeport of Bethell's case, Law Keport, 38 Ch. Div. p. 220. CHAPTER XXIX BUSHMEN AND HOTTENTOTS THE Bushmen and Hottentots appear to be members of the same family, and may be said roughly to have occupied what is now Cape Colony when it was first colonised by the Dutch. Mackenzie classes them with the Korannas and Griquas (the Griquas being mostly of mixed blood), as forming a race distinct in language and customs from the Kaffirs and other South Africans. 1 Amongst them the women are milkers of cows, whilst amongst the Kaffirs, Bechuanas, and other tribes of the Bantu family, women are not allowed even to enter a cattle- pen while the cattle are in it. 2 Galton says that there is no difference between Bushmen and Hottentots, and probably it may be safe to treat them as of the same race, although some portions have attained a higher position in the scale of civilisa- tion than others. 3 Moffat, however (p. 6), while treating the Bushmen as part of the Hottentot nation, says that they are distinct peoples, and cannot understand each other's language. Unfortunately confusion may occasionally arise in trying to distinguish these races from those of the Bantu family, owing to the fact that some travellers use the name Kaffir as synony- mous with Hottentot, or do not properly distinguish between the two. 1 Mackenzie, p. 490. Selous (p. 108) considers they were probably the earliest settlers in this part of the country of whom we know any- thing. 2 Mackenzie, p. 499. 3 Galton, p. 68 ; Moffat, pp. 3 et se%. 484 STUDIES IN ANCIENT HISTOR Y CHAP. The Hottentots, says Mackenzie, may be readily dis- tinguished from the Bantu races by their different manner of housebuilding. 1 The hut of the Hottentot has no wall distinct from its roof, whereas the Zulus and Bechuanas build walls of wattle and clay with a grass roof. Some writers, while admitting a connection between Hottentots and Bushmen, draw a great distinction between them as regards their manner of living, and certainly some accounts of the Bushmen or Bosjies- men represent them as being much wilder and ruder than the Hottentots, or at any rate than those Hottentots at the Cape who had come into contact with Europeans. The Bushmen, says Barrow, 2 differ in several respects from the Hottentots ; they are almost entirely naked and very short, the men not exceeding 4 feet 9 inches in height and the women 4 feet 4 inches. They live, says Campbell, 3 apart from each other a nomadic life, and are said to love their children, though there are frequent examples of the women killing their offspring if the fathers will not maintain them. Sometimes a young man steals a woman from her hut while she sleeps, and the theft may be committed without the parents' consent, and even without any information of the man's intention having been given. Most travellers concur in placing the intelligence of the Hottentots very low. They seem to be fast dying out. Thun- berg says that they, as well as the Kaffirs, were organised under chieftains who held sway in each village. 4 The few remains of them, he says, that still exist have in some instances retained the names of their tribes, but more frequently those names are retained which formerly distinguished each nation separately, and are applied to the district itself; and he gives the names of several such " tribes " with which he had come into contact, describing their geographical position. 5 They live together sometimes to the number of several hundreds in a village, and feed on roots and the flesh of wild beasts and of their own herds. 6 Their chief occupation is 1 Mackenzie, p. 499. 2 Walckenaer, vol. xvii. p. 308. 3 Id. vol. xviii. pp. 465 et seq. 4 Pinker-ton, vol. xvi. p. 37. 5 Id. p. 71. 6 Id. p. 83 ; Moffat, p. 54. xxix BUSHMEN AND HOTTENTOTS 485 hunting, at which they excel all other South African races. Livingstone describes the Bushmen as thin and wiry, and possessing no cattle or domestic animals but dogs. The Bakala- hari, who live under the same conditions as the Bushmen, he describes as being quite different, owning , ( cattle and tilling land, and belonging to the same race as the Bechuanas. 1 According to Thunberg the patriarchal form of government has from time immemorial existed amongst them, and this has been the origin of the many larger and smaller " tribes " and families into which they formerly were, and their survivors still are, divided. 2 Kolbe' says that all the property of a father descends to the eldest son, or passes in the same family to the nearest male, and never to a woman. The liberty and fortune of a younger son is in the hands of the eldest, if unmarried ; and the eldest son, if unmarried, is bound to provide for his father's wives. If married at the father's death the property is shared between all the brothers. 3 Thunberg says that the eldest son is sole heir to his father's property. 4 Thunberg's reference 5 to "nations," "tribes," and "families " leaves open the question whether the Hottentots were organised in clans upon the totemic principle. It is difficult to say whether any animals were strictly regarded as totems, though it appears from the testimony of several writers that the Hottentots regard some insect, generally said to be the mantis, and also the hare, with peculiar veneration. 6 Thunberg also mentions the turtle and tortoise as antidotes to poison and the dove as seldom eaten, but whether by the Hottentots or Kaffirs is not clear ; 7 and Livingstone says, " The animal they (some Bushmen under Horoge) refrain from eating is the goat." Kolbe says that they honour the moon, to which they sacrifice cattle, and the stag-beetle; 9 but Walckenaer notes that La Caille says that the Hottentots look on this insect as of evil augury, and that Le Vaillant says it is a mantis, and not 1 Livingstone, S. A., p. 49. 2 Pinkerton, vol. xvi. p. 139. 3 Walckenaer, vol. xv. p. 380. 4 Pinkerton, vol. xvi. p. 142. 5 Id. p. 71. 6 Id. pp. 37, 97, 142. 7 Id. pp. 11, 37, 74. 8 Livingstone, S. A., p. 165. 9 Walckenaer, vol. xv. p. 372. 486 STUDIES IN ANCIENT HISTORY CHAP. worshipped by the Hottentots. Kolbe adds that they will not kill their cattle except in case of pressing necessity j 1 but if so, their reluctance can hardly be ascribed to any veneration for cattle, as all the evidence points to their being cattle-eaters. 2 The gazelle called duyker was also observed with veneration, and not eaten except in the last extremity. 3 And Kolbe says that ancient traditions force them to abstain from certain food, such as pork and fish without scales, which are forbidden to both sexes. Hares and rabbits are forbidden to men only, and "le pur sang des animaux" and the flesh of the mole are forbidden to women only. 4 On the death of a Hottentot the corpse is taken out of a hole in the hut", after which some animal is killed and eaten. 5 The Hottentots appear to be a superstitious race, 6 and have a folklore and fairy stories, most of which turn upon the trans- formation of men into animals and birds. 7 They seem to believe in some sort of future existence. 8 The Damaras have a legend accounting for the origin of themselves and the Hottentots, which is stated by Galton as follows : 9 "In the beginning of things there was a tree (but the tree is somehow double, because there is one at Omaruru and another near Omutchamatunda), and out of this tree came Damaras, Bushmen, Oxen, and Zebras. The Damaras lit a fire, which frightened away the Bushmen and THE OXEN ; BUT THE ZEBRAS REMAINED. 10 Hence it is that Bushmen and wild beasts live together in all sorts of inaccessible places, while the Damaras and the Oxen possess the land." Mr. Galton adds that notwithstanding that everything came out of the tree, men have in some separate manner a special 1 Walckenaer, vol. xv. p. 323. 2 Moffat, p. 349. 4 Walckenaer, vol. xv. p. 324. 6 Walckenaer, vol. xv. p. 373. 8 Livingstone, S. A., p. 165. 10 This is an obvious error. It should be, " frightened the Bushmen and Zebras, but the Oxen remained. 7 ' [The correction has been made in the later edition, 1889, p. 115.] 3 Le Vaillant, vol. ii. p. 110. 5 Pinkerton, vol. xvi. p. 142. 7 Theal, passim. 9 Galton, p. 188. xxix BUSHMEN AND HOTTENTOTS 487 origin or " eanda." There are six or seven " eandas," and each has some peculiar rites. The tribes do not correspond with the " eandas," as men of every descent are to be found in each tribe. Independently of the tree and the " eanda " there is also Amakuru, hardly to be called a deity, though he gives and withholds rain. He is buried in several places, at all of which he is occasionally prayed to. The Hottentots are in the habit of abandoning aged persons of both sexes, and if a mother dies they bury with her her infant child, 1 and women often kill their infants if the father refuses to provide for them. 2 Kolbe* says that they rejoice at the birth of male twins, but if there are two girls, one of them is exposed, as is also the case with the girl if there are a boy and girl ; 3 and, according to him, they plead usage as a justifica- tion for infanticide, while they justify the killing of old people as being humane. 4 Kolbe's statements are substantially corro- borated by Thunberg. 5 Polygamy is allowed, but is said to be rare. The bridegroom goes to the bride's kraal, and the language used on the occasion of the marriage rather points to exogamy. Marriage, says Kolbe, is not allowed between cousins in the first or second degree, and any violation of the law is punished with death. 6 According to Thunberg, polyandry is also practised. 7 The Damaras are in the habit of stealing wives from the Namaquas, which, with the stealing of cattle, cause all their hostilities. 8 Klemm says, on the authority of Lichtenstein, that the ideas " maid," " virgin," and " wife," are not distinguishable. 9 The family bonds of relationship are not heeded. A member of a family separates from it and attaches himself to another circle according to caprice. Barrow mentions some Bushmen who had two wives each one past child-bearing, the other young and says that relation- 1 Walckenaer, vol. xvi. p. 103. Pinkerton, vol. xvi. p. 142. Moffat, p. 134. 2 Walckenaer, voL xviii. p. 465. 3 Id. vol. xv. p. 342. 4 Id. vol. xv. p. 314. 5 Pinkerton, vol. xvi. p. 142. 6 Walckenaer, vol. xv. p. 339. Pinkerton, vol. xvi. p. 141. 7 Pinkerton, vol. xvi. pp. 97, 141. 8 Walckenaer, vol. xviii. p. 464. 9 Klemm, vol. i. p. 336. 488 STUDIES IN ANCIENT HISTORY CHAP, xxix ship was no bar to marriage amongst them except between parents and children and brothers and sisters. 1 Theal says that the ceremony of making youth at a certain age into men is not yet laid aside, 2 a custom also noticed by Thunberg. 3 1 Walckenaer, vol. xvii. pp. 166, 167. 2 Theal, p. 22. 3 Pinkerton, vol. xvi. p. 141. APPENDIX THE WOKSHIP OF ANIMALS AND PLANTS 1 [THE matter of the following essay would have been all recast or redistributed had the author lived to complete his work ; indeed some of it has actually been so treated in the preceding pages. Thus those parts of the essay which deal with totemism among the Hindoos have been embodied in the chapter on fabricated genealogies, and are therefore here omitted. And the totemism of e.g. America has been rehandled and put into a new form in the chapters on America. But it appeared better to incur the charge of repetition than to alter the first part of the essay in any way. As to the ancients, Mr. M'Lennan was well aware that Bryant was a treacherous authority to rely on. He said him- self that nothing taken from him must be allowed to stand without being verified. And when his brother and Professor Robertson Smith entrusted me in 1885 with the task of supplying references to the ancient authorities, they instructed me to cut out all statements for which no such authority could be found, as well as anything else which might plainly be better removed. However, I have pruned this part of the text as sparingly as I could. A great many further notes on relics of totemism among the ancients were left by the author. But in the present state of opinion on the subject it is really unnecessary to add anything. A. P.] 1 Reprinted from the Fortnightly Review, 1869, 1870. 492 STUDIES IN ANCIENT HISTORY PART I. TOTEMS AND TOTEMISM Few traditions respecting the primitive condition of man- kind are more remarkable, and perhaps none are more ancient, than those that have been preserved by Sanchoniatho ; or rather, we should say, that are to be found in the fragments ascribed to that writer by Eusebius. They present us with an outline of the earlier stages of human progress in religious speculation, which is shown by the results of modern inquiry to be wonderfully correct. They tell us, for instance, that " the first men consecrated the plants shooting out of the earth, and judged them gods, and worshipped them upon whom they themselves lived, and all their posterity, and all before them, and to these they made their meat and drink offerings." They further tell us that the first men believed the heavenly bodies to be animals, only differently shaped and circumstanced from any on the earth. " There were certain animals which had no sense, out of which were begotten intelligent animals . . . and they were formed alike in the shape of an egg. Thus shone out Mot [the luminous vault of heaven ?], the sun, and the moon, and the less and the greater stars." Next they relate, in an account of the successive generations of men, that in the first generation the way was found out of taking food from trees ; that in the second, men, having suffered from droughts, began to worship the Sun the Lord of heaven ; that in the third, Light, Fire, Flame [conceived as persons] were begotten ; that in the fourth, giants appeared ; while in the fifth, " men were named from their mothers " because of the uncertainty of male parentage, this generation being dis- tinguished also by the introduction of " pillar " worship. It was not till the twelfth generation that the gods appeared that figure most in the old mythologies, such as Kronos, Dagon, Zeus, Belus, Apollo, and Typhon; and then the queen of them all was the IM-headed Astarte. The sum of the state- ments is, that men first worshipped plants ; next the heavenly bodies, supposed to be animals ; then " pillars " (the emblems of the Procreator) ; and, last of all, the anthropomorphic gods. Not the least remarkable statement is, that in primitive times APPENDIX 49* there was kinship through mothers only, owing to the uncer- tainty of fatherhood. 1 In the inquiry we are entering upon we shall have to contemplate, more or less closely, all the stages of evolution above specified. The subjects of the inquiry are totems and totem-gods, or, speaking generally, animal and vegetable gods ; and the order of the exposition is as follows : First, we shall explain with some detail what totems are, and what are their usual concomitants ; showing how far they have, or have recently had, a place among existing tribes of men ; and we shall throw what light we can on the intellectual condition of men in what we may call the totejn stage of development. Next we shall examine the evidence which goes to show that the ancient nations came, in prehistoric times, through the totem stage, having animals and plants, and the heavenly bodies conceived as animals, for gods before the anthropo- morphic gods appeared, and shall consider the explanations that have been offered of that evidence. The conclusion we shall reach is that the hypothesis that the ancient nations came through the totem stage, satisfies all the conditions of a sound hypothesis. 2 1 Sanchoniatho's Phoenician History, by the Eight Rev. R. Cumberland, pp. 2, 3, 23 et seq. London, 1720. Eusebius, Prcepar. Evangel., lib. i. cap. 10. 2 While the materials we have bearing on this subject are deemed worthy of being submitted for consideration, the investigation is yet far from being com- plete, and its completion will demand the co-operation of many. In the inquiry as here exhibited, it will be seen that several persons have given assistance. Did our hypothesis not seem sound, we should not propound it ; but, be it under- stood, it is submitted as an hypothesis only, in the hope that it may be tested by others better qualified for such investigations. The ancient mythologies have been so often crossed upon one another, interfused, and in appearance confounded with the intermixtures, intercommunications, and varying developments of the tribes of men who initiated them and modified them in successive generations, that it may appear a hopeless task to endeavour to throw new light upon them, still more hopeless to trace them to their beginnings. The only chance of dealing with them successfully, however, is to make them the subject of an hypothesis ; and though some may think the chance too small to justify the labour that this species of inquiry should be excluded from human endeavour we do not at all agree with them. Their opinion is opposed by the lessons taught by the history of scientific discovery. These show that the inquirer who has facts to go upon should never despair ; that in such a case as the present even a failure is a step of progress as demonstrating a line along which the truth does not lie one 494 STUDIES IN ANCIENT HISTOR Y TOTEMS. The first thing to be explained is the totem. The word has come into use from its being the name given by certain tribes of American Indians to the animal or plant which, from time immemorial, each of the tribes has had as its sacred or consecrated animal or plant. A proper under- standing, however, of what the totem is cannot be conveyed in a sentence, or reached otherwise than by studying the accounts we have of totems among different tribes of men; and, therefore, for behoof of those who are not familiar with these accounts, we must go somewhat into details. Unfor- tunately, totems have not yet been studied with much care. They have been regarded as being curious rather than important ; and, in consequence, some points relating to them are unexplained. As it is, we know that they prevail among two distinct groups of tribes the American Indians, already mentioned, and the aborigines of Australia. Many more instances of their prevalence, it may be believed, will yet be brought to light. In the meantime it is some compensation for the incompleteness of the accounts that we can thoroughly , trust them, as the totem has not till now got itself mixed up with speculations, and accordingly the observers have been unbiassed. 1. Totems or Kolongs in Australia. We have an account of these from the pen of Sir George Grey, who says the natives represent their family names as having been derived from some vegetable or animal common in the district they inhabited. Each family adopts as its sign, or kolong a word which is the equivalent of totem, and means, literally, a friend or protector the animal or vegetable after which it is named. The families here referred to are not families in our sense of the word, but stock-tribes, or tribes of descent, as appears from the following statement: The natives are divided into certain great families, all the members of which bear the same name as a family or second name. The principal V more key on the bunch, to be labelled as unsuited to the lock. A negative result may forward an investigation. Whether we have hit the truth or not, we trust we have at least been preparing the way for those who in the fulness of time will reach it. APPENDIX 495 branches of these families, so far as I have been able to ascertain, are the Ballaroke, Tdondarup, ISTgotak, Nagarnook, Nogonyuk, Mongalung, and Narrangur. But in different districts the members of these families give a local name to the one to which they belong, which is understood in that district to indicate some particular branch of the principal family. The most common local names are Didaroke, Gwerrinjoke, Maleoke, Waddaroke, Djekoke, Kotejumeno, Namyungo, Yungaree. These family names are common over a great portion of the continent ; for instance, on the western coast, in a tract of country extending between four and five hundred miles in latitude, members of all these families are found. . The family names are perpetuated and spread through the country by the operation of two remarkable laws : 1st, that children (boys as well as girls) always take the family name of their mother ; 2nd, that a man cannot marry a woman of his own family name. Sir George Grey elsewhere says that "the whole race is divided into tribes, more or less numerous according to cir- cumstances, and designated from the localities they inhabit, for though universally a wandering race with respect to places of habitation, their wanderings are circumscribed by certain well-defined limits." He further notices as " a most remarkable law," that "which obliges families connected by blood on the female side to join for the purposes of defence and avenging crimes." l From this statement it appears that we have in Australia certain great family or stock names, represented by persons in various local tribes ; that the marriage law prevents any local tribe coming to consist entirely of persons of one name or stock ; while the law of mutual defence and blood feud com- bines into what we may call gentes, within the local tribes, all who have the same totem and are of the same stock. This is clear from what follows immediately after the words last quoted, namely: "All their laws are principally made up of sets of obligations due from members of the same great family towards one another which obligations of family names are much stronger than those of blood." There are not only gentes within the local tribes, but the gentile bond is such as to constitute, in effect, a stock-tribe of all the gentes of the same family name, totem, or kobong, wherever they are situated. 1 Grey's Journals of Two Expeditions of Discovery in North- West and Western Australia) vol. ii. pp. 225 et seq., 230. 496 STUDIES IN ANCIENT HISTORY In the work just quoted, Sir George Grey refers to his Vocabulary of the, Dialects of South-Western Australia,, as giving under each family name its derivations, as far as he could collect them from the statements oY the natives. Unfortunately, he seems to have been able to collect the meaning in eight cases only, and we have been unable to enlarge the list. 1 Subjoined are the derivations in the eight cases : 1. Ballaroke. Ballar-wauk, Ballard, is given in the vocabulary as a very small species of opossum, with this note : " Some natives say that the Ballaroke family derive their name from having in former times subsisted principally on this little animal." Balla-ga-ra is also a species of opbssum. 2. Djin-le-nong-era, a species of duck. "The Ngotaks formerly belonged to this class of birds, before they were changed into men." 3. Karbunga, a species of water-fowl ; the mountain duck. " The No-go-nyuks are said to be these birds transformed into men." 4. Kij-jin-broon, a species of water-fowl. " The Didaroke family, a branch of the Ngotaks, are said to be these birds transformed into men." 5. Koo-la-ma, a species of water-fowl. "The Tdondarups are said by the natives to be these birds transformed into men." 6. Kul-jak, a species of swan. " The family of the Ballar- okes are said to owe their origin to the transformation of these birds into men." 7. Nag-harm, a species of small fish. " From subsisting in former times principally on this fish, the Nagarnook family are said to have obtained their name." 8. Nam-yun-go t an emu ; the local name for the Tdondarup family in the Vasse district. In this imperfect list we have eight families, or branches of families, derived from beasts, birds, or fishes ; and in five 1 The linguists of the United States Exploring Expedition seem not to have paid attention to this subject. APPENDIX 497 cases the statement that the tribesmen believe themselves to be of the stock of the bird or beast, and that their progenitors had been transformed into men. We have an Opossum tribe, an Emu tribe, a Swan tribe, a Duck tribe, a Fish tribe, and three water-fowl tribes ; and along with them, we have the general statement that all the tribes have kobongs or totems, animal or vegetable, after which they are named. The Opossums are bound together by what may be called a common faith and numerous mutual rights and obligations thence derived ; so are the Emus, Ducks, and Swans ; the stock names being thereby perpetuated, while the persons having them are diffused throughout the country by the law which makes it incest for an Opossum to marry an Opossum, a Duck a Duck, and so on. No one has yet taken the trouble of making the inquiry, but our persuasion is that this totemism, as it has been called, will be found to prevail, or have prevailed, through the whole of Oceania. It is mentioned in the Eeport of the United States Exploring Expedition 1 that the great Deity of the Tahitians, Taaroa, is named from the Taro plant ; and a legend is given which connects the Marquesan and Tahitian traditions, in explanation of the fact of as we infer the prevalence of vegetable names (presumably as tribal) both in Marquesas and Tahiti. The legend is, that the eponymous Oataia "named his children from the various plants which he brought with him from Vavau." The first king on the Tahitian list is Owatea, who is identified with Oataia of the Marquesans. His wife, in either case, is Papa " mother of the islands " and is the same with the wife of the great god Taaroa. 2 The 1 Vol. vi. p. 133. 2 This Papa appears in the New Zealand mythology as the mother of all beings. She is the earth ; her Imsband, Rangi, the heavens. The two clave together during 1000 divisions of time, each division a being called Po ; and their children, who "were ever thinking" what the difference might be between dafkness and light, after meditating their murder, resolved at last to rend them apart. In the family were the following gods : the father of forests, birds, insects, and all things that are in woods ; the father of winds and storms ; the father .of cultivated food ; the father of fish and reptiles ; the father of un- cultivated food ; and the father of fierce human beings. They all, in turn except the father of storms essay to rend their parents apart. Success at last 2 K 498 STUDIES IN ANCIENT HISTOR Y Koyal line is named from the Taro plant in this way : Owat and Papa had a deformed child whom they buried ; from it sprang the Taro plant, whose stalk is called haloa, which nam( they gave to their son and heir. This we must regard as sample of the legends which are formed on an advance froi totemism taking place, in explanation of its origin or relic Names taken from vegetables appear to prevail in the Sandwicl Islands. 2. Totems in America. Of these we have pretty full accounts. One is to be found in Dr. Gallatin's Synopsis of the Indian Tribes, contained in the Archceolagia Americana. He says : Independent of political or geographical divisions [i.e. of divisions of the native races into local tribes or nations], that into families or clans has been established from time immemorial. ... At present, or till very lately, every nation was divided into a number of clans varying in the several nations from three to eight or ten, the members of which respec- tively were dispersed indiscriminately throughout the whole nation. It has been fully ascertained that the inviolable regulations by which those clans were perpetuated amongst the southern nations, were, first that no man could marry in his own clan ; * secondly, that every child belongs to his or her mother's clan. Among the Choctaws, there are two great divisions, each of which is subdivided into four clans ; and no man can marry in any of the four clans belonging to his division. Amongst the Cherokees, the Creeks, and the Natches, the restriction does not extend beyond the clan to which the man belongs. There are sufficient proofs that the same division into clans, commonly called tribes, exists amongst almost all the other Indian nations [i.e. all the others as well as the southern Indians east of the Mississippi, of whom he is writing]. But it is not so clear that they are subject to the same regulations. According to Charlevoix, " most nations are divided attends the efforts of Tane-Mahuta, father of forests, who, with his head planted on his mother and feet against his father, thrusting, separated them. "Far beneath he pierces down the earth ; far above he thrusts up the sky." On the separation multitudes of human beings were discovered that had been begotten by Rangi and Papa, and lay concealed between their bodies. What follows introduces new gods, and explains the war of the elements. The whole of this mythology is scientific in this sense, that it is a series of hypotheses to explain phenomena. The part assigned to the forest god illustrates this. It may be believed the tree god was highly esteemed considering how much was due to him. Grey's Polynesian Mythology. 1 What is called the dan here is identical with the Australian family, as will presently appear. APPENDIX 499 into three families or tribes. One of them is considered the first, and has a kind of pre-eminence. Those tribes are mixed without being con- founded. Each tribe has the name of an animal. Among the Hurons, the first tribe is that of the Bear ; the two others, of the "Wolf and the Turtle. The Iroquois nation has the same divisions, only the Turtle family is divided into two, the Great and the Little." The accounts are not so explicit with respect to the Lenape tribes. Mr. Heckewelder, indeed, says that the Delawares were divided into three tribes, but one of them, the Wolf, or Minsi, had altogether separated from the others, and was a distinct nation or tribe [not ceasing, however, to be a clan in the sense now under consideration]. According to Mr. Johnston, the Shawnoes have four tribes : the Chillicothe, the Piqua, the Kiskapocoke, and the Mequachake. The first two, from having given names to distinct towns, would seem to be living in separate places ; but the fact that the Mequachake can alone perform the religious ceremonies of the nation gives it the character of a clan. Whether the totem or family name of the Chippeways descends in a regular manner has not been clearly explained. But Dr. James informs us that no man is allowed to change his totem, that it descends to all the children a man may have, and that the restraint on intermarriage which it imposes is scrupulously regarded. The Chippeways and kindred tribes are much more subdivided than the other Indians are into clans. Dr. James gives a catalogue of eighteen totems, and says many more might be enumerated. 1 The totems, and the restraints they impose, are found with the Iroquois as with the Delawares and Sioux tribes. The Omahaws (among the Sioux) are in two great tribes, the one divided into eight, the other into five bands. Each of these bands derives its name from some animal, part of an animal, or other substance, which is considered as the peculiar sacred object, or Medicine, as the Canadians call it, of the band. The most ancient is that of the red maize ; the most powerful, that of the Wase-ishta, (" Male-deer "). The Puncas are divided into similar bands. 2 We have made these long citations because they show us the totems or kobongs, as in Australia, descending as a general rule under the same system of kinship (through mothers only), and attended by the same law of intermarriage, namely, exo- gamy, leading to the interfusion of the stock tribes throughout the country ; and the constitution into gentes in the local tribes of all persons having the same totem. The laws of blood-feud, of mutual rights and obligations between those of 1 Archceologia Americana, vol. ii. p. 109. 2 Id. vol. ii. p. 110. 5oo STUDIES IN ANCIENT HISTOR Y the same stock, constitute stock -tribes of all having the same totem. 1 And we can see in the account cited how, at a stage considerably in advance of the Australian, the solidarity of the gentes in the local tribes has under these laws become so great as to enable the gentes, in some cases, to withdraw from the local tribes, in whic.li they were developed, and stand, like the Wolves of the Delawares, by themselves, in local tribes of one stock. On a change of kinship, which would permit the totem to descend from the father instead of the mother as it is said to do among the Chippeways the gentes would, even supposing exogamy to continue in force, become permanent homogeneous groups after their segregation. Let us obtain a list of the American totems. Nearly all, if not all, of the Indian nations upon this continent (says Mr. Lewis H. Morgan, of Rochester, State of New York) were anciently subdivided into Tribes or Families. These tribes, with a few exceptions, were named after animals. Many of them are now thus subdivided [so they have been advancing]. It is so with the Iroquois, Delawares, lowas, Creeks, Mohaves, Wyandottes, Winnebagoes, Otoes, Kaws, Shawnees, Choctaws, Otawas, Ojibewas, Potowottomies, etc. [We can supply from the Archceologia Americana, the Cherokees, Natches, and Sioux.] The following tribes [or families'] are known to exist, or to have existed, in the several Indian nations the number ranging from three to eighteen in each. The Wolf, Bear, Beaver, Turtle, Deer, Snipe, Heron, Hawk, Crane, Duck, Loon, Turkey, Musk-rat, Sable, Pike, Cat-fish, Sturgeon, Carp, Buffalo, Elk, Reindeer, Eagle, Hare, Rabbit, and Snake ; also the Reed-grass, Sand, Water, Rock, and Tobacco-plant. 2 To this list we may add from the Archceologia and other sources, the Tortoise, the Turtle in two divisions, the Great Turtle and the Little Turtle the Eed Maize, the Male Deer, the Wind, the Tiger, the Bird, the Eoot, the Birch-rind, the Thick-wood, the Sheep, the Brush-wood, the Moose-deer, the Cat, the Trout, the Leaves, the Crow, the Sun, the Eising Sun, and the Grey Snow, the Sun and the Snow being regarded as beings. There are thus forty-eight totems enumerated for American tribes, not counting the Male Deer or the Little Turtle, and we know there were others. 1 Archceologia Americana, vol. ii. p. 111. 2 Circular letter issued by Mr. Morgan, quoted in the Cambrian Journal for 1860, p. 149. APPENDIX 501 The following quotation from the Archceologia illustrates the effect of these names on narratives respecting the tribes and the actions of the tribes-men or tribes-women : Some superiority is everywhere ascribed to one of the clans : to the Unamis ("the Tortoise") among the Dela wares ; to the Wase-ishta (" Male-deer ") among the Omahaws ; to the Bear tribe among the Hurons and five nations. Charlevoix says that when the Mohawks put to death Father Jogues, it was the work of the Bear [clan] alone, and notwithstand- ing all the efforts of the Wolf and the Turtle to save him. 1 Of course the indefinite article would be employed, instead of the definite, in speaking of individuals. The Bear, is the tribe or clan; a Bear, a tribesman. In speaking of their marriages, it would be said, for instance, that " a Bear married y a Wolf," and " a Turtle a Beaver." In cases of nursing, a man's foster-mother might be a She-Wolf, a She-Bear, or a Tigress. 3. Relations between Men and Totems. Let us now see how those who have totems regard them ; and what, generally speaking, are their rejigious views. Grey says that " there is a mysterious connection between an Australian and his kobong, be it animal or vegetable." It is his " friend " or " protector," and is thus much like the " genius " of the early Italian. If it is an animal, he will not kill one of the species to which it belongs, should he find it asleep ; he always kills it reluctantly, and never without affording it a chance to escape. " The family belief," says Sir George, " is that some one individual of the species is their dearest friend, to kill whom would be a great crime. So a native who has a vegetable kobong may not gather it under certain circumstances, and at a particular period of the year." 2 We previously saw that the belief, in certain cases at least, is that the family were of the species of the totem before they were turned into men. It may be asked, What are their views of the power by which these transformations were effected ? We cannot answer this question ; but one thing seems to be clear, that their speculations have not carried them as yet beyond the 1 Archceologia Americana, vol. ii. p. 113. 2 Sir George Grey's Journals, vol. ii. p. 229. 502 STUDIES IN ANCIENT HIS TOR Y contemplation of the material terrestrial world they inhabit, and that in that world everything is to them at once material and spiritual, the animate and the inanimate being almost undistinguished. Like many races in Africa, they do not believe in death from natural causes, and think they would live for ever were it not for murderers and sorcerers. The latter they call Boyl-yas. A Boyl-ya gets power over a man if he obtains possession of anything that is his. 1 A Boyl-ya may cause death in many ways ; he may cause a man to be killed " by accident," or he may render himself invisible and come nightly and "feast" on his victim's flesh. He can transport himself through the air at pleasure ; and when he makes himself invisible, he can be seen only by other Boyl-yas. He enters his victim like a piece of quartz, and as such may be drawn out of him by the enchantments of friendly Boyl-yas. Pieces of quartz that have been so drawn out are preserved as the greatest curiosities. As some one is always the cause of death, the law is that when any one dies, some one else must be killed the Boyl-ya, or the murderer, or some relative of the one or the other. Of course the Boyl-yas are objects of great dread. They consume the flesh of their victims slowly, as fire would ; 2 they can hear from afar ; they come " moving along in the sky " ; and they can only be counteracted by other Boyl-yas. Besides the Boyl-yas, there is another object of terror the Wau-gul. It is an aquatic monster, residing in fresh water, and has supernatural powers. It also can " consume " the natives like the Boyl-yas ; but it confines its attacks mostly to women, who pine away almost imperceptibly and die. Nightmare is caused by an evil spirit that may be driven away by muttering imprecations and twirling a burning brand. Shining stones or pieces of crystal, called " Teyl," they respect almost to veneration. None but Boyl-yas venture to touch them. They believe in ghosts ; and on one occasion Sir George Grey was taken by an old lady to be the ghost of her son, who had lately died ! Such is the creed of this \l primitive race. They have no God in the proper sense of the 1 Sir George Grey's Journals, vol. ii. p. 323. 2 Id. vol. ii. p. 339. APPENDIX 503 word ; and the only benign beings they know are their totems. The Boyl-yas of course practise imposture, 1 but are probably self-deluded as well to a great extent. Speculation has not reached as yet among them to the heavens. Their super- naturals are all naturals, for even the Wau-gul no doubt a convenient fiction of the Boyl-yas for protection under the law of retaliation, and perhaps also in explanation to themselves of deaths they know they had nothing to do with is a living creature, the tenant of a stream or lake. Even their ghosts may return to them, if precautions are not taken to secure them in their burying -places their "houses," as they are called, and in which, even after death, they are not incapable of action. 2 It will have been seen that the totems are, as we may say, religiously regarded by the Australians, and that the Boyl-yas resemble the genii of the Arabian Nights, excepting that while they are genii they are also men. The Wau-gul might well grow into the water-kelpi, water-horse, or bull. It would be curious to know whether it is a fish or an aquatic kangaroo or opossum ! The American Indians, though they occupy a distinctly higher platform, have still much in common with the aborigines of Australia. Dr. Gay, who resided for several weeks among the Omahaws, states that among them the totem of each band "is considered as the peculiar sacred object (Medicine, the Canadians call it) of the band," 3 and all we know supports the view that in every case the totem is religiously regarded. 1 Sir George Grey's Journals, vol. ii. p. 218 : "The whole tendency of their superstitions is to deprive certain classes of benefits which are enjoyed by others." 2 Id. vol. ii. p. 336: "After burial, the dead man can insert a mysterious bone into each of three doctors, who sleep on the grave for the purpose. By means of this bone, the doctors can kill any one they wish by causing it to enter into his body." 3 Archceologia Americana, vol. ii. p. 112. The personification of inanimate objects, the animism, as Mr. Tylor calls it, of the Indians is nearly as complete as in Australia. See Archceologia Americana, vol. ii. pp. 25, 166, 169. No distinction between the animate and inanimate is made in the languages of the Esquimaux, the Choctaws, the Muscogee, and the Caddo. Only the Iroquois, Cherokee, and the Algonkin Lenape have it, so far as is known, and with them it is partial. 504 STUDIES IN ANCIENT HISTORY One author, Mr. Long, in a work published in 1791, describing the manners and customs of the North American Indians, holds totemism to be a religious superstition, and says the Indian believes that his totem, " or favourite spirit," watches over him. " The totem," he says, " they conceive, assumes the shape of some beast or other, and therefore they never kill, hunt, or eat the animal whose form they think the totem bears." l In illustration of the truth of this, he relates what once befell an Indian whose totem was the bear. The man dreamed he should find a herd of elks, moose, etc., at a certain place, if he went thither. Having a superstitious reverence for his dream, he went, unaccompanied, as he could get no one to go with him, saw the herd, fired, and shot a bear ! " Shocked at the transaction," says Mr. Long, " and dreading the displeasure of the Master of Life, whom he conceived he had offended, he fell down and lay senseless for some time." On recovering, and finding that nothing had befallen him, he hastened towards his home, when (according to his own report) he was met on the way by a large bear, who (he narrated) asked him what had induced him to kill his totem. On explaining the circumstances and his misfortune, he was for- given, but was dismissed with a caution to be communicated to the Indians, "that their totems might be safe, and the Master of Life not angry with them." "As he entered my house," says Mr. Long, who writes as if he saw the man immediately after his accident, " he looked at me very earnestly, and pronounced these words in his own language, ' Beaver, my faith is lost ; my totem is angry ; I shall never be able to hunt any more.'" Should one be surprised to find that admonitory bear of the man's imagination worshipped as a god further on in the history of Bear tribes advancing undisturbed by external influences, correlated with the Master of Life in the Olympus, or even preferred to, or identified with him ? The Master of Life of this story, we infer from other passages in the work quoted, is Kitcku Manitoo, a high rock in Lake Superior, which is worshipped as a god by the Chippeway Indians, and also by the Mathangweessawauks, whoever they 1 Long's Voyages, p. 86. APPENDIX 505 may be. 1 Is Kitchu Manitoo, it may be asked, the commence- ment of pillar- worship, of Siva-ism ? He is the Master of Life, and, in some tribes, the Great Spirit. The accounts of him are most vague, and show a faith shading up from the " great black man in the woods " of the Fuegians to the Master of Life, with a high rock for his representation, and thence to the Great Spirit who had no representation whose temple the Incas are said to have found standing and deserted on their arrival at Cuzco. In two cases only have we certain information of the ideas of God which the Indians entertained. (1) In Gookin's History of the Christian Indians is preserved a contract in the form of question and answer between them and our Government. It opens as follows : " Ques. 1. To worship the only true God, who made heaven and earth. Ans. We do desire to reverence the God of the English, because we see he doth better to the English than other gods do to others." (2) Of the Pawnees, whose "Great Spirit" is Wahcondu, Dr. Gallatin writes, " Like all other Indians, they put more faith in their dreams, omens, and jugglers, in the power of imaginary deities of their own creation, and of those consecrated relics (the totems) to which tha Canadians have given the singular appellation of Medicine." '" The American Indians, like the aborigines of Australia, regarded themselves, we have every reason to believe, as being of the breed of the totem. We know this was the view of the Sun tribes which we shall notice presently and of several Snake tribes. That the Caribs were of the stock of the Serpent we learn from Mr. Brett. 3 And on this point the regular authorities being silent we are entitled, we think, to found on evidence furnished by Mr. Fenimore Cooper. His view appears in The Last of the Mohicans. Magua, a Fox, 1 Long, pp. 68 et seq., 139. In Long's opinion totemism resembles the idea of Destiny, and he says it is not confined to savages, as " many instances might be adduced from history to prove." Very probably. The one instance he cites is that of a Jew banker, of the court of Louis XIV. of France, "who had a black hen, to which he thought his destiny attached. " They died to- gether. 2 Archceologia Americana, vol. ii. p. 130. 3 Brett's Indian Tribes of Guiana, pp. 390-393. 506 STUDIES IN ANCIENT HISTOR Y with a party of warriors, comprising a Beaver, happening to pass a colony of real beavers, the Beaver refused to pass with- out addressing his kinsfolk. "There would have been a species of profanity in the omission," says Mr. Cooper, "had this man passed so powerful a community of his fancied kins- men without bestowing some evidence of regard. Accordingly, he paused and spoke in words as kind and friendly as if he were addressing more intelligent beings. He called the / animals his cousins" and so on, concluding his address by begging them to bestow on his tribe " a portion of the wisdom for which they were so renowned." Uncas; again, Mr. Cooper represents as claiming to be of the stock of the Tortoise, " that great-grandfather of all nations " ; and, indeed, all his Indians appear to regard themselves, and one another, as inheritors of mental and physical qualities from their respective totems. One other and last relation between the totem and its owners, both in America and Australia, remains to be noticed. Grey tells us that the Australians use the totem as the family crest or ensign, and expresses the opinion that our heraldic bearings are traces of the totem stage lingering in civilised nations. It is w,ell known that the totem was also used as an ensign by the American Indians, who tattooed the figure of it on their bodies, and, not content with this, painted and dressed themselves so as to resemble it. Every reader of stories about these Indians must be familiar with the fact. Magua, for example, in the beaver scene, from the account of which we have just quoted, wore " his ancient garb, bearing the outline of a fox on the dressed skin which formed his / robe " ; while the Beaver chief " carried the beaver as his peculiar symbol." The accounts we have of the old Mexicans in war show that they had similar badges, every chief having his sign an animal, or animal's head, or a plant; and every company having a similar symbol on its standard. 4. Traditions of Totems in Central Asia. The totem stage appears to have been passed through by numerous tribes of Central Asia. MM. Valikhanof inform us that a heritage of the nomadic races in that part of the world is a profound regard for, and an abundance of traditions respecting old times, APPENDIX 507 preserved by their elders in legends and ballads, and that these traditions refer the origin of their tribes to animals as progenitors. The story of the origin of the Dikokamenni Kirghiz (they say 1 ) from a red greyhound and a certain queen with her forty handmaidens is of ancient date. A characteristic feature in Central Asiatic traditions is the derivation of their origin from some animal. According to the v testimony of Chinese history the Goa-qui (Kaotsche), otherwise known as the Tele or Chili people, sprang from a wolf and a beautiful Hun princess . . . who married the wolf. The Tugus (called the Dulgasses by Pere Hyacinthe) professed to derive their origin from a she-wolf; and the Tufans (Thibetians) from a dog. The Chinese assert that Balachi, hereditary chief of the Mongol Khans, was the son of a blue wolf and a white hind. 2 [The authority cited for this is Me'moires relatifs a I'Asie, by Klaproth, p. 204.] ... It is evident from these instances that this kind of tradition in Central Asia and America is the most ancient, and t even seems to be regarded as a descent to be proud of. The outspoken yet exalted tone of the Kirghiz legends, considered indecent by the present generation of Kirghiz, is a strong proof that they have descended in their original form. The tradition of the origin of the ninety-nine Kipebuk branches has been preserved among the Uzbeks and Kaisaks in such an indelicate shape that it is doubtful whether it will ever be possible to present it to the general reader. It is accordingly not given ; but surely the essence might have been, though not the shape. We learn from the same authorities that the genealogical tables of the Kaisaks, Uzbeks, and Nogais show that " they are a medley of different Turkish and Mongol tribes." The names of several tribes are given, but none have been examined etymologically to ascertain whether they comprise the names of animals or plants. The interfusion, or "medley," of the tribes (we are without a statement of the origin of it, but nearly all these tribes are exogamous, that is, prohibit marriage within the clan), and the general statement (though it is feebly supported by details) that they draw their origin back to animals, make it probable 1 The Russians in Central Asia. London, 1865. Translated by the Messrs. Mitchell. 2 In the Archceologia Americana, vol. ii. p. 112, it is noticed that among the Creeks the villages are divided into white and red, "distinguished from v each other by poles of those respective colours." Query Would a Deer in a white village be a White Deer, and a Wolf in a red village be a Bed Wolf? 5o8 STUDIES IN ANCIENT HISTOR Y that we have in the Kirghiz, Uzbeks, and Nogais a series of tribes that anciently passed through the totem stage. This view is confirmed by what was recently stated at a meeting of the Geographical Society by Captain T. G. Montgomerie ; namely, that round Cashmere, and among the aboriginal hill tribes on the Himalayan slopes, tribes of men are usually (or frequently, we have not the report before us) named, or we think he said " m'c&named," from animals now. If so, we may believe inquiry will bring to light a series of tribes in that quarter still existing in the totem stage. The statement was made in support of the hypothesis that an Ant tribe had existed to the north of Cashmere, put forward in explanation of what Herodotus relates that the gold-fields there were worked by ants. 5. The Sun as a Totem : Hints of a Totem Olympus. We saw that in the local tribes or nations in America, some one of the tribes of descent had a superiority ascribed to it that the Bear, for instance, was the leading tribe among the Hurons. This superiority infers subordination, of course ; in other words, a political system. It is stated in the Archwologia, that " it is among the batches alone that we find, connected together, a highly privileged class, a despotic government, and something like a regular form of religious worship." The Natches occupied three villages near the town that has pre- served their name, and were in four clans. What their totems were is not stated ; but " the privileged class " and the sovereign had for their totem the sun. This seems a legitimate inference from their being called Suns, and claiming to be descended from the sun the Sun tribe being so far like any other. " The hereditary dignity of Chief, or Great Sun," we are told, " descended as usual by the female line, and he, as well as all the other members of his clan, whether male or female, could only marry persons of an inferior [i.e. another] clan." * That is, the clan or tribe was in the same case with any other, except that it was dominant as the Bear was among the Hurons. 2 A Sun could not marry a Sun any 1 I.e. vol. ii. p. 113. 2 Are the accounts incomplete ? and is the dominant tribe among the Hurons APPENDIX 509 more than a Beaver could marry a Beaver ; and the Sun name was taken from the mother. If the sun could become a totem, why not the moon ? That they were both "beings we can see in numerous cases ; we have distinct proof of it among the Indians in the case of the Chippeways. 1 If they were totems, they will explain for us the solar and lunar races of the Aryans. "We have them in Peru as married persons, and also as brother and sister. The Incas were Suns, as their name and all the traditions imply a Sun tribe, nothing less or more ; their first parents children of the sun, sent to the earth to found society, as the reader may see in Prescott's opening chapters. Acosta tells us the brother of the Inca succeeded in preference to his sons, 2 and if so, this points to kinship among the Peruvian Sun tribe, having been at one time through mothers only a note of the totem stage. The pride of power led the tribe to give up exogamy and become a caste; but then, to keep the stock pure, the Inca always married a sister, and when a son succeeded, it was as heir of the Goya, the lawful sister-queen, showing the lingering preference for the mother's side. We infer the presence of Sun tribes among the Hurons, the Baya- goulas, now extinct, the Cherokees, the Choctaws, and the Caddoes of Eed River, all of whom there is reason to believe more or less formally worshipped the sun. The batches had sun temples and perpetual fires. 3 The Sun tribes may have been very powerful, but it is only what we should expect, also the Sun tribe ? The chief of the Hurons, Charlevoix states, is believed to have issued from the sun, and the dignity of chieftainship is hereditary through females only. It is a possible explanation that the chiefs of the Bear tribe may have invented for themselves a solar origin, in which case the chief would be a Bear, and yet a Sun. Peru presents us with an instance of a Sun that is yet a Serpent, for which a similar explanation would suffice namely, that the Snake tribe was dominant, and that its chief families assumed the Sun as their totem. 1 Archceologia Americana, vol. i. p. 352. The sun and moon were occasionally given to fighting, it appears. 2 Lib. vi. cap. xii., cited by Prescott. 3 ArclvKologia Americana, vol. ii. pp. 113, 114. Was Helios, who had herds of oxen on the island of Trinacria, chief of a Sun tribe there ? The Heliades are suggestive of a Sun stock. Max Miiller complains of Mr. Grote's disposition " to insist on the purely literal meaning of the whole of Greek mythology." "We shall see by and by that Mr. Grote's disposition is the right one. 5 10 STUDIES IN ANCIENT HISTOR Y among a race simple enough to believe anything, that a peculiar sanctity, and corresponding privileges, would readily be conceded to those believed to be descended from the great Lord of Day ; and that the supremacy in many groups should on this account be the more readily obtained by the solar stock. It is also apparent that this totem might well com- mand a general veneration the worship of all the tribes in the group ; but it is equally manifest that the Sun would not, any more than the Master of Life, where it took the first place in the State religion, interfere with the allegiance due from the stock tribes composing the nation to their respective totems. The Incas, as Mr. Prescott points out, had the good policy to collect all the tribal gods into their temples in and round Cuzco, in which the two leading gods were the Master of Life and the Sun. In the temples, Mr. Prescott tells us, " there were animals also to be found," but he does not specify them, stating only that " the llama with its golden fleece was the most conspicuous." Were these animals the totems, or their emblems, of our friends the Bears and Beavers ? l 6. Totem Gods a Totem Olympus. Among the Fijians we find a state of affairs such as may have preceded the con- solidation of the monarchy and the Olympus of the Incas. They are proud of their pedigrees, and Toki, one of their chiefs, claims to be the descendant of a Turtle. Others have fishes for their progenitors. Their greatest god, the Creator, who is omniscient, omnipotent, and so on, in the opinion of 1 The mythologies of Peru and Mexico have yet to be explored, and may be expected to prove a field worthy to be worked. The few facts we have yield a strong suggestion that the Toltecs, Chimenecs, Aztecs, and Tezucans were groups, compounded, like the Natehes, of tribes with totems the Sun dominant, in Peru at least. The legend of the founding of Tenochtitlan gives a prickly-pear, an eagle, a serpent, and the sun. The Mexicans had the eagle on their standard, and the serpent at least among their gods. The war god, Huitzilopochtli, means literally, " a humming-bird " and "left." He was figured with the feathers of the humming-bird on his left foot. If the humming-bird was a totem, this is the only case, excepting one serpent god in the same Olympus, of a totem be- coming a God of Terror that we are acquainted with. The years in the fifty-two years Mexican Cycle were named from plants and animals a list of them is not accessible. Their law of succession was polyandrous, from brother to brother, and to sister's sons, failing brothers. This demonstrates for them the stage of female kinship. We know nothing of the law of intermarriage. APPENDIX 5I i his special votaries, is Nuengei, " whose shrine is the Serpent." Some of their gods are " enshrined " in birds, fishes, or plants ; some, in the same way, in men. Their second god in import- ance is Tui Lakemla, who claims the Hawk as his shrine ; but another god disputes his right, and claims the Hawk for himself. The Shark is a great god; also the Crab. "One ^ god," says Mr. Williams, " is supposed to inhabit the eel, and another the common fowl, and so on, until nearly every animal becomes the shrine of some deity. He who worships the Eel-god must never eat of that fish, and thus of the rest ; so that some are tabu from eating human flesh because the shrine of their god is in a man. . . . The Land- Crab is the shrine of Eoko Suka, formerly worshipped in Tiliva, where land-crabs are rarely seen." When a land-crab favours them with a call, they make formal presents to him, "to prevent the deity leaving with the impression that he was neglected, and visiting his remiss worshippers with drought, dearth, or death." These gods are tribal, and no one can doubt but they are totems who have made such progress as we above suggested the Bear might make, and are become the objects of a more or less regular worship the Serpent tribe dominant, and the Hawk tribe in the second place. The Men gods are a new element ^ in the Olympus ; but they appear as " shrines " merely like the other animals, and were no doubt arrived at by an exten- sion to man of conclusions speculatively reached as to the nature of totem gods in general. The Fijians have filled the world with spirits and demons. They are incessantly plagued by ghosts, witches, or wizards. Vegetables and stones, nay, even tools and weapons, pots and canoes, have souls that are immortal, and that, like the souls of men, pass on at last to Mbulu, the abode of departed spirits. They worship pillars and rocks ; but, so far as we know, they do not worship the sun unless their men gods are of the solar stock. 1 7. The Mental Condition of Men in the Totem Stage. The state of mind of men in the totem stage is familiar enough, from the accounts we have of the lower races of men. The absence of scientific knowledge nowise implies an absence of specu- 1 Fiji and the Fijians, by Thomas Williams, vol. i. pp. 114, 123, 215 ff. 5 1 2 STUDIES IN ANCIENT HIS TOR Y lation ; it rather necessitates the presence of a great amount of it. Some explanation of the phenomena of life a man must feign for himself; and to judge from the universality of it, the simplest hypothesis, and the first to occur to men, seems to have been that natural phenomena are ascribable to the presence in animals, plants, and things, and in the forces of nature, of such spirits prompting to action as men are conscious they themselves possess. So far as we know, this has been at some time or other the faith of all the races of men; and again, so far as we know, it is a faith that has nowhere been given up as unsatisfactory otherwise than gradually on its being perceived, from case to case, that the behaviour of the forces of nature and of the bodies they act upon is not way- ward or wilful, but conformable to law; and until the law has been ascertained. This animation hypothesis, held as a faith, is at the root of all the mythologies. It has been called Fetichism ; which, according to the common accounts of it, ascribes a life and personality resembling our own, not only to animals and plants, but to rocks, mountains, 1 streams, winds, the heavenly bodies, the earth itself, and even the heavens. Fetichism thus resembles totemism ; which, indeed, is Fetichism plus certain peculiarities. These peculiarities are, (1) the appropriation of a special Fetich to the tribe, (2) its hereditary transmission through mothers, and (3) its connection with the jus connubii. Our own belief is that the accompaniments of Fetichism have not been well observed, and that it will yet be found that in many cases the Fetich is the totem. Be that as it may, we may safely affirm that as Fetichism dies slowly, withdrawing its spirits from one sphere after another on their being brought within the domain of science, so it grew slowly through various stages of development, bringing the realms of nature one by one within the scope of the hypothesis which is its foundation. Our information is incomplete ; but from all we know, the aborigines of Australia are, as theorists, far in advance of the Bushmen, Veddahs, Andamans, and Fuegians, while it 1 Himavat (the Himalayas) was a great Hindu god. He had goddess daughters; one, Ganga (the Ganges), another, Urna, "the most excellent of lesses." See Dowson, Diet. Hind. Myth., s.v. "Himavat." APPENDIX 513 appears they themselves have many steps to . take before reaching the fulness of the animism of some American Indians, i They have not yet, for example, so far as we know, vivified the heavenly bodies. The Indians, again, have not yet advanced so far as the New Zealanders, who assign spirits to groves and forests, 1 as did the Greeks and Eomans, while none of the peoples last mentioned reached that perfection of Fetichism allied to an ontology which is Pantheism. 2 The justification of the statement that there is no race of men that has not come through this primitive stage of specu- lative belief, will [not] be found in this exposition in its entirety. 3 We may here say that such a stage is demonstrated for the Hindoos and Egyptians by their doctrine of trans- migration. It is of the essence of that doctrine that everything has a soul or spirit, and that the spirits are mostly human in \ the sense of having once been in human bodies. All the spirits are of course ultimately divine detached portions of the Deity. We find in the Code of Manu that " vegetables, and mineral substances, worms, insects, reptiles some very minute, some rather larger fishes, snakes, tortoises, cattle, shakals, are the lowest forms to which the dark quality leads [the soul of a man]." 4 A man may after death, according to the shade of the dark quality, become an elephant, horse, lion, tiger, boar, or a man of the servile class ; while, in virtue of the good quality, he may rise to the rank of the genii, to be a regent of the stars, or even a god. This implies, of course, the existence of spirits resembling our own of various ranks, from those that dwell in minerals and vegetables up tcwjtjjat of Brahma. We have a similar implication from the Egyptian 1 See " The Adventures of Rate," and " The Children of Heaven and Earth," in Grey's Polynesian Mythology. 2 A striking illustration of the graduality of the evolution of Fetichism will be found in Fiji and the Fijians, p. 241. The Fijians are far in advance of the Tongans. 3 Two papers having a bearing on this matter, written by Mr. E. B. Tylor, i/ the one on "The Early Condition of Man," and the other on "Traces of Savage Thought in Modern Civilisation," both read before the Royal Institution, London, are well worthy of being consulted. 4 Code, chap. xii. w. 42, 43 ; and see chap. i. vv. 49, 50. 2 L 514 STUDIES IN ANCIENT HISTORY doctrine. Let us consider how such a doctrine could have arisen. 1 The doctrine connects itself at once with the Cosmogony and with ethics. Manu's account of the genesis, from the first divine idea to the seed and the golden egg and the waters ; to the Vedas milked out from fire, air, and the sun ; and to the final evolution of all Beings, animals, and vege- tables, will be admitted to be as unscientific, or foolish, as anything the Australians could devise, supposing them to have imagination enough to shape so grand, a theory ; and it is not a whit more ingenious than the Australian view of life, taking success in getting at the truth as the test of ingenuity. The truth, it may be said, is beyond the reach of speculation. No doubt ; this fantastical doctrine, however, may safely be assumed not to contain it. " A transmigratory soul " is not an hypothesis like phlogiston : the latter explained some facts ; the former, none. How then did it arise ? It resulted from ethical considerations, and the theory of the Cosmos. But whence came the latter ? Its source, we think, is in- dubitable. It was a speculation to explain the facts, real and imaginary, of existence. That is, in the order of events, Fetichism, which assigns " souls " to all things, came first, and afterwards the cosmical theory, which explained, inter alia, " the souls of all things," the ethical doctrine regulating their transferences merely. In other words, had the "souls" not been pre-existing we should not have had the theory an unquestionable product of human effort to explain facts nor anything resembling it. This, we submit, is the common-sense view. The doctrine supervened on a system of ideas com- prising all the elements with which it had to deal. The 1 The systems of transmigration have been various. In the Brahmanic the purified soul returns to Brahma ; in the Buddhistic it attains Nirvana. The Egyptian resembled the Brahmanic, as did the Grecian, which was neither indigenous to Greece nor a popular faith. The Jews may have had their system from the Greek philosophers. It is taught in the Kabbala, and resembles the Brahmanic. The soul of Adam reappeared in David, and was to reappear in the Messiah. Some early Christians held the doctrine, but it was never the creed of the Church. It was the creed of the Manichseans. Origen believed it ; so, lately, did Lessing. It was indigenous in Germany and in ancient Mexico. APPENDIX 515 windows in heaven, and the firmament separating the waters above from the waters below them, do not more clearly demonstrate the old theory of rain, than this doctrine demon- strates pre-existing Fetichism. That the doctrine of transmigration was invented at a pretty late date in the progress of the Hindu races we may be certain. There is but one sentence in the Rig- Veda (Hymn i. 164) which has even been supposed to imply transmigration, and it does not do so, we are assured, when the words are taken literally in their usual sense. Yet the belief in the soul's life after death may be traced in some of the hymns of the Veda. This belief, however, assumes many forms, and the present writer has no certain information as to its Vedic form. Of the forms it assumes many are highly curious. The Australian and Fijian we saw. Among the Tahitians human souls were supposed to be the food of their god, and they offered to him human sacrifices that he might be fed. The Khonds have a limited quantity of soul as tribal property, and they explain their female infanticide by saying that the fewer their women are the more soul there will be for the men. The customs of some tribes in Madagascar show that they think that one man may have several souls ; and not a few tribes, holding that the souls of the dead return in their new-born babies, bury in the houses or near the doors to facilitate the return. It is familiar that men everywhere in ancient times be- lieved spirits to inhabit trees and groves, and to move in the winds and stars, and that they personified almost every phase of nature. We have now seen that such beliefs cannot be regarded as having been deduced from the grander doctrines of the ancient religions ; but that the latter must be regarded as having been constructed upon such beliefs as their founda- tions. Demons and genii, and the spirits of plants and minerals, were older than Brahma ; let us hope they will not survive him. They are everywhere lively still, even in the most advanced nations ; and we have not to go very far back in time to find them playing a most important part in our medical theories. Demons a species of disembodied 5 1 6 STUDIES IN ANCIENT HIS TOR Y Boyl-yas were connected with diseases by the Jews and early Christians, and it is familiar how on one occasion when driven out of a man they entered into a herd of swine. The genii of the early Italians so like the totem are familiar, and even more so are the genii of the Arabian Nights. The Mahometans, if they are true to their prophet, must still believe in them. In that very curious book Mishkatu 'l- MasaUh, a record of the sayings and doings of the prophet, bearing to be made by those who knew him best his wives and disciples we find the following, which is pertinent to our subject : Ibn-Omer said, " I heard his highness say, * Kill snakes, and kill the snake which has two black lines upon its back, and kill the snake called abter, on account of its small tail ; for verily these two kinds of snake blind the eyes as soon as they are looked at ; and if a pregnant woman should see them, she would miscarry from fright.' " Ibn-Omer says, " Just as I was about killing a snake, Abu-Lababah-Ansari called out to me not to kill it. Then I said, * His highness ordered me to kill them ; why do you forbid ? ' He said, ' His highness, after giving the order for killing them, said, You must not kill the snakes that live in the houses, because they are not snakes, but a kind of genii.' " Abu- Sayib said, " We went to Abu-Said-Khud'hri ; and whilst we were sit- ting, we heard a shaking under his bedstead ; and we looked and saw a snake. Then I got up to kill it, and Abu-Said was saying his prayers, and he made a sign to me to sit down, and I did so. And when he had finished his prayers, he made a sign towards a room in his house, and said, ' There was a youth in my family lived there who had newly married.' Then Abu-Said said, 'We came out of Medinah along with the Prophet, to a trench which was digging for fighting, and this youth would ask the Prophet's permission to return to his house every day at noon, which was granted. Then one day the youth asked his highness's leave, who said, Put on your armour, because I am alarmed about you, from the evil designs of the tribe of Beni-Kuraidhah. Then the youth took his arms, and returned towards his house ; and when he arrived, he saw his wife standing between two doors ; and the youth was about piercing her with a spear, being seized with jealousy at seeing her stand- ing out of her room ; and she said, Withhold your spear, and come into the room that you may see what has brought me out. Then the youth went into the room, and beheld a large snake coiled up sleeping upon his bed, and he struck his spear into the snake ; then the snake attacked the youth, and bit him, and it was not known which of them died first, the snake or the youth. Then I went to the Prophet and mentioned the occurrence, and said, Supplicate God to give life to the youth. Then his highness said, Ask God to forgive your friend ; wherefore do you wish APPENDIX 517 a prayer to be made for his life ? After that he said, In these houses are the genii, some of them believers, and some infidels ; therefore when you see anything of those inhabitants turn them out, but 'do not hurry in killing them, but say, Do not incommode me ; if you do, I shall kill you. Then if he goes away, so much the better ; but if not, kill it, because it is an infidel genius. And his highness said to the youth's tribe, Take him away and bury him. And in one tradition it is thus, that his highness said, Verily there are genii in Medinah which have embraced Islam ; then when you see any one of them, warn him three days ; and if he appears after that, kill him, because he is none but an infidel.' " Omm Sharic said, " His highness ordered a chameleon to be killed, and said, ' It was a chameleon which blew the fire into which Nimrod threw Abraham.' "... Abuhurairah 'A.G.S. " An ant bit a prophet, and he ordered the ant-hill to be burnt, which was done. Then God sent a voice to the prophet, saying, ' Have you burnt, on account of one biting you, a whole multitude of those that remembered God, and repeated his name?'" 1 His highness's scientific views on other subjects were in keeping with his zoology. "The genii," he lays it down, "are of three kinds. One kind have wings and fly, another are snakes and dogs, and the third move about from place to place like men." 5 The third are not so unlike the Boyl-yas. In Mahomet's system the devil and bad genii are at the root of all diseases except fever, which results from the heat of hell-fire, an element of which the Australians are as yet ignorant. He believed, of course, in the evil eye, and in spells and amulets, as so many of us still do ; but perhaps he nowhere appears to more advantage than in his astronomy. Stars, he says, were created for three purposes to embellish the regions, to stone the devil, and for guidance in the forest and on the sea. Our poor Wolves, Bears, Beavers, and Opossums must be tenderly regarded, , and may, we think, be believed to be thoroughly earnest in their faith, when views like these appear as propounded by the founder of one of the greatest existing religions. Of the traces of Fetichism among the Greeks and Eomans, it would be waste of time to say anything. We have said enough to prepare the reader for the ex- amination we are about to enter upon, of the evidence of the 1 Vol. ii. p. 310. Book xviii. chap. iii. Part I. "In explanation of animals, lawful and unlawful, to be eaten." Calcutta, 1809-10. 2 Vol. ii. p. 314. STUDIES IN ANCIENT HISTORY worship of animals and vegetables among the ancients; to give him the feeling that it is not very improbable that in classical regions we shall find totems, or something like them. Let us, however, before proceeding with that examination, state the results we have reached. We have found that there are tribes of men (called primitive) now existing on the earth in the totem stage, each named after some animal or plant, which is its symbol or ensign, and which by the tribesmen is religiously regarded ; having kinship through mothers only, and exogamy as their marriage law. In several cases, we have seen, the tribesmen believe themselves to be descended from the totem, and in every case to be, nominally at least, of its breed or species. We have seen a relation existing between the tribesmen and their totem, as in the case of the bear, that might well grow into that of worshipper and god, leading to the establishment of religious ceremonials to allay the totem's just anger, or secure his continued protection. We have seen in the case of the sun, conceived as a being, and having his tribe like any other animal, a first place acquired and the honours of a regular worship among tribes still in the totem stage, and that it is not improbable the cultus of other totems became regular as sun worship advanced ; and in the case of the Fijians, where the serpent and not the sun in- troduced regular religious observances, we have a more or less regular worship of the other totems as we seem entitled to consider them advanced to the status of gods. 1 We have also seen that while the intellectual condition of men that accompanies totemism is well established for all the lower races of men now existing, there is much evidence that the higher races had anciently been in a similar condition. We have totemism in various phases attending that condition, and having reason to think that the higher races had once been in the same condition, we have a probability that they also may once have had totems. 1 In some quarters in America, images of animals have been found in excava- tions, and one view is that they were idols. It will be remembered there were such images in the Sun Temples of the Incas. APPENDIX 519 PART II. TOTEM-GODS AMONG THE ANCIENTS We now proceed to examine the cases of the ancient nations. Inasmuch as these had, before the dawn of their histories, advanced far in civilisation (otherwise their histories, which depend on monuments and literary records, could not have commenced), we should expect that in the interval which intervened between their being in the totem stage supposing they were ever in it and the beginning of authentic records, the totems, if they were to become gods, would be promoted to a distinct place as the gods of the tribes that possessed them, and be the objects to them of regular religious worship. ' Looking again to the results of exogamy and female kinship, we might expect that while here and there, perhaps, a tribe might appear with a single animal god, as a general rule tribes and nations should have as many animal and vegetable gods as there were distinct stocks in the population. Some one animal V we should expect to find in a first place among the animal gods of a people as being the god of the dominant tribe ; but we should not expect to find the same animal dominant in all quarters, or worshipped even everywhere within the same nation. Moreover, since if the ancient nations came through the totem stage their animal and vegetable gods must have been of more ancient standing than the anthropomorphic gods, such as Zeus, Apollo, and Poseidon, we should expect to find in the sacred legends some hints of that priority. If we find any great number of such gods worshipped by the ancients, and if we find hints of their priority; still more, if we find tribes named after the sacred animals, and having them for their ensigns ; and, lastly, should we find the worshippers believing themselves to be, or having traditions, such as the Kirghiz have, that they were of the stock or breed of the animal they worshipped, then we think we may safely con- clude that so many concurring indications of the totem stage having been passed through are not misleading that, in fact, the ancient nations had in the prehistoric times been in the same case as that in which we now find the natives of 520 STUDIES IN ANCIENT HISTOR Y Australia. It will be a confirmation of this conclusion should we find the hypothesis that they had been in the totem stage to make intelligible numerous legends that have hitherto appeared entirely without meaning. It will be a further confirmation should we find that there is evidence that the ancient nations had been exogamous, and had the system of kinship through females only. What evidence then have we to show that the ancient nations came through the totem stage ? If they did, it was in prehistoric times. About these we have some facts preserved in the signs of the Zodiac, the majority of which are animals, or compounds of human and animal forms. We have another set of facts in the fanciful forms of those constellations which were figured, prehistorically, in animal forms. 1 Some of the stellar groups, we know, were named after gods or deified heroes. Were the animal groups named after gods also, or how was it the animals came to be promoted to the heavens ? There is nothing in the grouping of the stars to suggest the animal forms : no one can seriously pretend to perceive materials for any such suggestion. The stars, we must believe, were long familiar objects of study and observa- tion before they were grouped and named ; that they were conceived to be beings we may say we know. How came the early students of the heavens to name the groups from animals, and even many of the individual stars ? The probability is, 1 The Zodiacal constellations figured on the porticoes of the Temples of Denderah and Esne, in Egypt, are of great antiquity. "M. Dupuis, in his Origine des Cultes, has, from a careful investigation of the position of these signs, and calculating precession at its usual rate, arrived at the conclusion that the earliest of them dates from 4000 B.C. M. Fourier, in his Recherches sur la Science, makes the representations at Esne 1800 years older than M. Dupuis. . . . The truth seems to he that nothing is as yet definitely known of these ancient representations ; for the manner in which the investigations have been mixed up with the Biblical question of the antiquity of man has prevented any truly scientific research." Chambers's Encyc. art. "Zodiac." The ancient Zodiacal figures of the Hindus, ancient Persians, Chinese, and Japanese, in some respects resemble those of the Egyptians. Mr. Williams, of the Astronomical Society, informs me that three of the Chinese signs are named from the quail. The symbols of the years in the Aztec Cycle were named after plants and animals. Neither these nor the two hundred gods in the Aztec Olympus have yet been examined. APPENDIX 521 that in ancient as in modern times, stars, when named, were given names of distinction, that commanded respect, if not veneration ; and the suggestion therefore is, that the animals whose names were transferred to the stars or stellar groups were on earth highly if not religiously regarded. The legends that have come down to us, explanatory of the transference to the heavens of particular animals, bear out this suggestion. It will immediately be shown that nearly all the animals so honoured were anciently worshipped as gods. Let us see what the animals are. There is first of all the serpent in the constellation Serpentarius, which some said represented ^Esculapius ; there are also Scorpio and Draco the scorpion and dragon ; there is the horse Pegasus ; the bull Taurus ; the lion Leo ; the dog Canis (major and minor) ; the swan Cygnus ; the doves (according to some) the Pleiades ; the ram Aries ; the goat Capricornus ; the fishes Pisces ; the bear Ursa (major and minor) ; the crab Cancer ; and the asses' colts the Aselli. There are others, but this selection will suffice for our purposes at present. 1. The Serpent. We take the case of the serpent first, because for several reasons it has been more studied than any other. The serpent faith was very wide-spread, and it has attracted special notice from the part assigned to the serpent in Genesis in connection with the fall of man. Faber and Bryant have both pretty fully investigated this subject, which has also been treated in a separate work by Mr. Bathurst Deane. 1 Lately (in 1864) M. Boudin handled it in what may be called a large pamphlet rather than a book, 2 and last year Mr. Fergusson's elaborate work 3 threw much light upon it, at the same time that it has done more than any previous work to draw public attention to this extraordinary religion. It is unnecessary to adduce the evidence which establishes the prevalence, in ancient and modern times, of this worship. 1 The Worship of the Serpent. London, 1830. 2 Culte du Phallus ; Culte du Serpent, fitudes Anthropologiques. Paris, 1864. 3 Tree and Serpent Worship, by James Fergusson, F.R.S. India Museum, London. 522 STUDIES IN ANCIENT HISTORY It is a fact conceded on all hands, and in Mr. Fergusson's book it is demonstrated. That work, also, is very important in this respect, that it abounds in photographic illustrations from the Buddhist Topes of Sanchi and Amravati, which enable the reader to realise the fact that the worship was real worship. Men and women are exhibited in the sculptures in the act of adoring the Serpent God, so that the actuality of the worship is, by the book, as vividly impressed on the mind as it could be by attendance at divine service in a Serpent Temple say at Cambodia. In Cambodia, indeed, one would have found the god to be a living serpent a totem whereas these sculptures show that the living serpent had, among the Buddhists, lost rank, the, god being a heavenly (Ophi-morphic) being whose symbol was a serpent of five, seven, or nine heads, such as never had been seen upon earth. In short, we are enabled to see from Mr. Fergusson's work that the serpent religion, starting from the worship of the living animal as its root, had grown into a refined faith, comprising a belief in a spirit world in which the Serpent God held high rank ; and in an Olympus in which other gods were combined with him, and in which, below the gods, were angelic beings of various orders of standing and power. It is remarkable that the divine nature of these angelic beings in human form is demonstrated by serpents springing from behind their backs or from their shoulders, as the divinity of angels and cherubs in our own symbolism is indicated by their wings. Mr. Fergusson's introductory essay shows that the worship of the serpent has, at some time or other, found a place in the religious system of every race of men. It had its place in Egypt and in Palestine, even among the Hebrews ; in Tyre and Babylon; in Greece and Eome; among the Celts and Scandinavians in Europe ; in Persia and Arabia ; in Cashmere and India; in China and Thibet; in Mexico and Peru; in Abyssinia, and generally throughout Africa, where it still flourishes as the state religion in Dahomey ; in Java and Ceylon ; among the Fijians, with whom, as we saw, it still prevails ; and in various quarters in Oceania. Not less well established is the fact that it was a terribly real faith, with its APPENDIX 523 priests and temples, its highly-organised ecclesiasticism and ritual, its offerings and sacrifices, all ordered according to a code. The code, the ideas of the divine government, the god himself even, varied from point to point, there being no more uniformity observable here than elsewhere in a matter of faith. In one place the god was a living serpent; in another a collection of serpents, as if the whole species was religiously regarded. Here, again, the object of worship was an image of a living serpent ; there, an image of a creature of the religious imagination a spiritual ideal the five-headed, seven-headed, or nine-headed Naga. The god in some systems stood alone, was the god God; in others he had associates, sometimes equal, sometimes even superior to himself, such as the sun, or fire, an anthropomorphic god, the emblems of the procreative power, some other animal, like the horse, or some tree or vegetable, or the ocean. But under all the varieties the fact is manifest of the serpent having attained divine honours : the character of being a good, wise, beneficent, powerful deity, to adore and propitiate whom was man's duty and privilege. We have cited no authorities in support of these statements, because the facts are indisputable and well known, and a general reference to the works of Bryant and Fergusson is therefore sufficient. 1 Two points, however, must be touched 1 As to the doctrines of the serpent faith, we have, unfortunately, but meagre accounts. The Dahomans have both an earthly serpent and a heavenly. The earthly serpent (called Danh gbwe) is the first person in their Trinity, the others being trees and the ocean. Burton says of this serpent, " It is esteemed the supreme bliss and general good. It has a thousand Danh-si, or snake wives, married and single votaries, and its influence cannot be meddled with by the two others [trees and ocean], which are subject to it." It is believed to be immortal, omniscient, and all-powerful. In its worship there are solemn processions ; prayers are addressed to it on every occasion, and answered by the snakes in con- versation with the high priest. The heavenly serpent is called Danh, and has for his emblem a coiled and horned snake of clay in a pot or calabash. He is the god of wealth. The priestesses, in this serpent system, are girls resembling the JSTautch girls in the temples of Southern India, and when of age they are married to the god, who himself sets his seal upon them, marking them with his image under circumstances and with mysteries that are undivulged. Ancestral worship is conjoined with that of the snake in Dahomey, as it has been and is in other places, and with it almost certainly, and not with serpent-worship, are connected the horrible human sacrifices that occur on the coast of Guinea. The 524 STUDIES IN ANCIENT HISTORY upon before we can advance with our argument. The first respects the antiquity of the faith ; and the second, the relations between the god and his worshippers. (1) Of the great antiquity of the faith there can be no doubt. Compared with it, all the religions are modern ; they imply it at their foundations, and their earliest history is the record of its more or less complete suppression or subordination. The cultus prevailed, for example, among the Hebrews before the true faith. " With the knowledge we now possess," says Mr. Fergusson, "it does not seem so difficult to understand what was meant by the curse of the serpent" [in Genesis]. . . . When the writers of the Pentateuch set themselves to intro- duce the purer and loftier worship of Elohim, or Jehovah, it was first necessary to get rid of that earlier form of faith which the primitive inhabitants of the earth had fashioned for them- selves." The curse, of course, was not on the serpent, but on the cultus. We find a similar story in Persia and in India, in both of which places this religion prevailed. "The serpent state of our information on the Dahoman religion is to be regretted, as a minute knowledge of the beliefs of the worshippers, and of their traditions regarding the history of their religion, would be valuable in this inquiry. It is equally to be regretted that we are without details as to the beliefs of the snake-worshippers of India, who, we learn from the Indian newspapers, are to be found throughout our Eastern empire. How much have we yet to learn of our contemporaries even under the same Government with ourselves ! As we write, a letter appears from Bishop Crowther, respecting serpent-worship at Brass, a station of the Niger mission. "No poultry," the Bishop says, "can be reared on account of the snake cobra, which is held sacred here. Not to be killed because sacred, they become possessors of the bushes, and prove a great nuisance to the country. They very often visited the poultry coops at night, and swallowed as many as they wanted ; in conse- quence ot which no poultry could be kept, either by the natives themselves, or by the supercargoes in their establishments on shore : neither goats, sheep, nor small pigs escaped them. Thus the country is literally impoverished by them." To support the superstition there are two articles in the treaty made and sanctioned by Her Britannic Majesty's Consul for the Bight of Biafra and the Island of Fernando Po on November 17, 1856, one of which runs thus : "Article 12. That long detention having heretofore occurred in trade, and much angry feeling having been excited in the natives from the destruction by white men in their ignorance of a certain species of boa- constrictor that visits the houses, and which is ju-ju, or sacred, to the Brassmen, it is hereby forbidden to all British subjects to harm or destroy any such snake, but they are required, on finding the reptile on the premises, to give notice thereof to the chief man in town, who is to come and remove it away." APPENDIX 525 that beguiled Eve," says Max Miiller, " seems hardly to invite comparison with the much grander conception of that terrible power of Vritra and Ahriman in the Veda and Avesta." l In the Avesta there is a great battle between Thra^taona and Azhi dahaka, the destroying serpent. 2 The greatest exploit of Indra was the slaying of the serpent Ahi. " Where, Maruts," he is made to say in one of the Vedic hymns, " was that custom of yours that you should join me who am alone in the killing of Ahi." In another song Traitana takes the place of Indra in this battle ; more frequently it is Trita who fights, but other gods also share in the same honour. 4 The result of Mr. Fergusson's investigations is to represent serpent-worship as the basis of the religions of India, ex- cepting Sivaism, in which the bull has had the first place. 5 In Africa we most probably have the faith as it existed before the dawn of history. We know from the Egyptian monuments (says Mr. Fergusson) that neither the physical features nor the social status of the negro have altered in the slightest degree during the last four thousand years. If the type was then fixed which has since remained unaltered, why not his religion also 1 There seems no d priori difficulty. No other people in the world seem so unchanged and unchangeable ; movements and mixtures of races have taken place elsewhere. Christianity has swept serpent- worship out of what were the limits of the Roman world, and Mahomedanism has done the same over the greater part of Northern Africa. Neither influence has yet penetrated to the Gold Coast ; and there, apparently, the negro holds his old faith and his old feelings fast, in spite of the progress of the rest of the world. It may be very horrible, but, so far as we at present know, it is the oldest of human faiths, and is now practised with more completeness at Dahomey than anywhere else, at least at the present day. 1 Chips from a German Workshop, vol. i. p. 155. 2 Ibid. p. 100. 3 Miiller's Rig- Veda Sanhitd, vol. i. p. 165. 4 The Vedic Ahi was three-headed, like the heavenly Nagas in Mr. Fergusson's photographs, or like the Persian Zohak, only one of Zohak's three heads had become human. 5 This, we shall see, is a very partial view. Besides the serpent and bull, the sun and moon ; the sheep, goat, and elephant ; and the tortoise, fish, boar, and lion, enter (as totems) into the bases of the Hindu mythologies. Fire also, we may believe, was a totem in India. The Piqua tribe (one of the tribes of the Shawanoese) are descended from a fabulous man generated in a fire. Arch. Amer. vol. i. p. 275. 526 STUDIES IN ANCIENT HISTOR Y (2) It was common for those who had this worship to believe that the serpent was their progenitor. They were called, and called themselves, Serpents, after and as being of the breed of their god. Whole peoples, says Bryant, had the serpent-name, and counted themselves as being of the Serpent- breed. 1 " In Phrygia and upon the Hellespont was a people styled 'Oc/>to7ez/et9, or of the Serpent-breed, who were said to retain an affinity and correspondence with serpents." * In Ehodes, an old name for which was Ophiusa ; 3 in Tenos, one of the Cyclades ; 4 in Cyprus, also of old styled Ophiusa and Ophiodes ; 5 in Crete, 6 and in the island Seriphus, 7 it is related there were Serpent-tribes, or, as fable put it, swarms of Serpents, the personality of the tribesmen being lost in their name, as derived from, the god. Similarly at this day in India there are numerous tribes of Nagas on the north-eastern frontier, i.e. literally, Serpents, who were undoubtedly so named from the Serpent- God, as the Snake Indians are named from their totem. The name Nag has passed into a family or stock name among Hindus generally. Colonel Meadows Taylor says, " It is a common name both for males and females among all classes of Hindus, from Brahmins down to the lowest classes of Sudras and Mle'chhas." 8 The Thebans were esteemed Serpentigense ; 9 JEgeus, one of the Athenian kings, was reputed of the Serpent-breed ; 10 and the honour of having been first king of Athens was assigned either to Apd/ccov, a dragon, or to Cecrops, who was half a snake n probably as being on the mother's side not of the Serpent stock. Sparta 1 Ancient Mythology, vol. i. p. 481. The references to Bryant are to the Second Edition. London, 1775. 2 See Bryant ut supra ; Strabo, L. xiii. 1, p. 880 ; Pliny, L. vii. c. 2. " Crates Pergamenus in Hellesponto circa Parium, genus hominum fuisse [tradit] quos Ophiogenes vocat." 3 Strabo, xiv. 2. 4 e K \r)dri '00toO<r<ra, Steph. Byz. 5 Ovid, Met. x. 229. 6 Antoninus Liberalis, xli. 7 Virg. Ciris, p. 477 ("serpentiferam," Scaliger coll. Strabone, x. 5). 8 Tree and Serpent Worship, Appendix D. We infer from the statement that Nag is the name of a gotra. 9 Schol. Soph. Antig. 126. 10 Tzetzes, Schol. Lycophron. 496. 11 Meursius de Reg. Ath. i. 6 ; Diod. Sic. i. 28 ; Aristoph. Vesp. 438. APPENDIX 527 is said of old to have swarmed with serpents ; l and the same is related of Amyclae, 2 in Italy, which was a Spartan colony, the meaning of the tradition being that the inhabitants in either case were what in India would be called Nagas, and in America Snakes. 3 The kings of Abyssinia put the Serpent first on their list of kings as the progenitor of the royal line. In Peru, where the worship of the serpent was conjoined (as in many other cases) with sun-worship, the principal deity in the Pantheon was the Sun-Serpent, whose wife the female Serpent or female Sun brought forth at one birth a boy and girl who became the first parents of mankind. So the Caribs a fact already glanced at relate that the first of their race was half a serpent, being the son of a Warau woman by a river-god. Being slain and cut in pieces by his mother's brothers, the pieces, when collected under a mass of leaves, grew into a mighty warrior, the progenitor of the Carib nation. 4 The legends of Cashmere throw not a little light on these beliefs. They show us a doctrine resembling that worked out in the story of Elsie Venner the serpent nature in the human body capable of being displaced by a proper human nature. An ancestor of Sakya-Muni, for example, fell in love with a serpent-king's daughter, and married her. She could retain her human body, but occasionally a nine-headed snake sprang out of her neck. Her husband having struck it off one time when it appeared, she remained human ever after. Others of these legends represent a serpent-king (Naga Eaja) as "quitting his tank," becoming converted, and building churches ; and a sinful Brahman as being turned into a Naga, and spending his life for some years thereafter in a lake. 1 Aristot. de Mirab. Auscutt. 23. 2 Pliny, Nat. Hist. iii. 5 ; viii. 29. 3 It is remarkable how many fables become intelligible when read in the light of this and similar facts which we shall produce. Take, for example, the 'case of Cadmus as interpreted in this light by Mr. Fergusson : "Cadmus fought and killed the dragon that devoured his men, and, sowing his teeth, raised soldiers for his own purpose. In Indian language, he killed the Naga Raja [Serpent-king] of Thebes, and made Sepoys of his subjects." 4 Brett's Indian Tribes of Guiana, pp. 390-393. 528 STUDIES IN ANCIEi It was a natural consequence of the serpent being believed, where he was dominant, to be the first father, that he should be believed to be the first instructor of men. Accordingly we find that it was " the feathered serpent " who taught the Aztecs a knowledge of laws and of agriculture, and the principles of religion ; and Cecrops (who was half a serpent) that introduced marriage to Greece, and taught the people laws and the arts of life. Let us now see the results we have reached. They are 1. That the serpent was in numerous quarters of the world worshipped as a god by the most diverse ' races of men. 2. That serpent- worship is of the highest antiquity. 3. That the worshippers, in many cases, believed themselves to be of the Serpent-breed, derived from a serpent ancestor. 4. That the worshippers were in numerous cases named after the god Serpents. We now notice (5) that the serpent was used as a badge in many cases among the tribes that had the cultus. It was so used, for example, in Egypt, where was the sacred serpent Thermuthis. The natives are said to have made use of it as a royal tiara (says Bryant) with which they ornamented the statues of Isis. 1 We learn from Diodorus Siculus that the kings of Egypt wore high bonnets which terminated in a round ball, and the whole was surrounded with figures of asps. The priests likewise on their bonnets had the representation of serpents. 2 Menelaus, a Spartan and Sparta, we saw, was " Ophite " is represented as having a serpent for a device upon his shield. 3 The deity might also be expected to find his place on the coins of his worshippers, and the ancient coins having the serpent are accordingly numerous. It appears on early Egyp- tian coins of uncertain towns, and also on other early African coins ; on early coins (all of date B.C.) of Heraclea in Lucania ; of Perinthus in Thracia ; of Homolium in Thessalia ; of Cassope in Epirus ; of Buthrotum and Corcyra in Epirus ; of 1 .Elian, Nat. An. x. 31. 2 Ancient Mythology, vol. i. p. 475 ; Diod. Sic. iii. 3. 3 Pausanias, x. 26. APPENDIX 529 Amastris in Paphlagonia ; of Cyzicus and Pergamus in Mysia ; of Dardanus in Troas ; of Cos, an island of Caria; and of Magnesia, Nacrasa, and Thyatira, in Lydia. 1 2. The Horse. The Horse figures in the heavens as Pegasus, and we find him on the coins of numerous cities. He is on the coins of various cities of Hispania and Gallia ; of Fanum in Umbria ; Beneventum in Samnium ; Nuceria in Campania ; Arpi, Luceria, and Salapia in Apulia ; Grumentum in Lucania ; Thurium in Apulia ; ^Etna in Sicilia, and also Camarina, Gela, and Panormus, in Sicilia ; of Syracuse ; Melita (malta) ; Panticapeeum in Taurica ; Cypsela, Maronea, ^Egospotami, and Cardia, all in Thracia ; Amphipolis, Bottiaea, and Thessalonica, all in Macedonia. On the coins of Thessalia in genere, and on those of Atrax, Crannon, Demetrias, Elatea, Gyrton, Larissa, Pelinna, Phalanna, Pharcadon, Pherae, Perrhaebia, Ctimene, Scotussa, and Tricca, in Thessalia ; of Alyzia in Acarnania ; Locri- Opuntii in Locris ; Phocians in Phocis ; Tanagra in Bceotia ; Pheneus in Arcadia ; Gargara in Mysia ; Parium in Mysia ; Alexandria in Troas ; Cyme in JEolis ; Colophon in Ionia ; Magnesia in Ionia ; Mylasa in Caria ; Termessus in Pisidia ; Antioch in Cilicia ; Adana in Cilicia ; Aninetum in Lydia ; Phrygia Epictetus ; Larissa in Seleucis ; Cyrene in Cyrenaica; Tarentum in Calabria, and (adds Mr. Sim) perhaps on many others. The coins are all of date before the Christian era. Was the horse, who was thus honoured, a god ? In the photographs in Mr. Fergusson's book we have* some evidence that he was a god among the serpent-worshipping Buddhists. The horse first occurs in Plate xxxv. Fig. 1. Mr. Fergusson remarks on it, " In this bas-relief the principal object is the Sacred Horse, richly caparisoned, who heads the procession, and towards whom all eyes were turned ; . . . behind him a chief in his chariot, bearing the umbrella of State, not over himself, but apparently in honour of the horse." It next occurs along with Siddhartha on Plate lix., but the worship in this case seems to be all given to the prince. It occurs again on Plates Ixxx. and Ixxxi. On these Mr. Fergusson observes : Figs. 2 and 3 of this Plate (i.e. Ixxx.) and Fig. 3 of Plate Ixxxi., 1 The lists of coins cited in this paper have been furnished to the writer by an accomplished numismatist, Mr. George Sim, Curator of the Coins in the Antiquarian Museum, Edinburgh. 2 M 530 STUDIES IN ANCIENT HISTORY instead of the emblems we are usually accustomed to, contain two medallions, the upper representing the worship of the Horse, the lower, Buddha, seated cross-legged, surrounded by listeners or adorers. As we have frequently had occasion to remark, the Horse plays an important part in the sculptures at Amravati. It is once represented as honoured at Sanchi ; but this form of worship occurs here several times, but nowhere so prominently as in those three Dagobas (and it is to be presumed that there was a fourth). It is not easy to say what we are to understand from the prominence of the Horse in such a position as this. Is it an importation from Scythia, brought by immigrants from that country ? Is it the Horse of the Sun or of Poseidon 1 Is it the Avalokiles'vara of the Thibetan fables ? Some one must answer who is more familiar than I am with Eastern mythology^ At present it will be sufficient to recall to memory how important a part the Horse sacrifice, or As'wamedha, plays in the Mahabharata, and in all the mythic history of India. What is still more curious is, that the worship of the Horse still seems to linger in remote parts of India. At least, in a recent work by Mr. Hislop, missionary at Nagpore, edited by Sir R. Temple, he [Mr. Hislop] describes the religion of the Gonds in the following nine words : " All introduce figures of the horse in their worship." Other instances might, no doubt, be found if looked for ; but the subject is new and unthought of. If Mr. Fergusson had looked further in Mr. Hislop's book he would have found that the fact of horse-worship is not left to inference or conjecture. In a footnote at p. 51, Sir E. Temple says : " The god Koda Pen, or Horse-god, is some- times worshipped by the Gonds, and sometimes there are sacred images of this animal." So we have in India a horse- god now. What tribes besides the Gonds have worshipped him? The horse occurs again in Mr. Fergusson's plates. In Plate xcv., Fig. 4, he is introduced in mid-air, alongside the wheel (a Buddhist idol l ), as an object of equal reverence ; and on a piece of sculpture, where the wheel just above him is the special object of worship. In Plate xcvi., Fig. 3, he issues from the portal with the umbrella of State borne over him, the hero of the representation. The same subject is repeated on another slab, Plate xcviii., Fig. 2. The opinion formed by Mr. Fergusson is that the bas-reliefs show that the horse was an object of reverence, if not exactly of worship, at Amravati, and 1 See Ezekiel, chap. x. vv. 8 ff. APPENDIX 531 that the reverence paid to him is the counterpart of the worship of the bull Nandi by the Shivites. Let us now see what evidence there is of this worship else- where. Mr. Bryant supplies a goodly array of facts. In his Essay l on Metis and Hippa, after disposing of the former as one of the most ancient deities of "the Amonians," represented under the symbol of a beautiful female countenance surrounded with serpents, he proceeds to say : Hippa was another goddess of like antiquity, and equally obsolete. Some traces, however, are to be still found in the Orphic verses, by which we may discover her original character and department. She is there represented as the nurse of Dionysus, and seeins to have been the same as Cybele, who was worshipped in the mountains of Phrygia, and by the Lydians upon Tmolus. She is said to have been the Soul of the World H fjiev yap"l7nra TOV Travros ovcra ^vx 7 ? 2 an( l the person who re- ceived and fostered Dionysus when he came from the thigh of his father. 3 This history relates to his second birth, when he returned to a second state of childhood. Dionysus was the chief god of the Gentile world, and worshipped under various titles, which at length came to be looked on as different deities. Most of these secondary deities had the title of Hippius and Hippia ; and as they had female attendants in their temples, these, too, had the name of Hippai. What may have been the original of the term Hippa and Hippus will be matter of future disquisition. Thus much is certain, that the Greeks uniformly referred it to Horses. Ares was Hippius ; 4 so was Poseidon, although a god of the sea, being so called from raising a horse out of the earth in his contest with Athene for the superiority at Athens ; 5 but Athene herself was Hippia, as were also Demeter and Hera. 6 Demeter, styled Hippa, the Greeks represented as turned into a mare ; 7 Hippius Poseidon, in like manner represented as a horse, they supposed in that shape to have had an intimate connection with the goddess. 8 The nymph Ocyroe was changed into a mare, and so was Philyra, whom Saturn, in the shape 1 Yol. ii. p. 27. 2 See, in proof of this, Orphic Frag. 43 ; Orpheus Gesneri Lipsise, 1764, p. 401. 3 Orph. Hymn. 48, 49. 4 Pausan. v. 15. 5 Serv. ad Georgic. i. 12. 6 Pausan. 7.c. 7 Pausan. viii. 25, 5 ; and see Smith's Did. s.v. " Arion." 8 Ovid, Metam. vi. 118. 532 STUDIES IN ANCIENT HISTORY of a horse, followed neighing over the mountains of Thes- saly ! l Bryant, who conceived that the ancients knew nothing of their own mythologies, and whose great discovery was that every mythological fact anywhere to be found related either to Noah, the ark, or the deluge, thinks the Greeks were quite wrong in fancying Hippa and Hippus to have had anything to do with the horse. These gods, he says, came from Egypt, and were one with the sun and Osiris, and ultimately with the ark. 2 He tells us, however, that the horse (like the ox and eagle, which we shall see were gods) was a sacred symbol in Egypt, where almost every animal, from beetles to bulls, was worshipped, so that the Egyptians made the mistake equally with the Greeks, if there was one. Mistake or not, there is no question of the reality of the faith that followed on it. The horse -gods and mare -goddesses had their temples and regular worship, and not only gods and goddesses, but places, and presumably tribes of men, were named from the horse. There were the Hippici Montes in Sarmatia ; "ITTTTOV K^firj in Lycia ; "ITTTTOV d/cpa in Libya ; and a town Hippos both in Sicily and in Arabia Felix. 3 The horse-name occurs frequently in composition, as in Hipporum, Hippouris, Hippana, Hip- ponesus, Hippocrene, and many others; and, indeed, horse- names are so frequent in Homer alone a fact observed by Mr. Gladstone as to suggest that there were horse-tribes in, and bordering on, Greece, as there were Nagas and Ophites. One of the twelve Athenian tribes was Hippothoontis, 4 their eponymous progenitor Hippothoon, who was nurtured by mares! 5 jEolus and his family were Hippotades, 6 and a village in the tribe (Eneis was Hippotamada. 7 There was a tribe, Hipporese, 8 in Upper Ethiopia, and the Hippopodes were a people of Scythia, who had horses' feet ! 9 There was a city Hipponesus 1 Ovid, Metam. ii. 668 ; Virg. Georg. iii. 92. 2 Vol. ii. p. 408. That the totem should be identified with the Sun is what we should expect. 3 Agathem. ii. 13 ; Steph. Byz. 4 Boeckh, Corp. Inscr. Att. 40, 45, 46, 59, etc. ; Pollux, viii. 110. 5 Hygin. Fab. 187. 6 Horn. Od. x. 2, etc. 7 Steph. Byz. 8 Plin. Hist. Nat. vi. 30. 9 Dionys. Per. 310 ; Plin. Nat. Hist. iv. 13, 27, etc. APPENDIX 533 in Caria, and another of that name in Libya. 1 There were two towns called Hippo in Africa, and a town Hippola jn the Peloponnesus ; also a Hippo in Spain, and a town of the Bruttii, now Monte Leone. 2 The horse appears on the coins of four cities of Thrace, where were the horses of Diomedes, that fed on human flesh 3 a suggestion that these horse-tribes men were cannibals. Bryant says these horses were the priests of Dionysus ; his theory also is that they were men. When we turn to Thessaly equorum altrix on the coins of fourteen towns in which we find the horse, we are in the country of the Centaurs, half men and half horses no doubt men who were yet called horses, after their animal god. Their battle with the Lapithae, springing out of a quarrel at the marriage of Hippodamia, is famous in fable. 4 Chiron, the most celebrated of the Centaurs, was a son of Saturn (by repute), who changed himself into a horse to avoid his wife Rhea. 5 Intimate relations these be- tween the horse and the oldest anthropomorphic gods. He was the instructor of mankind in the use of plants, the study of medical herbs, and the polite arts, having in these even the great serpent ^Esculapius for a pupil. 6 Finally, Jupiter made a constellation of him under the name Sagittarius. 7 Pausanias says that Demeter, worshipped by the Phigalians, was represented as a woman with the head of a horse. 8 Marus Balus, an old Italian god, who lived three times, was biform, half man, half horse. 9 In Pegasus we have a winged horse sprung from the blood of Medusa, that flew up to heaven immediately on being born. 10 He was the favourite of the Muses, figured in various exploits on earth, and was finally placed among the constellations. 11 He was the special insigne of Corinth, and occurs on ancient coins of that place, of Syra- cuse, and Corcyra. 12 A Gaulish coin belonging to the first 1 Steph. Byz. 2 Strabo, xvii. 3, 13 ; Steph. Byz. ; Plin. Hist. Nat. iii. 1, 3 ; Pompon. Mela, ii. 4. 3 Lucret. v. 29, etc. 4 Odyss. xxi. 295, etc. 5 Tzetzes ad Lycophron. 1200. 6 Find. Pyth. iii. 6, etc. 7 Ovid, Fast. v. 414. 8 viii. 42. 9 .Elian, Var. Hist. ix. 16. 10 Hes. Theog. 281, etc. u Ovid, Fast. iii. 457. 12 Spanhemii Numismata. vol. i. pp. 274 et scq. 534 STUDIES IN ANCIENT HISTOR Y century B.C. has the horse with a human head. We have heavenly horses in Homer; the horses, ordinary and winged, of Agni, Indra, and Soma, and the eight-legged horse of Odin. There is a controversy as to whether Agni himself was not a horse. 1 In Max Miiller's Rig- Veda Sanhita (p. 15) the reader will find the distinguished professor combating Messrs. Boehtlingk and Eoth over certain Yedic passages, in which these gentlemen, in their Dictionary, say : " He (Agni) him- self appears as a red horse." We cannot pretend to enter into the merits of the controversy, but the reader may already be satisfied that an Agni Hippius should create no more wonder than a Hippius Poseidon. 2 We conclude, then, that the horse had been anciently a god in India, in Egypt, in Greece, and many other quarters ; that it was such before most of the deities figuring in the Olympus appeared ; that it became the insigne of many tribes of men ; and that it is certain there were numerous tribes named after it. 3. The Bull. The Bull figures in the heavens ; and bulls, bisons, minotaurs, and parts of these on coins are too numer- ous for specification. A few will be found figured in the Numismata Sparihemii. As the bull and cow are well-known sacred animals, we may be brief with them. The living animal (says Bryant) was in many places held sacred, and reverenced as a deity. One instance of this was at Memphis, where 1 Whether he was a horse or not, he was certainly a goat, as we shall see. Like the other men-gods, he was in turn identified with the totem, whatever it was, of the tribe that took him up. 2 See Rig-Veda, Sanhita, pp. 14-18 ; and see p. 27. In the Padma Purana, Krishna in the form of a horse is represented as rescuing the vedas when " the worlds " were burnt up (Muir's Texts, vol. iii. second edition, p. 28) ; and in the Vishnu Purana we have the Sun as a horse teaching a horse-tribe men called Vagins (i.e. horses), from being instructed by the Sun-horse (Muir's Texts, vol. iii. second ed. p. 51; and see p. 52). The horse gives his name to a Brahmanic gotra. The Sun (Aditya) appears again as a horse in the Catapatha Brahmana (Id. vol. iv. p. 62 ; and see vol. i. second ed. pp. xii. and 12, where the horse is also identified with Yama and Trita.) We have no doubt that these partial con- tributions to ancient Indian literature were made by men of the horse stock. APPENDIX 535 they worshipped the sacred bull Apis ; l and another was to be found at Heliopolis, where they held the bull MneVis in equal veneration. 2 The like custom was observed in Mo-memphis, Aphroditopolis, and Chusa, with this difference, that the object of adoration in these places was an heifer or cow. 3 The animal was also worshipped under symbols, or as re- presented by images. We see this illustrated in the case of the Jews, who fell into the idolatry with the sanction of Aaron himself. 4 An apology made for Aaron is, that he adopted this image not from Apis or Osiris, but from the Cherubim having the faces of oxen ! 5 The idolatry was probably never fully suppressed. It was openly renewed under Jeroboam, who made two calves, and set one up in Bethel, the other in Dan. In this case, as in the preceding, the calf was recognised as the god that had brought the people out of the land of Egypt ! The calves of Jeroboam are spoken of by Hosea (x. 15) as young cows ; as also by the Septuagint and by Josephus, who says that Jeroboam made two heifers of gold, and consecrated to them two temples. The Bull Nandi is, at the present day, a quasi-god in India, worshipped by the Shivites ; while by all Hindus the cow is religiously regarded. Of course, in Bryant's system, the bull is Noah, while the crescent on the side of Apis is the Ark. Every one knows what cows are in the Dawn system of Mr. Max Miiller. As in the case of the serpent and horse, the religious imagination conjured into existence a variety of spiritual bovine beings bulls with men's bodies, men with bull's bodies, bulls with two heads, and so forth. Astarte, we saw, had, according to Sanchoniatho, a bull's head, and Diana was worshipped by the Scythae, under the title of Tauropolus 6 and Taurione. 7 In the Orphic fragments Dionysus is represented as having the countenance of a bull, and elsewhere as being a 1 Herod, ii. 153. 2 Plutarch de Is. et Os. 33. 3 Vol. ii. p. 415. Strabo, xvii. 1, 22 ; ibid. 35 ; Lilian, Nat. Anim. x. 27. 4 See a curious chapter on this subject in Lewis, Origines, vol. iii. p. 32. 5 The later Jews say that the insigne of the tribe of Ephraim was an ox. 6 Eurip. Iph. Taur. 1457 ; Aristoph. Lys. 447 ; Soph. Ajax, 172. 7 Suidas. 536 STUDIES IN ANCIENT HISTOR Y bull. 1 In Argos he was fiovyevrjs, the offspring of a bull ; 2 ravpoyevrfs is one of hfs epithets in the Orphic hymns. 3 Poseidon was Taureus 4 as well as Hippius, and so also was Oceanus. 5 The bull-faced people are frequent in the legends of India, where the bull is a god; and in Japan we find a deity, Goso Tennoo the ox-headed prince of heaven. 6 The people of the Tauric Chersonesus were named, accord- ing to Eustathius, from the bull Taurus ol Se Tavpoi TO eOvos CLTTO rov %o>ov Tcivpov, <acr/, KaKovvTai? So were the following mountains, places, and peoples :-^-Taurus, Taurania, Taurica, Taurenta, Tauropolis, Tauropolium, Taurominium, Tauri, Taurini, and Taurisci. 8 How far the god might be followed as giving names to other places and peoples by the process of etyrnologically analysing the names in different languages, we have not the means of ascertaining. We have found the bull figured in the heavens and on numerous coins, and giving his name to numerous tribes of men, worshipped as a god, and regarded as the father and first lawgiver by his worshippers. We have found him also in intimate relations with the earlier gods and goddesses, who either drew titles from him or wore his form, as if they supervened upon a system in which he had been chief, and from which, in the process of time, they displaced him. His case thus resembles that of the two animal gods previously considered. 4. The Lion. The Lion is in the heaven as Leo, and figures on the ancient coins of many cities, e.g.: On coins of Hispania and Gallia ; 9 Teate in Marrucini ; Capua in 1 Orph. Hymns, 30, 45 ; Lye. Cass. 209 ; et Schol. ibid. 2 Plutarch de Isid. et Osir. 35 ; Qucestiones Grcecce, 36. 3 Orph. Frag. 28. 4 Hesiod, Sc. 104 ; Hesych.; Eustatli. ad Iliad, ii. 381. 5 "Bull-headed." Eurip. Orest. 1377. 6 Ksempfer's Japan, p. 418, cited by Bryant. 7 Eustath. ad Dionys. Perieg. 306. 8 Steph. Byz. ; Cs. Bell. Civ. ii. 4 ; Strab. vii. 4, iv. 6, 12, etc. 9 Mr. John Evans, in his work on British Coins, p. 180, says the lion fre- quently occurs on Gaulish coins. APPENDIX 537 Campania ; Arpi in Apulia ; Venusia in Apulia ; Paestum in Lucania ; Heraclea and Velia in Lucania ; Rhegium in Bruttium ; Leontini, Panormus, Syracuse, and Messana, in Sicilia ; Chersonesus Taurica (Pan- ticapseum) ; Tomis in Mcesia Inferior ; Abdera, Perinthus, Cardia, Lysi- machia, and the Chersonesus, in Thracia ; Thasos ; Amphipolis, Macedonia ; Thessalia in genere; Corcyra in Epirus ; Heraclea in Acarnania ; Corinthus in Achaia; Gortyna in Greta; Adrianotherae vel Hadrianotherae in Bithynia ; Metroum in Bithynia ; Germe in Mysia ; Magnesia and Miletus in Ionia ; Smyrna in Ionia ; Acrasus, Apollonia, Attalia, Blaundus, Gordus, Hyrcania, Magnesia, Philadelphia, Saettae, Sardes, Silandus, Temenothyrae, Thyatira, all in Lydia, being thirteen towns; Acmonia, Cadi, Cibyra, Peltae, Sala, and Synaus, towns in Phrygia ; Pessinus in Galatia ; Gyrene in Cyrenaica ; in Libya in genere. The coins are all of date before the Christian era. Was the lion, who was thus honoured, a god ? He was ; but his worship must have early become obsolete, as we have only one well -vouched instance of it within the historical period, namely, in Leontopolis, the capital of a district of the same name in Egypt. .ZElian and Porphyry both say it was worshipped there was the deity of the place. 1 There is a considerable amount of evidence, however, that this animal had, in prehistoric times, been more generally worshipped, and that tribes of men had been named after him. We have become familiar with compounds of human and bestial forms in connection with the worship of the serpent, horse, and bull ; the serpent body with human head ; the female human form with one or more horse - heads ; the Minotaur ; and should expect that if the lion were a god, he should, by the same mental processes, be made to enter into similar compounds. Since we have him in one place as a god, and have him in the heavens and on numerous coins, and, what is familiar, as the symbol of many tribes, should we find such a compound of the human and leonine forms worshipped, venerated, or feared, or with a remarkable hold on the imagina- tions of men, it will not be unreasonable to infer that the compound had an origin similar to the others we have become acquainted with. Now we have such a compound in the 1 Bryant's Observations and Inquiries, Cambridge, 1767, p. 130. JElian de Animal, xii. 7. Porph. de Abst. iv. 9 ; cf. Strab. xvii. 1, 40. 538 STUDIES IN ANCIENT HISTOR Y Sphinx, which therefore may throw some light on the cultus of the lion. 1 In the Egyptian hieroglyphics, we are told, the sphinx bears the name of Neb or Lord, and Akar or Intelligence the form of it being a lion's body with human head. The Great Sphinx at Gizeh is colossal, and hewn out of the natural rock. 2 It is of great antiquity an age at least equal to the Pyramids. In front of the breast of this sphinx was found, in 1816, a small chapel formed of three hieroglyphical tablets, dedicated by Thothmes III. and Kameses II. to the sphinx, whom, it is said, they adored as Haremukhu, i.e. the sun on the horizon. 3 The fourth tablet, which formed the front, had a door in the centre, and two couchant lions placed upon it. "A small lion was found on the pavement, and an altar between its fore-paws, apparently for sacrifices offered to it in the time of the Bonians." 4 In 1852 discovery was made of another temple to the south of the sphinx, built at the time of the fourth dynasty, of huge blocks of alabaster and granite, and which was most probably, like the former, devoted to its worship. Numerous sphinxes have been found elsewhere in Egypt, as at Memphis and at Tanis. That found at the latter place is assigned to the age of the Shepherd dynasty. Sphinxes have also been found in Assyria and Babylonia, and they are not uncommon on Phoenician works of art. Mr. Layard mentions having dug out of the Mound of Nimroud " a crouch- ing lion, rudely carved in basalt, which appeared to have fallen from the building above, and to have been exposed for centuries to the atmosphere " ; also a pair of gigantic winged bulls, and a pair of small winged lions, whose heads were gone. Human- headed lions he found, of course ; also human figures with lions' heads. 5 1 The reader will find a long treatise on the Sphinx in the Numismata Spanhemii, where also the Sphinx is figured on several coins. It is hardly necessary to say it is common on coins. 2 It is upwards of 172 feet long and 56 feet high. 3 " Isis sub forma Leonis itidem et cum facie muliebri occurrit nonnunquam in nummis JEgyptiorum sicuti in quodam Antonini Pii quern seruat Gaza Medicea." Spanhemii Numismata, torn. i. De Sphinge in nummis. 4 See article "Sphinx," Charnbers's Encyc., and authorities there cited. 5 Nineveh, vol. ii. p. 463. APPENDIX 539 The Egyptian sphinx had the whole body leonine, except the face, and this would appear to be the most ancient form ; the sphinxes with wings are later, and are supposed to have originated with the Babylonians or Assyrians. The Greek sphinxes were still further from the primitive type ; they were all winged, and had other elements in their composition besides the human and leonine. Probably they were unrelated to the Egyptian as an original. The Theban sphinx, whose myth first appears in Hesiod (Theog. 326), had a lion's body, female head, and bird's wings a suitable emblem, we should say, for a composite local tribe comprising lions and, say, eagles or doves. 1 She was a supernatural being, the progeny of the two-headed dog of Geryon, by Chimsera ; 2 or of Typhon, by Echidna. 3 On either view she had lion kindred, for Typhon, although a Naga, had one celebrated lion among his offspring, the Nemean lion to wit, 4 who infested the neighbourhood of Nemea, filling its inhabitants with continual alarms. The first labour of Hercules was to destroy him, and the Nemean games instituted in honour of one who had fallen a victim to a snake 5 were renewed to commemorate the destruction of a lion ! A strong suggestion, this, of the new-comers, the Heraclidse, being alike antipathetical to the snakes and lions, to the tribes, as we read it, who had these animals as gods, and were called after them. Lion names were common, and the name remains. We believe the result of inquiry will be to establish, by etymo- logical evidence, that the animal gave its name to numerous tribes. Such evidence as we have to adduce of this fact, how- ever, will be better appreciated when produced further on in this exposition. 5. The Dog. The Dog gives its name to three constella- tions Canis Major, Canis Minor, and Canicula, as well as to the stars Canis Sirius (Cahen Sehur), the brightest in the heavens ; Procyon and Cynosura, " the dog's tail." It appears 1 Eurip. Phcen. 1023; Palsephatus de Fabulis ; Schol. Eur. Phcen. 45, 806; Apollod. Bill. iii. 5, 8. 2 Hes. I.e. 3 Apollod. Bibl. iii. 5, 8. 4 Ib. ii. 5. 5 lb. iii. 6, 4. 540 STUDIES IN ANCIENT HISTOR Y on various ancient coins ; for example, on uncertain coins of Etruria ; on coins of Pisaurum in Umbria ; Hatria in Picenum ; Larinum in Frentani; on the coins of Campania in genere, notably of Nuceria in Campania ; of Valentia in Bruttium ; Agyrium in Sicilia; of Eryx, Messana, Motya, Panormus, Segesta (very many), and Selinus, all in Sicily ; of Chersonesus Taurica ; Phalanna in Thessalia ; Corcyra in Epirus ; Same in Cephallenia ; Cydonia in Crete ; Colophon in Ionia, and Phocsea in Ionia. Besides these, which are all of date B.C., there are coins figured in the Numismata Spanhemii with the legend of the dog Cerberus, and one in Mr. Evans's British Coins, of which that learned author says : The reverse is very remarkable, and must be regarded as in some manner connected with the early British mythology, though I must con- fess myself entirely at a loss to offer any satisfactory elucidation of the device. The attitude of the dog [which has one of its fore feet placed on a serpent] is very like that in which it is represented on the small brass coins of Campanian fabric, bearing the name of Roma, but there is no serpent on those coins. The type is hitherto unpublished, and belongs to the third class of the coins of Cunobeline those with the name of his capital upon them. 1 With such facts before us, and the knowledge we have already attained to of their probable significance, it need not surprise us to find that the dog was a deity. Bryant, after doing all he could to work him into his Ark scheme, has to confess that his view, that the belief in the worship of the dog was derived from Cahen being the Egyptian name for a priest or sacred official, won't meet the facts. Though I have endeavoured to show (he says) that the term of which I have been treating was greatly misapplied in being so uniformly referred to dogs, yet I do not mean to insinuate that it did not sometimes relate to them. They were distinguished by this sacred title, and were held in some degree of veneration. 2 The facts are as follows : Juvenal states that dogs were worshipped in some places, " oppida tota canem venerantur " ; 1 Evans, British Coins, p. 316. 2 Vol. i. p. 351. 3 Sat. 15, v. 8. APPENDIX 541 Diodorus Siculus says the same thing ; l Plutarch relates that in Egypt they were holy, but not after the time of Cambyses, when they misbehaved themselves by devouring the bull Apis, whom that king slew ; 2 and Herodotus informs us they were so regarded by the Egyptians in his own time that when a dog died the members of the family it belonged to shaved themselves all over. 3 The dog was called Cahen and Cohen a title given by the Egyptians to the animal and vegetable gods they wor- shipped in general (query, an equivalent of totem ?) ; and while the living dog was thus esteemed, there were spiritual dog-beings or gods, such as Canuphis, or Cneph (Anuphis and Anubis of the Greeks and Eomans), some represented as having the human body and dog's head, and others conceived as having the full canine figure, with one, two, or more heads, just as in the case of the heavenly Nagas, bulls and horses. As the animals last named gave titles to the gods who super- seded them, so did the dog ; Hermes 4 was a dog. Hecate had three heads, one a dog's, one a horse's, and one a boar's, which suggests, on the system of interpretation we have been propounding, that she originated in a compromise of a local tribe, which contained gentes of the dog, horse, and boar stocks. The boar will be shown to have been a god, at least a totem. In a temple of Vulcan near Mount ^Etna was a breed of dogs that treated good men gently, and were ferocious to bad men, which is curious, as we have similar fables respecting serpents in Syria (given by Aristotle), 5 and birds in the islands of Diornedes (given by Pliny). 6 On these dog-beings Bryant has some remarks in which we are disposed to concur. " When I read of the brazen dog of Vulcan," he says, " of the dog of Erigone, of Orion, of Geryori [a two-headed dog], of Orus, of Hercules, of Amphilochus, of Hecate, I cannot but suppose that they were titles of so many deities, or else of their priests, who were denominated from their office." 7 1 i. 18. 2 Isid. et Osir. 44. 3 ii. 66. 4 Plut. de Isid. et Osir. xi. ; Porphyr. de Abst. iii. 16. 5 De Mirab. Auscult. 161. 6 Plin. Nat. Hist. x. 61. 7 Bryant, vol. i. p. 347. 542 STUDIES IN ANCIENT HISTORY There were dog-tribes as a matter of course. Such we must assume the Cynocephali in Libya to have been, whom Herodotus mentions as a race of men with the heads of dogs ; l named, as Bryant observes, from their god fable adding the physical peculiarity in explanation of the dog-name. ^Elian and Plutarch, besides bearing witness to the veneration paid to dogs in Egypt, relate " that the people of Ethiopia had a dog for their king; that he was kept in great state, being surrounded with a numerous body of officers and guards, and in all respects royally treated. Plutarcl} speaks of him as being worshipped with a degree of religious reverence." ^ No doubt they had heard something like this, and misunderstood it. The king was a dog, in the same way that a Naga Eaja is a serpent, and the reference is to a dog-tribe. What the lamented Speke tells of the traditions of the Wahuma in Central Africa suggests to us that inquiry may yet show that there was a tribe in that quarter with the dog for its totem, and it is probably there still. 3 6. The Swan. The Swan is in the heavens as Cycnus, and figures on the coins of Camarina in Sicilia ; Leontini in Sicilia ; Argissa in Thessalia ; Clazomenae in Ionia ; on the coins of other uncertain cities of Ionia, and of Eion in Mace- donia. The coins of Eion, says Mr. Sim, are of date 500 B.C., while the others having the swan are probably of date about 300 B.C. We have no direct evidence of the swan having been a god that is, having temples of his own ; but two great gods, Zeus and Brahma, wore his form, and the latter was named after him ; and there is a considerable quantity of myth and fable explainable on the supposition that the bird had been at least a totem. Mr. Evans inclines to think the swan was Phoenician. It is found figured on ancient Phoenician works of art. 1 Herod, iv. 191 ; Isidor. Hisp. Orig. xi. 3. 2 Bryant, vol. i. p. 329 ; Julian de Nat. Anim. vii. 40 ; Pint. adv. Stoicos, xi. 3 Speke's Journal, pp. 252, 257. APPENDIX 543 Three persons are named by Ovid as having been changed into swans : a son of Poseidon, who was killed by Achilles before the metamorphosis ; a son of Apollo, who in a fit of vexation committed suicide, and was changed into a swan ; and a son of Sthenelus, of Liguria, who in his affliction for the death of his friend Phaethon was changed into a swan. 1 Of the last story there is another version given by Lucian, who speaks of swans in the plural in his jocular account of an attempt to discover the sweet-singing birds when boating on the Eridanus. 2 In the Prometheus of ^Eschylus, lo is directed to proceed till she reaches the Gorgonian plains, where reside the three daughters of Phorcys in the shape of swans, with one eye and one tooth between them. 3 Socrates is represented as speaking of swans as his fellow - servants, and Porphyry assures us that he was very serious in doing so. 4 Calchas, a priest of Apollo, was called a swan ; 5 and at the first institu- tion of the rites of Apollo, it is said many swans came from Asia, and went round the island Delos for the space of seven days. 6 The companions of Diomedes, lamenting his death, were changed into birds resembling swans. 7 They settled in some islands in the Adriatic, and were remarkable for the tameness with which they approached the Greeks, and for the horror with which they shunned all other nations. 8 Lastly, the singing of swans was very celebrated, and spoken of not only by the poets, but by such men as Plato, Plutarch, Pliny, and Cicero as a thing well known. Their melancholy strains were never so sweet as when they were dying. The only instance of the form of this bird being assumed by a Greek god is in the case of Leda or Nemesis. Zeus, in the form of a swan, deceived the lady. She produced two eggs in con- sequence, from one of which sprang Pollux and Helena, and from the other Castor and Clytaemnestra ! Again, the inhabitants of islands who, though in the swan 1 Ovid, Met. xii. 144, vii. 379, ii. 377. Besides these there are two mythical persons of the name, both sons of Ares. 2 Lucian de Electro sen Cygnis. 3 ^Esch. Prom. 814. 4 Porph. de Abst. iii. 16. 5 Lye. Cass. 426. 6 Callim. Hymn to Delos, 249. 7 Ovid, Met. xiv. 509. 8 Plin. Nat. Hist. x. 61. 544 form, were yet human, like the birds of Diomedes, can only mean a Swan-tribe, we think. The fact of the swan figuring in the heavens and on ancient coins, taken along with the fact that it was a tribal name, makes it probable that the swan was a god, and highly probable that it was at least a totem elsewhere than in Australia, where it is a totem now. 7. The Dove. The Dove, or Pigeon, is figured on coins, in the Numismata Spanhemii, 1 of Eryx in Sicilia (where we shall see it was worshipped). Mr. Sim states that it is figured on the coins of Scione in Macedonia ; Halonnesus, an island of Thessalia ; Cassope in Epirus ; Leucas in Acarnania ; Seriphus and Siphnus, islands in the ^Egean Sea ; Antioch in Caria ; Side in Pamphylia ; and on uncertain coins of Cilicia, all of date B.C. It is a question whether the Pleiades derived their name from the doves direct, TreXetaSe?, the virgin companions of Artemis, who with their mother Pleione, when pursued by Orion in Bceotia, were rescued, changed into doves, and put in the heavens; or from the word 7rXetz>, to sail, the most favour- able season for setting sail being supposed to be the time of the heliacal rising of these stars. But there is no doubt that omens were taken from doves at the setting out on a voyage, and that the two accounts are reconciled by a third, namely, that these stars came to be called doves from the coincidence of their rising and the seasons esteemed most favourable for taking such auguries, and for setting sail. It is unnecessary, however, to found on the doves being a constellation ; as, whether they were or not, there is abundant evidence that the dove was a deity. The cultus is treated of at some length in Selden's De Diis Syris? and at great length in Bryant's work, the dove being very important to the Arkite scheme of that writer. 8 That there were persons called TreXetaSe?, or doves, in various places, is agreed upon. They were said to have been the most ancient prophetesses at Dodona, and also at Thebes ; and indeed the oracles at Dodona and in Libya were founded 1 Tom. i. p. 168. 2 Ed. Lipsise, 1672 ; Syntagma ii. cap. 3, De Dagone. 3 Vol. ii. pp. 281 et seq. APPENDIX 545 by two doves that came from Thebes. Herodotus' account of these black pigeons that flew from Egypt, and settled the one at Dodona and the other in Libya, is familiar. He states that, according to the priestesses of Dodona, the pigeon that arrived there spoke from a beech-tree in a human v.oice, directing a temple to be founded to Zeus ; but that the priests of Thebes, on the other hand, assigned the founding of Dodona to one of two of their sacred women who had been carried off by Phoe- nicians. 1 These women were called doves, as being ministers (says Bryant) to the dove-god. It is thus he explains the several narratives of women being, like the daughters of Anius, turned into doves. 2 They became priestesses. It seems certain that in some temples the deity had no representation but the dove. He was in the shape of that bird. Athenseus states that Zeus was changed into a pigeon, and this notion prevailed in Achaia, and particularly at ^Egium. 8 It was not merely Zeus, however, to whom doves were "ministers." They were sacred to Venus. "Ejusdem Deae quern admodum ministrse habitse fuerint, docet optime historia ilia de Columbis circa Erycem Montem in Sicilia volitantibus et diebus quas avaywyas et /caraycoyia nominabant incolae." 4 A dove, also, was the sole emblem of Semiramis, who was worshipped as a deity. 5 Selden quotes Johannes Drusius as follows : " Samaritanus circumcidit in nomine imaginis Colum- bam referentis, quam inventam in vertice Montis Garizim certo quodam ritu colunt ; " and says, " Aliam quam Semiramidis figuram hie non intelligo ; cujus etiam nomen Syris seu Baby- loniis Columlam Montanam denotare volunt nonnulli." The legend was that, on her death, Semiramis was changed into a dove, and under that form got divine honours ; 6 but Bryant, we think, is right in maintaining that she never existed, and that her title Samarim, or Semiramis, was a stock name. He says that it belonged to the Babylonians, and to all others as 1 Herod, ii. 54, 55. 2 Ovid, Met. xiii. 674. 3 Athen. ix. 51. 4 Selden, I.e. p. 274 ; Athenaeus, I.e. ; .ZElian, Var. Hist. i. 15. This temple of Venus at Eryx was celebrated. 6 Diodor. Sic. ii. 20. 6 Diodor. Sic. I. c. ; Athenagor. Leg. pro Chr. 26. 2 N 546 STUDIES IN ANCIENT HISTORY well who acknowledged Semiramis, the dove, and took it as their national insigne, i.e. totem. That the Babylonians did this seems to be well made out. One of the gates of their city, Herodotus mentions, was called Semiramis. 1 The Baby- lonians, according to Bryant (and Selden vouches that many have taken that view), were also called lonim, or children of the dove ; and their city lonah, the dove being the national ensign, and depicted on the military standard. Hence (says Bryant) the prophet Jeremiah, speaking of the land of Israel being laid waste by the Babylonians, mentions the latter by the name of lonah, which passage is rendered in the Vulgate, facta est terra eorum in desolationem a facie irce Golumbce. In another place the prophet foretells that the Jews should take advantage of the invasion of Babylonia, and retire to their own land, and he puts these words into the mouths of the people at that season : " Arm, and let us go again to our own people, and to the land of our nativity, from the oppressing sword." But the word sword, here is lonah, and [the passage] signifies from the oppression of the Dove the tyranny of the lonim. It is accordingly rendered in the Vulgate a facie gladii Golumbce. The like occurs in the 50th chapter of the same prophet. 2 Worshippers of the dove existed in Chaldea, among other districts in Babylonia. The Samaritans worshipped it, as the Jews alleged, and had a representation of it in Mount Gerizim, already noticed in a passage cited from Seldeu. The Assyrians worshipped it ; Sio Kal row? 'Acrcrup/ou? rrjv Trepiarepav rifjidv &><? Oeov, says Diodorus, 3 on which passage Bryant remarks, " It was, we find, worshipped as a deity." The worship prevailed in Syria, about Emesa and Hierapolis, 4 and " there were Sama- rim in those parts," says Bryant. The dove, in fact, was very generally received was almost as great a god as the serpent. Pausanias 5 mentions that JEsculapius, when exposed as a child, was preserved by a dove, which thus appears fostering a Naga. It became an emblem with the Hebrews, and is still, as every one knows, a symbol of the Holy Ghost, who once appeared in its shape. We have seen, however, that it was a reality long before it became a Christian symbol. To put this beyond 1 Herod, iii. 155. 2 Vol. ii. pp. 299 ff. 3 ii. 20. 4 Lucian de Syria Dea, 14, 54. 5 viii. 25. APPENDIX 547 doubt we must cite Clemens Alexandrinus, 1 who says its wor- ship was the basest idolatry, remarking that the people styled Syro-Phoenicians reverenced, some of them doves, others fish, as zealously as the people of Elis worshipped Zeus. 2 Xenophon, long before, noticed that in those parts divine honours were paid to doves. 3 Diodorus says the worship was universal in Syria. 4 It was most marked at Ascalon and Hierapolis, as we know on the authority of Philo Judseus 5 and Lucian, both of whom attest that the veneration of the people extended to the living bird. Lucian relates of the people of the latter city that the pigeon was the only bird they never tasted, as it was held by them to be particularly sacred. 6 We must believe it was so regarded by the Babylonians, who were named from it, and counted themselves to be its offspring; and we must believe that there were tribes elsewhere than in Babylonia that took its name and claimed the like descent. 8. The Earn. The Earn is in the heavens as Aries. It appears on the coins of many cities, as Panormus in Sicilia ; Perinthus, Hephsestia, and Samothrace, in Thracia ; Halonnesus, an island of Thessalia; Issa, an island of Illyria; Phea in Elis ; Cranii and Same, in Cephallenia ; Clazornense in Ionia ; of uncertain cities of Cilicia ; Antioch in Seleucis ; Damascus in Ccelesyria ; and of some other towns in Africa. These coins are all of date B.C. A coin of Panormus having the ram is very remarkable and suggestive. It is figured in the Numis- mata Spanhemii, torn. i. p. 204, along with the Yoni, at which it is staring. The ram was sacred to Jupiter Ammon, and probably had, at the Libyan oracle, a position not inferior to that of the Dove at Dodona. The story is, that Jupiter, in the form of a ram a ram incarnation relieved Hercules, or Bacchus, and his army when they were in straits, from thirst, in the deserts of Africa, who, out of gratitude, erected a temple to the god, re- 1 What follows here is abridged from Bryant, vol. ii. p. 312. 2 Clem. Alex. Cohort, p. 12, ed. Sylburg. 3 Xen. Anab. i. 4, 9. 4 Diod. Sic. ii. 4. 5 Apud Euseb. Prcep. Evang. viii. 14. 6 Lucian de Syria Dea, 14. 548 STUDIES IN ANCIENT HISTOR Y presented with the horns of a ram. 1 There were some three hundred Jupiters, as we know, 2 and if one of them got a place in a group in which the sheep stock was dominant, it would be a small tribute to the totem of the dominant tribe to give Jupiter ram's horns. So, where a horse or bull tribe was dominant, he might reasonably be Hippius or Taureus, and have, say, the head of a horse or bull, or some other element of the one or the other in his composition. 3 There are the usual stories indicating that there had been supernatural, if not divine, ram-beings. In the fable of Phrixus a ram with a golden fleece rescued the son and daughter of Athamas from their stepmother, Ino, carrying them through the air. This ram was said to be the offspring of Poseidon and Theophane. 4 The lady being changed into a sheep, the god took the form of a ram to woo her in. 5 The offspring of 1 Serv. ad <JEn. iv. 196. The god Ammon of Thebes was ram-headed.' 1 See Kenrick's Egypt of Herodotus, p. 44, and the note, p. 67, on the ram-sphinxes of Karnak. See also Lord Herbert of Cherbury's Religion of the Gentiles, p. 45, where the Ram-god is identified with the sun. His worshippers would not eat mutton ! 2 Tertull. Apol. 14. 3 We saw in America a considerable number of Suns and Sun-tribes, and we remember the policy of the Incas. There were far more Zeuses in Greek legend than Suns in America. Take the story of Endymion as handled in Miiller's Chips (vol. ii. p. 78). Endymion is son of Zeus and also of Aethlius, king of Elis an Inca who is, of course, himself a son of Zeus. Many cases resemble this. "The same custom," i.e. of taking the Sun for father (or, as we say, totem), says Miiller, ' ' prevailed in India, and gave rise to the two great royal families of Ancient India the so-called Solar and the Lunar races." 4 Hygin. Fab. 188. 5 Incarnations of gods in animal forms for such a purpose as we have here are feigned in many mythologies. Perhaps the most curious instance of the fiction is that which occurs in the Brihad-Aranyaka Upanishad (Muir's Texts, vol. i. pp. 24 ff.), where Parusha (the Procreator) having divided into male and female parts, the following incidents occurred. " He cohabited with her (i.e. his female division). From them MEN were born. She reflected, ' How does he, after having produced me from himself, cohabit with me ? Ah ! let me disappear ! ' She became a cow and the other a bull, and he cohabited with her. From them kine were produced. The one became a mare, and the other a stallion ; the one a she-ass, the other a male ass. He cohabited with her, etc. etc. The one became a she-goat, etc. etc." The speculation as to the origin of the different 1 Herod, ii. 42, etc. APPENDIX 549 the connection was thereafter by the gods gifted to Athamas, the father of Phrixus, as a reward for his piety. The recovery of the golden fleece from Colchis, as every one knows, was the object of the Argonautic expedition, an expedition of the most famous sort, ranking even with the hunt of the Calydonian Boar. These legends are intelligible if we conceive that there was a sheep-tribe, and an idol of the ram, believed to be a god, and an object of worship, that was stolen and sought to be recovered and restored to its shrine. The reader will remem- ber the Golden Fleeced Llama in the Temple of the Incas. Evidence, beyond what lies in these facts and legends, that the living animal was religiously regarded, we have none, except a few Vedic facts, and the fact that Sheep were wor- shipped in Egypt. There were numerous tribes of men in Egypt a land on which many races impinged ; and, in our view, we have in that an explanation of the multiplicity of the forms in Egypt of animal and vegetable worship. It was not that all Egyptians worshipped every creature, from bulls to beetles, and crocodiles to cats ; but that there were certain of them presumably of distinct tribes, gentes, or stocks, to whom one or other of the animals was sacred, and the others detestable. This is borne out by what Cunseus says (De Rep. He~b. lib. i. c. 4), as quoted by Lewis, in the close of the third volume of the Origines Hebrcece, in explanation of the saying that every shepherd was an abomination to the Egyptians. " That nation," he says of the Egyptians, " who reverenced, some sheep, some goats, some other four-footed beasts; being persuaded there was in them something of divinity." It is more forcibly borne out by what is stated by Wilkinson. " It frequently happened," he says, " in the worship of the sacred animals, that those which were adored in some parts of Egypt were abhorred and treated as the enemies of mankind in other provinces, deadly conflicts occasionally resulting from this worship and detestation of the same animal." : This is quite intelligible on the hypothesis that the animal gods were tribal, species of animals here contained is in several respects more primitive than that of the Khonds on the same subject, as given by Major MTherson. 1 Ancient Egyptians, vol. iv. p. 159. 550 STUDIES IN ANCIENT HISTOR Y or, more probably, gentile, i.e. totem -gods ; but how is it explicable on the supposition that they were emblems ? 9. The Goat. The Goat is in the heavens as Capricornus, and figures on many ancient coins, all of date B.C. ; on coins of Thermae, or Himera, in Sicilia ; JEgospotami in Chersonesus, Thracia; JEnus in Thracia; Macedonia in genere; Issa, an island of Illyria ; Pharos in Illyria ; JSgira in Achaia ; Elyrus in Greta; Syrus Insula; Antandrus in Mysia; Parium in Mysia ; JEgae in ^Eolis ; Ephesus in Ionia"; .ZEgse in Cilicia ; Cyzicus in Mysia ; Augusta in Cilicia ; Tralles in Lydia ; Commagene in genere. It appears on two British coins figured in Mr. Evans's book, and on some coins in the Gaulish series. 1 There is no doubt that the goat was a god, as the reader will find who consults any classical dictionary, art. " Pan." 2 The readiest to the present writer's hand is Lempriere, who has the following : " In Egypt, in the town of Mendes, which word signifies a goat, there was a sacred goat kept with the most religious sanctity. The death of this animal was always attended with the greatest solemnities, and like that of another, Apis, became the cause of a universal mourning." l Pan himself had a body compounded of the human and goat forms was a goat-being of the same order of beings as the Minotaur, Sphinx, Hippa, and others we have seen. Fable represented him as the offspring of various deities Mercury 4 and Jupiter 5 in particular. He took the complete form of a goat on some occasions, as once to woo Diana. 6 What form had she ? He was alive in the time of the wars with the giants, and when the gods fled from their enemies to Egypt he assumed the form of a goat, and they all immediately followed his example ! 7 The particular goat whom fable put in the heavens was Amalthsea, the daughter of a king of Crete, who fed Jupiter with goat's milk when he was a child. 8 So there 1 Evans, I.e. p. 114. 2 See also art. " Lupercalia. " 3 Herod, ii. 46. 4 Horn. Hymn to Pan, etc. 5 Apollod. Bill. i. 4, etc. 6 Philarg. ad Oeorgic. in. 391. 7 Hygin. Fab. 196. 8 Lactant. i. 21 ; Hygin. Poet. Astron. ii. 13 ; Callim, Hymn Jov. 49. APPENDIX 551 was a lady who was yet a goat, and a king, who was her father, in Crete when Jupiter was a baby. The goat was no doubt a totem-god long before Jupiter was thought of. We saw in Egypt a town named from the goat. Were there tribes named from it also ? It was also a stock name in India. Lewis, in his Origines (vol. iii. p. 21), points out that the Hebrews used to offer sacrifices to Seirim, who were demons in the form of goats. His explanation is that they did so in imitation of the ancient Zabii. " It seems more reasonable," he says than another hypothesis, which need not be cited here "to believe the old Hebrews worshipped the Demons adored by the ancient Zabii, who appeared in the shape of goats ; and this practice was universally spread in the time of Moses, which occasions that this kind of idolatry was so strictly for- bidden in his injunctions." In the Olympus of Mohammed are seven regions, and above the seventh, eight angels in the shape of goats. On their backs stands the throne of god. 1 10. The Fishes. The fishes may be rapidly disposed of. They are in the heavens, and very common on coins. They were worshipped in most places where doves were, as among the Syrians, Babylonians, and Phoenicians. In Egypt the fish had a prominent place in connection with Isis, who was figured with it on her head. The fishes in the heavens are spoken of by Hyginus as persons, and he quotes Erato- sthenes as saying that the fish was the father of mankind : " Eratosthenes ex eo pisce natos homines dicit." 2 The Phoenician god Dagon, also the Assyrian Oannes, was a man- fish, one of our familiar compounds. Dagon invented agricul- ture, of course, and many other arts, and was worshipped in many places. 3 Berosus, as quoted by Eusebius, 4 says Oannes had the body of a fish, and below the fish-head, placed upon the body, a human head coming out under the other. He had a man's feet coming out under the tail, and a human voice. 1 MishMtu 'l-Masabih, vol. ii. p. 652. 2 Hyginus, Poet. Astron. ii. 30. 3 Euseb. Prcep. Evang. i. 10. 4 Lewis, Origines, iii. 81. 552 STUDIES IN ANCIENT HISTOR Y He used to come every morning out of the sea to Babylon to teach the arts and sciences, returning to the sea in the evening. Derceto was another such compound a woman to the waist, for the rest a fish. 1 According to some she was human in the face only. 2 She was a Syrian goddess, and the Syrians, according to Diodorus Siculus, would eat no fishes, " but they worshipped fishes as gods." There is a story in explanation of this, to the effect that, ashamed of an indiscretion, the goddess plunged into a lake near Ascalon, where she had a temple, and became a fish. 3 Ovid calls her Dione, and "gives a somewhat different history of the plunge. He represents her as received in the water by two fishes, which afterwards became the Pisces of the heavens. 4 Fish were sacred to Venus. 5 A con- siderable variety of fishes are figured on ancient coins, the cetus and dolphin being the most frequent. We have no list of any number of them, but a few are figured in the Numis- mata Spanhemii. A variety of them will be found figured at p. 339 of vol. iii. of Mr. Campbell's Celtic Tales, being "all the fish figured on the sculptured stones of Scotland." Fish, in Mr. Campbell's opinion, " clearly have to do with Celtic mythology." We have seen fishes giving stock names to tribes of men now existing, and can understand how, having been totems, they should have become gods to the tribes that had them in that character. Of course in Bryant's system the fish is the Ark, while Dagon, Oannes, etc., are the Patriarch Noah. As to one fish we are able, thanks to Plutarch, to put his totemship beyond doubt. " The Egyptians in general," says that writer, " do not abstain from all sorts of sea-fish ; but some from one sort and some from another. Thus, for instance, the inhabitants of Oxyrynchus [Piketown] will not touch any that is taken with an angle : for as they pay an especial rever- ence to the pike, FROM WHENCE THEY BORROW THEIR NAME [i.e. they are Pikes], they are afraid lest perhaps the hook may be defiled by having been some time or other employed in 1 Lucian, de Syria Dea, 14. 2 Diodor. Sic. ii. 4. 3 Diodor. Sic. I.e. 4 Ovid, Fast. ii. 461. 5 Athenseus, vii. 18. APPENDIX 553 catching their favourite fish. The people of Syrene, in like manner, abstain from the Phagrus, or sea-bream." Can any one doubt that in Oxyrynchus there was a Pike-tribe ? l 11. The Bear. The Bear is in the heavens as Ursa Major and Ursa Minor, the former distinguished as early as the time of Homer by the name of Arktos. 2 He occurs on various Gaulish coins ; on coins of Urso in Spain, and on a coin of Orgetorix, chief of the Helvetii. 3 He probably occurs on other coins, but we have no list of them. 4 The constellation connects itself with the names of Callisto and Areas. Callisto was changed into a bear for a fault com- mitted with Jupiter, of which Areas was the fruit. Jupiter, to atone for the metamorphosis, made her a constellation along with her son. 5 This Areas, of the bear stock, reigned in Pelasgia, which from him took the name Arcadia. He taught the people agriculture, of course, and other arts, e.g. the spinning of wool. 6 The Greek name for the constellation enters into Arcturus, and there was another star near the Bear, called Arctophylax. The island of Cyzicus was called Arcton, and the Arctanes were a tribe of Epirus. 7 The suggestion is, that the bear gave its name to a stock, and was a god; that there were bear-tribes in Arcadia once as there are bear-tribes now in America. 8 The bear, as a god, probably became, in most places, obsolete very early, having no special claim to a place in the Keligion of the Life-powers the first great speculative faith that supervened on the primitive animal and vegetable worship, and with which most of the other animals we have been con- sidering undoubtedly came to be connected. It is curious that 1 Isis and Osiris, 7. The word translated pike is given in Liddell and Scott as meaning a species of sturgeon. 2 Iliad, xviii. 487 ; Od. v. 273. 3 Bevue Numismatique, 1860. Plate IV. 4 Mr. Sim's note is, " Bears are only to be found on uncertain coins of Gallia. Some of these have the wild boar on the obverse. Some have two bears. They are all earlier than Julius Caesar." 5 Ovid, Met. ii. 476-507. 6 Pausan. viii. 4. 7 Steph. Byz. 8 We have the bear as an object of worship in Athens, with a strange history in explanation of the fact. See Suidas, s.v. "Ap/crot. 554 STUDIES IN ANCIENT HISTOR Y we have him as an eponymous progenitor elsewhere than in Arcadia. For example, a bear was the progenitor of the kings of Denmark. In Olaus Magnus' History, 1 it is gravely related how this came about, the narrative being quoted, " ex historia charissimi ante-cessoris," of the author the Archbishop of Upsala. It opens thus : " Cujusdarn patrisfamilias in agro Suetico filiam, liberalis formse, cum ancillulis lusum egressam, eximise granditatis ursus, deturbatis comitibus, complexus rapuit." The lady, being carried off by the bear, had by him a son. " Ut ergo duplicis materiae benit/na artifex natura nuptiarum deformitatem seminis aptitudine coloraret, genera- tionis monstrum usitato partu edidit." She gave him his father's name. His grandson begat Ulfo, "a quo Hex Sueno et csetera Danorum Eegum stemmata, ceu quodam derivata principio, longo successionis ordine (teste Saxone) profluxerunt. Quomodo autem similes partus judicabuntur, August, de Civ. Dei plurima dicit de siinili propagine, utri sexui rnagis sit attribuenda." On which Olaus Magnus piously remarks, " Crediderim ego id a vindice Deo effecturn, ut Dani, qui de sanguinis nobilitate plus nimio gloriantur, Suetiamque fre- quentius quam felicius impugnare consueverant, Eegibus a fera Suetica genitis obnixos vertices inclinare cogerentur. Quam acer autem hie Ulpho Sprachaleg Suecus ursi nepos fuerit in bello, quam etiam astuti et vafri ingenii supra videre licet!" Joannes Scheffer mentions as one of the primitive gods of the Lapps, " Hyse," whose function it was " lupis et ursis imperare." Whether this king of the wolves and bears was a wolf or a bear, and what was his nature or functions, does not appear. In Scheffer's chapter, De Sacris Magicis et Magia Lapponum, we find the Bear on the Magic Tympanum along with Thor, Christ, the Sun, and the Serpent who were gods to them and some other animals, e.g. the wolf and reindeer ; and in his chapter on the wild beasts of the country, he tells us they call the bear the lord of the woods, "vocant eum dominum sylvarum," which is explained to mean that he is " herus omnium animalium reliquorum " ; so that the king of 1 P. 702, ed. Basilese, lib. xviii. c. 30. APPENDIX 555 the wolves and bears might well be a bear, and could not well be a wolf. 1 There is no clear evidence, however, of the worship of the animal by the Lapps. 12. The Crab. The crab stands next. He is in the heavens as Cancer, and on the coins of Cumse in Campania, Butuntum in Apulia, Bruttium in genere, Crotona in Bruttium, Terina in Bruttium, Agrigentum in Sicilia, Eryx in Sicilia, Himera in Sicilia, Panormus in Sicilia, Priapus in Mysia, Cos, island of Caria. The dates of these coins range from 300 B.C. to 100 B.C. There are probably many others having the crab. We do not know much of him in mythology ; but we saw him as a god now worshipped by a tribe in Fiji. The reason assigned for putting him in the heavens is of an intenser degree of silliness than that usually given for so promoting an animal. When Hercules was attacking the Hydra the many- headed Naga "Juno, jealous of his glory, sent a sea-crab to bite his foot. This new enemy was soon despatched, and Juno, unable to succeed in her attempt to lessen the fame of Hercules, placed the crab among the constellations, where if; now bears the name of Cancer." 2 It will be admitted that this story, read literally, is quite ridiculous. If we take Hercules to stand for a tribe the Heraclidae (what does this name mean etymologically ?), the Hydra for a Serpent-tribe or nation, and the sea-crab for a Crab-tribe, the story becomes intelligible. The Crabs, having come to the relief of the Serpents, when attacked by the Heraclidae, were defeated along with their allies. The introduction of Juno into the legend probably was of late date, and had for its object to explain why Cancer was a constellation a fact that would cease to be easily accounted for when, as a totem-god, the crab had become obscure or obsolete. 1 Schefferi Lapponia, ed. Frankofurti, 1673, pp. 59, 125, and 336. There was a wolf-man in Arcadia (and he was worshipped), namely, Lycaon, as well as a bear-man, who was king of the country ; and Pan's Greek name was Lyceus, from Xtf/cos, a wolf. 2 Article " Hydra" in Lempriere's Dictionary. This account is substantially the same with that given in the most recent Encyclopaedia. Apollodor. ii. 5 ; Hygin. Poet. Astron. ii. 23. 556 STUDIES IN ANCIENT HISTORY We have now examined the list of animals set down for consideration, excepting the Aselli the little asses, and them we must pass over, as they would take much space, and there are more important animals to attend to. The Jews said the Samaritans worshipped the Ass, and the Samaritans said the Jews worshipped it. The Komans and others joined chorus with the Samaritans. The reader who is curious on this subject, will find in Kitto's Encyclopaedia the edition before last sub voce "Ass," some guidance in his inquiries. The story Tacitus gives is well known, 1 and so is the contro- versy between Josephus and Apion as to whether the Jews had the cultus. Some light on the subject is thrown by the book of Zacharias in the Apocryphal New Testament, and in Hallam's Middle Ages, in which there are accounts of the Asinarii and the Festival of the Ass. The reader will recall the asses that helped Bacchus, 2 and that in Egypt the ass was the symbol of Typhon. 3 We may be pretty sure he was the totem of some tribes of men who were of importance, other- wise he would not have been promoted to the heavens. He furnished a stock name to the Arabs. 4 It would be out of place, even were we able to do it, to attempt to exhaust the subject in an article of this descrip- tion. There are two creatures, however, which it is as well we should notice before going on with our argument. They are the Bee and the Eagle. It is pretty certain, we think, that both of them were totems promoted to be gods. 13. The Bee. There was a goddess Melitta, or Melissa, who was represented by a bee, and there were tribes named after her, " Melittse," or " Melissse," that is, Bees. "The Grecians," says Bryant, "have sadly confounded the histories where they are mentioned by interpreting the Melissse Bees' 1 Tac. Hist. v. 3, 4. 2 Hygin. I.e. 3 Wilkinson, Ancient Egyptians, vol. iii. p. 124. 4 See MisMcdtu'l-MasaWi, vol. ii. p. 93, footnote respecting Himar, or the ass (that is, he was surnamed ass), " the last Khalifah of the dynasty of Ommiah." The ass was here in the royal line. APPENDIX 557 He admits the bee, however, to have been the hieroglyphic of Melissa. "It is to be found as a sacred and provincial emblem upon coins which were struck at places where she was worshipped (the italics are ours). But the Greeks did not properly distinguish between the original and the substitute, and from thence the mistake arose." The Greeks, we submit, knew quite well what they meant, and it is the moderns who should be reflected upon for misunderstanding them. They called them bees, as we, in referring to American tribes, would speak of bears, wolves, and eagles ; and the bee that had originally been a totem had become a totem-goddess. The following passage, from Bryant, we submit, is almost perfectly sensible when read in the light of our hypothesis : Philostratus mentions that, when the Athenians sent their first colony to Ionia, the Muses led the way in the form of Bees. And Herodotus says that all the northern side of the Danube was occupied by Bees. Jove also, upon Mount Ida, was said to have been nourished by Bees. When the temple at Delphi was a second time erected it was built by Bees. 1 There was, we may conclude, not only a Bee-tribe, but there were gentes of the Bee stock spread over a vast tract of country, as they should be owing to incidents of the totem stage. What Bryant says of the lee coins shows the im- portance of the sort of evidence ancient coins furnish. We have the bee on ancient coins of Athens, whence Philostratus says bees set out ; on coins of Elyrus in Crete, where Melitta, daughter of a king Bee, lived, and helped the goat Amalthaea to nurture Jove ; 2 on the coins of Coressia and Julis, towns in the island of Ceos ; on the coins of Praesus in Crete ; of Sicinus ; of Ephesus in Ionia, whose coins also give the bee and half- stag ; of Cyon in Caria ; Tabse in Caria ; Elseusa, island of Cilicia, and of Acrasus, in Lydia. These coins are all of date B.C. 14. The Eagle. This bird could perhaps be made as much of as the serpent, horse, or bull. We must dispose of it in a few sentences. Bryant says it was the ensign of the Egyptians, who were named after it; but more probably the 1 Philostr. Icon. ii. 8 ; Herod, v. 10 ; Callim, Hymn to Jove, 50 ; Pausan. x. 5. 2 Lactant. i. 22. ** 558 STUDIES IN ANCIENT HISTOR Y dominant tribe only was so named. The eagle was Msroch, the god of Nineveh. It was also the symbol of the kings of Chaldsea. Of course it got to be compounded with the human form," to have two and three heads, and so on. Mr. Layard remarks of these compounds of the eagle, bull, and lion, as follows : " It is worthy of observation that wherever they (that is, the human -headed lions and bulls) are represented either in contest with the man or with the eagle-headed figure, they appear to be vanquished." And he adds, "I have already ventured to suggest the idea which these singular forms were intended to convey the union of the greatest intellectual and physical powers ; but certainly their position with reference to other symbolical figures would point to an inferiority (that is, of the lions and bulls) in the celestial hierarchy." 1 Of the emblem hypothesis we shall have some- thing to say presently. Meantime, it suffices, as regards the eagle, to find a tribe named from it, and that in one quartei it was a greater god than the horse or bull. Among the Jewish tribes (the later Jews, say) the eagle was the emble of the tribe of Dan, an ox of Ephraim, and the lion of th( tribe of Judah, 2 the lion here appearing as belonging to the 1 Nineveh, vol. ii. p. 460. 2 See Lewis, Origines ; chapter on coins. It is Aben Ezra, says Lewis, wh< gives the tradition which assigns the Lion, Ox, and Eagle to the tribes of Judal Ephraim, and Dan respectively, as ensigns. In the Douay Bible (2nd edition, edited by Haydock and Hamill. Two vols. Dublin. No date) the reader wil see on a plate at p. 180, vol. i. the ensigns of the tribes according to, at least some authority esteemed by Roman Catholics. The plate illustrates v. 2, cap. ii. of Numbers, where the ensigns and standards of the Hebrews are referred tc On Judah's standard is the Lion ; on Dan's, the Eagle ; on Napthali's, the Hin( or Hart ; on Benjamin's, the Wolf ; on Manasseh's, the Horse (or Ass) ; 01 Ephraim's, the Bull or Ox ; on Asher's, a Tree ; on Issachar's, the Sun and Moon and on Gad's, a cone on an altar the Assyrian Linga ! In Jacob's dying speech, Genesis xlix., to the eponymous progenitors of the tribes in whicl their fortunes are indicated, Judah is spoken of as "a lion's whelp " ; Issachai as "a strong ass" ; Dan as "a snake in the way" ; Benjamin as "a ravenoi wolf"; Napthali as "a hind (or hart) let loose"; and Joseph as "a fruitf bough." Compare our version with the Vulgate. The wolf, hind, and lion onl] are the same in the speech and in the plate of the Douay Bible. In connectioi with the subject of this note, Ezekiel x. vv. 8-22, is worth looking at, it beinj kept in view what the faces of the cherubims were. And see Seder Olam Rabba, p. 58 ; Trans. Chron. Institute of London, vol. ii. part ii. ; and Carpzov's Apparatus Historico-criticus, etc., Frankfurt and Leipzig, 1748. APPENDIX 559 dominant tribe. The Eoman eagle will occur to every one, and in Eome eagle was a gentile name. A great many places were named from the bird, notably Aquileia, known as Roma secunda. We must say no more of the eagle, however. It is everywhere. The coins having it belong to all places and dates, and are far too numerous for enumeration. .The list of animals that were totems among the ancients might be extended, by evidence of varying degrees of force, to comprise the tiger, wolf, cat, panther, elephant, 1 stag, boar, fox, rat, and rabbit ; the raven, hawk, and cock ; the ant, butterfly, and grasshopper all the creatures, in short, that figure in heraldry. Strange as it may seem, there is a Lord of tigers now, and he is a good god, as a totem should be. 2 The Bygahs or Jogees regard him so much that they won't eat him a poor compliment, but it is significant. In the Bygahs' mythology a milch tigress was foster-mother to the first man. " Coeval with the creation of the world were created one Naga Jogee, and his wife, Mussumat Naga. One day they went into the forest to dig for roots, and from the earth they dug up a boy- child, who was nursed for them, under the direction of Mahadeo, by a milch tigress." 3 Major M'Pherson, in his paper on the religion of the Khonds, says that people believe " natural tigers to kill game only to benefit men, who generally find it but partially devoured, and share it; while the tigers which kill men are either Tari (a goddess), who has assumed the form of a tiger for purposes of wrath, or men, who, by the aid of a god, have assumed the form of tigers, and are called 1 The elephant is a totem-god now in Burmah, where the king is styled " King of the Rising Sun, Lord of the Celestial Elephant, and Master of Many AVhite Elephants, and Great Chief of Righteousness." (There is a Rising Sun tribe among the Cheppeyans, in North America, Archce. Amer. vol. ii. p. 18.) It occurs with totem marks in the Catapatha Brdhmana, and is there identified with Vivasvat (the Sun), the son of Aditi (see Muir's Texts, vol. iv. p. 13) a sun -elephant corresponding to the sun-serpent of Peru. Elsewhere we have Gunesh, an elephant-headed divinity, "the mother of the universe," an object of worship at this day. Diary of a Pedestrian in Cashmere and Thibet, 1863, p. 311. 2 He is mentioned in Mr. Justice Campbell's Ethnology of India, p. 9. 3 "The Bygahs' Mythology," p. 52 of the Report of the Indian Ethnological Committee, 1866-67. Nagpore, 1868. 560 STUDIES IN ANCIENT HISTOR Y ' Mleepa Tigers.' " 1 The way in which the beneficent nature of the totem is here, by fictions, put beyond suspicion, is delightfully simple. 2 " Mleepa " or " Were " wolves are also common, as every one knows ; and it is equally familiar that the wolf has often been a foster-mother, as she was to Komulus and Eemus. The tiger and wolf are totems in America, as are several others on the list above given. It is altogether out of the question, however, to attempt to deal here with such a list. Enough has been said to prove that the most savage animal may be accepted by a tribe, of men as a totem, and be thereafter developed into a great and benign god. 3 1 Religion of the Khonds, p. 25. 2 "Mleepa Tigers," Du Chaillu states, are to be found in Africa. They also occur among the Arawaks, who call them "Kanaima Tigers." See Brett, I.e. p. 368. 3 We may here, in a footnote, dispose of a few facts which, indeed, are those that, now four years ago, suggested this inquiry, though the writer has been unable to work upon it till recently. The fact of Serpent and Bull tribes being known to exist, and to have existed, seemed to offer an explanation of the myth of Cadmus, at Thebes, and of the cow that led him thither. 1 On the same suggestion it occurred that there might have been a Snake-tribe in Rhodes. Phorbas obtained the supremacy by freeing the island of snakes. 2 The myth of the Ants and .ffigina next strengthened the suggestion of the presence of tribes with totems. The ants in the island were miraculously turned into men the /j,6p/jM)Ks into the Myrmidons Ants, that is, quite on the level of the Australian opossums. 3 Then occurred the Calydonian boar hunt there is something like it in the Celtic Tales, and in the Highlands, we have no doubt, inquiry will yet establish the totem stage. It seemed incredible that the slaughter of a boar should have employed the whole chivalry of Greece an army of warriors and that the feat should ever after rank among the proudest exploits of the nation. The question rose, Was there a Boar- tribe ? The Oracle enjoined Adrastus to give his daughters in marriage, one to a boar, and the other to a lion. 4 This was complied with by their marrying Tydeus and Polynices respectively ! Tydeus came from Calydon, and was son of (Eneus, king of the country. He was therefore possibly a boar, if the question above put was to be answered in the affirmative. Was Polynices, then, a lion, and was there a Lion-tribe ? As he was the son of (Edipus, from the land of the sphinx, it seemed not improb- able, on the totem view, that he might be a lion. And so the matter appeared worthy of investigation. The facts here stated will, we think, be felt to add force to those in the text. Most of them were first noted by the writer in 1866, as challenging such an inquiry as the present. Since this note was in type the writer's attention has been called to The 1 Pausan. ix. 12. 2 D iod> Sic> v _ 58> 3 Hes. Frag. 67. 4 Apollodor. iii. 6. APPENDIX 561 We must also dispose of the worship of plants in a summary manner. This matters the less that the worship of a considerable variety of them is established in Mr. Fergusson's recent publication on Tree and Serpent Worship} Among these we have the Pear-tree, Oak, Asclepias a creeping shrub (the Soma, a great Indian god), the Pipal, the Fig- tree, the Bela, the Tulsi plant, the Tamarisk, and the Elapatia and Talok trees. To this list we may add the Olive, Laurel, Lotus, Palm, Pomegranate, and Poppy. A spiritual idea of a tree we have in Yggdrasil. Some of these became great gods, and got a place in the religion of the Life-Powers. In one or two cases the legends that give us the earliest accounts of plant worship, give us also a primitive mother for the tribe having the worship and the suggestion of kinship through the mother only having existed in the tribe. Thus in the legend of Athens, which introduces the Olive, as we have it from Varro (apud August, de Civi. Dei, xviii. 9), we learn that " a double wonder " having appeared springing out of the earth namely, the Olive-tree and Water the Oracle declared the Olive to signify Athene, and the Water Poseidon, and that the citizens must choose from which of the two they would name their town. Men and women voted together, and the latter carried the honour for Athene by a majority. Poseidon was thereon enraged, and to appease him women were deprived, among other privileges, of that of having their children named Antiquities of Heraldry, by Mr. W. S. Ellis, which has recently been issued, and which propounds a view which, at first sight, seems to resemble that in these papers insisted on. Some of the points made, and not a few of the facts founded on, in the chapter devoted to the Heraldry of Mythology are the same as those here given. His view of the order, and even of the nature of the evolution, will be seen, however, on a close inspection, to differ essentially from that of the present writer. Had Mr. Ellis more fully studied the totem he might have anticipated what is here being said. 1 Mr. Fergusson's book is, in our opinion, apt to mislead in several respects. (1) The reader gets the impression from it that the worship of the serpent is an exceptional phenomenon ; i.e. that it has been singular among animals in being worshipped. (2) It gives the impression that there is a special connection between the serpent and tree. (3) Its title gives the impression that trees only were worshipped, whereas its contents prove the worship as well of small shrubs and plants. All this notwithstanding, it is a valuable book, and one of the most beautiful ever issued. 2 o 562 STUDIES IN ANCIENT HISTORY after them. So that anciently, the story bears, children in Athens took their names from their mothers, as they do among the Australians and American Indians. The case of the loxidse again gives us the suggestion of female supremacy in a legend which also informs us that " they reverenced as holy and worshipped," a certain marsh plant, which no doubt was their totem. 1 With these few observations on plant worship we must pass on to our argument. We shall consider first the explana- tions that have been offered of divine honours being paid to such beasts as the serpent and lion, and to trees, etc. ; and after showing that they are unsatisfactory, we shall proceed to consider the weight of the evidence direct and indirect that goes to show the soundness of our own hypothesis. 1. The Emblem Hypothesis. Suppose we knew that all men were, as Bryant believed, derived from one family since 2348 B.C. the date of the Deluge that writer's Arkite system would still be worth nothing, either as an explanation of animal worship, or as evidence of the Deluge having occurred. He does not pretend to include nearly all the animals or plants that have been worshipped in his list of Arkite emblems ; and, accordingly, to give a reasonable colour to his hypothesis that there had been any Arkite emblems at all that had degenerated into gods, he ought to have excluded the possibility of those he includes having become gods through the operation of such causes as led to the worship of the others. Such causes, what- ever they were, being admitted to have been in operation, will explain all the cases ; and before an hypothesis of special causes in some cases can be entertained, the operation of the general causes as regards them must be shown either to be insufficient or to be excluded. This, however, Bryant has not attempted, or even thought of attempting, to show ; and, the] fore, even could we make the necessary supposition as to the history of human tribes, we must still conclude that this learned and, in a confused sort of way, ingenious man has succeeded in nothing not even in setting up a respectable 1 Plutarch, Theseus, chap. viii. APPENDIX 563 hypothesis. It is simply impossible, however, with our modern information the history of several nations having been carried beyond the point of time assigned to Noah and his family to make such a supposition as Bryant requires to set out with. Moreover his system demands not one, but a series of hypotheses, to support it, and they are all bad. (1) There is the hypothesis that the animals had been emblems. This is bad, as we have shown. (2) There is the hypothesis that the emblems degenerated into gods. This is not supported by one instance adduced of such degeneration having, histori- cally, taken place, or even by a fair analysis of the probable steps through which it could have happened. (3) There is the hypothesis that through the idolatry of some one animal of a species thus induced, a religious regard came to be extended to the species. This is subject to the same remark as we have made on the preceding hypothesis. The far- fetching processes by which even a poor appearance of a case has been made for the emblems as at all probable, we need not remark upon. At the same time, as we have amply acknowledged, we have profited much by Bryant's researches at one point. It was necessary in his scheme, as in ours, that it should be shown that the totems as we say ; the animal emblems, as he says were precedent to the gods of the mythologies. Another emblem hypothesis represents each animal as, in some way not now to be understood, typical of the nature of some one or other of the gods. This again is a fanciful ex- planation surrounded by the same sort of difficulties. How came men to think of taking animals and plants to represent their gods ? We can understand the selection only when we conceive their gods as spiritual ideals of animals or plants. Besides, the hypothesis assumes the deities as existing before the animal gods, and this is contrary to the evidence. And why should the selection of an animal to be the type of a god render its species sacred ? "We do not religiously regard the pigeon, though the dove is one of our most mysterious symbols. We can understand, on the other hand, how it decayed into a symbol, knowing it to have been a god that had grown obsolete. 564 STUDIES IN ANCIENT HISTORY The fish is a Christian symbol ; but we have not a religious regard for fishes. When the fish-god was a power, however, his worshippers religiously regarded the finny tribe. They would not eat them. It has been true of these as of most symbols ; facts come first, and symbols are facts in decad- ence. There is yet another form of the emblem hypothesis. It is that mentioned in a passage cited from Mr. Layard, and which, almost in a sentence, that author states and abandons. This is the hypothesis that the compounds of various animal and human forms " were intended to convey the union of the greatest intellectual and physical powers." This altogether fails to touch the fact of the real worship of living animals. Moreover, as an explanation of the compounds it is untenable. It simply won't hold of the Naga compounds. They are not intended to convey anything of the sort. Will it hold of the dog compounds ? As to the bull, lion, and eagle compounds, we saw Mr. Layard's opinion to be that it will not hold ; the evidence showing the creatures to have a place, and to be subordinated to one another in the celestial hierarchy. The fact is, though we now make use of lions, sphinxes, and so on, to convey such ideas as he refers to, we demonstrate in doing so only the poverty of the modern imagination and the feeble- ness of our art instincts ; inasmuch as being incapable of inventions, we mimic old forms derived from the religious faiths of long past and misunderstood generations. While no cases are producible in support of the emblem hypothesis of animals regarded as emblems merely, or illus- trating their transition from being emblems to being themselves objects of adoration, we are not without cases to show that the animal-gods were prolongations of the totems. We have such a case, for example, in Peru. The Peruvians, according to Acosta, worshipped the sun, moon, planets, and stars ; fountains and rivers ; rocks, great stones, hills, and mountains ; land (Tellus) and sea (Poseidon) all these objects being regarded as persons. They worshipped Thunder, believing him to be a man in the heavens with a sling and mace ! Of lesser objects APPENDIX 565 on earth, he tells us, they worshipped fruits and roots, some small stones, and the metals ; while among the animals they worshipped he makes special mention of the bear, lion, tiger, and snake. Now we are able from this author to see what were the speculations of a people in the stage in which, having animals as gods on earth, they also worshipped stars in heaven. Of his account of star-worship in Peru, we cite the following version from Lord Herbert of Cherbury : " They particularly adored that constellation which we call Cabrillas, or the goat, and they Colca ; and commanded that such offerings should be made to some stars, and such to others, those being particularly worshipped according as every one's necessity required. The Opisons adored the star Urchuchilly, feigning it to be a Ram of divers colours, who only took care of the preservation of cattle ; and it is thought to be the same which the astrologers call Lyra. Besides these two, they worshipped two others that are near them, and say that one of them is a Sheep and the other a Lamb. There are some who adore another star that ruled over the Serpents and Adders, from which they promised safety to themselves; others who worshipped the star called the Tiger, who they believed to preside over tigers, lions, and bears. They were of opinion that there was not any least or bird upon the earth WHOSE SHAPE OR IMAGE DID NOT SHINE IN THE HEAVENS, by whose influence its similitude was generated on the earth, and its species increased." * Thus we see that the beings in the stars were believed to have the animal forms, and to be powers in the celestial hierarchy. This case proves (1) a connection, such as we have been endeavouring to trace, to have existed between the worship of animals and the nomenclature of the heavens ; (2) that the celestial beings were conceived to be in the shape of the animals, and to have special relations to their breed on earth ; and (3) while it indicates the persistence of tribal preferences for particular stars as animal gods, it shows the process to have been in operation by which, on the consolidation of the political system, the divine functions are distributed among 1 Acosta, Histoire Naturelle, Paris, 1600, pp. 214, 217 (lib. v. chaps. 4 and 5) ; Herbert's Religion of the Gentiles, 1705, p. 86. 566 STUDIES IN ANCIENT HISTOR Y the tribal, or rather we should say gentile, gods of group. 1 Now of two things one. Either the Peruvians, as some maintain, independently achieved the civilisation they had starting from the totem stage in which their neighbours remained, or their civilisation, including the religious doctrines were derived by them from some one or other of those nations we call the ancient. On the former view, of course, the anima gods are the prolongations of the totems ; on the latter we have, in the case of the Peruvians, a reflection of the religious system of some one or other of the ancient nations. So that on the least favourable of the alternatives we have the fact that in some one at least of the ancient nations that wor- shipped animals and they all did the animals were not emblems, but the exact images of the gods. What was true in one case, the presumption is, was true in all. That is to say, there are not only no facts to support the emblem hypo- thesis in any of its forms, but the presumption derivable from the facts we have is against that hypothesis. 2. Mr. Fergusson's Explanations. So much for the emblem hypothesis. There is no other that we know of except in the special case of the serpent and tree, in regard to which views have been put forward by Mr. Fergusson. Tree worship he conceives to have sprung from a perception of the beauty and utility of trees. "With all their poetry and all their useful- ness," he says, " we can hardly feel astonished that the primi- tive races of mankind should have considered trees as the choicest gifts of the gods to men, and should have believed that their spirits still delighted to dwell among their branches, or spoke oracles through the rustling of their leaves." Of this it suffices to say, it does not at all meet the case of the shrubs, creepers, marsh -plants, and weeds, that have been 1 "We have seen in numerous cases the disposition of the tribesmen to identify their totem with the sun. It is highly probable that the identification of the totems with particular stars conceived as the sun's inferiors is, like the distri- bution of functions, a late phenomenon, posterior, that is to say, to the settled co-ordination of the tribes in the political system. APPENDIX 567 worshipped, and is obviously not the key to the mysteries of plant worship. , His account of the origin of serpent worship is, if possible, even more unsatisfactory. He ascribes it to the terror with which the serpent inspired men ; to the perception of his remarkable nature, the ease and swiftness of his motions, and his powers of quickly dealing death by sudden spring or mysterious deadly poison. To this the objection is, that the serpent religion is not a religion of fear but of love. The serpent, like the tiger and bull, is a benign god. He is a protector, teacher, and father. How came a religion beginning in terror to be transformed into a religion of love ? ' The terror hypothesis will, we submit, not meet the case even of the serpent. Arid no such hypothesis, it is obvious, can be extended to cover the run of cases to explain the worship, say, of the dog, the dove, or the bee. The hypothesis we put forward starts from a basis of ascertained facts. It is not an hypothesis explanatory of the origin of totemism, be it remembered, but an hypothesis explanatory of the animal and plant worship of the ancient ,- nations. It is quite intelligible that animal worship growing from the religious regard for the totem or kobong the friend and protector should, irrespective of the nature of the animal, be a religion of love. What we say is, our hypothesis explains the facts. It admits an endless variety of plants and animals to the pantheon as tribal gods; it explains why the tribes should be named from the animal or plant, and why the tribesmen should even, as we saw in some cases they did, esteem themselves as of the species of the totem-god. It explains why in Egypt, Greece, India, and elsewhere, there should be a number of such gods, by showing that there should be as many as there were stocks, counting themselves distinct, in the population ; and it also explains why in one place one animal should be pre-eminent and in another sub- ordinate, the gods following the fortunes of the tribes. It explains, moreover, on rational principles, for the first time, - the strange relations represented by the concurring legends of many lands as having existed between various animals and 568 STUDIES IN ANCIENT HISTORY the anthropomorphic gods ; it throws a new light on the materials employed in the so-called science of heraldry, showing whence they were drawn ; and, lastly, it enables us to see sense and a simple meaning in many legends, and in some historical narratives, that appeared to be simple nonsense till looked at in the light of this hypothesis. Since it is so simple and so comprehensive, and has a basis of facts for its founda- tion in existing totem-races ; since we have seen reason to believe that the mental condition of these races and the beliefs they entertain have been at some time the mental condition and beliefs of all the advanced races ; and since the only assumption we make is that all races have been progressive, which in other matters they undoubtedly have been, it seerns impossible to resist the conclusion that our hypothesis is a sound one that the ancient nations came through the totem stage. Some facts which make for our hypothesis cannot be too much insisted on. We have found in numerous cases what seems good evidence that from the earliest times animals were worshipped by tribes of men who were named after them, and believed to be of their breed. We have seen in several cases the oldest anthropomorphic gods having titles derived from the animals, or believed to be of their breed, or to have been fostered by them ; and the conclusion seems to be forced upon us that these gods were preceded by the animals as totems, if not as gods, and that the latter bore to them the same kind of relation that we know in India the serpent had to Buddha, and bears to Vishnu. On the rise of Buddhism among the Nagas, serpent worship was for a time repressed or sub- ordinated ; but the serpents were too strong. They reasserted themselves, and the old serpent faith revived with a human figure in the Olympus ! The heavenly Naga is even now the shield and protector of Vishnu. The hypothesis that similar occurrences had taken place among Horse, Bull, Earn, and Goat tribes will explain the peculiar relations which we have seen existed between these gods and these animals respectively, and we know of no other APPENDIX 569 hypothesis on which they can be, at least so wel}, explained. That Dionysus or Poseidon, for instance, should be ravpoyev^ is a fact presenting no difficulty on our hypothesis any more than that either of them should have been figured as a bull or with a bull's head. To what other hypothesis will the fact not be a stumbling-block ? Since these and all the other gods of their class were false gods that were gradually developed by the religious imagination, the fancy of poetical persons, and the interested imposture that is everywhere promotive of novelties in religion; since the whole of the facts we have been surveying demonstrate a progress in religious speculation from savage fetichism ; and since among the lowest races of men we find no such gods figuring as Zeus and his companions, we seem already, at this stage of the argument, to be justified in arriving at the conclusion that the ancient nations came through the totem stage, and that totemism was the foundation of their mythologies. II THE KAMILAROI AND KURNAI (By DONALD M'LENNAN) SINCE the preceding Essay 1 was written, Mr. Howitt has published an account of the Kurnai tribes of Gippsland (South- Eastern Victoria) ; 2 but, though in some other respects interest- ing, this adds nothing that is certain to our knowledge of the structure of Australian tribes. The Gippsland natives, like the Narrinyeri, called them- selves men (Kurnai), and the neighbouring tribes, with whom their relations were usually hostile, wild men (Bra-jerak). When the country was first settled, in 1839, they may have numbered (Mr. Howitt says) from 1000 to 1500, but by 18*77 there remained of them only 140 ; and their conversion to Christianity and settlement in missions had to a great extent broken down the force of the old customs among this remnant. The old tribal divisions had disappeared ; and the circumstances and mode of living of the survivors were entirely different from those of their predecessors. Mr. Howitt had an official position in Gippsland, and no doubt had good opportuni- ties of gathering such information as was still to be got ; but he had, in fact, very few too few informants, and he relied chiefly on the knowledge of a man who had lived from early youth 1 [I.e. the chapter on Australia, supra p. 278.] 2 Kamilaroi and Kurnai. By the Rev. L. Fison and A. W. Howitt. Mel- bourne, 1880. APPENDIX 571 among the settlers. 1 Though he had a slight knowledge of the Kurnai language, too, his communications with his inform- ants had to be carried on in the main in English, and the Kurnai English, however fit for other purposes, was not unlikely, without severe testing, to prove misleading or unin- forming in regard to such matters as kinship or tribal structure. Moreover, Mr. Howitt was chock-full of theories ; and it is clear that this helped to divert him from the line of inquiry which might possibly have yielded results of value. It follows that one has always carefully to scrutinise Mr. Howitt's facts, comparing them where possible with the slender accounts of the Gippsland natives which have come from other sources ; and (both on account of his theories and for other reasons) it is necessary still more carefully to scrutinise his inferences from facts, and as to the state of the facts in particular cases. In the list of the Kurnai divisions or local tribes, which he has laboriously collected, there appear three tribes or bodies of natives with names taken from men who were living when the country was settled by the whites (e.g. Bunjil Nullung's mob), and he offers these names to us as proper tribal names. This might of itself be enough to show that his facts are to be taken as somewhat doubtful ; and, as to his inferences, observe what his comment upon this mode of naming is. " Each of these divisions," he says, " re- ceived its designation from an eponym, who changed with each generation!' The Kurnai, we are told, claimed the whole of Gippsland from near Cape Everard to near Cape Liptrap, there being border lands, however, which were disputed between them and the Maneroo and Omeo tribes, their neighbours. The inhabitants of certain districts were known, as a whole, as Kroatungolung or east- country people ; and, similarly, there were Briakolung or west -country people, and Tatungolung or south-country people. Besides these there were, each with their separate country, the Bra-brolung tribes and the Bra-tauolung, to the 1 In Brough Smith's Aborigines of Victoria (vol. i. p. 57), Mr. Howitt men- tions that this man, Tulaba, " was caught as a young lad by the Macleods of Buchan, and thus got his name, Billy Macleod." 572 STUDIES IN ANCIENT HISTORY meaning of whose names a clue is given in the statement that Bra, like Kura, signifies man. The whole Gippsland popu- lation was comprised in those five divisions. This is what Mr. Howitt tells us ; but it should not pass unnoticed that the east, west, and south country people here spoken of were the Kurnai who lived to the east, to the west, and to the south of the Bra-brolung, among whom he found his chief informants. The Bra-tauolung lived to the south of his south-country people, and rather farther to the west than his west-country people. And while Bra-brolung and Bra-tauolung seem both to be simply variants of Kurnai (meaning men), east and west and south country people may have been designations in use among the Bra-brolung (such designations are so used among ourselves) for people who had besides other names more proper to them. There were diversities of language from district to district ; there were the usual blood-feuds and tribal quarrels, a frequent cause of quarrels being the stealing of women ; but the Kurnai seem to have been ready to unite against their Bra-jerak neighbours ; and the people of four out of the five districts were bound together by the ceremonies for the initiation of youths to manhood, which were performed by them in concert. In these the east-country people took no part. Initiation to manhood was unknown among them a proof this (the Kurnai, notwithstanding differences, being obviously one people) that it was formerly unknown among the Kurnai generally. While the whole country was divided (or divisible) into five districts as aforesaid, the population of each district was made up of small local tribes, each of which had its recognised boundaries. The structure or composition of these local tribes is what now concerns us, and, unfortunately, Mr. Howitt has not been able to tell us anything distinct about it. It may be said in passing that polygamy was practised among the Kurnai ; that fidelity was expected of the wife but not of the husband ; and that the husband did not give his wife to his guest. In general, too, the tribe had a name said to be taken from the principal place within its territory. There were no chiefs, but the ablest or most forward took the lead when leadership was wanted. The women had a voice, APPENDIX 573 and an influential one, in whatever concerned the common good. Mr. Howitt has not ascertained how kinship was taken in the Gippsland local tribe; nor has he collected many facts such as might help us in judging of this for ourselves. The transmission of family (totem) names is, among the Australians as among similar peoples elsewhere, the surest guide in this matter ; but, according to Mr. Howitt, family names were un- known among the Kurnai. He mentions only personal names and nicknames. And he believes that the Kurnai were not divided, as the Australians generally are, into totem clans or families. Eules for inheritance also throw light upon kinship ; but neither are these of any avail as regards the Kurnai. A man's personal property was in general buried at his death ; and there was nothing that could be inherited from him unless it were his wife or wives. The wives, Mr. Howitt says, where there were more than one, went to his brothers in order of seniority, one to each which, with polygamy, if taken literally, would prove male kinship, and might even suggest agnation ; but it is evident that Mr. Howitt did not think of inquiring whether uterine brothers had a preference, or whether it was uterine brothers who could thus inherit. Of the case in which one wife only was left, all he says is that the wife went to the deceased's brother, the statement being in no way particu- larised, and being therefore equally consistent with kinship being through males or through females. Even the blood- feud, as Mr. Howitt has been able to describe it, does not carry us far. The sister's son was foremost in the work of vengeance that is clearly brought out in one case ; and this shows beyond doubt that kinship was at one time counted through females only among the Kurnai (as Mr. Fison has remarked), and that kinship through females was still of importance, whether the exclusive kinship or not. Mr. Howitt's impression is that relatives both on the father's side and on the mother's side took part in the blood-feud, which would show that kinship through the father had come to be acknowledged. Unfortu- nately the cases he gives as examples of the blood-feud (possibly through the facts having been imperfectly explained to him) 574 STUDIES IN ANCIENT HISTORY suggest that the persons upon whom it placed a duty usually got all the friends they could to help them. Some other facts are given by Mr. Howitt which show the importance of the connection made through marriage all being consistent, however, with kinship through the father being also an established kinship. Marriage made a man free to live with his wife's local tribe as well as with his own. And it was a man's duty to share food with his father-in-law (it may be only when living with him, but it is puzzling to find this spoken of as an every-day matter).^ Young children were never killed (the distinction was made by Mr Hewitt's informants), but abandoned, as happens where the father is answerable for his children to the mother's relatives. The woman herself, too was under the protection of the blood-feud. As to the marriage law, Mr. Howitt has not tried to base it upon kinship, or to put it upon any definite basis. He has laboriously gathered information as to the localities to which the men of each local tribe went for their wives, but facts of this class cannot show whether there was any principle by which marriages were governed. It might be believed without them that men sought wives where women whom they were free to marry could be got most conveniently or easily, and that custom counted for a good deal in the matter ; and they do not carry us any further. Since, however, any Kurnai might marry a Bra-jerak or foreign woman, it may be in- ferred that, in general, where there was no kin there was no restraint upon marriage. And there was interdiction of marriage on the ground of kin. A man could not marry his " sister " a term which, as Mr. Howitt defines it, included all first cousins, whether on the father's side or on the mother's. Mr. Howitt does not tell us that it extended beyond first cousins. Nor has he been able to say that the interdiction of marriage on the ground of kin went further in one direction than in another. It was of importance, as it happened, in the case of those cousins only who belonged to different local tribes ; and it prohibited a man from marrying (among others)his father's sister's daughter and his mother's brother's daughter women who, with the totem system, would not have been of his own APPENDIX 575 totem, from whom an exclusive kinship either through males or through females would not have cut him off. A man could not marry any woman of his own local tribe. And in some cases, in addition to this, he could not marry any woman of his mother's local tribe. Mr. Howitt does not offer any explanation of these prohibitions. His statements might carry us, nevertheless, to a conclusion as to this, and as to tribal structure among the Kurnai also, but for one uncertainty. Mr. Howitt insists that wives were got among the Kurnai (except to a very 'trifling extent) by means of elopement or capture, which were both practised under the same conditions ; indeed, nearly all the information given to him was about elopements ; and one cannot be sure that he has not confounded limitations put by circumstances upon capture and elopement with limitations put by law or custom upon marriage. There might have been women both in a man's own tribe and in his mother's tribe whom he was not debarred from marrying, and yet it might have been too risky a thing or too gross an outrage, especially as the tribes were very small, for him to steal away with one of them, whether without her consent or with it. The Kurnai restric- tions, as stated by him, however, were those which prevailed among the Narrinyeri prohibition of marriage within the tribe, and prohibition of marriage with certain near relations not of the tribe with, in some cases among the Kurnai, the prohibition added of marriage within the mother's tribe. These are all restrictions which may occur where, after kinship through females only, the local tribe has become a proper clan established upon male kinship. The prohibition of marrying in the mother's tribe may be much more easily and completely explained upon this view of them than it can be upon any other view. And Mr. Howitt's facts are all consistent with it. Moreover, with fidelity exacted from the wife, there was no reason why there should not have been among the Kurnai the fullest acknowledgment of kinship through the father. And, on the other hand, the acknowledgment of kinship through the father is the best possible reason for fidelity being exacted from the wife. 576 STUDIES IN ANCIENT HISTOR Y We get incidentally another fact from Mr. Howitt which strongly suggests that the Kurnai were, as to kinship and the composition of their tribes, much in the position of the Narrin- yeri. He has been able to translate for us only a few of the Kurnai tribal names. And there is perhaps only one, the name of a tribe of the Bra-brolung, which he has translated with perfect confidence. But the name of that tribe meant Widgeon (Kamilaroi and Kurnai, p. 227). This is a name of the totem kind ; and if it was a totem name, it would show, consistently with what has otherwise been indicated to us, that one local tribe of the Kurnai was a proper clan with male kin- ship. Why, then, finding it among Australians, should we doubt that it was a totem name ? There would really, all the facts considered, be scarcely any reason for doubting, if any, but that our authority tells us he failed to discover totem divisions among the Kurnai. He had a theory, however, at this point, as will be seen by and by. And he has given us at least one other fact strongly suggestive of totemism, which occurs (where such a fact was likely to occur) in his account of the procedure at an elopement. 1 Moreover, the case in which totemism is most likely to escape discovery is that in which the tribe having male kinship, all the people of the tribe are usually of the same name and totem, while there is also (as there is said to have been, in general, among the Kurnai) a local name by which the tribe is commonly known. 2 1 "By and by," Mr. Howitt says, in describing an elopement, "one of the eligible young men met one of the marriageable young women ; he looked at her, and said, ' Djeetgun ! ' She said, ' Yeerung ! "What does the Yeerung eat ? ' The reply was, ' He eats so and so, ' mentioning kangaroo, opossum, or emu, or some other game. Then they laughed, and she ran off with him without telling any one " (Kamilaroi and Kurnai, p. 201). Mr. Howitt gives this without any sort of explanation, and totemism may give the explanation. The woman's question, if customarily made, must have been either still necessary or a traditional form ; there must have been a reason for it, whether a continuing reason or not. Now, when women eloped with strangers there might be need for inquiry as to marriageableness ; for the Kurnai men did not scar themselves that is, they did not bear tribal or family marks. The women did bear scars of some sort. And with totemism giving the rule for marriage, the man might marry the woman if he might eat her totem. 2 Dr. Gallatin was certainly not a less able or a less careful inquirer than Mr. Howitt, and his opportunities were incomparably better than Mr. Howitt's. It APPENDIX 577 Mr. Howitt thinks that the inhabitants of each of the five districts into which Gippsland was divided formed a single clan, of which each of the local tribes of the district was a section or division. But there appears to be absolutely nothing to support such a suggestion. On the other hand, with kin- ship through females only (of which it is clear they had had experience), the Kurnai, being exogamous, must have been in the common Australian case of having people belonging to different stocks, clans, or families mixed together, not only in every district, but in every local tribe. And a change in kin- ship need oust none of these clans from the district. It might lead to the people of each clan being collected into a single local tribe, as among the Narrinyeri; and the name of the Waiung or Widgeon tribe may indicate that it did this, to some unknown extent, among the Kurnai. The district names, too, so far as known to us (putting aside those which indicate geographical relation only) Bra-brolung, Bra-tauolung were mere variants of Kurnai, meaning the men or the people ; and are no more likely to have been clan names than Narrinyeri, which has the same meaning, and is, as we know, a general name for the people of many different clans, or than Kurnai itself. It appears, too, that the Bra-brolung, or, at any rate, some of them, could intermarry with one another, which makes against the view that they formed a single clan. Very likely, when not quarrelling among themselves, the inhabitants of each district of Gippsland were ready to unite against other Kurnai ; but from such a fact there can be no inference as to clanship. The Iroquois, notwithstanding their famous league, were made up of several clans or families, in which descent was taken through the mother. 1 seems, therefore, worth while to mention here, with reference to the scantiness of indications of totemism in Mr. Howitt's work, that Dr. Gallatin, in writing of the Southern Indian nations of North America, while describing them admirably in other respects, disclosed that they had totems only casually and in a footnote. That they were in totem clans is now known on other authority. Dr. Gallatin, however, was able to mention a number of totem names, while Mr. Howitt has given us only one, but with what may fairly be taken as an indication of there having been others. 1 Mr. Howitt's earliest statement about the Kurnai is to be found in Mr. Brough Smith's Aborigines of Victoria (vol. ii. pp. 323 et seq.), and is confined to the Bra- 2 P 578 STUDIES IN ANCIENT HISTOR Y Mr. Hewitt's facts which bear upon the tribal system of the Kurnai have now been all dealt with. But there are two other matters which figure largely in his book, and should not be overlooked. In a work which will be noticed hereafter, it appears that the natives of the Port Fairy district of Victoria were made up of five clans or families having totem names which were taken through the mother viz. Long-billed Cockatoo, Pelican, Banksian Cockatoo, Boa-snake, and Quail and that, as is usual in such cases, no man could marry any woman of his own totem name. It appears, too, that, like other people in the same social state, the Port Fairy natives had superstitions or practices which had to do with animals from which none of them took their names, and which were common to them all, of whatever totem name. And among other things, the common Bat, we are told, be- longed to the men of whatever totem, who protected it against injury, " even to the half-killing of their wives for its sake " ; so that it was protected by the men against the women. Similarly, the Fern-owl, or large Goatsucker, belonged to the women of whatever totem, and, though a bird of ill omen, creating terror at night by its cry, was jealously protected by them. If a man killed one, they were " as much enraged as if it were one of their children," and would strike him with their long poles ; so that the women protected it against the men. A fact such as this, wherever it occurs, is curious, and no doubt worth inquiring about. Among the Port Fairy people, brolung tribes. The name Kurnai is not mentioned in it, and it is not clear whether the Bra-brolung, as a people, called themselves Kurnai as well as Bra- brolung. In this statement Mr. Howitt said that the Bra-brolung were in three "classes," by which he meant local tribes, within each of which marriage was forbidden, while each might intermarry with the other two. Had this state- ment remained undisturbed, there could have been little doubt that the Bra- brolung local tribe was a true clan with male kinship, and exogamous (even without it being known that the name of one of the tribes meant Widgeon). In Kamilaroi and Kurnai Mr. Howitt gives the names of five local tribes of the Bra-brolung, and tells us something of the habits in respect of marriage of three of them. One of the three, he says, took wives from a neighbouring district only ; the second took wives from the Waiung or Widgeon tribe only ; while the Waiung took wives from those two, and also went abroad for them. As to the name Kurnai, it seems to have been added to the name of a local tribe when an individual of the tribe was to be designated. APPENDIX 579 whatever its import, it is clear that it had to do with sex ; and it is clear also that it had nothing to do with totem names or tribal divisions. It would not be surprising were one brought to conclude that it was connected with the initiation to man- hood, the effect of which was to take youths from the women and make men of them. It is upon a fact precisely the same as this, however, that Mr. Howitt and his collaborates, Mr. Fison, have based their whole view of the history and tribal condition of the Kurnai ; and, naturally, their view is a very extraordinary one. Among the Kurnai the Yeerung, or Emu-wren, belonged to the men, and was protected by the men against the women, just as the Bat was among the Port Fairy people ; and the Djeetgun, or Superb Warbler, belonged to the women, and was protected by the women against the men, as the Goatsucker was among the Port Fairy people. Both figured in the proceedings of the men and women at the ceremonies of initiation, and in these, in connection with the young men, another bird, the Nurt, a species of duck, also played a part. Mr. Howitt seems at an early stage of his inquiry to have become possessed with the notion that the Yeerung and the Djeetgun were all the Kurnai had for totems. (But why did he overlook the Nurt ?) And he concluded that the Yeerung was the totem of the Kurnai men, and the Djeetgun the totem of the Kurnai women. By parity of supposition perhaps the Nurt should have been the totem of Kurnai young men. Yeerung and Djeetgun, how- ever, dominate his book, turning up in the most unexpected places ; and Mr. Fison has devised a theory to account for the totem system being employed among the Kurnai, not to dis- tinguish one body of kindred from another, but to distinguish men from women. One need not say much of such an error, or of the ingenuity of theorising expended upon it. It is enough to repeat that the Port Fairy people had precisely similar sex preferences to those of the Kurnai, and had totem names and divisions besides, and that (unless as regards their origin) the two had nothing to do with each other ; and it may be added, for those who are interested in Mr. Fison's theory, that though the former 5 So STUDIES IN ANCIENT HISTORY had somehow to do with sex, it was the latter which regulated marriage ; so that the theory, with all its developments, in- cluding the explanation of " marriage by elopement," is abso- lutely chimerical. Obviously Mr. Hewitt's view could never have been thought of had it been known that the Kurnai had totem names or divisions. And it was just the view which might make the most candid inquirer rather slacken his search for such things. Very likely these (if the totem system pre- vailed among the Kurnai) were not so easily to be heard of as a fact in which all Kurnai men and women were interested. As it is, Mr. Howitt has found for us one Kurnai tribal name, which was an animal name, and has given us indications of others. And without a good knowledge of Kurnai speech, sup- posing that the Kurnai had the totem system, it may have been impossible for any one to do more. It need scarcely be added that, whatever totem names they may have had, we may be sure that Yeerung and Djeetgun were not among them. The other matter to be noticed (of which incidentally something has been said already) is the mode in which the Kurnai got their wives. Here once more there is uncertainty as to Mr. Howitt's facts. He tells us that, as a rule, a young Kurnai could get a wife only by running away with her that is, by her freely eloping with him, without the knowledge of her parents. " If the young man was so fortunate as to have an unmarried sister, and to have a friend who also had an unmarried sister, they might arrange with the girls to run off together ; " this is the first illustration of the system of elope- ment which he gives us. In other cases, a young man and young woman ran away together because they happened to fancy one another, and the first advances might come from either. The girl, in all cases alike, went freely, and without the knowledge of her parents ; and their consent, if asked, would be refused as a matter of course. By and by it comes out that there was a good deal of capture among the Kurnai, and that the stealing of women was a frequent cause of quarrels between tribes. There is some recognition also of exceptional cases, in which women did not elope, but were given away made, it would seem, APPENDIX 581 because a correspondent of Mr. Howitt's who had lived long among the Kurnai, Mr. Bulmer of the Lake Tyers mission, had expressed doubts as to the accuracy of Mr. Howitt's information. Mr. Bulmer's own account of marriage among the Kurnai, written some years before Mr. Howitt's, is to be found in The Aborigines of Victoria (vol. i. pp. 82 et seq.). When a girl is thirteen or fourteen, he says (p. 84), a yam-stick is given to her for protection ; " and this precaution is nearly always needed, for it would not be sufficient for her to say ' No ' to an important question. She drives away any young man who is smitten with her charms with her yam-stick." Mr. Bulmer here puts the risk to a young woman from violence into the foremost place, but it was not his purpose to speak of capture. " Matches," he continues, " are generally made up among the young men ; the women never initiate matches, though they have a good deal to say when it is known that a young woman is sought after by some young man. The match is mostly arranged between two young men who have sisters, or some female relative over whose fate they may happen to have control. They follow a system of barter in their matrimonial arrangements. The young woman's opinion is not asked." He speaks, it will be observed, as if this matchmaking, carried on by means of barter, were a common thing, as if it were a system indeed, as if it were the regular thing; and what he describes is marriage by exchange, which is exceedingly preva- lent among other Australians. He goes on to tell how the transaction was carried out. " When the young men have settled the business," he says, " they propose a time when one of them is to take a girl for his wife. The young man marches up to her equipped as if for war, with his club and club-shield in his hands ; and indeed these are needed, if he does not wish to receive a blow on his head from the yam-stick, which would perhaps prevent the further progress of his love-making. After a little fencing between the pair, the woman, if she has no serious objections to the match, quietly submits, and allows herself to be taken away to the camp of her future husband." It is plain that the young man marching up to the woman 582 STUDIES IN ANCIENT HIS TOR Y equipped for war does not go to propose an elopement to her, and that he does not go meaning to put up with a refusal. It is manifest that, in virtue of the bargain made, he goes to take her away, by force if necessary ; and, in short, there is a form of capture. Mr. Buhner tells us also that elopements took place, and he speaks as if they were pretty frequent. He shows that elopement was resented much as capture would have been, though, no doubt, in a less degree ; and also that the woman was nearly always brought home again, so s that elopement was far from leading immediately to marriage. After a couple had eloped again and again, however, and been punished for the offence, the girl's friends usually consented to a marriage. The man had got love tokens from her, Mr. Bulmer explains, which, in the hands of a sorcerer, would put her life at his mercy ; she became ill, and her friends grew anxious ; there was pity for the young man, too, who had braved so much for her ; and so, in the end, " the tribe give him to wife the girl for whose sake he has borne so many honourable scars." In the detail just given there is little difference between Mr. Bulmer and Mr. Howitt ; and (besides that the woman was given at last) it scarcely seems descriptive of a mode of getting married to which there was no sort of recognised alternative. Mr. Hewitt's admissions about capture show that he made his statement as to elopements much too strong; and, after reading Mr. Bulmer, it cannot be doubted that he mistook the place of exchange in Kurnai marriage. So much, as to exchange, might almost have been inferred, indeed, from the Kurnai having been polygamous. Men do not get many wives by elopement practised under difficulties; men do not get many wives even by capture polygamy cannot get on without more regular means of wife-getting ; and, allowing for inheritance, there can scarcely be any practice of it worth speaking about without wives being obtainable by some form of contract. No doubt the marriages arranged by brothers, which Mr. Howitt classes among elopements, were in general marriages by exchange. APPENDIX 583 What proportions, however, the three capture, exchange, elopement bore to each other among the Kurnai does not greatly matter. It is plain that the Kurnai arrived at marriage by exchange in a movement from capture, for what the man got by his bargain was permission to take away or carry off the woman. That elopement also was arrived at in a move- ment from capture is clearly shown by a curious fact, immediately to be noticed, which Mr. Howitt mentions. And if elopement and exchange both grew up upon capture, which of the three was the more in vogue becomes a question of subordinate interest. In an account of the capture of women as practised by the Australians, given in Mr. Brough Smith's Aborigines of Victoria (vol. ii. p. 316), it is stated that "in any case, where the abduction has taken place for the benefit of some one individual, each of the members of the party claims, as a right, a privilege which the intended husband has no power to refuse." Where a tribe carried off women from another tribe, it is added, " the unfortunates are common property till they are gradually annexed by the best warriors of the tribe." The general case and the particular agree in this, that, in the first instance, a woman was common to those who had made capture of her ; while she was, in the latter, given immediately after that to the man for whose behoof she had been carried off, and was, in the former, monopolised as soon as some particular man was influential enough to make his own of her. Similarly, Mr. Howitt tells us of a tribe in which (p. 346) "the female war captive was at first common to the men present at the capture, and then became the property of her captor " that is, provided she were a woman he was free to marry. And of a woman who had been taken in war, he was told (Kurnai practice this) that " before she was the wife of Tankli she belonged to all the Yowung men," which coin- cides with the second of the foregoing statements. It need not be said that the practice thus disclosed as following upon capture is not peculiarly Australian ; and no far-fetched theory is needed to account for it. With contract for marriage and capture remaining as a form, we sometimes 584 STUDIES IN ANCIENT HISTORY find it continued the bride carried off, though the carrying off has been only a form, the bridegroom's party treat her as in like case their predecessors would have treated a captive woman. Now, among the Kurnai, as Mr. Howitt shows us, this practice was a concomitant also of elopement. The man with whom a woman had agreed to elope summoned his friends, and the woman was treated by the party as she would have been if they had carried her off for him by force, or taken her in battle. Where we find it with marriage, this s practice reveals to us an association early formed between wife -getting and capture, which, with capture become a form, and possibly reduced to a meaningless formality, continues to manifest itself in marriage doings. It as clearly shows that the elope- ments of the Kurnai in which the man stole away with a woman, and was punishable much as he would have been for stealing her were an aftergrowth of their system of capture. Mr. Fison has suggested that this practice among the Kurnai illustrates what Sir John Lubbock has called expiation for marriage. But elopement, though so much akin to capture, and though it might prove a step towards marriage, did not make marriage among the Kurnai ; in general, it was not followed by the appropriation of the woman, and Mr. Fison has been puzzled to say what there was to be expiated. " The man," he says, in the only passage in which he has ventured to speak clearly (p. 303, footnote), "is one of a group each member of which has as much right to elope with the girl as he has. The secret meeting in the forest seems to be a com- pounding for that right." He might have said they had all as much right to capture her without being very paradoxical. And the same could equally well be said of all the men of many groups. But when he speaks of a right to elope with her, it becomes difficult to treat him seriously. In elopement the girl had choice as much as the man ; she was not restricted to men of any particular group ; there was no man who had any right to elope with her. Her consent even can scarcely be said to have given a right even to the man whom she APPENDIX 585 favoured, though no doubt it gave him an opportunity. More- over, the " expiation " was not offered to all the men of even one group, but only to those near friends of the man from whom he might have claimed help for her capture, the men who, with regular marriage, would have been his backers at the wedding. On Sir John Lubbock's theory, no doubt, a war captive, or foreign woman taken captive, belonged to the captor, and there was no need that her appropriation should be expiated; it was the appropriation of a native woman that needed to be expiated, that infringing a right of the tribe. It seems worth while to add, nevertheless, that before the practice which the Kurnai followed in their elopements, and which some other peoples have had in marriage with the form of capture, could be taken as showing expiation for marriage in any case what- ever, the same practice in all cases of actual capture would have to be explained as expiation for marriage. That fighting men are much the same, with women at their mercy, in early times and in late, and whatever the current views about marriage right may be, seems to afford a simpler view of its origin. But, apart from that, could it be said there was expiation for marriage in the case of those women captured by a tribe, who were common to the tribe at first, but were afterwards appropriated ? Did the women remain unappro- priated until their appropriation was sufficiently expiated ? The fact seems to be that the woman's first position in this case is that of undivided and unallotted booty unallotted because nobody has a paramount claim to it. She is common because she is not anybody's, but everybody's, and she ceases to be common as soon as some one man is able to make his own of her. Mr. Howitt gives, in appendices to his work on the Kurnai, some information procured by him respecting four other Australian tribes information too meagre in every case, but still worth a few words of notice. The Gournditch - Mara tribe (Western Victoria ; Gourn- ditch being a local name, and Mara = man) is said, on the 586 STUDIES IN ANCIENT HISTOR Y authority of the Eev. J. H. Stahle, to have been composed of four classes, that is, divisions, named Keriip (water), Bum (mountain), Dirck (swamp), and Gilger (river) ; and to have had " no exogamous rule affecting marriage " ; so that, though wives were got from neighbouring friendly tribes, and occasion- ally through capture, there was no rule which restricted marriages within the tribe itself. It is not stated whether the distribution of the tribesfolk into four divisions had any effect or served any purpose, but the names given are of the totem kind ; they may therefore have been totem or family names ; and the blood-feud may have depended upon them, though it is said they did not regulate marriage. Besides the statement as to marriage, the most interesting points in Mr. Stable's information are that children belonged to the division of their father, and spoke his language, when the mother was of another tribe, and not the mother's language (the mention of which seems to show that the mother was often of another tribe) ; and that, there being a headman of the tribe, son suc- ceeded to father in this office. It thus appears that kinship through males was thoroughly established among the Gourn- ditch-Mara. There is nothing to show to what extent there was kinship through the mother. Sisters were exchanged in marriage, that is, girls were bargained away not by their parents but by their brothers a point worth noticing, because the same thing happened among the Narrinyeri, who had male kinship, and among the Kurnai, who appear to have had male kinship ; but the consent of the parents was necessary. Fidelity was expected from the wife, and the husband did not lend her to his guest, both points tending to show a thorough establish- ment of male kinship. There was no objection to polygamy. As to the statement about marriage, of course, there is the question whether it can be trusted ; and it does not tell us, be it observed, of any restriction upon the marriage of relatives however near. On reasoning given in Primitive Marriage, however, a practice of polyandry might prevent the rise of exogamy ; and exogamy, where it had arisen, might disappear in the confusion following upon the change from kinship through females only to kinship through males. If the APPENDIX 587 Gournditch-Mara had the totem system, and were really not exogamous, the latter is the probable explanation of their not being exogamous. Is it still possible to ascertain whether they had any marriage law ? The mere statement that they were not exogamous, marking them as singular among Australian tribes, makes them more worth inquiring about than most of the numerous tribes which are known to have been exogamous. The information as to the Geawe-gal tribe (now extinct ; Hunter Eiver, New South Wales) communicated to Mr. Howitt by Mr. G. W. Eusden is so meagre, and has been so evidently supplied from imperfect recollection of facts imperfectly observed or understood, that it would perhaps be best to say nothing of it. Mr. Eusden tells us, however, that "it was absolutely necessary that women should be married according to tribal laws," and he states that these were laws " of forbidden degree or class," which may be taken as an indication of exogamy. He mentions also that "the Geawe-gal had a superstition that every one had within himself an affinity to the spirit of some bird, beast, or reptile," which (error or misunderstanding duly allowed for) may be taken as an indication of totemism. Mr. Eusden has nothing to tell as to the kinship system of the Geawe-gal. Some statements as to the Turra tribe (York's Peninsula, South Australia), which are very interesting so far as they go, are given by Mr. Howitt on the authority of the Eev. W. Julius Klihn (but apparently not in the language used by him). Of this tribe, according to Mr. Klihn, there were two main divisions Wiltu (eagle-hawk) and Multa (seal). It will be seen immediately that it is left somewhat in doubt whether these were not really separate tribes. The Wiltu sub- tribe contained people of ten, the Multa people of six, different totems. All the totems are mentioned. We are told that the divisions or sub-tribes were exogamous, but that any totem of one might marry with any totem of the other; and that children belonged to the totem of their father, and therefore, to his division or sub-tribe ; so that here once more we come upon kinship through males in an Australian tribe in con- 588 STUDIES IN ANCIENT HISTORY nection with totemism and exogamy. Daughters were given in marriage by their parents, and wives were also obtained by exchange of female relatives. Faithfulness, it is said, was expected both of husband and wife, which tends to show a thorough establishment of the family upon the basis of male kinship. Elopement was not unknown, but the penalties were serious. A girl who eloped would be put to death by her relatives, if they could recover her ; and they would kill the man also if they could, but he was generally protected, it is said, by the division of the tribe to which he belonged. What is mentioned to illustrate this, however, is that if a man of the Wortu (wombat) totem of the Wiltu (eagle-hawk) sub-tribe were to elope with a Multa woman, he would be protected by the Wiltu- Wortu men that is, by the men of his own totem ; and, to illustrate the marriage-law, it is added that a Wiltu- Wortu man would not be allowed to keep a Wiltu- Wortu woman that is, a woman of his own totem even if he captured her ; statements which suggest a doubt whether either the restraint upon marriage or the duty of affording protection really extended beyond the totem. We are not told into what small groups or local tribes the Turra people were subdivided, but by the conditions under which they lived they almost must have been subdivided. It might have been expected that the people of the same totem would in general be found living together; nor does the mention of the capture of a Wiltu- Wortu woman by a Wiltu - Wortu man, as if it were a case which might occur, make against this, for (besides that the woman, when married, would be living among strangers) the statement was evidently made in answer to a "leading question," and seems to be only a way of saying that a man might not marry a woman of his own totem in any case whatever. As to the sub-tribes, we are told that when a woman of one was captured in war by a man of the other, her sub-tribe would fight to recover her, and failing in that, would endeavour to capture a woman of the other ; and this tends to show that they lived as separate tribes. Mention is also made of their meeting together for a grand corrobboree. APPENDIX 589 In connection with the grand corrobboree, a statement (already referred to in a footnote) * is given which is of the greatest interest in relation to the origin of exogamy. At the corrobboree (though at ordinary times men did not lend their wives to brothers or friends) "the old men took any of the young wives of the other class [sub-tribe] for the time, and the young men of the Wiltu exchanged wives with those of the Multa, and vice versd, but only for a time, and in this the men were not confined to any particular totem." As men were not confined even in marrying to any particular totem, this can only mean that for the time the totem made no restriction ; that a Wiltu man might for the occasion take any Multa wife even a woman of his own totem. The Turra people habitually insisted on conjugal fidelity ; and the men of one sub-tribe were prohibited from marrying (so it is stated) any of the women whom the men of the other would in general have for wives, and, at any rate, they all were strictly prohibited from marrying within their own totem. Was their practice at corrobborees a tradition (observed at extraordinary occasions) of the ordinary practice of their early predecessors ? Did the exogamous prejudice apply at first to wiving only ? Of the Wa-imbio tribe, whose territory extended from the junction of the Darling and Murray rivers to the Eufus, a slight account, supplied by the Eev. J. Bulmer, is given by Mr. Howitt. Of the Wa-imbio, Mr. Bulmer says, there were two primary divisions, Muquarra (eagle-hawk) and Kilparra (crow) ; he implies that each of these contained people of several different totems, but he mentions two totems only Karnie (a large lizard), which belonged to the Muquarra division, arid Namba (the bone-fish), which belonged to the Kilparra division. Mr. Bulmer wrote from experience that he had of the Wa-imbio in his early days; he seems to have remembered no other totems, but he is " most confident " that children were always of the same division as their mother, which involves that they were of her totem also, and that kinship was taken through the mother. With respect to the conditions of marriage, he thinks the consent of parents was usually required, elope- 1 Supra, p. 287. 590 STUDIES IN ANCIENT HISTOR Y merits also occurring, however ; but beyond this his know- ledge seems to have failed him ; what he says scarcely amounts to more than that there were conditions involving exogamy, and that they were strictly enforced. A man would not be permitted, he tells us, to keep as his wife a woman of his own " class," and the people would not hesitate to kill a man who broke the rule ; but he does not say what " class " means in this statement whether a division of the tribe or the totem kindred only. In an earlier statement l he had said that " a Muquarra could not marry a Muquarra, ^nor a Kilparra a Kilparra " ; but in that statement no hint was given of the existence among the Wa-imbio of any totems except these two. The only marriage mentioned by him is a marriage, which followed upon betrothal, between a Wa-imbio man and a woman of the Tapio tribe. To the preceding notices of Australian tribes scarcely any addition of fact can be made from that portion of Kamilaroi and Kurnai for which Mr. Fison is responsible, that being chiefly occupied with theories and controversial matters. The brief citations from correspondents which occur in it (though his theories are mainly founded upon them) are, in general, too meagre to be of any use, even could they be trusted ; and it often seems quite obvious that they are not to be trusted. Of course it has to be borne in mind that Mr. Fison was unable to study Australian tribes for himself; and that the only means he had of adding to our knowledge of these fast- dying Aborigines (a most praiseworthy object, if pursued with proper care and without theoretical bias) was to spread schedules of questions far and wide, among all sorts of people who had been in contact with natives, and do his best with the answers he got. It must be said, however, that (while nearly all his new matter is, to speak mildly, doubtful) he has, in his eager- ness for new information, ignored a great deal that was already known on better authority than it was in his power to have recourse to. He has besides legitimately, no doubt, in setting forth his hypotheses, but to a much greater extent than justice 1 The Aborigines of Victoria, vol. i. p. 86, 1878. By R. Brough Smith. APPENDIX 591 to these demanded used all marriage and kinship terms, as he says himself, in an accommodated sense that is, in a non-natural sense, and in other ways mixed up fact and theory together, often speaking too confidently of his theoretical views, as if they were matter of fact. His work, while it displays an intrepid ingenuity which finds nothing impossible, nothing even difficult, would for those reasons prove misleading and mystifying if taken for a source of information as to the condition of the Australian tribes as they actually have been made known to us. Some account, however, must be given of the tribe, or population, by means of which (and not the Kamilaroi, who are, in fact, a difficulty for him, though their name is on his title-page) he has illustrated his principal theory, if only to illustrate his manner of dealing with facts. This tribe, when Mr. Fison received his information about it, consisted of seventeen persons ; twenty-eight years earlier it had num- bered 900. His informant (a Mr. Stewart) regrets that he had not been put upon close inquiry ten years earlier, " when the natives were numerous, when there were old people of intelligence to be found among them, when one might, without hesitation, accept their ideas and expressions as original." The name of this tribe is not mentioned, which is curious ; it is spoken of only as the Mount Gambier tribe. It had the totem system, but, as Mr. Stewart understood the matter, with this peculiarity, that while each body of kindred had one animal or plant which was its totem for example, pelican or tea-tree each regarded a number of other things as in- cluded with the totem, which apparently means as on the same footing with it. For pelican (to go on with the examples already given) these are said to have been dogs, black-wood trees, fire, and frost ; for tea-tree, ducks, wallabies, owls, and crayfish. A man did not kill, or use as food, any of the animals belonging to his own totem group (of the same sub- division with himself, are Mr. Stewart's words ; and the totem divisions are his subdivisions) unless compelled to it by hunger, and then with expressions of sorrow for having to eat his wingong (friends), or tuinanang (flesh). " When using the last word, they touch their breasts to indicate the close 592 STUDIES IN ANCIENT HISTOR Y relationship, meaning almost a part of themselves." The totem clan was called tuman. It is natural to suspect mis- conception on Mr. Stewart's part as to portions of this state- ment ; and of this something will be said immediately. The statement, nevertheless, gives unexceptionable evidence of totemism ; and those of the same totem were those of the same flesh, or, as we say, of the same blood that is, they were the body of acknowledged blood-kindred. Accordingly the tuman took "a prominent part in the blood revenge arrangement," " also in cases of uncertain death, the tuman of the slayer (slain ?) appear at the inquest." The life of the tribe was so far based on the totem system. It included eight different totem clans or tumans, the names of which are mentioned. So much is clear. There was also a division of this population (for which as a whole we know no name) into Kumite and Kroki (feminine, Kumitegor, Krokigor) ; and five totems are mentioned as having been Kumite, three as having been Kroki. The meaning of these words is unknown; it is not known that they are either totem names or local names. The tribe is described by Mr. Fison as exogamous, and as taking kinship through the mother. Were it worth while to make surmises, it would be not unreasonable to surmise that at Mount Gambier two separate local tribes containing different totem kindreds had, through the operation of exogamy and female kinship, become welded into one community. As to the peculiar extension of the totem system in this tribe which has been mentioned, there is a general statement made which goes vastly beyond the particulars which are given for the several totems. Mr. Stewart is quoted as saying that "not only mankind, but things in general, are subject to these [the totem] divisions " ; and to this is appended as a specimen, Mr. Fison says the full list of those " things in general " which Mr. Stewart felt able to mention for each totem (averaging three or four for each). The reference to mankind in that passage is in no way explained, and it is of uncertain meaning; it may mean only that the tribe was composed of totem clans, and yet something more seems APPENDIX 593 intended; surely it cannot mean that strangers to the tribe were necessarily considered as belonging to one or other of its clans. Things in general, however, according to the state- ment, were distributed between the clans. On what principle of distribution then ? " I have tried in vain," Mr. Stewart says, "to find a reason for the arrangement." How then did he find out so much as he did ? and did he jump from that to his general statement ? or did the general statement come to him independently of the particulars? Mr. Fison has here left us entirely to our conjectures. He has given us, however, one little glimpse of Mr. Stewart pursuing his inquiries. " I asked," Mr. Stewart says, " c To what division does a bullock belong ? ' After a pause came the answer, ' It eats grass ; it is Boortwerio [tea -tree].' I then said, 'A crayfish does not eat grass [the crayfish had already been described to him as Boortwerio] ; why is it Boortwerio ? ' Then came the standing answer for all puzzling questions : That is what our fathers said it was.' " Here we see the poor savage anxious to please, indifferent to truth, ready to give answers such as are expected of him (so observers concur in describing him) put to the question with a result which perhaps might have been anticipated. He finds a totem or tuman for the bullock on the spot. Mr. Stewart, be it ob- served, was by this time fully possessed of the notion (in whatever way he got it) that things in general were distri- buted between the totems. It may be assumed that he had been put upon his inquiries by Mr. Fison ; and it will be found that Mr. Fison had got a hint of something similar from another quarter (not, however, without warning as to the risk of error; see The Aborigines of Victoria, vol. i. p. 91). The only things it seems necessary to say about this re- markable disclosure are, first, that it, at any rate, shows us again very clearly the position of the totem clans in this tribe (as Mr. Stewart understood the matter) that they were the units, or separate components, of the community; and, next, that Mr. Fison should not have published, as illustrating savage speculation, a statement one branch of which, as it stands, looks like mere nonsense, and which may be altogether 2Q 594 STUDIES IN ANCIENT HISTOR Y a product of self-mystification, without giving the necessary explanations, and not without giving us in some part of his work every word which his correspondent had given to him. We need in such a case all the testimony, and that in the very form in which it has been given, if only to enable us to form our own opinion of the witness and of his actual meaning. To complete the account that can be given of the Mount Gambier people, it remains to notice what is disclosed to us as to their marriage law and their system of kinship. This is very brief; it is given as on the authority of Mr. Stewart, but with only a single line cited from that informant to illustrate or support it. Kumite, the male, Mr. Fison tells us, married Krokigor, and the children were Krokigor (that is, of the Kroki division) ; Kroki, the male, married Kumitegor, and the children were Kumitegor (that is, of the Kumite division). Kinship was taken through the mother, according to this statement ; that is clear and need not be doubted. As to the marriage law indicated, when, at another part of the work, we get incidentally, and for a different purpose, a brief citation from Mr. Stewart (that which has already been noticed), we find that Mr. Stewart, with the knowledge he had, had to speak of the totems almost exclusively. It was the totems which were prominent in the blood-feud ; the totems which determined what each man ought not to eat ; the totems between which there was the alleged distribution of mankind and of things in general. Mr. Stewart tells us nothing here of Kumite and Kroki, but that either might eat the other's totems ; and this was probably stated in answer to a question, for it was superfluous coming after what he had said previously of the totems in same connection. This evidence prepares us to find the totems regulating marriages in the tribe, as they did all other things of importance. Now Mr. Fison tells us that they did not affect marriage in any way; but that marriage was prohibited between all Kumite men and women, and between all Kroki men and women, while any man of the one name might marry any woman of the other. It appears that he expected to find the totems affecting marriage by putting a further restriction upon it (pp. 41, 42); he was APPENDIX 595 prepared to find that a Kumite totem did not intermarry with all the Kroki totems, but only with one or more of them ; and the single line as to marriage which he cites from Mr. Stewart gives a negative answer to a question on this point. What other inquiries he made, and how far Mr. Stewart's knowledge extended, we cannot know. Though treating this matter of the marriage law as of great import- ance (and it is the sort of matter which necessarily is of interest), Mr. Fison has produced no statement about it from Mr. Stewart except the one line just mentioned ; and so there is nothing more to be said of it. It is clear, at any rate, that the Mount Gambier people had the totem system, and also that they had exogamy and female kinship. "What Mr. Fison descries among the Mount Gambier people (and this it is which makes Kumite and Kroki import- ant in his eyes) is a division of the tribe into two intermarry- ing classes having kinship through the mother, by which he does not mean two divisions within each of which marriage was prohibited, but two divisions which intermarried with each other, and at first made no other marriages ; or, to be more accurate, in which the men of one had conjugal rights of a communal sort over the women of the other of their own generation. It need not be said that this is hypothesis merely, and there is nothing adduced to support it for this particular tribe or population. It is as intermarrying classes neverthe- less that Kumite and Kroki are introduced to us ; and by stretching out to sea as far as New Britain, Mr. Fison finds four other tribes fit to be put in the same category as illustrat- ing what he believes to have been the earliest division of tribes among the Australians. Two only of the four are Australian, and Mr. Fison's information about these seems to consist of a few sentences for each division, which, as de- scribed to us, is each subdivided into two, but with a system of naming which suggests that the two are one a system of naming which is a difficulty for Mr. Fison ; not an insur- mountable difficulty, however, for he has been able to offer two entirely different solutions of it. How far imperfect information is the true solution it is bootless to inquire. Of the 596 STUDIES IN ANCIENT HISTORY first, the Mount Gambler people, we have already spoken. The other (on the Darling Eiver) is also in totem divisions, being made up of the Muquarra and the Kilparra (eagle-hawk and crow), of which we have learned something already on the authority of Mr. Bulmer. Each division includes several totems, but Mr. Fison's informant, a Mr. Lockhart, has, like Mr. Bulmer, been unable to give a list of the totems, and mentions only three. The little Mr. Lockhart says of marriage, by the way (p. 41), is expressly said to illustrate how the divisions Muquarra and Kilparra are continued through the system of female kinship, but has been unaccountably mis- understood by Mr. Fison, who finds in this the main, if not the only, foundation for a theory of his which has been already referred to, that where a division includes several totems, and there is no marriage within the division, the totem puts a further restriction upon marriage ; so that, for example, a particular totem of Muquarra could intermarry only with a particular totem of Kilparra. For the Banks Islanders and the natives of New Britain, who complete his list, Mr. Fison's information is more scanty still a sentence or so for each. Such is the basis of fact for the hypothesis of intermarrying classes on one side ; to note how very slight it is, besides being otherwise very dubious, is all that is here intended. On what basis of fact, then, does Mr. Fison place the other branch of his hypothesis, the communism predicated of the men and women of the same generation ? First and chiefly, he relies upon the fact that in a large proportion of Australian tribes a husband lends his wife to his guest. This he takes to be not a part of hospitality, a kindness customarily rendered on the side of the husband, but proof of a right to the woman on the part of the guest, than which there scarcely could be any view more novel or sur- prising. It seems to involve that there was a right to the hospitality to which this was at any rate incident ; and yet Mr. Fison has to admit that by many tribes a stranger might be not entertained but eaten. No doubt the practice referred to was very common. No doubt, too, where it prevailed it was usually ascertained that the men and women might cohabit APPENDIX 597 without incest (this is stated, at p. 66, to have been done in one district even in cases of forcible rape) ; and no doubt, that point being clear, the woman was not refused to the guest, but that is a very different thing from the guest having a right to her, which he could assert even in her husband's dwelling. Plainly the facts show nothing more than that the Australians acted on their views of hospitality, when they happened to have conceded it. Mr. Fison relies also upon the license allowed at corrobborees, where the polygamous elders who monopolised the women waived their rights for the time in favour of the body of tribesmen. This license (without which the monopoly of the elders might have been intolerable) he takes to be proof of an ancient right of tribesmen ; but what it discloses manifestly is a state of things in which there is no right. He adds, on the authority of Mr. Lance (of whose competency as a witness we have already had opportunities of judging 1 ), that, among the Kainilaroi, when Kubbi and Ippata met, they saluted each other as spouse ; that the Kubbi thus meeting a stranger Ippata would treat her as his wife ; and that his right to do so would be recognised by her tribe, and this is capped in a footnote by a statement from Mr. Cyrus E. Doyle to the effect (to go on with the same names) that any Kubbi could take any Ippata as his wife and keep her, and that his right to her would not be questioned by her family (Kamilaroi and Kurnai, p. 53). Mr. Doyle's statement, made of a polygamous district where wives must have been eagerly sought for, is obviously sheer nonsense. Mr. Lance's, when his misconceptions of fact are allowed for, may show that considerable license was practised without much objection from the men interested among the Kamilaroi ; based upon essential error as it is, it cannot show anything more. As to the use of the word which is rendered spouse between entire strangers, what it would show is that husband and wife addressed each other by a term which was proper to be used between many other people than husbands and wives, that they had no special term of address for each other. Father and son, brother and sister, all kinsfolk indeed among the Australians, were as to terms of address in this position 1 Supra, p. 299 note. 598 STUDIES IN ANCIENT HISTORY (see Studies in Ancient History, first series, pp. 310, 311). A similar use of the word for husband is noted by Mr. Bulmer in his account of the Wa-imbio tribe (Kamilaroi and Kurnai, p. 289), and in that case it carried no implication of rights or duties, for fidelity was expected from the wife. It is clear that Mr. Fison has found no basis in fact at all for this second branch of his hypothesis. He has evidence which, no doubt, is strongly suggestive of ancient promiscuity, by which is meant a social condition in which conjugal rights did not exist, but nothing whatever which indicates the modi- fied communism for which he wanted a basis, which is a com- munism involving marriage right the " communal marriage " of groups of men in the one division (of a tribe in two divisions) to the groups of women which were of the same generation with them in the other division. He has failed to show that men had some right of a conjugal sort over women other than the right which was acquired by marriage, and that is all he has attempted. And he has not, even from Mr. Lance, any suggestion of such a right being limited to women of the man's own generation. Moreover, he has no fact which suggests that marriage (or license either) was ever, among the Australians, confined to the tribe. And it is well known that the Australians, since they have been known to us, have everywhere taken women, by capture or otherwise, from strange tribes as often as they could get them. The hypothesis appears, then, to be " in the air." To examine Mr. Fison's method of establishing it would not be an agreeable task, but that seems to be unnecessary. It was devised to account for what Mr. Lewis Morgan has called the Turanian form of the classificatory system of relationships ; and evidence has already been adduced which shows clearly that Mr. Morgan (whom Mr. Fison follows) entirely misconceived that system and its uses. 1 If the misconception has to be admitted, there is no need to take trouble about the theories formed to account for it. And it will be found when the actual use of the classificatory terms is considered that if we are to seek the origin of those terms in some system of marriage 1 Studies in Ancient History, first series, pp. 305-312. APPENDIX 599 and the form of the family consequent upon it, there is no need to think of any forms but those which are well known by observation. It has seemed proper to notice what Mr. Fison adduces as favouring his hypothesis, because, from Mr. Fison's mode of working, there appears to be a little danger of intermarrying classes and the semi-conjugal rights ascribed to men over women whom they were merely free to marry being accepted by the unwary as Australian facts. Mr. Dawson's book, 1 which has already been mentioned incidentally, gives an account of the natives of the Port Fairy district, Western Victoria, a considerable population once (for Mr. Dawson calculates that the twenty-one tribes which met together once a year must have numbered 2500, and there were tribes on the coast which did not come to those meetings), but represented in 1880 by only fourteen survivors. Mr. Dawson tells us that he got his information from natives only ; and that he avoided putting suggestive or leading questions as much as possible, because " the natives, in their anxiety to please, are apt to coincide with the questioner, and thus assist him in arriving at wrong conclusions " a fact which many other observers have noted, and which should always be kept in view in reading Kamilaroi and Kurnai. To this work his book, from which theories are absent, offers a striking contrast. Of course it is not to be supposed that all the information he gathered is equally to be depended upon. The population consisted of local tribes, which Mr. Dawson believes to have numbered, on an average, about 120 apiece. Throughout all tribes, it was made up of five classes, which took their names from animals the names being Kuurokeitch (long -billed cockatoo), Kartpoerapp (pelican), Kappatch (the Banksian cockatoo), Kirtuuk (the boa-snake), and Kuunamit (the quail). Every one belonged to the class of his mother ; and marriage being forbidden between those who were con- sidered to be of one flesh, no man and woman of the same class could marry one another, however remote from each other their tribes might be. Moreover, the Kuurokeitch and the 1 Australian Aborigines. By James Dawson, Melbourne, 1881. 6oo STUDIES IN ANCIENT HISTORY Kartpoerapp classes were regarded (on what grounds there is nothing to indicate) as being so closely related that marriage between them could not be permitted ; so also the Kappatch and Kirtuuk classes ; while a Kuunamit was free to marry into any class but his own. The marriage law, therefore, as in all works about Australia except Mr. Fison's, was a law of prohibition only not a law which, while prohibiting men from marrying certain women, gave them rights over certain others. And Mr. Dawson gives a tradition which illustrates it admirably (it may be worth while to compare with this Mr. Fison's Theory of the Kurnai System), and which also shows that it was by exogamy (or prohibition of marriage between persons of the same flesh) and female kinship that these natives explained to themselves the fact that their tribes were made up of persons belonging to several different classes, to use Mr. Dawson's word that is, stocks, or bodies of blood (or flesh) kindred. The tradition ascribed the origin of all their tribes to a Kuurokeitch who had for wife a Kappaheear (feminine of Kappatch), and whose children therefore were Kappatch and Kappaheear, and could not marry each other. This made it necessary to introduce "fresh flesh," and wives were got from a distance ; and, children following the mother, five different stocks came by and by to be represented in the community. It is also stated that every person was considered to belong to his father's local tribe, and could not marry within it ; nor could a man marry a woman of the tribe from which his mother came, nor of his grandmother's tribe (by which the mother's mother's tribe seems to be meant), nor of an adjoin- ing tribe, nor of any tribe speaking his own dialect. It may be taken that these prohibitions (some of which may merely indicate what was usual) were accretions which, with time and change, were made to the marriage law, and that the original prohibition was that which is illustrated by the tradition just spoken of prohibition of marriage between persons of the same flesh kindred or female kinship stock. It was the duty of the chief (for there were chiefs, who are said to have had great power, and to have been treated with much observ- APPENDIX 601 ance) to ascertain that there was no flesh relationship between persons about to be married or betrothed to one another ; and that, even when this was not doubtful, his permission had to be " rewarded with presents." That a process of change had gone pretty far among these people is shown by many circumstances which Mr. Dawson mentions. He tells us, as to children, that " if the infant is a boy, the nearest relative is the father; if it is a girl, the nearest relative is the mother" (p. 38); and agreeably with this, that the first child of either sex is called after its father, and the second, if a daughter, after its mother (p. 41). This would show, what has not been noticed elsewhere in Australia, and is found very rarely anywhere, kinship to have been in a stage of transition a man belonging still to the widely diffused female kinship stock of his mother, and being bound not to marry a woman of its "flesh," and being nearer, neverthe- less, to his father than to his mother (while a woman was still nearer to her mother than to her father) ; and it is consistent with this that the tribal bond should have become so important that he was also forbidden to marry in the tribe of his father (which was his own) on the one hand, and in the tribe of his mother, and the tribe of his mother's mother, on the other. What Mr. Dawson tells us of the blood revenge, of inherit- ance, and similar matters, is also consistent with this, but can hardly be made to throw additional light upon the matter. There seems sufficient reason for believing, however, that these natives were in a transitional phase of kinship. Children were usually betrothed when just able to walk, the father of the girl making the proposal. After a betrothal the girl's mother and aunts (which aunts not indicated) might not look at, or speak to the boy ; and they used the " turn- tongue " language when they had to speak in each other's presence. The father being dead, the brother could give a girl away with consent of the uncle (father's brother, no doubt) ; while, if a girl had no male relative, the chief could give her away; but it is said there was danger (p. 35) of his keeping her to himself which, if it be taken as correct, would show that the rule against marrying in the tribe might be 602 STUDIES IN ANCIENT HISTOR Y waived, in the case of the chief, though it was enforced by him upon the others. Polygamy was carried so far by the old chiefs that " many young men are compelled to remain bachelors, the native word for which means to look out; while an old warrior may have five or six of the finest young women of the tribe for his wives." Both men and women among these people were orna- mented by cicatrices, " arranged in lines and figures, according to the taste and custom of the tribe " ; but Mr. Dawson takes it that these were ornamental merely. The bodies of relatives of either sex who had lost their lives by violence were eaten, " as a mark of affectionate respect, in solemn service of mourning for the dead." The flesh of enemies was never eaten, nor that of members of other tribes. Having given a tradition preserved by Mr. Dawson to show that the Port Fairy natives, assuming the existence of totem families or clans, having female kinship, in which marriage, otherwise free, was forbidden on the score of kindred, explained to themselves the composition of their tribes as being the result of their exogamy and their system of female kinship, it seems worth while to point out that it is the same constitution of society which is disclosed in what Mr. Fison has termed the Murdu legend (Kamilaroi and Kurnai, p. 25). This was first published in a pamphlet by Mr. Samuel Gason on the Dieyeri tribe (Cooper's Creek), and is quoted therefrom by Mr. Fison. It sets forth that, " after the creation, brothers, sisters, and others of the closest kin intermarried promiscuously " ; that " the evil effects of these alliances " became manifest ; that a council of chiefs was held to consider how to get rid of them ; and that the result was a petition to the Muramura (good spirit), who ordered " that the tribe should be divided into branches, and distinguished one from another by different names, after objects animate and inanimate, such as dogs, mice, emu, rain, iguana, and so forth ; the members of any such branch not to intermarry, but with permission for one branch to mingle with another. Thus the son of a dog might not marry the daughter of a dog, but APPENDIX 603 I either might form an alliance with a mouse, rat, or other family." This, like the Port Fairy tradition, is an attempt to account for the constitution of a population composed of totem clans interfused with one another, within each of which marriage, otherwise free, was forbidden. The difference between the two is that this is purely irrational; while the Port Fairy explanation is rational as far as it goes and it was no proof of irrationality to take totem families, with exogamy and female kinship, for granted as a basis to start from. The point to be noticed, however, is that both take account of that structure of society which has been shown above to be the prevailing one in Australia, and which, with variations in kinship and changes consequent thereupon, is found throughout Australia, wherever we have trustworthy evidence. The Murdu legend, equally with the Port Fairy tradition, is dead against the theories of Mr. Fison. It shows us no " intermarrying classes," with a marriage law which, while prohibiting marriage in the one, gave men conjugal rights over the women of the other, which rights, when the " class " came to include totems, were restricted by the totems. It shows us a community made up of totem clans, with a marriage law of prohibition merely, marriage being prohibited between all persons of the same totem. It seems worth while here to repeat our information about the Kamilaroi being still, and being likely to continue, imperfect that it is simply this marriage law which was indicated to Mr. Lance, to explain how an Ippai could have an Ippata for his wife. "This Ippai is not a Blacksnake, but an Opossum ; that explains it," Ippai Opossum might not marry Ippata Opossum, but there was nothing to prevent him from marrying Ippata Blacksnake. It will be found that it is, in general, perfectly clear that what Mr. Fison and his correspondents speak of as " classes " are totem clans or families merely, and that, wherever this is not clear, the constitution of the " class " is involved in obscurity. It is interesting to note that Mr. Lewis Morgan found in the opening part of Mr. Gason's story a " basis of probability " for his hypothesis of the consanguine family, taking that as 604 STUDIES IN ANCIENT HISTORY giving " an accepted and perpetuated native legend " of what happened after the creation (Kamilaroi and Kurnai, p. 4). It does not seern, however, to countenance that hypothesis, which is an hypothesis of limited and systematised communism sub- sisting in a body of kindred, all the men and women of the same generation being married to one another " in a group." It is interesting, too, to note that Mr. Fison who does not accept the hypothesis of the consanguine family, upon which Mr. Morgan's other theories are built up, and prefers to begin with " intermarrying classes," having the system of communism which had been thought out by Mr. Morgan inclines towards the Muramura account of the origin of totem kindreds as affording the best explanation of the origin of his " classes " (Ibid. p. 161). Mr. Gason goes on to mention that the marriage system shown in the legend was still observed among the Dieyeri, and that the first question asked of a stranger was, " What murdu ? i.e. of what family are you ? " And we learn from another authority (the Eev. H. Vogelsang information procured by Mr. Howitt Kamilaroi and Kurnai, p. 52) that the question Minna Murdu was connected with " eating and hospitality." " For instance, when a stranger blackfellow arrives here, the question is Minna Murdu ? What are you ? Kangaroo, or Eat, or Mouse, or whatever else it may be. All those of the same name go to the same camp, eat together, live together, even lend each other their women. Even alien blackfellows, from a distance of 300 or 400 miles, are thus hospitably entertained." " Our tribe, the Dieyeri," Mr. Vogelsang con- tinues, "have different names for their Murdus from those of the neighbouring tribes, but they can always understand each other " that is, they do not fail to find out what the Murdu is. It thus appears that, among the Dieyeri, men acknowledged a stranger from whatever distance, and though speaking a strange language, to be of their kindred as soon as it was ascertained that he was of the same Murdu or totem family with them ; that thereupon they entertained him hospitably ; and that the lending of women was a part of hospitality. Of course a woman whom a man of the stranger's totem might have APPENDIX 605 for wife was a woman with whom the stranger was free to cohabit. It may be added that the tribes consisting of Muquarra and Kilparra (eagle-hawk and crow, each comprising several totems), which Mr. Fison ranks as intermarrying classes, have a rather interesting totem tradition of their own (The Aborigines of Victoria, vol. i. p. 423). "They believe that the beings who created all things had severally the form of the Crow and the Eagle. There was continual war between these two beings, but peace was made at length. They agreed that the Murray blacks should be divided into two classes the Muquarra or Eagle-hawk, and the Kilparra or Crow. The conflict that was waged between the rival powers is preserved in song . . . the meaning of which is : Strike the Crow on the knee ; I will spear his father. The war was maintained with great vigour for a length of time. The Crow took every possible advantage of his nobler foe, the Eagle ; but the latter generally had ample revenge for injuries and insults. Out of these enmities and final agreement arose the two classes, and thence a law governing marriages amongst these classes." It is a tradition, obviously conceived in the Eagle interest, of a state of war between rival powers having been followed by peace and peaceful relations, which has got from the narrator, as most things have done in the work from which it is taken, a tinge of Mr. Fison's theories. But what it suggests is, not that Crow and Eagle agreed to divide one tribe into two, with a view to the better regulation of marriage, but that Crow and Eagle or Eagle-hawk were tribes (and they might have been constituted in the ordinary Australian way) which long waged war against each other, and that at length there came peace, and then their complete interfusion by means of friendly marriages. And, whatever such traditions may be worth, this account of Muquarra and Kilparra seems vastly preferable to Mr. Fison's. Printed by R. & R. CLARK, LIMITED, Edinburgh. THE UNIVERSITY OF CALIFORNIA LIBRARY