BSSi 
 
 
 
ESSAYS 
 
 HILOLOGICAL AND CRITICAL 
 
 SELECTED FROM THE PAPERS 
 
 OF 
 
 JAMES HADLEY, LL.D. 
 
 Professor of Greek in Yale College ; President of tJie American 
 Oriental Society ; Vice-President of the American Phi 
 lological Association ; Member of the National 
 Academy of Sciences, etc., etc. 
 
 NEW YORK 
 
 HOLT & WILLIAMS 
 1873 
 
Entered according to Act of Congress, in the year 1873, by 
 
 ANNE T. HADLEY, 
 In the Office of the Librarian of Congress, at Washington. 
 
 POOLE & MACLATJCHLAN, 
 
 PRINTERS AND BOOKBINDERS, 
 
 205-213 East Twelfth St., 
 
 NEW YORK. 
 
PREFACE. 
 
 THE articles contained in this volume have been selected 
 rom among the papers, published and unpublished, left by 
 3 rofessor Hadley, and are put forth by his family and friends, 
 rle died on the I4th of November, 1872, in the fifty-second 
 /ear of his age. He had been for twenty-seven years an in- 
 >tructor in Yale College, entering its service as tutor in 1845, 
 Deing appointed assistant professor of Greek in 1848, and 
 succeeding President Woolsey in the principal charge of that 
 department in 1851. A sketch of his life, with an estimate of 
 lis character as a man and as a scholar, is given by President 
 Dorter in the New Englander for January, 18/3, and has also 
 Deen issued as a pamphlet, accompanied by an account of his 
 sarly studies, drawn up by himself many years ago, and by a 
 nearly complete list of his literary productions, with the times 
 ind places of their composition and publication. 
 
 Of the papers here given, the following have already ap- 
 )eared : Art. I., on the Ionian Migration (first part), in the 
 ournal of the American Oriental Society, vol. v., 1856; 
 Art. IV., on Bekker s Digammated Homer, in the Transac- 
 ions of the Connecticut Academy, vol. i., 1866; Arts. VI. 
 and VII., on Greek Accent and Greek Pronunciation in the 
 .enth century, and Art. XV., on English Vowel Quantity, in 
 .he Transactions of the American Philological Association, 
 
 59-72 ; Art. XIV., on Ellis s Early English Pronunciation, 
 n the North American Review, vol. ex., 1870; Arts. XVI. 
 and XVII. , on Tennyson s Princess and the Number Seven, 
 n the New Englander, vols. vii., 1849, and xvi., 1858; and 
 Art. XX., on the Language of Palestine at the Time of 
 !hrist, in Messrs. Hackett and Abbot s edition of Smith s 
 Bible Dictionary, vol. ii., 1868. Thanks are due, and arc here 
 
iv PREFACE. 
 
 publicly given, to the proprietors of the several works named, 
 for kindly permitting the republications made in this volume. 
 Art. VI., it may be added, has been reproduced, in German 
 translation, in Professor G. Curtius s Studien Z2ir Griecliischen 
 und Lateinischen Graunnatik, vol. v., 1872. To Art. XV., 
 on English Vowel Quantity, there is added in the original a 
 " list of all the words, not already noticed, which appear in 
 the Ormulum with the same vowel quantity which they have 
 in modern English," occupying twelve pages; this list it w r as 
 thought better to omit here. 
 
 Of the remaining essays, the larger number (together with 
 some of the preceding) were prepared for the American Ori 
 ental Society, and read at its meetings ; such are Arts. I. 
 (second part), II., III., VIII., IX., X., XIII., and XIX.; 
 others namely, Arts. V. and XII. were presented to the 
 Classical and Philological Society of Yale College ; and one, 
 Art. XL, was brought before the Connecticut Academy, 
 but at a meeting of which the record is lost ; its precise date 
 is therefore uncertain. The XVIIIth article, the Class De 
 cisions, calls for a word of explanation. As one of the Col 
 lege instructors and officers, Professor Hndley was accustomed 
 to preside at weekly "disputes" in the Junior class the 
 reading of essays upon selected questions and to review and 
 sum up the arguments brought forward, giving his own 
 opinions on the subject discussed. In preparation for this 
 exercise, it was his habit often to write out beforehand his 
 views, not for reading to the class, but as part of the ground 
 work of his " decision ; " and of such written arguments he 
 has left a considerable mass (about a hundred and fifty in 
 number). These exercises were always greatly enjoyed by 
 the class ; and it is partly in response to wishes warmly ex 
 pressed from various quarters that a selection from the de 
 cisions is here included. They must be taken for what they 
 are thoroughly off-hand productions, written ciirrcntc calaino 
 and without special preparation, and in no case revised or 
 corrected. Their author, doubtless, would never have thought 
 of such a thing as making any of them public ; but it seemed 
 highly proper that, in a volume intended as a memorial and 
 illustration of his scholarly life, they should not be passed 
 
PREFACE. v 
 
 over. They will show, on the one hand, the zeal and devo 
 tion with which he performed his College duties, and, on the 
 other, the rapidity and precision with which he thought and 
 wrote ; they will be valued especially by his numerous pupils, 
 and by all the graduates and friends of Yale College. 
 
 There are perhaps other essays in the volume which their 
 author might not himself have chosen to publish. There are 
 certainly some which he would not have let pass from his 
 hands without a thorough revision. This is especially true of 
 the essays in comparative philology (Arts. II. and IX. -XI.) ; 
 the progress of investigation and deduction in this department 
 is so rapid that no one s views can remain long absolutely 
 unmodified. But although these essays need not be taken as 
 representing in every item Professor Hadley s final opinions, 
 it has not been thought worth while to alter or annotate them ; 
 as popular summaries they are authoritative, and by the 
 special student who shall criticise them in detail they need 
 only to have their date duly considered. 
 
 The plan of arrangement of the volume has been to put 
 first the Greek articles, next those in general philology, then 
 those upon English, and last the more miscellaneous and 
 lighter papers ; each division being arranged nearly chrono 
 logically. Taken together, they will measurably illustrate the 
 wide range and varied direction of their author s studies but 
 with the exception, especially, of three departments : mathe 
 matics, where he early displayed an ability that bade fair, if 
 he had continued his devotion to it, to place him among the 
 foremost men of the day in that branch of science ; Celtic 
 philology, in which he was a proficient ; and the Roman Law, 
 his academical lectures on which are published in a separate 
 volume, simultaneously \vith this. 
 
 And yet it may also be -said with truth that the specimens 
 of his work here presented are far from fully exhibiting his 
 powers and acquirements. He was a man who put a larger 
 share of himself into his personal teachings, and a smaller into 
 what he gave to the world at large, than most others. He 
 was absolutely without the desire to shine, and he needed the 
 impulse of a more imperative call than he ever received to 
 draw him fully out of his modest retirement. It is no proper 
 
vi PREFACE. 
 
 place here to extol his abilities before the wider public which 
 will, as it has the right to do, judge them by his recorded 
 work, and which will do them justice ; and yet I cannot close 
 this preface without a word of recognition of them, of appre 
 ciation of his worth and sorrow at his loss. In extent and 
 accuracy of knowledge, in retentiveness and readiness of 
 memory, in penetration and justness of judgment, I have 
 never met his equal. Whatever others may have done, he 
 was, in the opinion of all who knew him most fully, America s 
 best and soundest philologist, and his death, in the maturity 
 and highest activity of his powers, is a national calamity, a 
 calamity to the world of scholars. Especially painful and 
 irreparable to me has been the loss of a fellow-student to 
 whom I had for twenty-three years looked up as a teacher, a 
 colleague and friend whose counsel and sympathy I had ?o 
 long enjoyed, and the purity and elevation of whose, character 
 had been to me a model of human excellence ; and I have 
 found it a very sad pleasure to assist his family in arranging 
 and publishing this memorial of his high and varied scholar 
 ship. 
 
 W. D. WHITNEY. 
 
 YALE COLLEGE, NEW HAVEN, April, 1873. 
 
CONTENTS. 
 
 PAGE 
 
 Preface Ill 
 
 I. The Ionian Migration : 
 
 1. The lonians before the Ionian Migration I 
 
 2. Recent Discussion and Opinion concerning the Ionian Migra 
 
 tion 20 
 
 II. The Root prach in the Greek Language 37 
 
 III. The Greek Genitive as an Ablative Case 44 
 
 IV. On Bekker s Digammated Text of Homer 56 
 
 V. On Ancient Greek Rhythm and Metre Si 
 
 VI. On the Nature and Theory of the Greek Accent no 
 
 VII. On the Byzantine Greek Pronunciation of the Tenth Century, as illus 
 trated by a Manuscript in the Bodleian Library 128 
 
 VIII. Ross on Italicans and Greeks , 141 
 
 IX. On Indo-European Aspirate Mutes 168 
 
 X. Oh the Formation of Indo-European Futures 184 
 
 XL On Passive Formations 199 
 
 XII. Remarks on the Uses of the Latin Subjunctive 215 
 
 XIII. On the Origin of the English Possessive Case 221 
 
 XIV. Ellis s Early English Pronunciation 240 
 
 XV. On English Vowel Quantity in the Thirteenth Century and in the 
 
 Nineteenth 263 
 
 XVI. Tennyson s Princess 296 
 
 XVII. The Number Seven 325 
 
 XVIII. Decisions of College Class Disputes : 
 
 1. Are the Writings of Lord Byron Immoral in their Tendency?.. 346 
 
 2. Is Ancient Eloquence Superior to Modern ? 349 
 
 3. Is a Reform Desirable in the Method of Writing the English 
 
 Language ? 351 
 
 4. Was Civil Liberty in Europe promoted by the Career of Napo 
 
 leon ? 356 
 
 5. Is Europe tending to Republicanism ? 3^1 
 
 6. Should Day-Dreaming be indulged in ? 369 
 
 7. Can Immortality be shown from the Light of Nature ? 373 
 
 8. Is an exclusively Vegetable Diet Advantageous? 379 
 
 XIX. On the Hebrew Chronology from Moses to Solomon 385 
 
 XX. On the Language of Palestine at the Time of Christ 403 
 
 Index 419 
 
I. 
 
 THE IONIANS BEFORE THE IONIAN MIGRATION.* 
 
 1856. 
 
 THE name of Ernst Curtius is well known to American 
 scholars from his excellent volumes on the geography of 
 Peloponnesus, as well as several smaller works. His essay, 
 published last year under the title above given, presents novel 
 and interesting views in regard to the earliest times of Greece. 
 I propose in this article to give a brief statement of those 
 views, with some criticism of the arguments by which they 
 are supported. It will appear, as I proceed, that the subject, 
 though belonging to Greek history, is one which has its claims 
 upon the attention of an Oriental Society. 
 
 At the outset of authentic Greek history, we find the west 
 ern coast of Asia Minor, with the neighboring islands, occu 
 pied by Greeks, undoubted members of the Hellenic body. 
 Of these the largest portion, extending on the mainland from 
 the mouth of the Hermus to that of the Mreander, and hold 
 ing the important islands of Chios and Samos, called them 
 selves lonians a name which belonged to them in common 
 with the inhabitants of Attica and Eubcea on the west of the 
 yEgean, as well as the island group of the Cyclades in the 
 centre of that sea. The Asiatic lonians, after passing through 
 a long career of independence and prosperity, were incorpo 
 rated about 550 B. C. into the kingdom of the Lydian Croesus, 
 along with which they came only a few years later into the 
 more comprehensive and permanent empire of the Persian 
 Cyrus. This was the close of their independent existence. 
 For its commencement we must go back to the mythic period 
 at least to a period lying on the debatable ground between 
 history and mythus. In the traditions of the Greeks as to 
 their own early times, we find the origin of the Asiatic lo- 
 
 * Die lonier vor der Ionise hen Wanderung, von Ernst Curtius. Berlin, 1855. 
 8vo. pp. 56. 
 I 
 
2 IONIAN MIGRA TION. 
 
 nians traced up to an ancient colonization from the west, by 
 emigrants who came from European Greece. This emigra 
 tion is represented as one consequence, among many, of the 
 great event which stands on the threshold of Greek history, 
 itself obscurely seen, but sufficiently recognized as the cause 
 or occasion of almost all we see in early Greece the invasion 
 and conquest of Peloponnesus by the Dorians. The story is 
 briefly this: I condense from Grote. "A multitude of 
 refugees from various parts of Greece, fleeing before the Dori 
 an invaders, sought shelter in Attica. Alarmed by the grow 
 ing population of that territory, the Dorians of Peloponnesus 
 marched against it with a powerful army ; but finding that 
 victory had been assured to the Athenians by the generous 
 self-devotion of their king Codrus, they gave up the enterprise 
 and returned home. The Athenians on the death of Codrus 
 abolished the kingship ; but his descendants for several gener 
 ations held the supreme power as archons for life. His two 
 sons, Medon and Neileus, having quarreled about the succes 
 sion, the Delphian oracle decided in favor of the former ; 
 whereupon the latter, affronted at the preference, resolved to 
 seek a new home. There were at this moment many dispos 
 sessed sections of Greeks, and an adventitious population, 
 accumulated in Athens, who were anxious for settlements 
 beyond sea. The expeditions which now set forth to cross 
 the /Egean, chiefly under the conduct of members of the 
 Codrid family, composed collectively the memorable Ionic 
 Emigration, of which the lonians, recently expelled from 
 Peloponnesus, formed only a part ; for we hear of many quite 
 distinct races, some renowned in legend, who withdrew from 
 Greece amidst this assemblage of colonists. The Kadmeiahs, 
 the Minyae of Orchomenus, the Abantes of Eubcea, the Dry- 
 opes ; the Molossi, the Phokians, the Boeotians, the Arcadian 
 Pelasgians, and even the Dorians of Epidaurus are repre 
 sented as furnishing each a proportion of the crews of those 
 emigrant vessels. At the same time other mythic families 
 beside the Codrids, the lineage of Neleus and Nestor, took 
 part in the expedition. Herodotus mentions Lykian chiefs, 
 descendants of Glaukus, and Pausanias tells us of Philotas a 
 descendant of Peneleos, who went at the head of a body of 
 
IONIAN MIGRA TION. 3 
 
 Thcbans. Prokles, the chief who conducted the Ionic emi 
 grants from Epidaurus to .Samos, was said to be of the lineage 
 of Ion son of Xuthus. The results were not unworthy of 
 this great gathering of chiefs and races. The Cyclades were 
 colonized, as also the large islands of Samos and Chios near 
 the Asiatic shore, while ten different cities on the coast of 
 Asia Minor, from Miletus on the south to Phokasa on the 
 north, were founded, and all adopted the Ionic name. Athens 
 was the metropolis or mother city of all of them : Androklus 
 and Neileus, the CEkists of Ephesus and Miletus, and proba 
 bly other QEkists also, started from the Prytaneium at Athens 
 with those solemnities, religious and political, which usually 
 marked the departure of a swarm of Grecian colonists." Such 
 is the traditional account. The main fact contained in it may 
 be regarded as certain that after the Dorian conquest of 
 southern Greece, and in consequence of that event, large 
 bodies of Greeks, the most important part of them lonians, 
 set forth, chiefly from the coast of Attica, to cross the ./Egean 
 Sea. The time of this migration may be set down by a loose 
 approximation at IOOO years before our era. 
 
 Now the principal thesis of Curtius in his Essay is this : 
 that in the migration just described, the lonians of Greece 
 were going home, to their own country and kindred. It was 
 the returning emigration to a land from which, ages before, 
 their fathers had passed over into Greece and not only that, 
 but a land which had never ceased to be occupied by the 
 same race, by a people of Ionian name and lineage. They 
 found, on arriving in Asia, not only Dardanians, Carians, Ly- 
 cians, and other tribes, which Curtius regards as differing 
 not very widely from lonians in language and culture : but 
 they found there lonians, identified with themselves by virtue 
 of the common name, origin, and traditions. They found in 
 fact the lonians the principal branch as well as the elder of 
 their race who in these Asiatic seats had risen to a height of 
 achievement and reputation not yet equalled by any Greeks of 
 Europe. Let us, however, trace the theory more in detail, 
 going back to its remote starting-point in the past, beyond 
 the reach of history, beyond the reach even of mythus, where 
 only ethnographic science can furnish any glimmering of light. 
 
4 IONIAN MIGRA TION. 
 
 The primitive Aryan colonization, flowing westward from 
 Armenia into Asia Minor, filled the elevated plateaux of that 
 peninsula with Phrygian races. Here the Greeks, long iden 
 tified with the Phrygian stock, first begin to be distinguished 
 as Greeks, with a stamp and nationality of their own. Here 
 they develop what must be considered as the common type 
 of Hellenism in language and character. But almost from 
 the beginning they divide themselves into two great sections. 
 The one is that afterwards known in history as the Ionian. 
 The other includes the remaining fractions of the Greek na 
 tion : we might call it Hellenic in a narrower sense, as being 
 first to assume the Hellenic name : it is sometimes called 
 ^Eolo-Dorian, from the designations of its leading members 
 in the historic period. After a time these sections part com 
 pany. The latter or Hellenic section break up from Asia, 
 cross the Hellespont and Propontis, and find new seats in 
 the mountains of Thrace and Macedonia. Here they remain 
 in isolated Alpine valleys, forming their separate local con 
 stitutions, until, dislodged by new movements of population, 
 and pressed southward, they make their appearance in dif 
 ferent masses, as ^Eolians, Dorians, Achaeans, in Northern 
 Greece. Here again in the course of time new causes arise, 
 which carry portions at least of these tribes still further in the 
 same direction, into Central and Southern Greece. Hence 
 the occupation of Peloponnesus by the Achseans, whom the 
 Homeric poems represent to us as seated in that territory 
 and exercising full ascendancy. And hence too the later and 
 far more important conquest of the same territory by the 
 Dorians and their auxiliaries. 
 
 The lonians meanwhile remain in Asia Minor, but no 
 longer in the highlands of the interior. Descending gradu 
 ally along the great river-valleys, they at length reach the 
 ^Egean sea, and then, spreading themselves northward and 
 southward, occupy the whole western coast possessing thus 
 a territory distinguished alike for the richness of its soil and 
 the genial beauty of its climate. They are closely connected 
 here by proximity and by intercourse with other tribes, such 
 as the Dardanians, Lycians, Carians, Leleges, from whom in 
 fact they are not separated by any broad lines of ethnical 
 
IONIAN MIGRA TION. 5 
 
 distinction. Under these circumstances they enter upon a 
 career of activity and culture, which appears to have received 
 its impulse from the Phoenicians, and to have been shared in, 
 more or less, by the other tribes just mentioned. Visited at 
 first by the Phoenicians for the purposes of trade, they soon 
 learned from them the art of navigation, and set up business 
 on their own account, as the rivals of their late masters. As 
 sociated with the Phoenicians in many parts of the ^Egean, 
 and supplanting them in others, they have become inextricably 
 confused with them in the traditions of the Greeks. The 
 Ionian myth which represents Byblus, one of the oldest 
 Phoenician cities, as the daughter of Miletus, shows perhaps 
 that the lonians gained a foothold even on the coast of Syria ; 
 at any rate, it is a proof of close connection between these 
 two maritime peoples. There is clearer evidence to show 
 that the lonians visited the coast of Egypt, and even estab 
 lished settlements, more or less permanent, in the marshy 
 Delta of the Nile. This was regarded by the Egyptians 
 themselves almost as foreign territory ; since we find that 
 Psammetichus the same prince, who, perhaps a thousand 
 years later, opened the whole country to the Greeks when 
 banished from Egypt, took refuge in the Delta. And the 
 men of brass who were announced to Psammetichus, while 
 there, as having just made their appearance, and who proved 
 to be a party of Ionian rovers recently landed, were but a 
 specimen of their own countrymen, who, a thousand years 
 earlier, made repeated descents upon the same coast for the 
 mingled purposes of traffic and plunder. But the attention 
 of the lonians was naturally directed more to the west, 
 Crossing the ./Egean Sea, they occupy first the Cyclades, and 
 then Euboea and Attica. They establish their settlements on 
 the Pagasaean Gulf, and on both sides of the Euripus. Traces 
 of them are found along the whole eastern coast of Pelopon 
 nesus, in Corinth, Epidaurus, Trcezen, Argos, and even in 
 the island of Cythera. Passing over the Isthmus, they ap 
 pear in the Corinthian Gulf, where we find them in southern 
 Phocis, and much more in northern Peloponnesus, in the dis 
 trict afterward called Achaia. From thence they spread 
 southward over Elis and Messene in western Peloponnesus; 
 
6 IONIAN MIGRA TION. 
 
 and having thus reached the Ionian Sea, they occupy the 
 Ithacan islands, and extend themselves northward to the 
 island of Corcyra, and the coasts of Epirus and Illyria. 
 More than this : in the mythic wanderings of yEneas, Curtius 
 would recognize a traditionary representation of Ionian set 
 tlement, which must then have stretched along the western 
 coast of Italy from Eryx to the mouth of the Tiber. Even 
 in Sardinia, he considers the name of a people called the 
 lolaeans, and of their founder lolaos, as giving evidence of 
 early Ionian colonization. 
 
 Throughout the course of these migrations, the lonians carry 
 with them the culture of the vine and the worship of the wine- 
 god Dionysus, Everywhere we find them settling along the 
 coasts, and showing an especial preference for the rich, though 
 marshy, alluvium at the mouth of rivers. Occasionally, how 
 ever, they follow up a river-valley quite into the interior of a 
 district, as in Bceotia, where the Asopus leads them to the in 
 land city of Thebes. Everywhere wandering in ships, they 
 wander without women ; and hence their colonization appears 
 as the establishment of a few foreign settlers among a native 
 population, whom they do not attempt to dispossess, but ex 
 ercise over them the natural ascendancy of superior ability 
 and civilization. Thus in Attica there is no change of popu 
 lation : the primitive people, whom Greek tradition names 
 Pelasgi, remain in their old seats, unchanged except as they 
 are civilized, Ionized by trie foreigners from Asia. The Egyp 
 tian Cecrops, the mythic author of civilization in Attica, is 
 no proper Egyptian, but an Ionian, who had become domi 
 ciled in Egypt, A similar view is taken of Danaus the 
 Egyptian founder of Argos. These traditions of early con 
 nections between Egypt and Greece are, in the view of Curtius, 
 too deeply rooted and too widely ramified to have sprung up, 
 as K. O. Miiller assumed, after the comparatively recent pe 
 riod when the Egyptians under Psammetichus came into 
 closer relations with the Greeks. Yet, on the other hand, it 
 seems equally evident that no influence strictly and properly 
 Egyptian could have had a leading part in moulding the civi 
 lization, substantially homogeneous and independent, of early 
 Greece. The difficulty finds its solution in the view that these 
 
IONIAN MIGRA TION. J 
 
 Egyptian settlers, who figure in tradition, were lonians, who 
 had found a residence in Egypt and came from thence to 
 Greece. The Phoenician Cadmus and his colonization of 
 Thebes are treated in the same way. Curtius does not deny, 
 indeed, that there were in Greece, to a greater or less extent, 
 ancient settlements of native Phoenicians ; but he maintains 
 confidently that no such alien Semitic settlers could have 
 gained historic importance as founders of royal or sacerdotal 
 families. It is of course still easier to connect the Phrygian 
 Pelops, and his immigration into the peninsula which took his 
 name, with the colonial extension of the Ionian race. The 
 Argonautic expedition is a story of Ionian adventure. Its 
 leader, who comes into Thessaly an unknown wanderer, bears 
 a name, Jason ( Ido-cov), which stamps him as Ionian ; and 
 its Thessalian starting-point, lolcos or laolcos, is with great 
 probability explained as meaning the naval station of the 
 lonians. 
 
 Here on the coast of Thessaly the lonians are again brought 
 into contact with their brethren of ^Eolo-Dorian descent. 
 After a local separation of generations and centuries, these 
 long-sundered sections of the Grecian people are brought once 
 more into local connection. The most conspicuous result is 
 the formation of the celebrated Amphictyonic League, the 
 oldest and largest and most influential of the Grecian Amphic- 
 tyonies. It is a religious association of Thessalian tribes 
 (neighbors to one another, ^Ajju^ncriove^) for the common 
 worship of the god Apollo. The lonians, after being for a 
 long time worshippers preeminently of the god Poseidon, of 
 whom the western Greeks at that time knew as little as 
 of the element he ruled, had in their eastern home re 
 ceived the Apollo-worship a new religion, as Curtius calls 
 it, which everywhere exercised a transforming and inspiring 
 influence on its converts. Zealously devoted to its propa 
 gation, they introduced it among their brethren of Thessaly. 
 Thus in the Amphictyonic deity we find a proof of Ionian in 
 fluence ; which appears further in the frequently-recurring 
 Ionian number twelve, as that of the confederate tribes. The 
 Amphictyonic League, though primarily a religious organiza 
 tion, expressed political aspirations, and worked toward po- 
 
8 IONIAN MIGRA TION. 
 
 litical results. It produced a feeling of closer union and of 
 common brotherhood among its members, which led to the 
 adoption of the Hellenic name as a common designation for 
 the united Amphictyonic people. Hellen in the myths is 
 either father or ^brother of Amphictyon. Hence the tribes 
 of Macedonia and Epirus, however closely resembling the 
 Hellenes, never received the Hellenic name, which belongs 
 only to the Amphictyonic tribes and the districts which come 
 under their control or influence. 
 
 Although Ionian influence, as we have seen, was predomi 
 nant in the origin of the Delphic Amphictyony, that first re 
 union and organization of the Greek races, yet the relative 
 weight of parties did not always remain the same. A re 
 action at length commenced a reaction of the older tribes in 
 the interior against the newer occupants of the sea -board of 
 the western Greeks against their emigrant brethren from the 
 east. The ruder tribes of Thessaly, receiving the imported 
 civilization of the lonians, come at length to feel themselves 
 the equals of their late instructors, and can no longer brook 
 the ascendancy to which they at first submitted. Hence a 
 decided revolution in the political state of Greece, proceeding 
 from Thessaly, and having for its ultimate result the almost 
 complete expulsion of the lonians from European Greece. 
 But this revolution is the work of ages, and has its different 
 epochs, according to the different races who successively ap 
 pear to carry it forward. 
 
 First, the ^Eolians, who are represented in the traditions as 
 arising from a mixture of the inland tribes with the maritime 
 population of the sea-board. Though in fact supplanting the 
 lonians, they do not appear as their opponents or even as 
 their rivals. The yEolids are themselves bearers of Ionian 
 cultivation and the worship of Poseidon ; their royal seats, as 
 lolcos and Corinth, are stations of Ionian colonization ; their 
 mythic heroes, as Jason and Sisyphus, are representatives of 
 Asiatic culture. 
 
 Second, the Achaeans are likewise in many ways closely 
 connected with the lonians, as the mythus intimates, when it 
 makes both Ion and Achaeus sons of Apollo. Yet the mili 
 tary exaltation of the Achaeans is the first great blow to Ionian 
 
IONIAN MIGRA TION. 9 
 
 preponderance in Greece. While the Achaeans of Phthiotis 
 press on toward the sea-coast of Thessaly, the other branch 
 of that people conquer the Peloponnesus, form new states 
 there hostile to the lonians, whom they expel from Troezen 
 and other parts of Argolis, and with fleets of their own begin 
 those struggles with the tribes of Asia Minor which are com 
 memorated in the legends of the Trojan war a war in which 
 the Ionian peoples, as the Athenians, take scarcely any part, 
 while heroes akin to the lonians, as Palamedes and Odysseus, 
 enter into it with reluctance. 
 
 Third, the Dorians, a people much more alien to the lonians 
 and much more independent of their influence ; a people who 
 adhere with tenacity to their original peculiarities of life and 
 character ; in them was first seen the full native vigor of the 
 mountain tribes. Breaking up from their scats in Mt. CEta, 
 they cross the Corinthian Gulf, by a gradual conquest over 
 throw the Achaean power, and make themselves masters of 
 nearly all Peloponnesus. As they advance, the lonians every 
 where lose ground ; on all sides they are driven back to their 
 ships : and now begins a great retreat of the lonians from 
 their settlements in the west ; a great return to their mother 
 country on the east of the ^Egean Sea. Only in Attica do 
 they at last succeed in making an effectual stand : thus main 
 taining a foothold in European Greece, and preventing Hel 
 lenic history from being again divided, as it had been, ages 
 before, between two distinct races upon opposite sides of the 
 ^Egean. Even in Asia Minor they are not by themselves. 
 Achaean and Dorian colonies reproduce there the collisions of 
 western Greece, keeping up a restless activity of mind, by 
 which Ionian art is stimulated to a rapid development, until 
 it puts forth its fairest blossom in the Homeric Epos. Still, 
 in the Dorian and ^Eolian districts of Asia Minor, the basis 
 of population remained essentially Ionian : and in the Ionian 
 revolt, as it is called, the whole people of the western coast, 
 from Lycia to the Propontis, rose as one people against the 
 barbarian conqueror. 
 
 Such is the theory of this ingenious and strikingly written 
 essay. Before taking up any points in the argument on which 
 it rests, we must observe that this idea of lonians in Asia pre- 
 
I O IONIAN MIGRA TION. 
 
 vious to the Ionian Migration is wholly foreign to the mythic 
 or semi-historical traditions of the Greeks themselves. It may 
 be shown, perhaps, that in those traditions there are state 
 ments which imply the existence of a primitive Ionian people 
 in that region ; statements which cannot be explained on any 
 other supposition. But it is confessedly true that the tradi 
 tions conveyed no such idea to the ancient Greeks who had 
 them ; certainly not, after they had assumed the forms in 
 which they have come down to us. 
 
 In looking at the evidence on which our author relies, to 
 sustain a proposition of which no memory is found in the 
 most ancient literature and tradition of Greece, it is natural to 
 inquire, first, whether any testimony can be gleaned from early 
 Oriental sources. Here Curtius finds a confirmation of his 
 views in the name given to the Greeks by all the ancient 
 nations of the East. It is well known that the common form 
 "Icovei is made by a contraction of the earlier Idoves ; and there 
 is great reason to believe that this latter form had originally a 
 medial digamma, and was pronounced laFoves, sing. larcov. 
 Now the Greeks are called by the Indians Javanas, by the 
 Hebrews J avail, by the Persians Juna or Jauna, in Aramaic 
 Jaunojo, in Arabic Jaumini, in Armenian Juin, and in Coptic 
 Uinin. It can hardly be doubted that these are all forms of 
 one and the same name ; and that this is no other than larwv 
 or Ia/<we?, the special name of the Ionian Greeks. We may 
 not unreasonably suppose that it was the Phoenicians who first 
 applied this name as a common designation for the whole 
 Greek people, and that the widely-extended commerce of the 
 Phoenicians was the means of its diffusion throughout Asia. 
 It is further probable that the Phoenicians had the name in 
 this use of it before the time of the Ionian Migration. We 
 find it in the tenth chapter of the book of Genesis, in the list 
 of Noachids, where it undoubtedly refers, not to a part of 
 the Greeks, but to the whole people. This document, if of 
 Mosaic origin, is at least thirteen centuries older than the 
 Christian era : while, even among those who deny its Mosaic 
 origin, it is allowed by all the sounder critics to be older than 
 the division of the Hebrew monarchy. But this occurred 
 about 1000 B, C,, perhaps at the same time with the Ionian 
 
IONIAN MIGRA TION. 1 1 
 
 Migration, probably not later than that event. What shall 
 we conclude, then, from this early use of the Ionian name as 
 a designation for the whole Hellenic people ? Curtius replies 
 the fact is inexplicable unless we assume that, of all the 
 Grecian tribes, the Ionian was the first which became known 
 to the Orientals ; it must have existed as their neighbor and 
 carried on intercourse with them by land and water, not 
 simply as early or a little earlier than yEolians and Dorians, 
 but long before all other Greeks. It appears to me that this 
 lancfuaee overstates the case. On the coast of Syria at the 
 
 o o -^ 
 
 present day all Europeans are Franks. Yet other nations of 
 Europe beside the French were represented in the first cru 
 sade, and still more in the second, which followed only a half 
 century later. On the other hand, the Europeans have given 
 the common name of Tartars to the nomadic tribes east of the 
 Caspian Sea. Yet it is certain that even the first invading 
 hordes, which entered Europe under the successors of Gen- 
 ghiz Khan, were not composed wholly or principally of Tartars 
 properly so called. Because the French give the name of 
 Allemands to all the Germans, it surely does not follow that 
 their ancestors for a long time were acquainted with no Ger 
 mans except those included in the Alemannic confederacy. 
 As to the case in hand, we can only say (assuming that the 
 Phoenicians were the first who used Ionian for Greek) that 
 either the lonians were the first Greeks known to the Phoeni 
 cians, or they were somehow, from greater proximity, or 
 closer intercourse, or some one of many other possible rea 
 sons, more prominently present to the view of the Phoeni 
 cians, when this use of the name originated. 
 
 A second testimony is supposed to be furnished by early 
 Egyptian records. On the celebrated Rosetta stone, and on 
 other monuments of the Macedonian and Roman periods, the 
 idea Greek is represented by a hieroglyphic group consist 
 ing, first, of three papyrus plants standing side by side, and 
 secondly, of three baskets placed one above another. These 
 elements, it is said, give the meaning Lords of the North. 
 The pronunciation of the group, as determined by a compari 
 son of the demotic characters in the Rosetta inscription, is 
 said to be unquestionably Uinen^ which we have just seen to 
 
1 2 IONIAN MIGRA TION. 
 
 be the Coptic name for the Greeks. Now the same hiero 
 glyphic group is found upon a series of monuments belonging 
 to the early Pharaohs, and always in reference to a people 
 described as subject to the kings of Egypt. Of these kings, 
 some as Amenophis II, Sethos I or Sesonchis I belong to 
 the great heroic dynasties of Thebes, the eighteenth and nine 
 teenth dynasties, in the fifteenth and fourteenth centuries : 
 others to the twenty-second dynasty and the tenth century, 
 as Sesonchis, the Shishak of the Old Testament, the con 
 queror of Jerusalem. It would seem, therefore, that several of 
 the early Egyptian sovereigns claimed to be masters of the 
 Uinen, lonians or Greeks. Curtius docs not suppose, what 
 indeed would be in the highest degree improbable, that these 
 records refer to expeditions by sea or land sent out from 
 Egypt to the western border of Asia Minor, and there sub 
 duing or pretending to subdue the Ionian population of the 
 country. He considers them as referring to lonians of Egypt, 
 settled in the Delta of the Nile, who may at various times 
 have been attacked and perhaps reduced to submission, more 
 or less complete, by native sovereigns of the country. It 
 appears from the researches of Lepsius that this name belongs 
 to a group containing nine names of nations, which recur in 
 the same fixed order, the supposed Uinen standing first among 
 them, and Egypt itself, upper and lower, being included in 
 the series. That all the others beside Egypt belong to foreign 
 nations is inferred from the fact that in a Theban tomb the 
 bearers of the two Egyptian shields are plainly distinguished 
 from the other seven by their red complexion and peculiar 
 hair-dress. These are the statements. If they really prove 
 that Ionian settlements were made in Egypt as early as four 
 teen or fifteen centuries before Christ, they doubtless serve to 
 confirm the theory of Curtius. It does not appear, indeed, 
 that the monuments give any direct indication as to what part 
 of the world these Uinen (if they are rightly read so) come 
 from. But it is certainly more probable that such Ionian 
 settlements, if actually made in Egypt, should have been made 
 from Asia Minor than from European Greece. But we seem 
 to have here what may eventually turn out to be a good argu 
 ment, rather than what we can now receive and rely upon as 
 
IONIAN MIGRA TION. 1 3 
 
 such. Even Curtius does not appear to expect that it will 
 produce general conviction. " Every first attempt," he says, 
 " to connect Greek and Egyptian history with one another, 
 to supplement the beginnings of one by materials drawn from 
 the other, must, however cautiously undertaken, encounter 
 manifold objection, consisting either in a vague and general 
 want of confidence, or in scientific doubts as to the correctness 
 of the method and the certainty of the facts made use of." In 
 the present case our suspicions are stronger from the ob 
 scurity which rests on other names of conquered nations 
 found upon the monuments of these ancient Pharaohs ; hardly 
 two or three of them, it is said, have been identified with cer 
 tainty. We must add, however, that Lepsius accepts without 
 hesitation the views of Curtius upon this point : he has no 
 doubt that the name in question refers to Ionian Greeks 
 settled in Egypt, " so that, as early as the sixteenth and fif 
 teenth centuries, lonians that is, at least a part, a considera 
 ble colony of that people were dependent on the Egyptian 
 sovereigns." 
 
 We turn now from Oriental testimonies to inquire how far 
 the known facts of Grecian history support the theory in ques 
 tion. Curtius asserts that in particular localities on the coast 
 of Asia and the neighboring islands there are traces of Ionian 
 occupancy before the time of the Ionian Migration. It is to 
 be regretted that he has not drawn out more at length this 
 part of his argument. As it is, the few brief indications which 
 he gives hardly suffice to make a definite and satisfactory im 
 pression. Miletus and Ephesus, he says, were even in name 
 nothing but renewals of older settlements : and the same fact 
 is expressly attested in regard to Erythrae, Chios, and Samos. 
 Admitting now the correctness of these traditionary notices, 
 granting that the places mentioned were inhabited before the 
 Ionian Migration, are we authorized to assume, what is not 
 contained in the traditions, that these earlier occupants were 
 lonians? What more natural than to find that, among the 
 numerous places settled by these colonists from Europe, some 
 had been previously occupied by the natives of the country, 
 who may have abandoned them before the time of the Ionian 
 colonization ; or, in other instances, may have been dispos- 
 
1 4 IONIA N MIGRA TION. 
 
 sessecl and driven out by the colonists themselves ; or, again, 
 may have remained where they were, submitting to the new- 
 corners and fusing with them into one community. 
 
 Again, he urges that the worship of Apollo Didymceus in 
 his sanctuary near Miletus a worship common to all the 
 lonians appears in tradition as older than the planting of the 
 Ionian colony in Miletus. In like manner, the Delian sanc 
 tuary of Apollo was the Mother-sanctuary for all the stations 
 of Apollo-worship in Greece, and must therefore have existed 
 earlier than the Ionian Migration, though tradition very dis 
 tinctly represents the island of Delos as having at that time 
 received its Greek population in place of the Carians, its 
 earlier inhabitants, I would not say there is no force in the 
 argument derived from these facts. Yet the question must be 
 raised : granting, in accordance with the tradition, the primi 
 tive antiquity of these places as stations of Apollo-worship, 
 how far may we infer, what is not expressed in the tradition, 
 that the primitive worshippers were lonians ? Curtius him 
 self does not suppose that the worship of Apollo was confined 
 to the Greeks : he will not venture to say that it originated 
 with them ; he believes it to have been extensively diffused 
 among the non-Hellenic tribes of western Asia. There is no 
 strong improbability against the supposition that the lonians, 
 instead of founding the establishments referred to, were only 
 the successors of their founders. It is well known that the 
 nations of antiquity regarded it as a point of great importance 
 to keep up local rites of worship, even in conquered places. 
 Curtius mentions that when the lonians were driven out by 
 the Achaeans from northern Peloponnesus, some of their 
 families were retained in Helice, in order to continue there 
 the former \vorship of Poseidon. And, apart from this gen 
 eral feeling, the lonians were little likely to neglect any old 
 and celebrated sanctuary of Apollo, a divinity whom they 
 honored with peculiar veneration. 
 
 For further proof of primitive Ionian occupancy, we find 
 our author referring to the city of lasus, situated on a small 
 island near the coast of Caria. Ng tradition, he observes, 
 was able to refer this Carian place to any settlement proceed 
 ing from the west ; and yet lasus, with its entire environment, 
 
IONIAN MIGRA TION. 1 5 
 
 was, in more than name alone, a genuine, primitive portion 
 of Ionia. Now the Greek character of this place, and even 
 its Ionian character, will be readily admitted. But we know 
 not how to explain the statement that no tradition could refer 
 it to a settlement proceeding from the west. For Polybius 
 (xvi. 11), in a passage which we can imagine no reason for 
 discrediting, tells us expressly that lasus, according to the 
 assertion of its people, was settled by a colony of Argives, 
 though, having afterwards lost a considerable portion of its 
 citizens in a war with the Carians, it received a large rein 
 forcement from Miletus, headed by a son of Neilcus, the 
 Ionian founder of the latter city. lasus, then, appears in the 
 same class with other Greek cities of Asia, which referred 
 their origin to European Greece : there can be no reason why 
 it should be distinguished from the rest, as furnishing clearer 
 evidence of a primitive Ionian population in western Asia, 
 
 Curtius argues from the immediate and great prosperity of 
 the settlements established by the Ionian Migration that they 
 could not have been planted among an alien people, on coasts 
 before occupied only by barbarians. But the Greek cities of 
 Sicily and Southern Italy were founded centuries later, in 
 regions where the previous inhabitants were entirely and 
 unquestionably barbarian : yet, notwithstanding this original 
 disadvantage, such was their progress that, in the time of 
 Xerxes, Hicro of Syracuse was the greatest power in the 
 independent Grecian world, and perhaps a match for all others 
 put together. And later, we find the Greeks of Sicily main 
 taining their ground, though with difficulty, in a long-con 
 tinued struggle against the Carthaginians, a power which 
 proved almost an overmatch for Rome, when mistress of all 
 Italy. Our author evidently feels that this parallel progress 
 of the Italiot Greeks tells against his argument ; and, to 
 weaken its force, asserts that the progress of the Asiatic 
 lonians was different and more remarkable in three particu 
 lars : I. They established a confederacy of their cities. But 
 the want of cooperation in the other case serves rather to 
 increase the marvel. 2. They developed a civilization more 
 purely Hellenic. This, however, may be accounted for by 
 the fact, which probably all would admit, that the barbarians 
 
1 6 IONIAN MIGRA TION. 
 
 of western Asia Minor were much more like the Greeks than 
 the barbarians of Southern Italy and Sicily ; so that the ex 
 traneous influences were more nearly Hellenic in the former 
 case than in the latter. Nor does this general similarity of 
 Carians, Lycians, Phrygians, etc., to the Greeks require us to 
 suppose that they had been in previous uninterrupted com 
 munication with Greeks on the same shores, as our author 
 assumes. He maintains, in fact, that the two sections of the 
 Greek people preserved their essential identity notwithstand 
 ing a separation for centuries by the waters of the yEgean. 
 3. The lonians of Asia made higher attainments in art and 
 literature. True : but would the colonists of Sicily have gone 
 higher in these respects, if on their first landing they had found 
 the island half peopled by their countrymen ? Their attain 
 ments, in fact, if inferior to those of the lonians, may compare 
 with the attainments of Dorians and yEolians in Asia, though 
 these latter, as Curtius supposes, had the advantage of settling 
 among an old established population of their countrymen. 
 
 The strong point of this theory is the fact of its affording 
 an explanation for the peculiar position which the lonians 
 appear to have had in early Greece. The argument may be 
 stated thus. A people scattered far and wide along the sea- 
 coast, and found in the interior only where they might have 
 come by following a river-course back from the sea such a 
 people are not likely to have reached their seats by an over 
 land emigration. The lonians in Greece, then, must have 
 come there by sea, and in all probability from the east ; im 
 mediately from the ^Egean islands, remotely from Asia 
 Minor. But it is not likely that a whole people settled on 
 the Asiatic coast would float over the sea in this way. Their 
 wide diffusion in Greece makes it probable that there were 
 successive expeditions, with a considerable interval of time 
 from first to last. As they w r ere thus established in large 
 numbers and for a long time on the coast of Asia, it is likely 
 that a numerous people remained there, after the last expedi 
 tion set sail toward Greece ; enough to maintain themselves 
 in that position until, after the lapse of centuries, they wel 
 comed back their returning brethren from the west. I will 
 not stop to criticize the probabilities in this argument. But 
 
IONIAN MIGRA TION. I / 
 
 I must not close without observing that, whatever advantages 
 the theory under consideration may give us in explaining the 
 early times of Greece, they are not gained without drawback : 
 we encumber ourselves with some new and serious difficulties. 
 One of these has been already alluded to ; the complete for- 
 getfulness of Greek tradition as to the existence of these 
 primitive lonians of Asia. If the tradition, as our author 
 holds, has preserved some memory of their names and actions, 
 it has at any rate forgotten that they were lonians. This is 
 the more strange, as the national pride of lonians, living and 
 flourishing in the same seats, might naturally have clung with 
 more tenacity to the ancient renown of their ancestors. Why 
 should they give up their own Cecrops and Danaus and Cad 
 mus to the Egyptians and Phoenicians ? Why should they 
 remember so much about their early neighbors, and nothing 
 about their early selves ? Why should they remember so 
 much about Dardanians, Phrygians, Lycians, Carians in 
 western Asia, and nothing about lonians there ? Or why 
 should they remember so much about lonians in Attica and 
 Peloponnesus, and nothing about that people in their own 
 Asia Minor ? Why should a people whose forefathers, born 
 on the same soil, had run a career of wide-reaching activity 
 and enterprise, forget its connection with those forefathers, 
 and attach itself instead to the distant and less distinguished 
 ancestors of a part only of its members ? Athens, according 
 to this view, was the daughter of an Asiatic mother. So long 
 as there were lonians in Asia, the Athenians must have looked 
 to them as colonists to the inhabitants of the mother country, 
 with feelings of respectful attachment, which were peculiarly 
 strong in the ancient Greek mind. Why then should Athen 
 ians, returning to that mother country, forget the respect and 
 attachment which they had before cherished ? why should 
 they forget their original connection with a country which 
 had now become their own home ? If in everything else the 
 tradition lost its hold upon these primitive lonians, we should 
 expect that it would have retained them in connection with 
 the Ionian Migration. How could it carry these wanderers 
 across the ^Egean, without remembering the capital circum 
 stance that they w r ent, not to aliens or enemies, but to their 
 
1 8 IONIAN MIGRA TION. 
 
 own friends, countrymen, and kindred ? There is a singular 
 unanimity in this forgetfulness. Among a large number of 
 cities, scattered along a wide extent of sea-coast, we might 
 have expected that some one at least would remember a fact 
 so important in its early history. But there is no single ex 
 ception to the general obliviousness. It has a greater extent, 
 indeed, than we have yet noticed ; the Cyclades share in it. 
 If the view of Curtius be true, these islands must have received 
 their Greek population from the East, from Asia Minor. But 
 here again tradition is no less distinct and uniform in referring 
 the beginnings of Greek occupancy to colonization from the 
 West, from European Greece. 
 
 I will only notice further some particulars in the early 
 Epic literature which seem inconsistent with this theory. 
 Almost all critics are agreed now in referring the Homeric 
 poems to a date earlier than the year 800. They were com 
 posed, then, within two centuries from the Ionian Migration, 
 perhaps not more than a century after that event. If we 
 were to put the Ionian Migration at about 950, and the com 
 position of the Iliad and Odyssey at about 850, these dates 
 would perhaps correspond as nearly to the collective proba 
 bilities of the case as any that could be assigned. Now the 
 remarkable absence of allusions to Ionia, its places and 
 people, in the Iliad and Odyssey, which does not seem to be 
 fully accounted for by the Achaean subjects and /Eolian 
 scenes of those poems, is naturally explained by the recent 
 arrival of the lonians in that country. Their beginnings in 
 Asia were still matters of historic recollection ; there was 
 still a conscious newness about their places and their doings, 
 which interposed a wide gulf between them and the ancient 
 traditions of Achaeans and Dardans. But the theory of 
 Curtius supplies an immemorial past for the lonians in Asia, 
 and thus renders the phenomenon in question far more diffi 
 cult of explanation. Again, a people who had for centuries 
 followed the Phoenicians in a career of maritime enterprise, 
 competing with them and in many places supplanting them 
 as traders, must have become familiar with the use of letters : 
 and this, if true, would render still more unaccountable the 
 fact, already sufficiently perplexing, that these two long 
 
Uiuv ::.T7 
 
 7CWMAT ^f/^^f^ . 19 
 
 K, f -^H!^^ 
 
 poems, with their innumerable r~f~r~rirrT- lii i HiiUnjitTiiii^ in 
 
 the public and private life of the Homeric age, contain but 
 one disputed and doubtful allusion to the art of writing. And 
 once more, a people who had wandered for ages almost 
 around the Mediterranean must have acquired a stock of 
 geographical information more extensive and accurate than 
 that represented in the poems of Homer. If, for instance, 
 the lonians were conversant with the Delta of the Nile for 
 several centuries, and as late as the time of Shishak, about 
 950, how could the author of the Odyssey place the island of 
 Pharos, which stood close to the Egyptian coast, a full day s 
 sail away from it? And what shall we say of the " speciosa 
 miracula" which Horace admires, " Antiphaten, Scyllamque, 
 et cum Cy elope Charybdin?" How could such notions pre 
 vail among a people who had colonized western Sicily and 
 western Italy as far up as the Tiber, and even the remoter 
 island of Sardinia ? 
 
 We are aware that the foregoing discussion does very im 
 perfect justice to a theory, the strength of which, in its 
 author s own view, lies not in a few decisive arguments, but 
 in the simple, natural connection which it gives to many 
 scattered facts. We wish, also, to acknowledge, in the fullest 
 manner, the ability and learning with which it is supported. 
 We admit that it throws light upon important points in 
 Greek antiquity. We cannot, however, help feeling that the 
 case is not yet made out in its favor, and that it would be 
 unsafe to accept it, until further discussion and the progress 
 of knowledge shall have weakened the objections which now 
 present themselves, and set the evidence for it in a clearer 
 light. It is just to add, that this theory is propounded by 
 its author with all becoming modesty. He recognizes the 
 obscurities and perplexities which environ his subject, and 
 declares that his object in publishing his views is to determine, 
 from the discussion they call out, how far he can himself hold 
 fast to them as established truth. His views may be imper 
 fectly supported by the evidence ; but they are not put 
 forward with that offensive dogmatism which is perhaps 
 nowhere more common than in fields like this, where hardly 
 anything whatever can be known with certainty. 
 
20 IONIAN MIGRATION. 
 
 2. RECENT DISCUSSION AND OPINION CONCERNING THE 
 IONIAN MIGRATION. 
 
 1863. 
 
 The dissertation of Curtius on the Ionian Migration ap 
 peared in 1855. In 1856 came out the third and fourth vol 
 umes of Max Duncker s (Prof. Extr. at Halle) excellent 
 GcscJiiclite des Alterthums. The author, in a note to vol. iii., 
 p. 242, alludes to the new theory, "that the lonians, prior 
 to the Ionian Migration, occupied the coast of Anatolia, and 
 setting out from thence, planted colonies in Greece, from 
 which colonies at a much later period some noble families 
 wandered back again to Asia Minor. The only sure way 
 (he acids) of establishing this theory would be to show that 
 ancient Oriental sources prove, in opposition to the Greek 
 tradition, the settlement of the lonians on this coast prior to 
 B. C. 1000. The proofs which Curtius endeavors to draw 
 from the name of the lonians on monuments of Sethos and 
 Ramses can convince no one who is aware of the fact that, 
 among the names of nations whom the Pharaohs profess to 
 have vanquished, hardly two or three have yet been made 
 out with certainty. If the name Yavana occurs in the Laws 
 of Manu and the Epos of the Hindus, this circumstance may 
 probably be used to determine the age of those works, but not 
 that of the Ionian settlements. The mention of Javan in the 
 lists of Genesis, loth chapter, proves nothing against the 
 Greek tradition, as the composition of that document falls into 
 the lOth century, into the times after Solomon. The theory 
 in question is further contradicted by the unanimous testi 
 mony of Greek authorities, Herodotus and Thucydides at the 
 head, declaring that the inhabitants of the ^Egean islands, be 
 fore the Greeks occupied them, were Carian or Phoenician. 
 If the lonians had come from Asia Minor to Greece, they 
 must have taken previous possession of these islands." This 
 last argument is somewhat obscurely stated. It may be ex 
 pressed in this way. According to Curtius, the lonians, at 
 an immemorial period, crossed the yEgean from Asia to 
 Greece. But we cannot suppose them to have done this 
 
IONIAN MIGRA TION. 2 1 
 
 without occupying the ygean Islands which lay in their way. 
 Those islands, then, must have been occupied from an imme 
 morial period by Greek inhabitants. But the universal opi 
 nion of the Greeks in the historical period was to the contrary 
 of this that those islands, down to a comparatively recent 
 time, were occupied by barbarians, Carian and Phoenician. 
 
 The essay of Curtius was reviewed in Jahn s Jahrbucher, 
 early in 1856, by J. Classen of Frankfort. The reviewer 
 adopts in the main the views of his author, declaring that 
 " he cannot escape from their internal evidence and their 
 clearly-established connection," and avowing his belief that 
 " in them is found the key to one of the most difficult enigmas 
 of ancient history." He specifies four particulars in the ar 
 gument of Curtius which have made the strongest impression 
 on his own mind. I. There are no traditions which refer the 
 lonians of Greece, like other Greek tribes, to any primitive 
 home in the interior of the country ; while their earliest habi 
 tations in scattered localities along the sea-coast "show 
 plainly that they are settlements of a sea-faring people, who 
 never feel themselves at home except when they can breathe 
 the air of the sea-coast." 2. On the coast of Asia Minor, 
 from the Hermus to the Maeander, we find the lonians in a 
 compact and united body, such as they show nowhere else ; 
 so that, comparing their position here with that which they 
 had in Greece, we cannot doubt that they were only colonists 
 in Greece, but had their proper home in Asia. 3. The wide 
 diffusion of the Ionian name at an early period among the 
 nations of the East is only to be accounted for by assuming 
 that this race was in closer proximity and more intimate 
 intercourse with the Oriental nations than were the other 
 Greek races. 4. This theory furnishes the only plausible ex 
 planation for the stories of Cecrops, Cadmus, and the rest, 
 whom we cannot regard as wholly alien to the Greeks, but 
 may well believe to have been men of Ionian descent, who 
 came to Greece from settlements in the East, and were there 
 fore represented in the myths as Egyptians, Phoenicians, etc. 
 
 But, while he adopts the theory of Curtius, Classen calls 
 attention to some points in which his views are defective and 
 unsatisfactory. He complains particularly of the vagueness 
 
22 IONIAN MIGRATION. 
 
 of Curtius in reference to the connection between the Greeks 
 and their neighbor-tribes in Asia Minor, the Dardanians, 
 Phrygians, Lycians, and Carians. Curtius often speaks of 
 these as if he regarded them as having a very close affinity 
 with the Greeks in race and language, so as not to be really 
 a foreign population, but to coalesce readily with Greeks 
 whenever they are brought together, and to form with them 
 a homogeneous union. And yet he does not expressly affirm 
 that they have this character ; and indeed an affinity so close 
 is not only inconsistent (in appearance, at least) with 
 various indications, historical and philological, but, if it were 
 true, would make the theory of Curtius unnecessary ; for the 
 facts which he seeks to explain by his primitive lonians might 
 then be explained by the agency of Phrygians, Lycians, etc. 
 
 In regard to the Ionian name, Classen makes an ingenious 
 suggestion, which has since been adopted by a number of 
 inquirers, and even by Curtius himself. May it not be, he 
 says, that the name lonians (whatever was its primitive mean 
 ing) was not at the outset applied to themselves by the Greek 
 population of Asia, but applied to them by the nations of the 
 Orient, and used as a collective designation, embracing the 
 sea-coast tribes of Dardanians, Maeonians, Carians, and Ly 
 cians ? May not this name have been carried into European 
 Greece by the Phoenicians, who everywhere preceded and 
 served as pioneers for the wanderings of the Ionian tribes ? 
 May it not have been received in Greece under a Grccized 
 form, and applied to. the wanderers from Asia Minor who 
 settled on the Greek coasts and became fused with the earlier 
 population ? And when the counter movement began from 
 Greece to Asia Minor, may not the wanderers (or rather, a 
 part of them) have carried with them the Ionian name, as a 
 domestic designation, which became established on the coast 
 between that of ^Eolian on the north and that of Dorian on 
 the south, the application being determined by the fact that 
 the leaders of emigration to these regions were Ionian, ^Eolian, 
 and Dorian respectively ? 
 
 Classen suggests, also, that the name Leleges may have 
 been used from a very early period by the Greeks of Asia as 
 a general designation for themselves that is, for all who 
 
IONIAN MIGRA TION. 2 3 
 
 could speak their language, in opposition to the 
 that is, the collective tribes whose speech was unintelligible 
 to them. This, too, is a very ingenious thought, and has 
 since been reasserted by various other writers. 
 
 The views of Curtius were discussed about the same time 
 by a more distinguished scholar, G. F. Schomann, of Greifs- 
 wald, in a paper entitled Animadversiones de lonibus. Scho 
 mann is ready to admit that there were lonians in Asia Minor 
 long before the Ionian Migration. But he objects strongly 
 to the proposition that the lonians of Attica were derived 
 from these Asiatic lonians by colonies crossing over to the 
 western coasts of the yEgean and there establishing them 
 selves among a non-Ionian people. He sees no reason to 
 believe that there ever was a non-Ionian people in Attica. 
 And he sees no reason for supposing that the lonians came 
 into Greece at a different time or in a different way from the 
 other Greeks. If the lonians are found only on the sea-coast, 
 the same may be said of their brethren in Asia ; and in both 
 cases they may have been driven down by the pressure of 
 other tribes from an earlier home in the interior to these 
 maritime abodes. He goes into a somewhat lengthened ex 
 amination of the legends concerning Ion and his father 
 Xuthus, to show that they present no trace of a colonization 
 from beyond the sea. This result, indeed, is of inferior im 
 portance, as he proceeds to prove that the legends in question 
 are of no very high antiquity, but must have arisen after the 
 great Dorian invasion and the new relations which this estab 
 lished in Central and Southern Greece. The derivation of 
 the name Ion, from et/u, to go, which Curtius had spoken 
 of with some favor, making the lonians to be wanderers 
 in name as well as in fact, Schomann proves to be improbable 
 on etymological grounds. In regard to Cecrops, Cadmus, 
 and other leaders of colonies from the Oriental world, he 
 avows that, if it be forbidden to consign them to the realm of 
 fable, he would rather accept them for lonians than for Phoe 
 nicians and Egyptians. He remarks, however, that by the 
 treatment of Curtius great vagueness is given to the designa 
 tion of " lonians," which comes to be like that of Franks in 
 the modern Orient. So that when, in this theory, a multi- 
 
24 IONIAN MIGRA TION. 
 
 tude of things are ascribed to the lonians wanderings, set 
 tlements, establishment of religious worships and festivals, 
 communications of the most various arts and industries it 
 becomes a question in each case to what lonians they are 
 attributed. And though many things thus ascribed to the 
 lonians are not improbable in themselves, he denies that they 
 are more certain than what we believed before, and holds 
 that such conjectures can avail nothing at all toward a more 
 accurate knowledge of the Ionian race. 
 
 To the objections of Schomann, Curtius replied at once in 
 the Gottingcr Gelehrte Anzcigcn for 1856. I regret that I 
 have been unable to obtain a sight both of this article and of 
 another by the same author, which appeared in 1859 in the 
 same periodical. In 1857, Curtius published the first volume 
 of his elaborate and attractive History of Greece. He here 
 brings out in a popular form the views developed in his essay, 
 and defended in his reply to Schomann. He does not, in 
 deed, put them forward as ascertained historic truths, estab 
 lished by documentary evidence, and entitled to unquestioning 
 assent. He speaks of them as an attempt to connect the 
 Hellenic people with the Indo-European family of nations, 
 and to make intelligible their wanderings in the earliest period. 
 But he makes them the basis of his whole treatment of the 
 primitive Greek history, and thus gives them a position which 
 no hypothesis should receive unless it is pretty clearly re 
 quired by the known facts of history. His work being of a 
 popular character, he naturally abstains from polemical dis 
 cussions. He does, however, vindicate his theory from the 
 objection that it is at variance with the tradition of the Greeks. 
 To this he replies, first, that his view is not opposed to any 
 positive statement of Greek tradition ; for in reference to the 
 primitive diffusion of the Ionian race the ancients tell us 
 nothing : they are simply silent on the subject. And, second, 
 if his view is not attested by any positive statement of Greek 
 tradition, it is easy to assign reasons for this fact. "The 
 Greeks were so proud a people that they regarded their land 
 as the central land, the point of departure for the important 
 affinities of nations. And when, in Asia, the barbarians had 
 pressed on to the borders of the Archipelago, it became, under 
 
IONIAN MIGRA TION. 2 5 
 
 Athenian influence, a general feeling that European Greece, 
 still free and independent, was the proper land of the Hellenes. 
 Athens itself was assumed to be the metropolis of all lonians. 
 Under this influence all opposing traditions were more and 
 more thrust into the background, or with haughty boldness 
 set aside altogether. Even of the Carians it was maintained 
 
 o 
 
 that they had been driven to Asia from Europe, while, accord 
 ing to their own well-established belief, they were at home in 
 Asia. The Lycians, in like manner, were supposed to have 
 come from Attica to Lycia. Nay, the entire connection of 
 the Greeks with the races of Asia Minor was completely re 
 versed, and the consciousness of an original affinity of the 
 Greeks with the Phrygians and Armenians was so expressed 
 that the Phrygians were represented as having migrated from 
 Europe to Asia, and the Armenians, again, as having derived 
 their origin from the Phrygians." 
 
 This answer, it appears to me, gives undue weight to the 
 Athenian influence, at least for the earlier times of Grecian 
 history. The prominence of Athens in the political world of 
 Greece dates from the Persian wars, long after the beginnings 
 of literary production and documentary record. We have a 
 multitude of notices, of a historical or traditionary character, 
 which were composed in the sixth and seventh centuries before 
 our era ; and some, doubtless, of even older date. But in 
 those times the Greek cities of Asia were independent and 
 prosperous. Miletus and Ephesus were then more splendid 
 and powerful than Athens, and were not inferior to her in 
 literary culture. Why should the Greeks of those cities sur 
 render the honorable consciousness of their own ancient sub 
 sistence on the same sites, and be content to trace the origin 
 of their political being to a colonization, not then very ancient, 
 from a less powerful city on the other side of the /Egean ? In 
 the literature of the sixth, seventh, and eighth centuries before 
 Christ, we find many legends in regard to things that befell in 
 Thebes and Argos and Sparta prior to the Ionian Migration. 
 Why do we find none such for Ephesus and Miletus, which 
 Curtius supposes to have had a long history as Greek cities 
 prior to the same event ? It is certainly quite conceivable 
 that cities should have existed on the sites of Miletus and 
 
26 IONIAN MIGRA TION. 
 
 Ephesus at the time when the colonists from Attica landed on 
 the coast of Asia. But I see not how we could then account 
 for the obliteration of all legends relating to earlier personages 
 and fortunes of those cities, without supposing that their old 
 population was viewed as an alien race, so that a wholly new 
 order of things, a wholly new nationality for these cities, was 
 regarded as commencing with the colonization from Attica. 
 
 An appendix to vol. i. of Curtius s History contained re 
 marks by Lepsius in defence of his belief that the Ionian name 
 is to be recognized on Egyptian monuments of the fifteenth 
 and fourteenth, centuries B.C., among the nations described 
 as subject to the great Pharaohs of the iSth and iQth dynas 
 ties. There is no doubt that such monuments contain a group 
 identical in form with that used for the lonians, or rather the 
 Greeks, in the period of the Ptolemies. But De Rouge had, 
 some years previous, proposed a different interpretation for 
 the group as found on the earlier monuments. He noticed 
 that it occurs generally in the first place among a list of given 
 nations ; and the ideographic value of the signs composing 
 it appeared to give the meaning the Northerners all, i. e. 
 all the tribes of the North ; which would thus be a general 
 designation for the eight tribes enumerated after it. This 
 view appeared to be confirmed by the fact that in some in 
 stances the nine names are followed by another series begin 
 ning with Kesh (i. c. Cush) and containing apparently the 
 names of southern tribes. It was supposed further that, after 
 the Greeks or Macedonians became masters of Egypt, the 
 group was held to signify the northern lords, and applied 
 as a flattering designation to the now dominant people. 
 
 Bunsen, in the last part of his Acgyptcns Stcllc in dcr Wclt- 
 gcscliicJitc, had taken up this view of De Rouge s, and main 
 tained it in opposition to that of Lepsius. In doing so, 
 he was uninfluenced by any prepossession against the new 
 Ionian theory ;. for to this, in the same place and in the fullest 
 terms, he signified his own adhesion. Lepsius, in the appen 
 dix referred to, brings forward objections to the interpretation 
 of De Rouge and Bunsen, and reiterates his clear conviction 
 that the group in question can only refer to the Greeks of 
 the islands and coasts of the yEgean. He thinks it probable, 
 
IONIAN MIGRA TION. 2J 
 
 too, from the monumental evidence, that some of this people 
 had formed permanent settlements in northern Egypt, on the 
 Delta of the Nile. The force of his arguments can hardly be 
 appreciated by any but professed Egyptologists ; and Curtius 
 himself admits that, as matters now stand, the evidence from 
 this source is not such as to demand the full confidence of 
 scholars in general. 
 
 Conrad Bursian, since known as a zealous student of Greek 
 geography, published in 1857 a dissertation entitled Qitces- 
 tionum Euboicarum Capita Sclccta, in which he came for 
 ward as an adherent of the Ionian theory ; and not long 
 afterwards, reviewing Duncker s History in Jahn s JaJirbiicJicr, 
 he censured that writer for his skepticism in reference to it. 
 To the objection of Duncker, already noticed that if the 
 lonians at an immemorial period had crossed the yEgean 
 from Asia to Greece, they would at the same time have oc 
 cupied the islands on their way, whereas Greek tradition 
 represents these islands as occupied down to a much later 
 period by Phoenicians and Carians he replies, that the lonians 
 passed by these islands without taking possession, for the 
 very reason that they found them already held by the Phoe 
 nicians. This answer is hardly in accordance with the 
 spirit of the theory, which makes a previous Phoenician oc 
 cupancy to be everywhere the condition and occasion of 
 Ionian settlement ; and in fact, Curtius, in a subsequent 
 article (which we shall notice further on), gives a different 
 reply. The latter asserts, first, that the barbarians might be 
 supposed, without improbability, to have thrust themselves 
 into these islands after they had been settled by the lonians ; 
 second, that there is good reason for believing that these 
 islands, long prior to the Ionian Migration, were not wholly 
 in barbarian possession, but were occupied, at least in part, 
 by a Greek population ; and third, that the Carians them 
 selves were closely akin to the Greeks, and perhaps origi 
 nally spoke Greek, but were afterwards barbarized Semitized 
 by communication with the Phoenicians. 
 
 The Ionian theory was further discussed, in 1857, by^Alfred 
 von Gutschmid, in his Beitragc zur Gcschiclite dcs alttn 
 Orients. This work was an extended and far from favorable 
 
2 8 IONIAN MIGRA TION. 
 
 critique of the fourth and fifth volumes of Bunsen s Egypt. 
 But, as Bunsen in his closing volume had taken up the Ionian 
 theory, Gutschmid also was naturally led to take notice of it. 
 His remarks are very interesting, though not free from a 
 tone of dogmatism, and an asperity towards those from whom 
 he differs, which appear too often in the writings of this 
 learned and able scholar. He denies, as my own article had 
 already done, that the great and rapid progress in wealth, 
 power, and literary culture, made by the Ionian cities .of Asia 
 Minor after the colonization from Athens, is any proof that 
 those colonies were not planted among an alien people. He 
 brings forward many examples to show that such a progress, 
 surpassing that of the mother country in the same time, is a 
 frequent phenomenon in the history of colonization. He 
 refers especially to North America, and the uncommonly 
 rapid increase of population and the growth of ne\v states 
 which have been witnessed here. To the great argument of 
 Curtius, founded on the position of the lonians in Greece, he 
 replies as follows : 
 
 " The assertion that the lonians in Greece nowhere appear 
 in a compact mass is of a subjective nature. On both sides 
 of the Isthmus, we find them in connected seats : in the South 
 they form the population of all Achaia and the northern 
 coast of Argolis ; in the North they have Attica, Southern 
 Bceotia and the neighboring parts of Phocis, and Euboea. 
 What is most important in this region, beside the towns on 
 the coast, those of the interior also are Ionic, which is not 
 the case in Asiatic Ionia. The fact which Curtius notices (p. 
 4), that just here no trace is found of migrations made by 
 the lonians from one habitation to another, might be regard 
 ed as an indication of their primitive settlement in this region. 
 But if these seats appear to any one insufficient to be the home 
 of so great a race, let him hold to the statement of Hero 
 dotus a statement oftener evaded than really refuted that 
 the lonians were originally Pelasgians. Curtius himself has 
 taken a step in this direction, by claiming for his lonians the 
 names Argos and Larissa, which have been universally re 
 garded as Pelasgian. True, he conceives the relation between 
 the lonians and Pelasgians to be this, that the former came 
 
IONIAN MIGRATION. 
 
 2 9 
 
 by sea in single parties, that they connected themselves with 
 the native inhabitants, and especially the Pelasgians, and be 
 came gradually so lost in the general mass as to be no longer 
 distinguishable from it. Still, there is nothing to forbid our 
 regarding the lonians as a name which at a more recent 
 period separated itself from the Pelasgian. To show how far 
 Curtius is warranted in his hypothesis concerning the home 
 of the lonians, I will bring forward an analogy from a time 
 of authentic history, in which we have the advantage of docu 
 mentary data. I refer to the Malays, who, as a people of 
 coasts and islands, arc well adapted to the comparison, and 
 who have, in common with the lonians, not only a decided 
 capacity for the sea, and a free restless nature, but also the 
 fact that in the diffusion of Islam they have been the carriers 
 of culture in the Indian Archipelago, just as the lonians were, 
 according to Curtius, in the Greek Archipelago, by transport 
 ing the ancient civilization of the East. The Malays the name 
 is said to mean the same thing as that of the Parthians, viz. 
 fjueravda-rat,^ emigrants are spread over all the Indian islands, 
 but only the smaller belong wholly to them ; the larger are 
 occupied in their interior by primitive inhabitants, who are 
 of the Ethiopian race. Whence the Malays came, nobody 
 can say ; it appears inconceivable that one of the islands 
 should have been their home. On the continent there is but 
 one land where the Malays are found in large masses, the 
 peninsula of Malacca, which has its name from them, as Ionia 
 has from the lonians. Nowhere do the Malays form states to 
 the same extent as here (precisely what Curtius insists on in 
 reference to the Asiatic lonians). For here was the seat of 
 the most powerful Malay kingdom, that of Singhapura, Ma 
 lacca, and Gohor ; here are still found numerous Malay states, 
 while, e.g. in Java, the most important of the Indian islands, 
 the older states were founded by Brahman wanderers from 
 Further India, and that of Bantam by Arab Sayyids. If 
 now we should apply the argument of Curtius to this case, 
 we must infer that Malacca was the primitive seat of the Ma 
 lays, and that from thence they occupied the coasts of the 
 islands. But history teaches us that in the year 1160 A.D., 
 the Malays, under their king Qri Tribhuvana, first passed 
 
30 IONIAN MIGRA TION. 
 
 over from Sumatra to the mainland, and founded Singhapura ; 
 that in 1253, under their fifth king, Qri Skandar, they founded 
 Malacca ; that in 1511, under their twelfth king, Sultan Mah- 
 mud Shah, they founded Gohor : while we have similar 
 exact dates for the successive diffusion of the Malays on the 
 peninsula. This example may teach us the necessity for 
 caution, in cases where we have no authentic history to de 
 pend upon." 
 
 In his celebrated essay on the Book of Nabatha^an Agricul 
 ture (ZtscJift. d. dcntscJicn uwrgcnl. Gcs. vol. xv., 1861), Gut- 
 schmid again found occasion to speak of the Ionian theory, 
 which Chwolson had laid hold of to account for the amazing 
 fact that a Babylonian author, twenty-five centuries before 
 Christ, should combat the botanical opinions of Greek or 
 Ionian writers. He docs not here enter upon any formal dis 
 cussion of the theory. He only says : " Chwolson, as was to 
 be expected, tries to make out of the universal lonians a 
 little capital for his client Out ami ; " and then, after quoting 
 a passage in which Chwolson thanks heaven * that the old 
 naive chronology of primary schools, with its dates of 1697 for 
 Phoroneus, 1377- for Deucalion, etc., has passed away forever," 
 he adds : "The fact is just the other way, that the Asiatic and 
 Egyptian wanderers Pelops, Danaus, Cecrops, Peteos, Erec- 
 theus, whom one had thought to have passed away forever, 
 have been dragged by Curtius out of the rubbish-hole, and 
 turned to profitable account as Orientalized lonians or Ionized 
 Orientals. The lofty air with which the author of the hypo 
 thesis warns off objectors, preferring this course to an at 
 tempt at refuting their objections, and the eagerness with 
 which enthusiastic philologs in Jahn s Jahrb dckcr claim the 
 admiration of their readers for E. Curtius s studies on the older 
 Greek history, may certainly lead outsiders to suppose that 
 they have here a made-out fact, to which only personal enmity 
 refuses its due acknowledgment. But in truth, the very per 
 sons who are most competent to judge., historians and geo 
 graphers (I mention only Dunckcr, in his Greek History, 
 and Kiepert, in his Researches on the Genealogical List of 
 Nations in Genesis), have declared themselves very decidedly 
 against the Ionian hypothesis, and this is pretty generally re- 
 
IONIAN MIGRA TION. 3 1 
 
 garded as a thing gone by. In reference to every hypothesis, 
 the first question must be, not is it good/ but is it neces 
 sary ? is it on the whole the most satisfactory solution of 
 existing difficulties ? He then maintains that the only 
 positive fact alleged in support of the theory (if fact it be) 
 the appearance of the Ionian name on the early Egyptian 
 monuments is not really explained by it, the occurrence of 
 lonians in Egypt at that remote period being about equally 
 marvellous whether you suppose them to have come there 
 from the eastern or the western shores of the ^Egean. 
 
 Among " the enthusiastic philologs of Jahn s Jahrbucher" 
 Gutschmid must have included August Baumeister, who in 1860 
 reviewed the Greek History of E. Curtius, and spoke particularly 
 of the Ionian theory in the warmest terms of commendation. 
 
 The same question was treated, with greater thoroughness, 
 in another work which appeared in 1860 a gymnasial pro 
 gram by H. DonclorfF, entitled the lonians in Eubcea. 
 Dondorff agrees with Curtius in recognizing the existence of 
 lonians (so called) in Asia, and their wide-spread activity as 
 a seafaring people long prior to the Ionian Migration. But 
 he does not therefore accept the theory of Curtius, against 
 which he acknowledges the force of the objections raised by 
 Duncker and Gutschmid. He holds that the early lonians 
 of Asia and the yEgean were not the same people as the 
 lonians of Attica : the latter were a Hellenic tribe, while the 
 former, he thinks, were wholly or mainly Semitic barbarians, 
 including Carians, Cypriots, and even Philistines. The con 
 nection between these and the lonians of Greece is nothing 
 but a coincidence in name. 
 
 This description of DondorfFs views I give, having never 
 seen his dissertation, on the faith of Curtius himself, who re 
 viewed it in Jahn s Jahrbucher for 1861. The review, as might 
 be expected, is mainly occupied with a vindication of the Ionian 
 theory against the objections of dissenting critics. It is evi 
 dently prompted in part by the sarcastic remarks of Gut 
 schmid in the essay on the Nabathaean Agriculture, although 
 that essay is nowhere mentioned. Curtius refers to his own 
 former articles as proof that he has not shunned a discussion 
 of the subject : he may have refused the challenge of an over- 
 
3 2 IONIAN MIGRA TION. 
 
 bearing and contemptuous polemic, but he has been always 
 ready to meet candid and courteous opposition. He takes 
 pains to show that his views have not failed to commend them 
 selves to a number of well qualified judges. " Leonhard 
 Schmitz," he says, " even before the appearance of my His 
 tory, had made this view the basis of his own. I find my 
 own view again fully set forth in Lorenz Diefenbach, Orig. 
 Eiirop., p. 78, and if he has come to it independently (for he 
 makes no mention of me), this is only the more welcome 
 guaranty for the truth of my hypothesis. W. Vischer, in his 
 Er inner nngcn aits Griechenland, p. 301, agrees with me en 
 tirely as to the Ionizing of Argos ; and more recently, in the 
 Schweizerisches Museum, he expresses only a wish that the 
 inferences drawn from the hypothesis might be kept within 
 narrower limits." 
 
 It is evident from this article of Curtius that three argu 
 ments which, as we have already seen, were used in the essay 
 of 1855 to support his theory, are now no longer relied on for 
 that purpose. These are : I. the wide diffusion of the Ionian 
 name throughout the East as a designation for the Greeks ; 
 2. the mention of lonians on the old Egyptian monuments of 
 the fifteenth and fourteenth centuries ; and 3. the rapid pro 
 gress of the cities which the colonists from Attica, in the 
 so-called Ionian Migration, were said to have planted on the 
 coast of Asia Minor. These arguments, which were all 
 objected to as unreliable in my previous communication on 
 this subject, are now practically abandoned : it may be said 
 that they are withdrawn, and no longer form part of the case. 
 Curtius now rests his cause mainly on two points. These 
 are: i. the position of the lonians in Greece, confined from 
 the earliest time to certain scattered localities on the sea- 
 coast : this he still maintains against the opposition of Scho- 
 mann and Gutschmicl ; 2. the traditions in reference to the 
 colonists from Attica in the Ionian Migration : these he ex 
 amines in some detail, to show that they imply a close affinity 
 between the new-comers in Asia and the previous inhabitants 
 of the places where they settled. In this examination, I must 
 think that he relies too much on the fullness and authenticity 
 of the traditional accounts, and thus draws from them con 
 clusions which have no sufficient vouchers. 
 
IONIAN MIGRA TION. 3 3 
 
 As to the first point, which I have always regarded as the 
 stronghold of the Ionian theory, its strength, it seems to me, 
 is in a measure compromised by the new views which our 
 author has taken up in reference to the Ionian name. I 
 translate the highly interesting paragraph which relates to this 
 subject. " As regards the history of the name lonians, I am 
 quite ready," he says, " to adopt the same points of view as 
 Dondorff, and to recognize in it a collective name, which 
 gradually came to be restricted, and to assume a more definite 
 application. That view was proposed by Classen, and de 
 fended with much penetration in this Journal, as far back as 
 1856, and I have never ventured to maintain an opposite 
 position. According to this view, we should have to assume 
 some three principal epochs in the history of the name : I . 
 the name Yavanim, as diffused among Aryans and Semites 
 through the whole Eastern world, embraces all the seafaring 
 tribes of the yEgean except the Phoenicians, including both 
 pure Greeks and various tribes of mixed origin, Carians, 
 Citians, Philistines, etc. It is a name which may correspond 
 somewhat to that of the Leleges, an aggregate name, which 
 does not, any more than that of Franks in the modern East, 
 designate a whole connected by community of language, a 
 definite etJinos. This name, first brought by the Phoenicians 
 to the European Greeks, was adopted by the latter to desig 
 nate their kinsmen w T ho were gradually settling on their sea- 
 coast, whom they became acquainted with first as intermixed 
 with Phoenicians, and afterwards in their purer nationality. 
 Thus, 2. the foreign name Yavanim, adapted to the Greek 
 mouth as lao^e? or "IWe?, became established on the soil of 
 Europe, and especially in Attica and ^Egialeia. Here Ionian 
 history first develops itself. Here, therefore, the name also 
 first obtain^ a historical significance, which causes the prior 
 existence of the people, their transmarine origin, to be wholly 
 forgotten. Then follows the Trepan a is rwv Icovcov et? Aalav 
 (the Ionian Migration). And now, 3. the name fixes itself 
 in Asia Minor, where the Carians, barbarized kinsmen of the 
 Greeks, are overcome, the nobler germs existing among the 
 Greeks who had remained in their primitive home on the 
 Asiatic coasts are roused to new life, and a city-culture estab- 
 3 
 
34 IONIAN MIGRA TION. 
 
 lished, which, under influences from the Asiatics of the in 
 terior, soon takes such a direction that the Athenians dis 
 claimed the very name of lonians." 
 
 Curtius is careful to mark these views as relating only to 
 the name " Ionian : " yet it is plain that they are of no little 
 importance with reference to his whole theory. On the one 
 hand, they appear to weaken in some measure the founda 
 tions on which it rests. Evidently they destroy whatever 
 residue of weight any one might recognize in the diffusion of 
 the Ionian name through the Eastern world, and in the sup 
 posed mention of lonians on the ancient Egyptian monuments. 
 But further than this : if it is admitted that the Ionian Greeks 
 received that name in Greece itself, having previously been 
 otherwise designated, then the confinement of that name to 
 certain limited localities becomes a matter of less consequence, 
 for it does not necessarily imply that the race was confined to 
 the same localities : it is quite conceivable that the race itself 
 may have had a much wider extension in Greece, and that 
 only a particular part, which in some way came to be distin 
 guished from the rest, received the name of lonians. Yet, 
 on the other hand, these new views give some aid to the 
 theory. They help to account for the fact that no traditions 
 represent the lonians in Greece as having come there cither 
 from the opposite coast of Asia or from the islands of the 
 ^Egean. They also help to account for the fact that no tra 
 ditions represent the Ionian emigrants from Attica to Asia as 
 having settled there among a population whom they regarded 
 as kinsmen and brethren. This last point I have alluded to 
 before, and it is one which I look upon as of great conse 
 quence. It is not at all improbable that the Attic settlers in 
 Ephesus and Miletus may have found in those regions a 
 population not widely diverse from themselves in language 
 and ethnical affinities. But there is no convincing reason to 
 believe there is much reason for declining to believe that 
 they found a population which they recognized as having the 
 same nationality with themselves, as standing in a relation 
 to themselves similar to that of the Achaeans, Dorians, or 
 ^Eolians. There is no reason for supposing that the earlier 
 inhabitants of the land were any nearer to the Greeks in origin 
 
IONIAN MIGRATION. 35 
 
 and language than were the Pelasgians. Curtius, in his His 
 tory, describes the Pelasgians as an earlier wave of population 
 which streamed over from Asia to Europe, just in advance 
 of that which brought the Hellenes to their Grecian home. 
 He describes them as distinct from the Greeks, yet closely 
 resembling them, and readily coalescing with them. That a 
 population of this kind, whether called Pelasgic or designated 
 by other names a population standing in this sort of relation 
 to the Greeks may have existed in Asia Minor long before 
 the Ionian Migration, is certainly not improbable. That the 
 Lycians and Phrygians stood thus near to the Greeks, I can 
 not believe. The language of the Lycian monuments is so 
 far from resembling the Greek that some have doubted 
 whether it was even Indo-European. As for the Phrygians, 
 we have in Homer the proper names of Dardanians or Tro 
 jans, a people of the Phrygian stock. George Curtius has 
 shown that many of these names imply a language different 
 from the Greek. In regard to the Carians, though the 
 Semitic origin often asserted for them is not sufficiently 
 proved, we yet see, from the Homeric epithet, fBapfBapofy&vot, 
 applied to them, that their language was unintelligible to the 
 Greek ear. But we hear of Pelasgians in Asia Minor, and 
 yet more of Leleges, whom we may believe to have stood in 
 a near relation to the Greeks. While, therefore, we hold it 
 improbable that the colonists of the Ionian Migration found 
 on the Asiatic coast any people whom they felt to be so near 
 them as the lonians or even the Dorians and ^Eolians of their 
 own land, we think it not indeed a historic certainty, but by 
 no means improbable, that they found populations who 
 seemed as near to them as the Pelasgians in Europe. And 
 we think it still less a historic certainty, yet also not improba 
 ble, that these populations of the Asiatic coast may in earlier 
 centuries have had communication with the western shores 
 of the ^Egean, and that the development of the Ionian 
 character and name in Greece may be somehow connected 
 with such communications. So much as this we should, 
 according to our present lights, consider to be the net result 
 of the discussions which have been set on foot by the disser 
 tation of E. Curtius. But it seems highly important that 
 
36 IONIAN MIGRA TION. 
 
 these conclusions should not be invested with more of cer 
 tainty than properly belongs to them that they should be 
 recognized as historic speculations rather than historic verities, 
 as probabilities or possibilities rather than facts. 
 
 Curtius insists that we must have some theory which shall 
 enable us to comprehend the facts that lie on the threshold 
 of Hellenic history. Such a theory he has labored to con 
 struct, and he claims the right to maintain it until a better 
 one is offered him. It is unquestionably right that he should 
 show his theory to be superior to any other which has been 
 proposed : but it is quite conceivable that he should succeed 
 in this without thereby proving it to be true. There are 
 many problems in history, and especially in ancient history, 
 of which the solution is more to be desired than looked for ; 
 many in which the deficiency or uncertainty of the data is 
 such as almost to preclude the hope of a satisfactory solution. 
 It may be that the problem which Curtius has discussed with 
 so much ingenuity and learning, and with no less candor and 
 courtesy, will have to be ranked in this category. 
 
II. 
 
 THE ROOT PRACH IN THE GREEK LANGUAGE. 
 
 1855- 
 
 THE Sanskrit language has a root prach, in very common 
 use, signifying to ask a question. This radical appears to 
 have been preserved in all the great subdivisions of the Indo- 
 European class. We find it in the Gothic frah, he asked/ 
 A. S. fregnan, modern German fragen. Bopp, in his San 
 skrit Glossary s. v., gives a Lithuanian praszau, rogo, pre 
 cor, and perszu, procus sum, uxorem mihi deposco : a 
 Russian/r^ ^, rogo, precor : an \ris\ifiafrach, inquisitive, 
 which Bopp explains by a supposed reduplication which, 
 however, is more probably to be explained as a compound 
 of the root with the old preposition fia, before : so fia- 
 fraighe, question, fiafruighim, I inquire, ask. In Latin 
 we find the verb precor, to pray, and the noun procus, a 
 suitor, both certain derivatives of this root. Bopp also refers 
 to it the verb posco, as if for prosco, supposing that r is omitted, 
 as in the Greek preposition TTOTI for Trpori. irpori or 777)09 is 
 a derivative from Trpo : its p therefore is original, and the form 
 TTOTL is a corruption, as also the corresponding Latin form pot } 
 which never indeed occurs as a separate word, but appears as 
 a prefix in possideo (=pot + sedeo), porrigo (=pot + rego), and 
 other compounds. Bopp further adds the Latin rogo, as if for 
 progo, and the comparison is certainly plausible ; but I am 
 not able to produce any fully satisfactory analogy for this 
 assumed loss of the initial/ before a following r. 
 
 As to the Sanskrit prach itself, Pott has explained it as a 
 compound, made up of the prefix pra, l before, and the root 
 ichy to desire, which appears only in the special tenses (cor 
 responding to the Greek present and imperfect in the different 
 modes). It is a curious fact that this compound of pra and 
 m ich does really occur in Greek, with a distinctness which ex- 
 
33 ROOT PR A CH IN GREEK. 
 
 eludes all doubt as to its nature and origin. The Sanskrit/; ^ 
 is the Greek TT/QO, and the Greek equivalent of ich would natu 
 rally be IK, as the Sanskrit palatals are for the most part de 
 generate gutturals. Hence pra + ick^TrpoiK, which we find 
 already in the Homeric Trpo l/CTr]?, a beggar. See Odyssey 
 xvii. 352, where Ulysses, appearing at the banquet of the 
 suitors in the garb of a beggar, is thus addressed by the 
 swineherd Eumaeus, who brings him bread and meat from 
 Telemachus : 
 
 TOL, % elve, Mot raSe, real ere 
 alri^eiv iJid\a irdvra^ e 
 
 The corresponding verb Trpotcro-oyLtat, fut. 7rpoL%ofjiai 7 to beg/ 
 is found in Archilochus, and in the compound KaTaTrpoiaaofjiai 
 occurs with considerable frequency. We have likewise the 
 noun Trpoij;, Attic Trpoli;, a free gift, gratuity (literally, ob 
 tained by begging, had for the asking ), the genitive of which, 
 Trpoiitos, and the accusative, TrpoiKa, are used very commonly 
 to signify as a free gift, gratis. 
 
 Now the Sanskrit pracJi is not by any means so clearly the 
 compound of pra and ich as is the Greek Trpo l K. According 
 to the constant law of Sanskrit euphony, pra + ich ought to 
 give us prccJi and" not pracJi. There is therefore room to 
 doubt whether the term prach is really a compound. At all 
 events, if it be so, it has undergone such change as to assume 
 the appearance of a simple root it has completely lost the 
 consciousness of its composition, has passed as an uncom- 
 pounded radical through the whole class of Indo-European 
 languages, and has received thus an extension far wider than 
 the simple ich, from which it is perhaps derived. We have 
 seen it already in the Latin, the Slavonic, the Lithuanian, the 
 Teutonic, and the Celtic languages. Does it occur also in 
 Greek ? Does this language furnish us, besides the distinct 
 compound of pra with ich, which we have been considering, 
 with any representative for the more common and extensive 
 prach, which is perhaps only the result of their fusion ? My 
 belief is that it does so, in the noun OeoTrpo-n-os. This word 
 
ROOT PRA CH IN GREEK. 39 
 
 occurs frequently in Homer, and signifies one who prophe 
 sies, a diviner or soothsayer. In Herodotus we find a different 
 meaning, that of a public envoy sent to consult an oracle. 
 We might give examples of the word in both these uses ; but 
 this is not necessary. I suppose it to be a compound of #eo<? 
 with the root praclt, and to signify, according to its etymol 
 ogy, one who asks a god or the gods. 
 
 This derivation explains perfectly the two meanings which 
 we have just mentioned as belonging to the word. The pro 
 phet who pre-announces the events of the future does so, not 
 of himself, not from his own stores of knowledge, but from in 
 formation which he seeks and obtains of a divinity -he pro 
 fesses himself and is understood by others to be a questioner 
 of the gods. And the propriety of the term is no less clear 
 in its application to the public envoy sent to consult an 
 oracle : it is his office to interrogate the god. 
 
 This derivation is equally satisfactory as regards the form 
 of the word. The Sanskrit palatals, as already intimated, 
 correspond generally to gutturals in the cognate languages, 
 and the interchange of gutturals and labials is one of the most 
 familiar phonetic phenomena. No one could hesitate about 
 accepting the Latin procus as a derivative of the Sanskrit 
 pracli, and there is as little difficulty in admitting a Greek TT^OTTO? 
 as the equivalent of the Latin procus. A case exactly analo 
 gous may be found in the Sanskrit vac, to speak : from 
 which root we have the Latin vox, vocis, vocare, and the 
 Greek cnjr, OTTO? (originally FOTTS. POTTOS, compare 7^77-09, 
 Fenreiv). The compound name Ka\\io7rr), beautifully speak 
 ing, bears the same relation to the Latin vox and Sanskrit 
 vac, that $eo7rpo7ro?, according to our view, bears to the Latin 
 procus and the Sanskrit prac/i. 
 
 Let it not be thought strange that in Trpotfcr^, l a beggar, 
 the Greek should retain the primitive K, and yet in 7773077-0?, 
 * an asker, coming perhaps from the same ultimate theme, 
 should change it to the labial TT. The language vacillates be 
 tween these sounds in one and the same word, as in the Ionic 
 /coto9, #60-09, /core, /ecu?, etc., which present the primitive gut 
 tural, while the common Greek has 7roto9, 7700-09, Trore, 7ro>9, 
 etc., with the labial. So in the root which signifies to see, 
 
40 ROOT PR A CH IN GREEK. 
 
 Sanskrit iksh, perhaps originally aksh, from which come 
 aksha and akshi, eye/ Latin dim. oculus, ocellus. In Greek 
 the corresponding radical is OTT, found in cnfro/zai, I shall see/ 
 OTrwTra, I have seen/ o<#aAyno9, eye/ etc., and yet we find 
 6We, the two eyes/ which presupposes a form o/ae, while the 
 Bceotian o/craAAo?, eye/ gives us the primitive guttural 
 without change. 
 
 The etymology here proposed has the advantage of account 
 ing perfectly for the accent of the word. It is a well-known rule 
 for compounds of the second declension, such as TratSo/rroi/o?, 
 in which the second part (KTOVOS) is a verbal of two syllables 
 with short penult that if this verbal has a transitive relation 
 to the first part of the compound, the accent is on the penult; 
 thus, 7rat^o/cToVo9, son-killing, murderer of a son ; but if 
 the verbal is intransitive or passive in its relation to the other 
 part, the accent goes back to the antepenult ; thus 
 TraiSo/cTovos, son-killed, killed by a son. According to this 
 rule tfeoTr/ooTro?, derived from prach and accented on the 
 penult, could only have the sense of questioning the gods 
 (or a god) a sense which, as we have seen, eiccounts per 
 fectly for the actual uses of the word. 
 
 The arguments in favor of this derivation will gain addi 
 tional force, if we look at the other attempts which have been 
 made to assign the etymology of the word. The old estab 
 lished traditional method makes the TrpoVo? a compound of 
 irpo with the root of eVo?, efaeiv, as though TT/DOTTO? were for 
 TrpoFOTros. As a transitive compound, this would signify 
 foretelling the gods/ which, of course, could hardly have an 
 intelligible meaning. As an intransitive compound, it would 
 signify foretold by the gods/ a meaning intelligible enough, 
 but wholly inapplicable here. To account for its actual use, 
 we should have to render it foretelling from the gods/ i. <?. 
 by communications derived from them/ or foretelling by 
 the gods/ i. e. by their counsel or assistance. Now there 
 is some latitude in regard to the relations expressed by com 
 pound words, and it is not unlikely that instances might be 
 found which would go to support the rendering just given. 
 Still, they are inconsistent with the general analogy which 
 prevails in the composition of words, and ought not, of course, 
 
ROOT PRACH IN GREEK. 4-1 
 
 to be adopted without clear and convincing evidence. But 
 the form of the word supplies an insuperable objection to 
 this etymology. The primitive form, as we have seen, would 
 have been TrpoFoiros. To make TT^OTTO? out of this, we must 
 suppose that the weak letter F fell away, and that the o of the 
 preposition was then elided before the o of the root. Now 
 there are two circumstances which render this elision in the 
 highest degree improbable: i. that in other compounds of 
 this root the Homeric language holds on with markecTtenacity 
 to the F t allowing no elision to take place before it : so in the 
 compound with airb, which occurs very frequently, but al 
 ways as aTTociTrelv, never as aTreiTreiv : so in aftapToeTTijs, not 
 afj,apT7nj<; ; aperpoein]?, not a^eTpeirt^ ; and, 2. that even 
 before words which began with an original vowel, the prepo 
 sition Trpo always retains its o, suffering it indeed to be con 
 tracted with the following vowel, but never surrendering it 
 altogether: thus typovpos, a guard, from Trpo and opdco, not 
 <popo?. So strong are these reasons that, since Buttmann in 
 his Lexilogus declared himself against this derivation, it has 
 been generally abandoned : and it is matter of surprise that 
 Benfey in his Wurzellexikon should not only have retained 
 it, but retained it without the slightest expression of doubt, 
 without intimating that it either had been or could be called 
 in question. 
 
 Buttmann himself proposed to take the word from 6eo$ and 
 TrpeTra). This etymology would account well for the form, 
 which would thus correspond (for instance) to #60X0709 from 
 0eo? and \eya). How is it as regards the meaning ? In its 
 Homeric use, Trpiirw signifies to be prominent or conspicu 
 ous, to appear. It is said of things that strike the eye, and 
 by later poets is occasionally applied to objects that impress 
 other senses the hearing or even the smelling. From this 
 idea of appearing, the word came to mean, I. to appear 
 as something, be like it; 2. to appear well, be seemly or be 
 coming which in later writers is by far the most frequent 
 use of the word. But neither of these uses is to be found in 
 Homer : and, of course, neither of them can well be relied 
 upon for the explanation of OeoTrponros. For this purpose, we 
 must go back to the first meaning, to be conspicuous, to ap~ 
 
42 ROOT PR A CH IN GREEK. 
 
 pear. But this meaning is intransitive, and hence unsuited 
 to the accent of Oeoirpoiros, which, as we have seen, indicates 
 a transitive compound. Buttmann indeed suggests that 
 TrpeTTM may have been originally transitive, to make con 
 spicuous, to make appear, to show, and he finds an instance 
 of this earlier use preserved in a line of Euripides (Alcestis, 
 515, TL -fcprjiJLa rfjSe rfj Kovpa TrpeTret?) ; but his explanation of 
 the line, as President Woolsey has remarked in his note on 
 the passage, is not only inconsistent with the settled usage 
 of the tragic poets, but at variance even with the immediate 
 context. It must, however, be admitted as a thing alto 
 gether possible, that a word which in its separate use is 
 constantly intransitive may appear as transitive in a com 
 pound. Tips, if we assume it in the present case, would 
 give us for QeoTrpoTros the meaning one who makes a god ap 
 pear, manifests or reveals a god. Such a designation for the 
 diviner, though certainly quite conceivable, would belong 
 rather to the style of rhetorical description than to the simpler 
 and more obvious views of ordinary etymology. He is a re- 
 vealer of the will, purpose, or knowledge of the divinity ; but 
 to call him a revcaler of the divinity, one who makes the god 
 appear to men, seems unnatural and overstrained. Nor does 
 Buttmann himself appear to have understood the origin of 
 the term in this way. After explaining the primitive use of 
 TrpeTrcL), according to his view of it, as meaning to make a 
 thing appear, show it, he goes on to say that " probably the 
 old expression was #609 TrpeTrei, a god sends a sign (causes an 
 appearance) the sign sent was called OeoTrpoiriov and the in 
 terpreter of it ^eovrpoTTo?." In this explanation, it would seem, 
 he takes OeoTrpnTnov, the neuter noun, directly from $eo? and 
 TTyoeVft), with the meaning that which a God shews, the sign 
 or omen (a meaning, by the way, which the word appears 
 not to have in actual use) ; and then, from this deoTrpoTriov, 
 derives the masculine 6eo7rpoTro$, as if it were one who deals 
 in OeoTrpoTTia. It is hardly necessary to say that such a pro 
 cedure is wholly inadmissible, being in fact the reverse of the 
 real one : OeoTrpoTriov, instead of coming before OeoTrpoTros, 
 must be formed from it, so as to mean that which belongs to 
 his utterance or oracle. Compare fiavreiov 
 
ROOT PR A CH IN GREEK. 43 
 
 , formed from [iavrevs, a prophet/ with the mean 
 ing prophecy or oracle. 
 
 It is hardly worth while even to mention a third explana 
 tion of the word, according to which it would mean one who 
 speaks things Sew Trpenrovra, fitting for a god. 
 
 It is certainly a curious fact that a root so widely extended 
 in the cognate languages, and so fully represented in the 
 Latin, should be confined in Greek to a single compound, a 
 compound evidently formed after the language had assumed 
 its distinctive character, and yet one which may be referred 
 with probability to the earliest period of the language, and 
 which even in the classic time of Athenian literature had be 
 come an antique expression. Perhaps the accidental coinci 
 dence of sound between Trpevro), to ask, and TrpeTra), to ap 
 pear, may have had something to do with the early disap 
 pearance of the former. At all events, we have a somewhat 
 similar phenomenon in the Celtic languages. The Irish re 
 tains the root, but only in composition with a proper Irish 
 prefix a compound, therefore, which must have been formed 
 after the language had assumed its distinctive character. 
 From the Welsh it seems to have disappeared altogether, 
 
III. 
 
 THE GREEK GENITIVE AS AN ABLATIVE CASE. 
 
 1858. 
 
 IT is well known that some eminent grammarians have re 
 garded the Greek genitive as primarily a front-cast, and 
 have therefore sought to explain its various uses as being all of 
 them at bottom /r<? //^-relations. Thus Kuhner, in his largest 
 Greek Grammar (ausfuhrliche Grammatik dcr griechischen 
 Sprache), commencing the discussion of this case, says : "The 
 Genitive expresses, I. in local relation, the proceeding from 
 an object, or removal and separation from it, inasmuch as it 
 assigns the object or the point from which goes out the action 
 of the verb : yafeaQai Ke\ev0ov, to retire from the way ; 
 2. in causal relation, the cause, origin, author, in general the 
 object which calls forth the action of the verb, produces it 
 (gignit, hence the significant name genitivus)* excites it, 
 occasions it : iinQv^w Trjs aperrjs, I have a desire for virtue 
 (a desire awakened by virtue). As both in the local and the 
 causal relation the direction whence lies at the basis, the geni 
 tive may be named also the whence-cvse" Thus Kuhner, 
 who, in accordance with this exordium, proceeds to develop 
 the uses of the case as follows : (I beg leave to present a 
 brief outline of his development, and to solicit attention to it, 
 though the detail, I am afraid, may appear a little tedious) 
 A. In LOCAL relation, with words which express more or 
 less distinctly the ideas of separation and removal : to these 
 Kuhner adds verbs of beginning, but with questionable pro 
 priety ; vTrapxeiv abiKwv epycov is not to begin from unjust 
 actions, as a starting-point, but rather to make a beginning 
 
 * Kuhner perhaps means only that the term genitive is capable of bearing this 
 significance : not that those who first used it understood it thus. With them the 
 
 nominative case (^ bvo^affriKri} was the one used to name a person ; the genitive 
 (T\ yei/z/d?), the one used to express his yews, family connection, origin, or 
 descent : thus AA.e lai/5/w 6 Qih nnrov, Alexander the son of Philip. 
 
GREEK GENITIVE AS ABLATIVE. 45 
 
 of, /;/ unjust actions ; the special relation here would seem 
 to be partitive more than separative. Then B. In CAUSAL 
 relation, which is drawn out to much greater length than the 
 local, and is arranged under three principal heads, according 
 as the cause is conceived to be an efficient, an occasioning, or 
 a conditioning cause. Under efficient cause comes, I. Geni 
 tive of origih and author; 2. Genitive of possessor the owner 
 being regarded as in a manner producing what he posses 
 ses, making his money, making his property, as we phrase it ; 
 certainly, making it his own ; 3. Genitive of the ivkole or 
 genitive partitive the whole being conceived as in a sense 
 giving rise to its parts, producing them from itself; 4. Geni 
 tive of place (where) and time (when) the action being re 
 garded as evolved from and by the place and time of its 
 occurrence; 5- Genitive of material either the material 
 which makes, composes something, as a house of stone ; or 
 the material which makes a fullness of something, as full of 
 corn : i. e. genitive of plenty. Under this head Kiihner puts 
 constructions such as eaOieiv /cpewv, to eat of flesh-meat/ as 
 if the meat were the material which makes the eating (where 
 Prof. Crosby, with better reason, recognizes a partitive rela 
 tion) ; also a/covew dopvftov, to hear a noise, as if the noise 
 were the material of the sensation (which again is explained 
 differently, and, as it seems to me, better by Prof. Crosby) : 
 rov KaaiyviJTOv TI (/>?;? ; what say you of your brother ? (where, 
 however, the genitive seems really to depend upon the pro 
 noun what report, what account of your brother do you 
 give ? ) etc., etc. So much for the efficient cause. Next the 
 occasioning cause: this occurs, i. with words of emotion: 
 epwv e/wrafeo pvdwv, be heedful of my sayings the sayings 
 themselves are to serve as a cause, an occasion for heedful- 
 ness ; 2. with words of punishment, accusation, condemnation 
 these being consequences occasioned by the crimes to which 
 they relate; 3. in some other connections, among which we 
 find to? TroSoji/ efyov, (as they were in respect of feet, i. e.) ac 
 cording to their swiftness of foot, the feet being looked upon 
 as occasioning the state of swiftness, more or less, which they 
 happen to be in. Third and last, the conditioning cause. 
 This is used especially for expressing mutual relations, the 
 
46 REEK GENITIVE AS ABLA TIVE. 
 
 existence of the relation on the one side being a necessary 
 condition to its existence on the other. Thus, I. Genitive of 
 superiority and inferiority: Ticro-a^epvrjs ap%ei, TWV iroXecov, 
 Tissaphernes governs the cities to a governor subjects are 
 an indispensable condition; 2. Genitive with comparatives: 
 fji6L^o)i> rov a$e\<pov, greater than his brother to any greater 
 a something less is an indispensable condition. This, how 
 ever, Prof. Crosby explains more probably as a genitive of 
 distinction, i. c. greater (in distinction) from his brother ; 
 3. Genitive of price, value, merit the price of a commodity 
 being the condition of its sale ; 4. Genitive with substantives, 
 to express almost any mutual relation : <P/\wr7ro? o Trarrjp 
 A\e];dv$pov, Philip the father of Alexander Alexander here 
 is the conditioning cause of Philip s paternity. 
 
 Such is Kiihner s theory of the Greek genitive. Our dis 
 tinguished countryman and associate, Prof. Crosby, adopts 
 the same leading view, but develops it in his own way, with 
 characteristic ingenuity and elegance, with much more sim 
 plicity and naturalness than the German grammarian with 
 all the simplicity and naturalness, I believe, which the neces 
 sities of his theory admit of without the Beckeritc tendency, 
 too frequently apparent in K uhner, to impose a meaning on 
 language rather than educe the meaning out of it. 
 
 I propose to offer some remarks on this theory of the Greek 
 genitive, which regards it as always a. from-ca.se, and finds in 
 the idea of departure or derivation its primary and pervading 
 significance. That the Greek genitive is a from-case, in fre 
 quent use as such, is conceded at the outset : the question 
 will be, whether this was its primitive and universal character. 
 This question might be discussed from the standpoint of 
 philosophy if philosophy has any point that stands ; float- 
 point would perhaps be the more proper word. I propose, 
 however, to treat it as a question of history ; to inquire 
 whether in point of fact the uses of the case had the genesis 
 ascribed to them. It is only the possibility of treating it thus 
 in a historical way that makes it proper for me to bring it be 
 fore the notice of the Oriental Society : it is the historical 
 aspect of the subject that looks toward the Orient. 
 
 The Greek language, as all know, has not preserved the com- 
 
GREEK GENITIVE AS 
 
 plete case-system of the Indo-European class. In this respect 
 it is inferior to the Latin, which, in addition to the five cases 
 of the Greek, has an ablative, and, to some extent, a locative: 
 while the Latin is inferior to the Zend and Sanskrit, which 
 have a locative throughout, and besides that an eighth case, 
 the instrumental. It will be observed that this full system of 
 cases includes both ^genitive and an ablative ; and that these 
 are clearly distinguished from each other both in form and 
 meaning. The ablative singular in Sanskrit is made by adding 
 / or d to the stem : the rules of euphony in reference to final 
 consonants are such in that language as to leave it uncertain 
 whether the proper affix was t or d. But a comparison of 
 the Zend makes it probable that the original form was /. In 
 Latin the ablative singular ended in d. This ending, it is 
 true, does not appear in the classical Latin : but in the oldest 
 monuments of the language, such as the inscriptions of the 
 Columna rostrata (260 B. C.), and the Senatus-Consultum de 
 Bacchanalibus (186 B. C.), it is found as the regular and con 
 stant ending : thus we find prceda-d, scnatu-d as ablatives of 
 prceda y sen at us ; in a I tod niarid=in alto mari ; presented, 
 dictatored, from prcesens, dictator. The Oscan inscriptions 
 show the same ending d through all the declensions : e. g. 
 dolud malludy corresponding to Old Latin dolod malod, later 
 Latin dolo malo, with evil art, fraud. The ablative plural 
 in all these languages coincides with the dative plural, and 
 has in all of them substantially the same ending: in Sanskrit 
 bkyas, Zend byo (with 6 for as by a regular euphonic law), 
 Latin bus. If now we look at the meaning of the case thus 
 formed, we find it to be, what the name we give it imports, 
 a/n?;//-case. It is appropriately used in expressing removal, 
 separation, or distinction. It is used in a transferred or 
 metaphorical sense to express cause not cause in the all-com 
 prehending extension given to that idea by the theory just 
 described but cause that is cause (if I may be allowed the 
 expression), in which the character of origin, procession, 
 derivation, is obvious and palpable. As it may thus express 
 not only the producing cause, but also the means, it en 
 croaches on the borders of the instrumental : in Latin, in 
 deed, which has lost the instrumental form, the whole domain 
 
48 GREEK GENITIVE AS ABLATIVE. 
 
 of that case has been seized upon and appropriated by the 
 ablative. Still further, the Latin ablative covers in a great 
 measure the ground of the original locative, and is thus much 
 broader in its extension, and more various in its character, 
 than the corresponding case in Zend and Sanskrit. Still it 
 remains the fact, that in all these languages the ablative is the 
 from-ca.se, the proper case for expressing zfrom, whether 
 literally in space-relation, or metaphorically in those relations 
 which most distinctly and unequivocally imply a/r<?;/7-idea. 
 
 From the endings of the ablative, as just given, those of the 
 genitive are widely different. The Sanskrit genitive singular 
 adds s to the stem, often with a connecting-vowel a, making 
 as, Zend o, Greek 09, Latin is. But to stems in short a the 
 Sanskrit adds sya: thus vrikd, wolf, Nom. vrika-s, Gen. vrika- 
 sya. To this corresponds the Homeric genitive \VKO-IO (for 
 \VKO~O-IO) in which the a is dropped ; in the common form i is 
 also dropped and the two o s contracted : \VKO-O, \VKOV. In 
 Latin thcjtts of hn-jns, en-jus, c-jus, perhaps corresponds to 
 the same ending sya. In the genitive plural, the Sanskrit has 
 dm, Greek aw, Latin um. Some of the pronouns in Sanskrit 
 have sam instead of dm; thus from ka, interrogative, come Gen. 
 Plur. kc-sham, fern, kd-sdm. These reappear in Latin, with 
 the usual change of s to r, as quo-rum, qua-rum. The Latin 
 even extends this formation (in rum) to all words of the first, 
 second, and fifth declensions. The genitive is the only case of 
 the eight which remains to us in English (leaving out of view 
 the pronouns) ; and it is not at all unlikely that the meaning 
 indicated by our common name for it, the possessive, was in 
 fact the primitive meaning of the form. If, however, we look 
 at the general range of its use in the languages that have it, 
 we may describe it as the case of appurtenance, which means 
 belonging to something, pertaining to it. Its most frequent 
 use, and by far the most important, is with substantives, to 
 mark one thing as in some way belonging to, or connected 
 with another. When used with verbs, it represents the action 
 of the verb as belonging to or connected with the object, 
 rather than as falling directly upon it : as, in English, / know 
 of the man differs from I know tJie man. 
 
 Now in these simple facts we have the materials for an 
 
GREEK GENITIVE AS ABLATIVE. 49 
 
 argument, the grounds for a probable conclusion. The San 
 skrit, Greek, and Latin have a case common to all the three, 
 formed by all in the same way, by the same additions to the 
 stem, and employed by all in the same way (as regards its 
 leading use), i. e. as a case of appurtenance. The Sanskrit 
 and Latin have another case, common to both, formed alike in 
 both, and employed alike in both (as regards its leading use), 
 i. e. as a case of departure, a/V0;/z-case. This latter case is want 
 ing in Greek, its meaning being expressed by the one before 
 described. What is the natural inference ? Is it not, that the 
 primary use of the genitive in Greek is that which belongs to 
 the same form in the other languages, and that its ablative 
 use has been superadded to this, through the loss or aban 
 donment of a proper ablative form ? If so, then, instead of 
 explaining those uses of the Greek genitive in which it coin 
 cides with the Sanskrit and Latin by its ablative use, we 
 ought rather to reverse the process, to show how the abla 
 tive use can be explained from that of the proper genitive. 
 
 The only way, perhaps, of escaping from this conclusion, 
 would be to set up a hypothesis of extremely hypothetical 
 character, something like the following: "That the ablative 
 form as distinct from the genitive is a later thing in the his 
 tory of the Indo-European languages ; that the primitive stock 
 had but one form, the genitive ; that this form was used 
 originally in the meaning of an ablative, but in process of 
 time developed from that a great variety of uses, such as we 
 find in the Greek genitive ; that afterwards, for the sake of 
 distinctness or for some other reason, a new form was in 
 vented, an ablative form ; and that a portion of the uses of 
 the genitive, in fact its most original uses, were set off to 
 this new ablative case." 
 
 It is hardly necessary to point out the assumptions involved 
 in this hypothesis. If the ablative form is of later origin 
 than the genitive, it does not follow that its meaning was 
 before expressed by the genitive. If before the invention of 
 an ablative form its meaning was expressed by the genitive, 
 it does not follow, nor is there evidence to prove, that this 
 ablative meaning was the primitive use of the genitive, out 
 of which all its other uses were developed. But it is far from 
 4 
 
50 GREEK GENITIVE AS ABLATIVE. 
 
 certain that the ablative form is actually more recent than 
 the genitive. It is true that the ablative singular in t or d 
 does not occur, so far as we know, in the languages of Europe, 
 with the exception of the Latin and its Italican kindred. 
 But the non-appearance of a case-form in particular languages 
 does not prove that they never had it. Our vernacular has 
 in nouns but one inflected case, the genitive ; but our ances 
 tors a thousand years ago had a nominative, dative, accusa 
 tive, and vocative. In pronouns they had the trace of an in 
 strumental. Could we make out the Teutonic as it was a 
 thousand years earlier, we might perhaps find a complete in 
 strumental form, with an ablative and a locative, the perfect 
 apparatus of Indo-European noun-inflection. The ending of 
 the dative and ablative plural (Sanskrit bJiyas, Eatin bus) 
 lingers still on the extreme western verge of Indo-European- 
 ism, in the Irish bh. The case which has this ending is used 
 in the Irish as a dative plural : that it was once used as an 
 ablative is not unlikely, though we have no certain proof of 
 the fact, since the from-r elation is always expressed by means 
 of prepositions ; and though these prepositions are regularly 
 followed by the form in bli, yet nothing positive can be in 
 ferred from that, as other prepositions show the same con 
 struction. It must be confessed, however, that this discus 
 sion is somewhat irrelevant, with reference to our present ob 
 ject. Eor that, it matters little whether the genitive was or 
 was not earlier in its origin than the ablative. Whether or 
 not any members of the Indo-European family broke off from 
 the common stock, carrying with them a genitive, before the 
 ablative form had made its appearance, is of little conse 
 quence to us in this inquiry. Eor the Greeks, it is quite cer 
 tain, did nothing of the kind. On that point the Eatin abla 
 tive is an unequivocal and unimpeachable witness. The 
 Italican languages may have assumed a distinct existence 
 about the same time with the Greek : that they are not of 
 later origin than the Greek, will be admitted on all hands. 
 As they have a genitive and ablative completely distinct from 
 each other, we are forced to believe that the distinction of 
 these two cases in form and meaning existed in the common 
 language before the separation of the Greek, and was not the 
 
GREEK GENITIVE AS ABLATIVE. 5 1 
 
 result of any later development. It follows that the Greeks 
 must have given up an existing and established ablative form, 
 transferring its functions to the genitive. Why they did so, 
 it may not be possible for us to ascertain. We should find 
 it equally difficult to show why they gave up the instru 
 mental form, which is found in the Slavonic, the Lithuanian, 
 -and to some extent even in the Teutonic idioms; or why 
 they gave up the dative form, which is almost as widely re 
 tained among the languages of Europe as the genitive, while 
 they transferred its functions to the locative. It is important, 
 however, to observe, in regard to the ablative, that the San 
 skrit, even in its earlier Vedic form, shows us to some extent 
 the same condition which we find in Greek. The ablative 
 singular in /, already described, belongs to only one class of 
 Sanskrit nouns, though a most numerous and important 
 class words which have stems in short a, corresponding to 
 the Greek and Latin substantives and adjectives of the second 
 declension. For all other nouns in Sanskrit, the form of the 
 genitive singular is used also as an ablative, precisely as in 
 Greek ; though in the other numbers the Sanskrit invariably 
 distinguishes the two cases. The Greek, then, only carries 
 out consistently, through all words and all numbers, a ten 
 dency which the Sanskrit shows in most classes of words in 
 the singular number. Bopp conjectures that the use of the 
 genitive for the ablative in Sanskrit rests upon a merely 
 phonetic interchange of consonant-sounds ; he supposes 
 that nans, i ship (for instance), stem ndu or ndv, made origi 
 nally an ablative ndv-at, and that this by a sibilation of the t 
 became ndv-as, like the genitive. Now the assumed muta 
 tion of t to s is perhaps hardly borne out for the Sanskrit by 
 the euphonic analogies of that language. But for the Greek 
 there is no difficulty in such an assumption ; the euphonic 
 law of the language proscribes the final r ; and as we have 
 Trpo? for TTpori or TTpor, and XeXu/co? for \e\vrcor, Gen. XeTuwor- 
 09, so we might have vaos, VUFOS for an earlier vapor. The 
 identity thus arising between the genitive and ablative forms 
 in most singular words would naturally contribute much to 
 bring about a universal substitution of the one case for the 
 other. According to a plausible conjecture of Bopp s, the 
 
52 GREEK GENITIVE AS ABLATIVE. 
 
 same change of r to 9 has occurred even in adjectives of the 
 second declension, producing the adverbs in &)? (for COT) ; thus 
 , badly, would be for /caKwr, an original ablative of 
 OS, bad, though used in an instrumental rather than a 
 proper ablative meaning : compare Sanskrit papa, evil, 
 Nom. papas, Abl. papat, from evil, by evil means, in evil 
 manner. 
 
 I think it must be evident, from /these considerations, that 
 there is no historical reason for regarding those uses of the 
 Greek genitive in which it coincides with the common use of 
 that case through the other Indo-European languages as 
 arising from the idea of the ablative ; but, on the contrary, 
 that there are historical probabilities of considerable force 
 against such a supposition. In the view of comparative 
 philology, the theory of Kiihner, who deduces the idea of 
 appurtenance from that of departure, is less warranted than 
 Madvig s, who takes the idea of appurtenance as the proper 
 meaning of the case, and deduces from that the idea of de 
 parture ; since departure (he says), removal, separation, im 
 ply a previous connection or appurtenance. Thus the first 
 idea of the genitive would be in connection with ; the second 
 idea, out of connection with. Why the genitive should not 
 equally have come to mean into connection with, it might be 
 hard to explain. In fact, this view of Madvig s seems 
 scarcely tenable, and he himself does not insist upon it, or 
 attempt to give it that completeness of development which 
 Kiihner and Crosby have given to the opposite view. The 
 truth appears to be, that the Greek genitive combines in 
 itself the functions of two cases originally distinct, functions 
 associated in Greek not so much from any perceived internal 
 connection between them, as from accidents affecting the 
 outward forms of inflection. The Greek language itself fur 
 nishes another and more striking instance of this accumula 
 tion of offices, originally distinct, in one form. The Greek 
 dative, it is well known, both in singular and plural, has the 
 form of a locative case, denoting the place where or in which ; 
 but, as actually used, it combines, with the meaning of a 
 locative, those of the dative and the instrumental. These, in 
 the full system of the Indo-European languages, make three 
 
GREEK GENITIVE AS ABLATIVE. 53 
 
 distinct cases. In this instance, Prof. Crosby, guided by his 
 finer appreciation for language, has recognized the diversities 
 of meaning as fundamental, and abstained from any attempt 
 to connect the locative and instrumental uses with the proper 
 dative. Kiihner, on the other hand, whether that his sensi 
 bility to language was less delicate from the first, or because 
 it was blunted by his Beckerite philosophy, or for both rea 
 sons together, has been less abstinent ; he makes no scruple 
 to identify all these uses. The fundamental use of th e case 
 (he says) is the locative, to express the where : from this 
 comes, on the one hand, the proper ^dative, the whither, and 
 on the other, the instrumental, the whence t With equal pro 
 priety, as the genitive and dative have but one form in the 
 Greek dual, he might for that number derive all the uses of 
 the genitive also from what he recognizes as the fundamental 
 idea of the dative. This could give him no difficulty : for 
 as the locative where of that case has developed an instru 
 mental zvhence, and as all the relations of the genitive are in 
 his view W^;^^-relations, they could easily be connected with 
 the instrumental dative. To a grammarian whose where by 
 natural evolution develops a whither and a whence that is, 
 gives at once all possible space-relations it cannot be difficult 
 from any given relation to derive all other possible relations. 
 If the genitive were not only a genitive and ablative, as it is 
 in the Greek singular and plural, and to a great extent in the 
 Sanskrit singular, but a genitive and locative, as it is in the 
 Sanskrit dual, or a genitive, locative, dative, instrumental and 
 ablative, as it is in the Greek dual and even if to all these it 
 superadded the offices of nominative and accusative such 
 modes of explanation would be quite sufficient to demonstrate 
 the fundamental identity of all these uses. 
 
 Let me not be misunderstood. I do not regard it as ab 
 surd in itself (however unhistorical it may be) to derive all 
 the uses of the Greek genitive from a single root the one 
 idea of departure. If the historical argument were as strong 
 in its favor as it actually is against it, I should not reject it 
 as a thing incredible. The historical analogies sometimes 
 brought up in its support have, as it seems to me, but little 
 force. It is said that the Greek genitive can almost always 
 
54 GREEK GENITIVE AS ABLATIVE. 
 
 be represented by the English of/ and that of means from/ 
 so that father of Alexander must have meant originally 
 father/;w// Alexander/ house of Miltiades must have meant 
 ( house from Miltiades/ and so in all those uses where the 
 from-idea seems at first view most alien from the proper 
 meaning of the genitive. Similarly the French genitive is 
 made by the preposition de, the original meaning of which is 
 from. But the original meaning of a preposition is not of 
 necessity to be regarded as having once belonged to it in 
 every phrase or class of phrases which contains it. Compare 
 the English preposition t,o, as used with the infinitive. There 
 can be no doubt that this preposition originally meant 
 towards. There can be no doubt that, when first .used with 
 the infinitive, it was in this sense of direction or tendency. 
 But it is equally certain that with the infinitive, when used as 
 the subject of a sentence, it has not, and never could have 
 had, such a meaning. To err is human never meant, or 
 could have meant, toward erring is human. One who should 
 attempt to derive this use of the preposition by internal con 
 nection from its original meaning would have to subject lan 
 guage to unwarrantable force. The to was, no doubt, first 
 used with the infinitive in cases where the meaning towards 
 was appropriate either literally, / sent him to do it, i. c. to 
 ward doing it/ or metaphorically, I exhorted, commanded, re 
 quested him to do it. In this way it extended its use, until it 
 came to be regarded as the proper concomitant or exponent 
 of the infinitive, without reference to its origin : and only 
 then, when its original force had vanished, when it had taken 
 on a wholly different character, could it be used in such ex 
 pressions as to err is human. Now something like this may 
 have been true in reference to the genitive in of or de. The 
 preposition was perhaps applied first in those cases of the 
 genitive which may without violence be regarded as ablatives : 
 as John of Salisbury (where we may say either belonging to/ 
 or arising from Salisbury ), one of t lie peers (one belonging to, 
 or one taken from the peers), the son of Alfred (belonging to 
 him, or springing from him, as his son), etc. Having thus es 
 tablished itself in these and similar cases as an equivalent for 
 the genitive, the form with <?/may then have been extended, 
 
GREEK GENITIVE AS ABLATIVE. 55 
 
 without reference to its proper meaning, so as to represent the 
 genitive in its other uses. If this course of things be admitted 
 as probable, or even possible, it cannot be asserted that all the 
 uses of the genitive in of or de were ever actually thought of 
 as/r<?;/2-relations, in any development or modification of the 
 /n?;//-idea. And of course the English and French construc 
 tions cannot be relied on as parallel cases to support the 
 theory which makes all uses of the Greek genitive to befrom- 
 relations, developments or modifications of the/r0;#-idea. 
 
 It may be thought, however, that there is some practical 
 advantage in making the student refer all uses of the same 
 form, even if originally distinct, to one common root ; that 
 the one form and one universal meaning will thus stand 
 together in his mind, mutually supporting and supported, and 
 the phenomena of the language present a unity and harmony 
 which must be lost under a different treatment. Whatever 
 advantages may be gained in this way are perhaps hardly an 
 equivalent for the sacrifice of truth, at least of historic proba 
 bility, which this method involves. But there is another 
 point to be considered a question of pedagogical ethics 
 is there not some danger of blunting the philological con 
 science of the student ? While I admit the abstract possi 
 bility of deducing all uses of the genitive from this single 
 root, I cannot help believing, and my brief analysis of 
 Kuhner s scheme may bear me out in the belief, that the pro 
 cess can hardly be accomplished without a good deal of 
 straining and forcing. It cannot be well for the learner that 
 he should be accustomed, in the first stadium of his gram 
 matical course, to subject the facts of language to a process 
 of screwing and twisting ; that he should be taught to desert 
 the simple, obvious, natural interpretation of language for 
 that which is constrained or arbitrary. 
 
IV. 
 
 ON BEKKER S DIGAMMATED TEXT OF HOMER. 
 
 1863. 
 
 IT is more than forty years since Richard Payne Knight pub 
 lished in 1820 his famous digammated Iliad or rather 
 Vilviad of Homer. The book has taken its place among 
 the curiosities of literature. Its author was an ingenious and 
 elegant scholar ; but he had his hobby, and he rode it unmer 
 cifully. The horse of Phidippides, the spendthrift son in 
 Aristophanes Clouds, was marked with a KOTTTTO. (KOTTTT arias). 
 Payne Knight s hobby was branded with another lost letter 
 of the primitive Greek alphabet, the Digamma : wherever 
 he goes, he bears the digamma with him. 
 
 It is one of the most remarkable circumstances about Payne 
 Knight s Iliad, that, more than twenty years after its publica 
 tion, a distinguished American scholar should have thought 
 it worth while to reproduce three books of it on this side the 
 ocean (see Anthon s Home*r, New York, 1844). A page or 
 two by way of specimen might have been amusing at least, 
 even if uninstructive : but to take up in this way more than 
 fifty pages of a school-book was to make the joke somewhat 
 ponderous. 
 
 It might have been expected that the example of Payne 
 Knight would deter succeeding editors from repeating an ex 
 periment which in his hands had turned out in a way at once 
 so unfortunate and so ludicrous. But Immanuel Bekker, the 
 Coryphaeus of recent textual criticism, has not shrunk from the 
 hazard. In 1858 he brought out an Iliad and Odyssey in 
 which the lost letter is admitted to a place in the Homeric 
 text. This work embodies the results of many years minute 
 and laborious study. In 1809, after the appearance of Wolf s 
 Homer in its third edition, Bekker, then a young man, re 
 viewed it in the Jena Litter at ur-Zeitung. The review is said 
 
BEKKERS DIGAMMATED HOMER. 57 
 
 to have shown great mastery of the subject, and great aptitude 
 for those critical labors which were to form the life-work of its 
 author. In 1843 he published a new recension of the Homeric 
 text, which was immediately and universally recognized as a 
 marked advance on that of Wolf. For the last five or six 
 years, he has been giving out in the Monatsberichte of the 
 Berlin Academy a highly remarkable series of observations 
 and researches in reference to Homer. With great ingenuity 
 and acuteness, they evince an amount of patient labor which 
 is absolutely marvellous. Thus, he goes through the whole 
 extent of the poems to note and collect the verses in which 
 the third foot is without a caesura. In the 15, 694 verses of 
 the Iliad, he finds only 185 which have no caesura in the third 
 foot; in the 12,101 of the Odyssey, only 71. Again, he goes 
 through the whole extent of the poems to mark the cases of 
 bucolic caesura, and observe whether the fourth foot, the one 
 which precedes the caesura, is a dactyl or a spondee. Thus, 
 in the fifth book of the Iliad he finds 531 bucolic caesuras, of 
 which 470 are preceded by dactyls, 61 by spondees ; in the 
 eleventh book, 575 bucolic caesuras, 478 preceded by dactyls, 
 97 by spondees ; and so on for the other books. These are 
 only specimens of the tasks which this conscientious and in 
 defatigable critic has imposed upon himself. The results of 
 these protracted investigations appear in his last edition of 
 the Iliad and Odyssey, that of 1858. This edition shows a 
 great advance upon his first, of 1843. It is, in fact, con 
 structed on a different principle, and aims at a different ob 
 ject. The aim of Bekker in his first edition, like that of Wolf 
 before him, was in general to reproduce the Homeric text as 
 it was settled by the great critic Aristarchus, about two cen 
 turies before Christ, and handed down without intentional 
 variation by subsequent copyists. It was the rule with Wolf, 
 and with Bekker in his first edition, to give the readings which 
 certainly or probably belonged to Aristarchus, except in oc 
 casional instances where there was unequivocal evidence to 
 show the priority of a different reading. But in his second 
 edition Bekker has taken a wider range. He has adopted as 
 his guide the principle of analogy, and by the help of it has 
 sought to go back beyond Aristarchus. Relying on analogies 
 
58 BEKKERS DIGAMMATED HOMER. 
 
 presented by a careful study of the Homeric poems, he has 
 departed in many cases from the readings of the manuscripts, 
 even where these could be traced with more or less certainty 
 to Aristarchus himself. The general propriety of this method 
 has been disputed in many quarters. It is, indeed, rather 
 singular in a critic like Bekker, who strenuously maintains 
 the fragmentary origin of the poems, and who finds evidence 
 of such an origin in the varieties and inconsistencies which 
 they show, both as to grammatical forms and as to the use 
 of words. He expects departures from analogy ; he re 
 gards them as having an d priori probability ; and yet the 
 tendency of his criticism is to sweep them away from the text, 
 wherever this can be done by gentle means : for he abstains 
 on principle from changes of a violent or extreme character ; 
 he does not treat his text with the despotic ingenuity of a 
 Bentley. 
 
 But our object at present is to consider only one feature of 
 the work its introduction of the digamma. The objectors 
 generally admit that the digamma-sound (the v, or rather, the 
 w-sound) belonged originally to the Homeric poems, and that 
 it is proper in commentaries and other philological works to 
 point out the traces of its existence and to discuss the extent 
 of its use. But they object to a cligammated text. They 
 maintain with much plausibility that the poems from their 
 first reduction to writing have never shown this letter ; and 
 that the attempt to go back, not only beyond the first manu 
 scripts that we have, but beyond the first that ever existed, 
 can have no reasonable hope of success. At any rate, they 
 say, the case is not yet ripe for a digammated text. In re 
 gard to many words, it is still uncertain whether they were or 
 were not sounded with a digamma in the Homeric time ; and 
 in regard to many which certainly were so, it is doubtful 
 whether they were uniformly sounded with this letter, or 
 whether it was not sometimes omitted in pronunciation. If 
 we take words which certainly had a digamma in the Homeric 
 language, and attempt to represent them uniformly with this 
 letter, we must make many violent and arbitrary changes of 
 the text. If we adopt the principle of giving them with di 
 gamma wherever it can be done without such changes, we have 
 
BEKKER S DIGAMMATED HOMER, 59 
 
 to draw an*uncertain line between changes which are violent 
 and changes which are not so. And whichever of these 
 courses we take, we can have no assurance that we are repro 
 ducing the genuine Homeric usage. It is impossible to deny 
 the force of these objections. But their force would be much 
 greater, if by the decree of fate the world were restricted to 
 one printed text of Homer, just this and no more. In that 
 case, we should say without hesitation, give us a text which 
 comes as nearly as possible to that which Aristarchus follow 
 ing, as we know that he did, with great soberness and cau 
 tion, the testimony of the best manuscripts that he could find 
 fixed upon as the true one ; or, if you depart from that, do 
 so only when there is decisive evidence to warrant the de 
 parture. As a basis for Homeric study, as a standard for 
 general use and reference, a text thus constituted is the best 
 that we can have. But we are by no means restricted to a 
 single text. For general purposes, we may continue to use 
 Bekker s first edition, or we may take, what differs very little 
 from it, the text of Dindorf in T.eubner s Bibliotheca, or any 
 better one which we can find constructed on the same prin 
 ciples. But Bekker s second edition will still have its value 
 as a tentative, to show how far the principle of analogy, in the 
 hands of a consummate critic, will serve to correct and im 
 prove the text of Homer which has come down to us by tra 
 dition from the ancient Alexandrine editors. And especially 
 with reference to the digamma it will have a value of this kind, 
 as showing what results can be secured by an intelligent, 
 moderate, and cautious attempt to reinstate the long-lost 
 Homeric letter. On this point the editor himself says, in his 
 brief and pithy preface : " The yEolic letter, after it had dis 
 appeared through time and negligence, was by the mar 
 vellous sagacity of Bentley reclaimed from oblivion ; but lay 
 thus for a long time, ridiculed by wits, by scholars invidiously 
 assailed or unintelligently defended [the last evidently a hit at 
 Payne Knight]. By Heyn it was admitted, at least into his 
 commentary. The indispensable uses of this letter I could 
 no longer treat with contempt. I have therefore restored the 
 digamma, but so far only as I had the power and right to do, 
 proceeding cautiously and with moderated step : I have re- 
 
60 BEKKER S DIGAMMATED HOMER. 
 
 stored it to its own place, as indicated by manifest" traces, not 
 by eager wishes or by hasty assumptions of my own." 
 
 This language is fully borne out by an inspection of the book. 
 The carefulness and conscientiousness of the editor are every 
 where apparent. In deciding what words are to be regarded 
 as having the digamma, he relies mainly on the indications of 
 the Homeric verse. From this it follows as a natural con 
 sequence that he recognizes only an initial digamma. Thus 
 he writes Jto?, of Zeus, /cX^t?, * key, ot?, sheep, not AIFQS, 
 K\r]Fis, OFW, though there is strong reason, derived from in 
 scriptions or from later dialects or from cognate tongues, for 
 believing that these words had digamma in the Homeric lan 
 guage. He does not, however, reject the initial digamma of 
 a word when it is brought by a prefix or by composition into 
 the middle. Thus the digamma of peiicoa-i, twenty, is re 
 tained in epeiKoai,, that of Fayvvfii,, to break, in erdyrj, was 
 broken, that of WX-Tro/zat, to hope, in FeFO\7ra, I hope/ 
 that of FiSetz/, to see, in Arl8r}^ unseen god, Hades, that 
 -of Fr)8v$, * sweet, in yueTu/^S?;?, fieXcF^B^a polvov, honey sweet 
 wine. He rejects all combinations of digamma with another 
 consonant. Thus Srfv, long, which in numerous passages 
 has the appearance of beginning with two consonants, and 
 has been supposed by many to have the digamma-souncl after 
 the 8 (8Fi )v), is by Bekker always written with a simple 8. 
 In 8ec8ia, I fear, many, since Buttmann, have recognized a 
 Homeric Se^Fta ; in e SSetera, * I feared, a Homeric eSfeiaa : but 
 Bekker always writes them according to the traditional way, 
 only omitting one 8 from eSSeio-a. Nor does he recognize any 
 lost letters beside digamma. Curtius, in the second part of 
 his Principles of Greek Etymology, has endeavored to show 
 that a consonants-sound has in some instances given rise to 
 the same appearances in the Homeric verse as those oc 
 casioned by the v- or w sound (the digamma). But the words 
 which are thus supposed to show traces of initial y Bekker 
 either writes with digamma, or leaves them with a vowel-ini 
 tial. We shall refer again to this point before closing. 
 
 We have said that, in determining what words had initial 
 digamma in the Homeric language, our editor relies mainly on 
 the indications of the Homeric verse. Having satisfied him- 
 
BEKKEFS DIGAMMA TED HOMER. 6 1 
 
 self in the case of any particular word that it did have the di- 
 gamma, he proceeds to write the word, as also its derivatives 
 and compounds, with that character : and this he does, not 
 only where metrical reasons favor or require the introduction 
 of a consonant, but wherever metrical reasons do not abso 
 lutely forbid the introduction of a consonant. In very many 
 instances where the verse as we have it in our traditional text 
 will not allow the digamma to come in, the difficulty can be 
 removed by changes of the text which are more or less obvi 
 ous. Bekker s principle, it is evident, has been to write the 
 word with digamma, whenever this is consistent with the 
 verse as it stands in the ordinary text, or can be made so con 
 sistent by some slight and easy change of reading. He shows 
 his judgment and moderation as a critic by refusing (at least, 
 in general, with only rare exceptions) to make any consider 
 able or arbitrary change of reading for the sake of getting in 
 his digamma. Rather than do this, he will allow the word to 
 appear in a particular case without the initial consonant which 
 usually belongs to it. I may illustrate his mode of procedure 
 by a more particular statement of what he has done in the 
 first book of the Iliad. In the 611 lines of which it consists, 
 there are found, if I have counted right, 162 which show the 
 digamma. But some of these contain it more than once, so 
 that 184 words are written with this character. In 36 of these, 
 it is found, not at the absolute beginning of the word, but 
 after a preposition or other prefix. Of the 184 words there 
 are only 31, or about a sixth, in which the introduction of the 
 diagamma has required any further change of text ; and in 18* 
 of these 31, the only change required has been the omission 
 of a movable v from the end of a preceding word. Thus in 
 verses 
 
 14. crrf/xjotar e^cov eV ^epcrtv Kr)j36\ov A.7r6\\o)vo<s. B. ^epcrt Fe/c. 
 
 96. TOVVZK ap aXye e SwKev e/<7^/3o/Vos r/S ert Scocret. e ScoKe Fe/c. 
 
 294. et Sqtroi TTO.V tpyov V7reio//,ai, orrt 
 
 KZV 
 
 In 8 instances, a slight change has been made in the gram 
 matical form of the preceding word, though in 2 out of the 8 
 the change was not necessary in order to the introduction of 
 the digamma. The other six are as follows : 
 
BEKKEFS DIGAMMATED HOMER. 
 
 vna 
 
 21. dojaeroi Aios t tov ei<r)/3<)\ov ATroAAwra. P>. via 
 230. 8tup a.Troa.ipe icrOaL, os ns crtOev avriov tiTry. 
 
 288. Tra^TWT/ ^tev Kpareetv e$eA.ei, Travrecrcrt 8 di/dcnreiv. TTMTLV oe Fav. 
 
 365. olcrOa Tit] rot raOr etSm ry TraVr dyopcixo. ravra Fiovirj. 
 
 482. are/ p?7 Trop^upeov yiteyaA.* ta^e r^os tovcnys. /xeya 
 
 576. I(r6\.rjs ecrcrerat ^Sos, eTret ru. ^cpetoi^a vi/<a. eVrrat 
 
 In three instances a particle which seemed unnecessary has 
 been omitted to make room for the digamma : 
 
 64. 05 K /, TTOI ort rocrcrov e^wcraro ^ot)8os A/roXXooi/. 13. o? FetTri^. 
 548. cure ^(7)7/ Trporepos rovy etcrcrat oirr dr^peoTrcov. TW Fetcr. 
 
 582. a/\A.a cru rovy e/reeo-crt KaOaTrrtcrOai /xttA.aKotcrtv. TOV FCTT. 
 
 In two instances one particle has been substituted for an 
 other : 
 
 19. eKTre porai Hpta/toto TroXtv t! 8 oiVaS tKecr^at. B. Kal Fot/caS*. 
 
 395- ^y eVret wi/?ycra; KpaSr^v Aios ?ye /cat epyw. ^e Tt Fep/\w. 
 
 I do not find in thq first book any instance in which words 
 are transposed for the sake of bringing in the digamma. I will 
 add one or two instances from other parts of the poems : 
 
 X. 341. /u.cr<r>7yus Kp^rr/pos tSe Opovov dpyvpor/Aov. B. Kp-^Tr}pos /xecro-^yu FtSe. 
 ^. 37- coraorav ev St^poujt * 7nrracr<je 8e 6*v/xos cKacrrov. 
 
 8 eTrdracrcre Fe/c. 
 
 There are several instances in the first book where a word 
 usually digammated is compelled by stress of metre to forego 
 this addition. The lines which I have noted are 
 
 216. xpry IATJV o-(a/irepov ye, $ea, FeVos etpt cro-acr^at. not Feipvcrcracr^cu. 
 
 239. Trpos Atos etpmrai o 8e TOI /xe ya? to-crerat opKOS. Fetpvarai. 
 
 294. et 8^ aot TTtti/ epyov V7rei^o/xat, OTTI Ke FetV^?. VTT(,Fto/>tat. 
 
 438. CK 8 eKaro/x/3^1/ fir/crav eK7y/5o/\a) A7roAA vi, FeKfjfioXia. 
 
 555- v ^ v ^ a ^^5 8et 8otKa Kara c/>peVa /xi^ ere 
 
 My search was a hasty one. It is most likely that careful 
 looking would bring out a few more cases of this kind. 
 
BEKKER S DIGAMMATED HOMER. 63 
 
 Before proceeding to notice and criticise the treatment of 
 particular words, there are two remarks of a general nature 
 which it seems important to premise. They relate to the evi 
 dence in favor of a digamma, and the evidence against a 
 digamma, in the case of any particular word. 
 
 First, as to the evidence in favor, it must always be borne 
 in mind that there are cases in which hiatus was more or less 
 freely allowed in the epic verse, so that hiatus occurring in 
 such cases furnishes little, if any, presumption for an initial 
 digamma. In my Grammar (67 D) are mentioned three 
 cases of this kind: I. when the first of two words ends in a 
 close vowel (i, v), and seldom or never suffers elision : this 
 applies especially to the dative singular of the third declen 
 sion, as TraiSl oiracrcre ; 2. when the two words are separated 
 by a clearly required mark of punctuation, as fcdOrjao, e/ny> S 
 eTTiTreiOeo jjivOw ; 3. (the most important case) when the vowels 
 which make hiatus are two short syllables of the third foot, or, 
 in other words, at the feminine caesura of the third foot, as 
 TWV ol ef eyevovro em fjueydpoicrL jeveOXr/. In this place it has 
 been proved that hiatus is allowed with much the same free 
 dom as at the end of a verse. There is another case which 
 ought to be added to these three a case in which hiatus is 
 easily excused, if not freely allowed and that is, after a long 
 vowel or diphthong in arsis, and particularly the arsis of the 
 third or the fifth foot. The first line of the Iliad is an instance 
 in point : 
 
 I. /tvjvLv aeiSe, Otd, n^X^taSecL) A^i/Y^cs. 
 
 which shows hiatus after the arsis of the fifth foot : after the 
 arsis of the third, we find it in A. 
 
 24. d/VA. OVK ArpetSr; Aya/xe/xvovt r/i/Save Ov/Jiw. 
 42. Ticreittv Aayaot e/xa Sd/cpua croicn 
 
 Here also, after the long arsis of the third foot, as well as in 
 the feminine caesura of that foot, we find something of the 
 same freedom as at the end of a verse. This appears in such 
 lines as A. 
 
 153. otvpo fjLa^rjcrop.vos^ 7ret ov n /JLOL airtot 
 
64 BEKKERS DIGAMMATED HOMER. 
 
 where o?, the last syallable of /^a^cro/xero?, stands in the 
 third arsis before eVet , which certainly did not begin with 
 digamma. 
 
 The other remark relates to the negative evidence, that 
 which goes to disprove a digamma. It is well known that, 
 for every digammated word, even the best ascertained, there 
 is some evidence of this character : there are some passages 
 in the poems as we have them, in which the digamma cannot 
 be written without violating the metre. It is obviously desir 
 able that we should have some idea as to the range of these 
 exceptional cases, their numerical ratio to the whole number 
 of passages in which the word occurs. On this subject there 
 are some good remarks in Kuhn s Zeitschrift, x. 60 ff . , in an 
 article by H. L. Ahrens on e/cacrro?, one of the words which 
 Bekker has digammated. Ahrens enumerates all the instan 
 ces, 110 in number, where this word occurs in the Iliad, and 
 states that there are 44 of these in which the digamma could 
 not be written into the traditional text without a violation of 
 metre. But as 16 of the 44 can be made to admit the 
 digamma by simply dropping a movable v from the preceding 
 word, they are left out of the count, which reduces the un- 
 conformable cases to 28 in 110, or about 25 per cent. This 
 ratio, Ahrens says, is so large as to throw discredit on the ini 
 tial digamma. To prove it so, he takes the word ava% 
 (originally Famf), king, which he finds to occur 151 times 
 in the Iliad. Here, setting aside, as before, the cases in 
 which only a movable v stands in the way of the digamma, he 
 makes 1 1 out of 151 to be the number of unconformable cases, 
 or about 7-^ per cent. He then proceeds to say that as to 
 other words which have an unquestionable digamma, as epyov, 
 ISelv, ol/cos, ot^o?, etc. , any one may satisfy himself by his 
 own observation that the percentage of unconformable cases 
 is not larger, that it is rather smaller, than for ava%. It could 
 not justly be inferred from these observations of Ahrens that 
 no words had the digamma in Homer for which the uncon 
 formable cases exceed 7 or 8 per cent. ; but only that in such 
 words the digamma must be regarded as more or less doubt 
 ful, and, if the proportion is very much greater, as improb 
 able. It is also evident that the weight to be given to this 
 
BEKKEFS DIGAMMATED HOMER. 65 
 
 test will depend somewhat on the absolute number of instan 
 ces in which a word is found. If the word occurs but seldom, 
 the ratio of conformable and unconformable cases may be in 
 a measure accidental ; but the influence of accident dimin 
 ishes as the numbers we are dealing with increase. 
 
 We come now to some criticism of particular words, as 
 written by Bekker with or without digamma. All the words 
 which have been generally agreed upon as showing evidence 
 of this initial receive it here. The list of digammated words 
 given in my Grammar (23 D) was not designed to be complete, 
 but only to include the most important roots in which traces 
 of the consonant-initial have been generally recognized. It 
 contains about 33 distinct roots ; and in all these, without ex 
 ception, Bekker has admitted the digamma. Besides these, 
 he admits it in some 2O or 30 more, for I have not been able 
 yet to make out an exact list. In many of these additional 
 words, the real existence of the digamma is beyond all rea 
 sonable doubt. This is true, for instance, of ero?, year/ 
 which connects itself naturally with Skt. vatsara, year, 
 Lat. vctus, old (i.e. full of years, annosus ). Out of 19 
 Homeric passages which show the word, only two resist the 
 introduction of the consonant, and these allow it if we only re 
 insert an elided vowel : thus, rooraavr erea (B. 328) may be 
 changed to roo-a-avra rerea, vroXX erea (T. 255) to 7ro\\a rerea, 
 the last two vowels of Ferea being taken as one syllable by a 
 frequent synizesis. But, as might have been expected, there 
 are words written by Bekker with digamma in which there is 
 room for doubt and for difference of opinion. A striking in 
 stance of this kind is found in the word eXcop, prey, which 
 occurs 8 times, with the connected e\wpa and eXcepta, each of 
 which occurs once (the last in A.. 4). In three of these 10 
 passages (P. 667, A. 4, E. 684) there is a hiatus before the 
 word, but it is at the feminine caesura of the third foot, where 
 hiatus has scarcely any weight in proving an initial consonant. 
 In 5 passages (E. 488, P. 151, 7. 271, e. 473, w. 292), a v 
 movable precedes, which neither hinders us from assuming 
 digamma, nor furnishes any proof of its existence. The re 
 maining two passages give evidence of an initial vowel. In 
 one, IlaTpQKXoLQ 8 eXcopa (2.. 93), the e of Be is elided before 
 5 
 
66 BEKKER S DIGAMMATED HOMER. 
 
 eXcopa : in the other, ^77 TTW? aoi e\cop (v. 208), the diphthong 
 of poi is made short. In the first case, Bekker gets rid of the 
 difficulty by reading TlarpoKKov 8e rekcopa ; in the second, h e 
 yields to the difficulty, and writes e\cop without digamma. 
 We cannot here lay much stress on the proportion (20- per 
 cent.) of unconformable passages, the whole number being so 
 small. But as digamma is not required in any single passage 
 of the 10, and is excluded by 2, it is certainly hazardous to 
 assume it without other proof of its existence. Such proof 
 one might perhaps find in its derivation. It is natural to take 
 it from the root which appears in the second aorist of aipeco, 
 infin. e\eiv, indie. el\ov, where the augment affords evidence 
 of an original consonant initial. If el\ov is for e^eXov, eXetz/ 
 for Fekeiv, we might connect them with Latin vello , to pluck. 
 But the 2d aor. of aipeco is never written with digamma by 
 Bekker, and it is quite clear that it was not digammated in 
 the Homeric language. It furnishes, therefore, a very fee 
 ble presumption for the digamma of the substantives ; and we 
 cannot but conclude that it would have been the safer and 
 wiser course to leave the substantives also without di 
 gamma. 
 
 Another word in which we must question the propriety of 
 the digamma that Bekker gives it is the deponent verb 
 epvofjicii, to watch, guard, preserve. This verb, in many of 
 its forms, is apparently identical with epvco, to draw ; and it 
 has been assumed almost universally that they come from 
 the same root. Buttmann in his Lexilogus argues the ques 
 tion at length, maintaining their essential identity. Apart 
 from the indications of a digamma, there are other reasons 
 for separating the two verbs. Thus, as to form, epvco, to 
 draw/ shows ei only where it would arise from augment or 
 reduplication; while epvofjiai, to guard, shows elpva-aovrai in 
 the future, elpvaaaa-Oai, in the aorist infinitive, and other like 
 forms. Again, epvo/ncu, to guard, is sometimes inflected ac 
 cording to the /u-form, as in epvro, elpvvro, etc., which is 
 never the case with epvco, to draw/ And yet again, with 
 epvo^ai, to guard, there is a verb pvofjLai,, with initial p, which 
 has the same meaning : with epvco, t to draw, there is no such 
 side-form. Of these points of difference, Buttmann does not 
 
BEKKERS DIGAMMA TED HOMER. 67 
 
 notice the first two : as to the last, he says with much force 
 that the substantive pvrr)p, pole of a wagon (drawer) gives 
 proof of a verb pvw = epvw, to draw. If we turn from the form 
 to consider the meaning of these verbs, we find something 
 of a step from drawing to watching/ Buttmann, however, 
 bridges over the gulf: from drawing to one s self (the proper 
 sense of the middle form) comes the idea of rescuing ; from 
 rescuing that of guarding : from guarding that of watch 
 ing and even of * watching against. The development is cer 
 tainly possible, and, if it stopped at the point of rescuing, we 
 might regard it as probable : but from rescuing to watching 
 against, watching to injure, there is still along journey, which 
 we cannot assume without hesitation. But these reasons for 
 separating the words gain almost irresistible confirmation 
 from a circumstance which Buttmann has not noticed viz., 
 that the indications of the Homeric verse show very clearly 
 that e pvco, to draw, began with a consonant, and almost as 
 clearly that epvop.ai^ to guard, began with a vowel. I have 
 collected the Homeric passages which show middle forms of 
 epvco, to draw : I find 60 in all, many of which give strong 
 proof of an initial consonant, while only 3 (/. <?. 5 per cent, 
 of the whole number) oppose its introduction. Of the de 
 ponent epvofjiat, I find 43 instances in all, among which 23, or 
 more than half, resist the introduction of a digamma. Forms 
 which begin with e followed by pp I of course do not reckon, 
 as they obviously belong to pvoftcu, not epvo^ai,. It might be 
 said, however, that some forms in which e is followed by 
 single p, such as epvro, was guarding, epvcro-aro, guarded, 
 could also be taken from pvopat,, the p being left single after 
 the augment, as often happens in the aorist of />ef&>, to do. 
 Assuming this, we shall have in all 29 instances of epvofjiai, of 
 which 12, or more than 40 per cent., will resist the introduc 
 tion of digamma. Again, it might be said that such a form 
 as elpvo-aaro is to be explained from eFepvaro-aro, by omission 
 of digamma and contraction of the vowels, so that we could 
 not expect to see in elpvaa-aro the digamma which belongs 
 to the verb-root. If we admit the justice of this reasoning, 
 we shall still further reduce the number of instances to be 
 considered, bringing them down to 19 ; of which, however, 
 
68 BEKKER S DIGAMMATED HOMER. 
 
 9, or nearly half, will still oppose the digamma. It is possi 
 ble that two or three of the cases which I have regarded as 
 middle forms of epiw might be assigned to epvopai, in the 
 sense of rescuing ; but if we should transfer them accordingly, 
 this would not materially affect the numerical relations just 
 exhibited. Observe, then, that in the middle of epra) only 
 one twentieth of the instances resist the insertion of digamma, 
 while in epvopai, to guard, nearly or quite half of them resist 
 it. This is a very great difference, and cannot possibly be 
 imputed to accident. I hold, therefore, that Ahrens is fully 
 justified in separating the two words, as he has done in his 
 Grammar of 1852 ; and I regard as highly probable his con 
 jecture that pvo/j,ai began originally with cr (compare the 
 Greek a conjunctive, which was originally sa), and that it is 
 connected with Latin servo. The primitive sense may have 
 been that of watching/ which we see in the compound ob- 
 scrvo, and from which we readily derive the ideas of guarding 
 and preserving. But whether it once began with a or not, 
 we must in any case disapprove the procedure of Bekker in 
 writing it, wherever he can, with digamma. There are in 
 fact only 4 places out of more than 40 which give any sign of 
 an initial consonant. Two of these are in the 23d book of 
 the Odyssey (82, 229), which has in it much that is peculiar, 
 while the others are in a line that occurs twice (i. 194, K. 
 444) : 
 
 avrov Trap vr]i re ^ikveiv teal vr}a epvadai. 
 
 This shows another metrical irregularity, the short i of vrjt 
 being used for a long syllable. Apparently it is only a varia 
 tion of the perfectly regular verse 
 
 avrov Trap vrjeacn fjueveiv KOI vt)a$ epvcrdai (. 260, p. 4 2 9) 
 
 the two plurals being changed to singulars, with little regard 
 to metrical exactness. 
 
 In speaking of epvofjiai, I have touched incidentally upon 
 the question whether digamma should be prefixed to the 
 augmented forms of digammate verbs. Wherever the aug 
 ment makes a syllable by itself, Bekker, no doubt with cor- 
 
BEKKER S DIGAMMA TED HOMER. 69 
 
 rectness, writes the digamma after it : thus erdyrj, l was 
 broken, from Fdyvv/M] eVaA/^ was pressed, from pei\w ; 
 FL7rov, 1 1 said, tense-stem FGITT, from repeir. But when the 
 augment coalesces with the root in the same syllable, he places 
 digamma at the beginning, and of course before the augment : 
 thus he writes relSov, I saw, originally eriSov ; F^vaaae, 
 was ruling, originally epavacrcre ; prjvSave, was pleasing, 
 originally epavbave. Now the temporal augment of ijvao-cre 
 must either have come from a stem which had already lost 
 the digamma, or it must have arisen from a stem with digam 
 ma, by dropping that consonant between e and a, with con 
 traction of these vowels. In either case, the augmented form 
 should be without digamma, which could only appear by what 
 must be regarded as an improbable transposition : eravaaae, 
 peavaaa-e, pijvacrcre ; epiSov, pelbov, pelSov. It is, at any rate, 
 a transposition which we should not accept without clear 
 indications in the Homeric verse. I must own that I have 
 not looked up the evidence myself on this point. But a 
 writer in Jahn s JalirbucJier (Ixxxi. 68 1), Heinrich Rumpf, 
 professes to have done so. He says, at least, that he has 
 looked at all the 2d aorist forms of the root iS which by the 
 augment begin with ei, also at the forms of dvaao-co, avbdvw, 
 and ayvv/j,i, which by the augment begin with 77. The num 
 ber of these, taken together, would be about 60, and there 
 are 6 of them (-v/r. 392, 7. 305, i. 182, K. 373, X. 162, r. 539) 
 which resist an initial digamma. The proportion here is not 
 decisive. But it is more important that he finds not a single 
 case which requires digamma, and only one which on Bekker s 
 principles can be regarded as yielding it any particular sup 
 port. We must conclude, then, that there is no sufficient 
 warrant for Bekker s writing of these forms. 
 
 In this connection I may speak of the form TJLKTO, a pluper 
 fect middle of the stem IK or eirc. It occurs four times in the 
 Odyssey in the expression Se/ms 8 TJLKTO yvvai/ci, which does 
 not allow an initial digamma. "Hi/cro is most naturally ex 
 plained as being equivalent to epepiKro, the first e being the 
 augment of the pluperfect, which after the loss of digamma is 
 contracted with the e of the reduplication : eperiKTo, eet /rro, 
 r/t/cro, like epavacrae, eavaaae, ijvacrcre. Now it is remarkable 
 
70 BEKKER S DIGAMMATED HOMER. 
 
 that the form e i/cro, with e instead of 77, occurs once, in i|r. 107 
 /cat fjioi Gfcacrr eVereXXei/ ei /cro Se 0ea-/ce\ov avra). This form 
 is naturally explained as being for Fepucro, with the reduplica 
 tion, but without the augment, of the pluperfect. It will be 
 seen that the passage allows, though it does not require, an 
 initial digamma. Bekker writes it without ; in our judgment 
 he should have inserted it : thus, KCLI fjLOi eicaa-T eVereXXe 
 F6FLKTO Se OeaiceKov avra). 
 
 If we have complained of Bekker for prefixing digamma to 
 the augmented forms of digammate verbs, we have to com 
 plain of him for omitting digamma in some instances from 
 their reduplicated forms. The word just mentioned, in which 
 he writes eFt/cro, not reft/cro, is a case in point. Another is 
 seen in %. 348 : 
 
 iravTolas evefyvorev FOi/ca Be TOL Trapaeibeiv. 
 
 On first looking at this, I thought that perhaps the pause 
 (colon) before eoi/ca might have had something to do with 
 Bckkcr s retention of the preceding v movable. But I found 
 afterwards a passage (I. 70), which in this respect is exactly 
 similar, but is differently treated by Bekker : 
 
 Satvu Salra yepovcri FFOi/ce TOL, ov TOL aFGt,K$, 
 
 This inconsistency, I suspect, must be the result of inadver 
 tence. In all other cases, so far as I have observed, Bekker 
 writes the perfect and pluperfect active of this verb with di 
 gamma, where the verse allows it. The number of instances 
 is very large 125, if I have counted right and the uncon- 
 formable cases only 10, or about 8 per cent. The perfect of 
 av^avw (root FO&) occurs but twice (I. 173, a. 422). Bekker 
 both times writes TTCLCTIV eraBora. We hold that he should 
 have written TTCLO-L FFaB6ra : the presumption is that the di 
 gamma was regularly repeated in the reduplication, as Bek 
 ker gives it in F&FOiica. The perfect middle of ei\co (root F\) 
 occurs four times, twice after v movable, and twice after a 
 hiatus, which however is at the feminine caesura of the third 
 root. Bekker everywhere writes eVe A/^e&i, eVeXyaez o? : we hold, 
 as before, that he should have written FeFe\fJie6a, F6F\/Aeves. 
 The next case to be considered that of eKTro^ai (root F\TT), 
 
BEKKERS DIGAMMATED HOMER. *J\ 
 
 to hope is attended with more difficulty. The perfect 
 eo\7ra and pluperfect ea^ireiv occur twelve times in all : 3 
 times with hiatus, 4 times with v movable before them : there 
 fore / times where digamma is admissible; leaving 5 cases 
 which resist it. This large proportion of unconformable cases 
 might make us doubt whether we ought to recognize digamma 
 at all in these forms. But the F of the root is unquestionable, 
 and gives a strong presumption for F in the reduplication. 
 And besides, the three cases of hiatus occur in a part of the 
 verse (at fern. caes. of 2d foot) where hiatus is inadmissible. 
 We hold, therefore, that there is sufficient evidence of Ho 
 meric F6Fo\7ra and FF(*)\7riv (or FFO\7reiv), and that these 
 forms should have been given, according to Bekker s princi 
 ple, wherever the verse allows them. He has in fact given 
 them only in the 3 cases of hiatus, while in the 4 of v movable 
 he retains that letter, and writes eVoA/Tra, Fco\7rei,i> ; thus con 
 travening both his general method and his procedure in the 
 parallel case of eoiica, ewiceiv. In eopya, dapyeiv, we find very 
 much the same state of things 12 passages in all, of which 5 
 resist digamma. The v movable, however, occurs here in only 
 one case (f. 289) : 
 
 Tp<w/eT?79, 05 S?) 7ro\\a Kate dvOpcoTrotcriv eaypyei. 
 
 Here, from the analogy of his procedure in reference to 
 (i>\7rei,v, we may presume that Bekker would have written 
 dvOpMTTOicnv epcopyeiv, if he had not followed Voss in making a 
 greater change, altering the dative to an accusative, in ac 
 cordance with the usual construction of the verb, making 
 dv6p(*)7rovs elFO)p<yiv. It might be questioned, however, 
 whether we ought not in this case to have rjFopyeiv for 
 FFop<yeiv, in the same manner as TJFLKTO for CFCFLKTO. In 
 the perfect middle of epya) (root Fepy), to shut, we find a 
 different state of things. Here we have ep^arai, and ep^aro 
 occurring 7 times. They are evidently forms without redu 
 plication, like olSa, I know (i. <?. FolSa, not F6Foi$a), 
 elfiai, am clothed (i. . Fea^ai, not FeFeor^aL) ; and, in 
 stems beginning with other letters, Se^arat, have received 
 (for SeSe^arat), avwya, I command (for tfvtoya). Hence, 
 when we find eepy^evai and eep^aro occurring each of them 
 
72 BEKKERS DIGAMMATED HOMER. 
 
 once, we must presume that the first e is not part of a redu 
 plication, but the same common prefix which we find, for 
 instance, in eeXSo//,at (i. . eF\$ofj,at) for Fe^o/mai, to wish, 
 and the aorist participle eeio-dfievos (i. c. e^etcra/ae^o?) , hav 
 ing likened one s-self. We shall not then be surprised to see 
 that erepry/jievat, and erep-^aro do not admit initial digamma. 
 We come now to consider the question whether our editor 
 does right in recognizing only one lost consonant, the di 
 gamma, or whether he should not have recognized others as 
 producing similar appearances in the Homeric verse. Cur- 
 tius, in the concluding part of his Principles of Greek Ety 
 mology, maintains in the case of several words that the epic 
 hiatus was occasioned by a consonant j/-sound. He holds 
 this to be true in reference to eoi/ca, ecpfceiv, which we have 
 just considered. He remarks that dialects and inscriptions 
 give no evidence of digamma in this word ; that no root vik 
 in the sense of likeness is to be found in the cognate lan 
 guages ; and that it is therefore very hazardous to write 
 FeFOitca, FeFoo/ceiv, in the text of Homer. He observes that 
 there are clear traces in Herodotus and elsewhere of a word 
 Se/#??Xos or Set/eeXo? having the sense of (e)ftfeXo?, like, simi 
 lar. He is therefore led to adopt the conjecture of Bopp, 
 that the root of eoi/ca is formed from that of BeiicvvfJLi, Lat. 
 dico, Skt. dig (i. e. dik), to show. He conceives that the 
 B assumed a parasitic j, and then dropped away itself thus, 
 dik t dyik, yik and that from yik thus formed came by redu 
 plication yeyoiKa, yeywiceiv. I cannot think that there is much 
 plausibility in this explanation. If the transition from dik to 
 yik was made in the formative Indo-European period, we 
 might expect to find a root yik having the sense of likeness 
 somewhere in the cognate languages, which Curtius does not 
 pretend is the fact. It is evident, indeed, that he regards 
 the evolution of yik from dik as taking place in the Greek, 
 after the Indo-European time. We must think, then, of the 
 root dik as already provided with inflection, making a redu 
 plicated preterite SeSoi/ca, from which would come first, by 
 adding y to both S s, SyeSyoi/ca, and then, by dropping both 
 the S s, yeyoiica. But the change from 8 to y is confessedly a 
 rare one in the Greek language ; how hazardous then to as- 
 
BEKKEKS DIGAMMATED HOMER. 73 
 
 sume that it has occurred twice in the same form ! It might 
 perhaps be said that the change occurred first in some such 
 form as $uce\os, meaning like, which passed into yuce\os ; 
 that this gave the suggestion of a root jn/c, meaning to be 
 like, and that yeyoucd was formed independently from this 
 suggested root, and not by phonetic change from a pre-ex 
 isting SeSoifca. This is indeed possible : but we should 
 scarcely expect that a root arising at this comparatively late 
 stage of linguistic development would take the more primitive 
 formation seen in eouca, with its interchange of L/C, eac, one. 
 For such reasons, the rise of eouca from a rookdik seems to me 
 scarcely more than a possibility. It must be observed, too, 
 that unless the connection of eouca with a root dik is rendered 
 probable, there is no more reason for writing it with y than 
 with F. And if no more reason, we may justly say that there 
 is less reason for y than for F, because the very regularity with 
 which this word in the Homeric text gives evidence of a con 
 sonant initial is a circumstance in favor of F. In this respect 
 it ranks, as we have seen, among the most regular words, the 
 unconformable cases being only 10 in 126, or about 8 per 
 cent. If we consider how much earlier and more complete 
 was the disappearance of y from the language when com 
 pared with that of F, we shall be slow to believe that the 
 former should maintain itself in the Homeric verse with the 
 same constancy as the latter. 
 
 Another case in which Curtius recognizes traces of an 
 initial y in Homer is the deponent tefjuai,, to be eager, to de 
 sire, to long. It occurs in6i instances, of which 22 by hiatus 
 give evidence of a consonant initial, and the unconformable 
 cases are only 3, or about 5 per cent. Bekker writes it with 
 digamma. To this Curtius objects, asserting that the verb 
 f lr)[LL is a reduplicated form of the root yd, which appears in 
 Sanskrit, and is itself an extended form of the root /, to go. 
 Thus lt]iJbi=yi-yd-mi ) to cause to go, to send. In the middle 
 this would mean to send one s self, and hence to hasten, to 
 pursue eagerly, to aim at, to long for. To this no objection 
 can be made as regards the meaning. But it is a remarkable 
 circumstance that t ^/u, to send, shows no traces in Homer 
 of anything but a vowel initial. Of simple forms in the pre- 
 
74 BEKKER S DIGAMMATED HOMER. 
 
 sent and imperfect these I take for comparison because 
 terete, to desire, is confined to those tenses I find in the 
 sense of sending 29 (all active except A. 77, M. 274, ^. 304, 
 which show the middle or passive) ; and of this 29, not less 
 than 24, or more than 80 per cent., refuse to admit a conso 
 nant initial. Compound forms, such as a^irjfjii, yue^/^/xt, etc., 
 I have not taken into the account : I believe, however, that 
 all of them which are capable of furnishing any evidence on 
 the point testify against a consonant initial for the simple 
 verb. It may be regarded as perfectly certain that vrj/M, to 
 send, was sounded by Homer with an initial vowel, and lepai, 
 to desire, with an initial consonant. We have here a dis 
 tinction of the same kind as that which we before proved to 
 exist between epvay, to draw, and pvo/j,cu, to guard. 
 Now it might be said by one who maintained the original 
 identity of T?7/u, to send, and le/mai, to desire/ that the y 
 which once belonged to both alike was retained, and that with 
 uniformity, in the sense of desiring, after it had been lost, 
 and that with uniformity, in the sense of sending. That this 
 is a possibility we admit, but it is nothing more. The pro 
 bability is that the two words are radically distinct ; and if 
 so, then, for the same reason as before, the ^has more in its 
 favor than the_^. The fact that later dialects furnish no sup 
 port to it is of little significance, as the deponent te/Aai be 
 longs only to the early language. 
 
 The remaining case in which Curtius recognizes initial y as 
 exercising the power of a consonant in the Homeric verse is 
 the relative stem, which appears in 09, ?5, o, o>9, olo9, ocro?, 
 OTTO)?, O7roto9, o(f>pa, ^09, <W, 6W9, etc. Savelsberg, in the 
 eighth volume of Kuhn s Zeitschrift, has given a series of cita 
 tions from Homer showing traces of an initial consonant for 
 this class of words. Unfortunately he has not furnished 
 exact numerical data, by which we might see the comparative 
 frequency or infrequency of the phenomenon. I have not 
 myself had time to supply the deficiency. I have only run 
 hastily through the first book of the Iliad, noting all words which 
 show the relative stem omitting, however, the relative ad 
 verb 0)9, which in this respect stands by itself. The whole 
 number of instances noted was 72, of which 30, or about 42 
 
BEKKEKS DIGAMMA TED HOMER. 7 5 
 
 per cent., testify against a consonant initial. Of the remain 
 ing instances, many were found at the beginning of a line, or 
 in other indecisive positions ; and in fact there are only 8, or 
 one ninth of the whole number, which give any indication of 
 a consonant initial. Even of these, the majority, either from 
 the part of the verse which they occupy, or from the pause 
 which precedes them, are of but little \veight : only 2 or 3 
 give decided indications of an initial consonant. I strongly 
 suspect that a more extended comparison would not essen 
 tially change the proportions derived from this first book. 
 They seem barely sufficient to give plausibility to the con 
 jecture that the relative stem did once begin with a consonant, 
 but had nearly or quite lost it in the Homeric time. I say 
 " nearly or quite : " for if the letter had wholly died out from 
 common use shortly before Homer s time, the force of epic 
 tradition would probably have caused some traces of it still 
 to appear in his verse. But the adverb &>? differs in this re 
 spect very remarkably from the other forms of the relative 
 stem. According to Bekker, as cited by Curtius, the in 
 stances which indicate a consonant initial are three times 
 more numerous than those which indicate a vowel initial. It 
 can hardly be doubted that this word, as pronounced in the 
 Homeric time, began frequently, if not generally, with a 
 consonant. 
 
 It must be owned that our condition as regards the ety 
 mology of the Greek relative is an unsatisfactory one. We 
 are less confident and comfortable than we were ten years 
 ago. Then we had no hesitation about connecting it with 
 the Skt. yas, yd, yat, assuming a change of y to the rough 
 breathing, as in fjTrap (i. e. ^Trapr), liver/ Skt. yakrt (i. e. 
 yakart). But now, if we do not surrender this conviction, 
 our faith in it has become less full and sure. A Locrian in 
 scription, published by L. Ross in 1854, presents the form 
 JFOTT, with digamma, for ort. A digammated form of the 
 relative stem is also seen in a gloss of Hesychius, quoted by 
 Savelsberg : .Bo-Xt/acor^?, <rtW(j5>?7/3o9, Kpijres i.e. for "youth 
 ful companion " the Cretans use /SaTu/acor???, (i.e. ^aAt/aam??, 
 equivalent to vJXt/awr?;?) . To these testimonies, Savelsberg, 
 in the article referred to, adds the indications of digamma in 
 
76 BEKKER S DIGAMMA TED HOMER. 
 
 the Homeric verse, and concludes that the Greek relative was 
 ^6?, FIJ, FO, or FOS, pa, FOT. These he supposes to have been 
 later forms of KFOS, KFCL, KFOT, Latin qui, quac, quod. He 
 thus identifies again the Greek and Latin relatives, though 
 in a very different way from that of the old-fashioned ety 
 mology, which held that the original h of the Greek relative 
 was in Latin hardened to a / -sound (</?/). The omission of 
 the / -sound in the Greek 1^09 would be something like that in 
 the Latin nbi t undc, for cnbi, cnndc, which remain in the 
 compounds sicubi, alicunde. The stem kva t which would 
 thus underlie the relatives of these two languages, Savelsberg 
 supposes to have been developed out of ka, the stem of the 
 Sanskrit interrogative. He goes yet further, and from the 
 same origin derives even the Sanskrit relative : ya, he thinks, 
 is for kya ; and kya, like kva, is only an altered form of ka. 
 But Schleicher and Curtius are not yet prepared to admit 
 that the Greek relative-stem began with digamma, still less 
 that it was ever kva. The former touches on the subject in his 
 Compendium of Comparative Grammar, p. 180: the latter, 
 more at length, in his Principles of Greek Etymology, ii. 
 177-8. In respect to the FOTI of the Locrian inscription, 
 they say that, when the digamma-sound had nearly vanished 
 from the Greek dialects, its sign was sometimes used improp 
 erly by scribes or grammarians for other spirants, and espe 
 cially for the y, which had no sign of its own even in the 
 earliest Greek alphabet ; and they appeal to a Corcyraean in 
 scription, which shows a genitive singular masculine of the 
 first declension in -AFO, where all analogy would lead us to 
 expect -ayo or -ahyo, Skt. asya. As regards the Homeric 
 usage, they say that the phenomena which seem to indicate 
 digamma could equally well be produced byjj/. In this there 
 is no intrinsic improbability, though we should be glad to 
 have the support of some parallel case which we could look 
 upon as clear and certain. The parallels which Curtius 
 brings forward are the verb (e^ai, l to aim at, desire/ and the 
 root of eoi/ca ; neither of which, as we have seen, is much to 
 be relied on. As to the derivation of a relative ^09 from KFOS 
 by omission of K, Curtius remarks that " the only phonetic 
 analogy which could be called in to support it is that of the High 
 
BEKKERS DIGAMMATED HOMER. 77 
 
 German ivcr (for /twer)= Gothic Jruas, cf. Eng. what. But 
 the loss of the feeble h proves little for that of k ; and how 
 improbable that of the two consonants the Greek would give 
 up the perfectly familiar K. in favor of the unstable digamma, 
 wavering even from the earliest time ! " It may be observed, 
 in passing, that Curtius own derivation of eoi/ca from dik y 
 dyik, yik is liable to the same objection : it makes the Greek 
 give up the familiar 8 in favor of the unstable and perishing 
 y. "Still less," he continues, "can it be proved that Skt. 
 yas has come from kyas, and that ka, with the secondary 
 kva, kya, is the common root of all these widely ramified 
 pronouns. Finally, the demonstrative meaning of the Greek 
 o? in real o? e^rj speaks against this derivation, and recom 
 mends the assumption that the originally demonstrative stem 
 /, with the secondary formj/tf, lies at the basis of the Greek 
 relative." As the demonstrative use in KOI o? e^fyrj is confined 
 to the nominative, while in the accusative we have KOI TOV, it 
 seems to me quite possible that the o? is for o, by confusion 
 of the two forms o? and o, so much alike in appearance, 
 though so diverse in origin. Curtius then adds, as Schlei- 
 cher also does, that, if the Greek relative did really begin with 
 F, it would be preferable to explain it from a stem sra, which 
 appears with relative force in Gothic svc, as/ whence the 
 German so in its relative use. This relative stem sva was long 
 ago recognized in Greek by Curtius himself (Kuhn s Zeit- 
 schrift, iii. 75, 76), though only in the merest relic, the adverb 
 <?;, as, which the Alexandrine critic Zenodotus read in two 
 passages of the Iliad (B. 144, H. 499). Lottner afterwards, 
 in Kuhn s Zeitschrift, ix. 320, proposed to derive all the 
 forms of the Greek relative from this same stem. But the 
 traces of a digamma in the Greek relative are much less fre 
 quent and decided than we should naturally expect to find 
 them if this were its real derivation much less so than in the 
 forms of the possessive o?, r/, 6V, * his, her, its, which come 
 from a stem of the same sound, sva, though of widely different 
 import. Possibly the fact which we have noticed, that the 
 adverb o>? differs so much from the other forms of the relative 
 in the indications which it shows of a consonant initial, may 
 warrant the conjecture that they are of different origin that, 
 
78 BEKKER S DIGAMMATED HOMER. 
 
 in fact, &)? came from the digammate stem sva, while 09, oto?, 
 ocro9, and the rest, are akin to the Sanskrit yas, yd, yat, and 
 came from a stem with initial _y-sound. 
 
 Ahrens, in Kuhn s Zeitschrift, x. 65 ff. , has sought to show 
 that appearances in the Homeric verse similar to those pro 
 duced by digamma are in some instances to be ascribed to a 
 lost sigma. An instance of this kind in which, however, 
 the initial a is not lost, but retained in the written text is 
 presented by the word 9, hog. The simple form of this 
 word appears in 55 passages with initial <r, as 0-1)9, 0-^09, etc. 
 It occurs also in 21 passages where an initial consonant would 
 be incompatible with the metre ; and in all these places the 
 a is dropped ; we have 9, 09, etc. In the later language, 
 both forms of this w 7 ord were in use, though the one with 
 consonant initial was comparatively rare. If, when the poems 
 were reduced to writing, the form with a had been wholly 
 lost from use, it is probable that our written text would have 
 shown 9 only, with initial vowel, in all the 76 passages, 
 though in many of them the metre would have shown traces of 
 a lost consonant. And it is quite conceivable that in other words 
 this may actually have been the case. Such an occurrence 
 Ahrens recognizes with no little plausibility in the words v\rj, 
 wood, eo9 possessive, and e/cao-ro9, each. In regard to 
 v\?1, Lat. silva, it is certain that it began originally with a, 
 and equally certain that in Homer s language it usually began 
 with a vowel. But there are two cases of a remarkable 
 hiatus before the word (oreiero 1^X77, H. 285, and eTre^evaro 
 V\TJV, e 257, where v\rj in each forms the sixth foot), which 
 seem to show that the initial cr was not wholly forgotten in 
 the Homeric time. The possessive 609, in its relation to 09, 
 is explained by Ahrens in a w r ay which has been quite gene 
 rally received as probable. He assumes in the earlier period 
 two forms, a fuller 0-6^69, and a shorter OTOS : from 0-6^69 
 came regularly the 609, from orbs the 09 of our common text. 
 This explanation is supported by the analogy of the posses 
 sive forms T609 and 0-09 for the second person ; of which, in 
 all probability, reo 9 is for T6F09, and 0-69 for TFO?. Now in 92 
 instances of the pronoun 09, there are 52 which do not allow 
 an initial consonant, and it is therefore certain that in the 
 
BEKKER S DIGAMi 
 
 Homeric time it generally began ^&S&f^&Ji:Jfc^ there 
 are 4 instances (A. 533, I. 420, 687, 0/524) oiremarkable 
 hiatus in the first foot, after the first short of a dactyl (Zet>9 
 Se edv, x W a *n v > ?re fc 9 ^)> which seem to present traces of 
 the primitive initial a. And moreover, this word makes hiatus 
 14 times in the feminine caesura. We have seen that hiatus 
 is readily allowed in that place ; but its relative frequency 
 is so great in the case of this word (three times greater than 
 in the analogous case of e ^09, my ) as to warrant the sus 
 picion that it arises from a peculiar cause, and is connected 
 with the primitive initial <r. Bekker writes the pronoun, 
 wherever he can, with initial F : he appears to suppose that 
 from FOS came ew by a prefixed e, and then, by transposi 
 tion of digamma, peos. But such a transposition is a more 
 hazardous assumption than he seems to think ; and the form 
 reos has little support, either in the Homeric text or in the 
 suggestions of comparative philology. 
 
 It remains to speak of the pronoun e/cao-ro^, each, and 
 the kindred adverb e/cdrepde, on each side : the pronoun 
 e/cdrepos itself is incapable of appearing in the heroic hexa 
 meter. The derivation of kfcdrepos and ettaa-ros is as yet far 
 from certain. It is probable, however, that the -/carepos and 
 -KacrTos are a comparative and superlative form from the in 
 terrogative stem KCL that they are, in fact, identical with the 
 interrogatives jrorepos and TTOCTTO?, which in their Ionic forms 
 are fcdrepos and KOCTTOS. It is probable also that the first syl 
 lable e is the same as in the numeral e/carov, Lat. centum, Skt. 
 catam (i. . katam). If so, it is probably for ev, the root of 
 the numeral et?, one : thus, eKarov 1 ONE hundred, e/cdrepos 
 = ONE which-more, one which of two. Here now we stum 
 ble again upon an uncertainty : but of all the explanations 
 proposed for the numeral et?, ev, the most probable is that 
 which connects it with Skt. saina, our Eng. same, Lat. scmcl, 
 simplex , singuli. It thus appears that cr may probably have 
 been the primitive initial of e/cdrepos, e/caaros. We have 
 already observed that, according to Ahrens s enumeration, 
 e/cacrro? occurs no times in the Iliad. Now 66 of these are 
 cases of hiatus, some of them easily admissible, but many 
 others giving strong indication of a consonant initial. It is 
 
So BEKKER S DIGAMMATED HOMER. 
 
 not therefore surprising that Bekkcr should have written 
 Fe/cao-ros and FettdrepOe wherever the verse allows it. Out of 
 28 unconformable cases, he makes 17 conformable by various 
 conjectures, several of which belong to the most hazardous 
 that he has ventured. In II cases he has left the initial 
 vowel untouched. Here the proportion of unconformable 
 cases, 25 per cent., throws suspicion on the digamma, which 
 is much increased by the fact that comparative philology has 
 no plausible explanation for the forms pe/cdrepos, re/cao-TOs. 
 Such forms as a-e/cdrepos and oreKacrros are much more proba 
 ble on grounds of comparative philology. Practically, then, 
 the case as to eWcrro? stands in this way. It cannot well be 
 doubted that the word, sometimes at least, began with a con 
 sonant in the Homeric language. If we assume that digam 
 ma is the only initial consonant of the Homeric language 
 which has failed to appear in our text, then we must recog 
 nize a Homeric Fe/cacrro^ : such, doubtless, was the reasoning 
 of Bekkcr. But the assumption is an unsafe one : there is 
 reason for suspecting that other initial consonants of the 
 Homeric language have had, though to a far less extent, the 
 same fortune as digamma ; and in this particular word there 
 is reason for suspecting that it began with some other conso 
 nant. Yet we would by no means advise either that the 
 relative-stem should be written in Homer with initial y, or 
 that e/cacrro? and eo? should be written with initial cr. A lost 
 digamma manifests itself in the Homeric verse in many words 
 with much clearness, and with considerable approach to uni 
 formity : it may therefore with propriety be inserted in an 
 edition having the character and aims of the one under re 
 view. But the case is widely different with a lost y and a 
 lost a : these, if we make the most of them, are only rare 
 and doubtful. 
 
V. 
 
 ON ANCIENT GREEK RHYTHM AND METRE. 
 
 1864. 
 
 RHYTHM consists in a regular succession of times of 
 proportional times so marked off and distinguished 
 that the proportionality of the times and the regularity of 
 their succession shall be obvious to human sense. These 
 times may be marked by movements of the body, as in danc 
 ing ; by tones of various pitch and stress, as in music ; by 
 the syllables of uttered words, as in poetry. We have here 
 the three principal applications of rhythm, three principal 
 domains in which rhythm manifests its nature and power 
 dancing, music, poetry. They were recognized as such by 
 ancient writers, as arts which are all alike under the sway of 
 rhythm, in which the same principles of rhythm find applica 
 tion and illustration. These principles do not require words 
 for their manifestation ; they do not require even sounds : the 
 silent art of orchestic has its arses and theses, its trochees and 
 iambi, its dactyls and anapaests, not less truly than music and 
 poetry. In fact, the name feet for rhythmic elements, arsis. 
 (raising of the foot), thesis (setting down of the foot), have 
 primary reference to orchestic. It is apparent from these 
 remarks that rhythm may be treated in two different ways. 
 Its principles may be set forth in a general, abstract manner : 
 not without illustrations drawn from the concrete forms of 
 dancing, music, or poetry ; but in such a way as to give 
 prominence to those abstract, general principles which per 
 tain alike to these three arts. The subject thus treated is a 
 true and proper rhythmic. But, on the other hand, a writer 
 may take up one of the arts in which rhythm manifests itself, 
 making it his leading aim to set forth the forms of that art as 
 they have been developed under the influence of rhythm. If 
 poetry, for instance, is subjected to this process, the result 
 will be a system of metric. Metric is a description, a scienti- 
 6 
 
82 GREEK RHYTHM AND METRE, 
 
 fie description, of verse according to its rhythmical forms. 
 Ancient literature had its works both of rhythmic and of me 
 tric, agreeably to the distinction which we have here traced 
 Those on metric were the more numerous, and have been 
 more largely preserved. The most important is the Eyxei- 
 piSiov, or Manual, of Hephsstion, a grammarian who is sup 
 posed to have flourished about the middle of the second 
 century after our era. Unfortunately these metrical writers 
 had only an imperfect conception of the work which they 
 were taking on themselves. In treating of metric, it was their 
 business to point out in a systematic way everything pertain 
 ing to the rhythmic enunciation or delivery of verse ; so that 
 from a study of their writings one might form a distinct con 
 ception of the way in which an ode of Pindar or a chorus of 
 Sophocles was actually intended to sound I mean, as regards 
 its rhythm. To do this it was not enough to note the succes 
 sion of long and short syllables. It was necessary to mark 
 also the rhythmic accent, and to distinguish a heavier accent 
 from a lighter. It was necessary to note the cases in which 
 rhythmic times were occupied by a prolongation of long 
 syllables ; for we know that these were sometimes so far pro 
 longed as to become equal to three shorts, four shorts, or even 
 five shorts. It was necessary to note the cases in which 
 rhythmic times were occupied by pauses intervals of silence, 
 like the rests in music for we know that such pauses were 
 frequent, taking the time of one short, two shorts, three 
 shorts, or even four shorts. It was necessary in treating of 
 lyric compositions, such as Pindar s Odes, or the choruses of 
 the dramatists, in which the verse changes from line to line, 
 to point out the principles of symmetry and the laws of suc 
 cession by which these changes were doubtless regulated and 
 controlled : for we have the strongest reason to believe that 
 here, as everywhere else, the Greeks exercised their prodi 
 gious power of invention in subordination to very definite 
 canons of taste and beauty. Now of all this we find very 
 little in the ancient writers on metric. They confine them 
 selves, for the most part, to noting the succession of long and 
 short syllables. It is true that they have something to say- 
 though in a quite imperfect and unsatisfactory manner on 
 
GREEK RHYTHM AND METRE. 83 
 
 rhythmic accent, on thesis and arsis, the accented and unac 
 cented parts of the foot. But beyond this they scarcely go. 
 I may illustrate their deficiencies by a single instance. It is a 
 fact that there were spondees in which each syllable had the 
 length not of two shorts, as usual but of four. It is a fact 
 known to us from other sources ; but in the writings on metric 
 now extant there is no hint of it whatever. This imperfection 
 in the works on metric is not hard to account for. Their 
 writers were mere grammarians : they were much interested 
 in syllables, and their long and short quantity but they had 
 not the theoretical* and practical knowledge of music which 
 was indispensable to the proper execution of their task. I 
 say that a theoretical and practical acquaintance with music 
 was essential to a proper treatment of the ancient metric. It 
 is a very important circumstance, and one that should never 
 be lost sight of, that the ancient poetry was much more closely 
 bound up with music than the modern. Even the simplest 
 kinds of verse, the epic hexameter, the dramatic trimeter, 
 were pronounced they were intended to be pronounced in 
 a kind of recitative, a sort of semi-musical utterance, with 
 musical accompaniment. But lyric compositions, such as the 
 odes of Simonidcs and Timotheus, the choruses of ^Eschylus 
 and Euripides, were designed to be sung. The poet was a 
 musician also ; his contemporaries were accustomed to think 
 and speak of him as not only poet but musician. His musical 
 characteristics formed an essential part of the critical judg 
 ment passed upon him by the public for whom he labored. 
 The great variety and complexity of the rhythms which he 
 used depends upon the fact that they were intended to be 
 sung. Rhythmic structures so various and complex were not 
 fitted for reading, and would never have been produced for 
 it. They require musical utterance for the development and 
 appreciation of the rhythm. German translators have often 
 overlooked this fact. Hartung endeavors to reproduce the 
 metres of Pindar line for line in his German version. It is 
 impossible that an attempt so difficult and constraining should 
 not operate disadvantageously on the force, aptness, and clear 
 ness of his translation. And what is the gain ? No one could 
 give the German verses their intended rhythm without a care- 
 
84 GREEK RHYTHM AND METRE. 
 
 ful and painful scrutiny of the annexed metrical scheme. And 
 when all is done, when with much effort you can accent the 
 heavy German lines on the right syllables, the rhythmical 
 result is one which Pindar himself would never have regarded 
 as legitimate or desirable. 
 
 We may take yet another instance, an extreme case, to 
 show how the writers on metric some of them, at least 
 while counting longs and shorts, could overlook the most 
 obvious facts of rhythm. The so-called elegiac pentameter, 
 which in the elegiac distich alternates with the dactylic hexa 
 meter, is in fact only a variety of the hexameter a hexameter 
 in which the third and sixth feet are reduced to a single 
 syllable, the remaining times in those feet being made up 
 by pauses. If we take the first line of the Iliad M.r]viv aeiSe, 
 Bed, n^XtjldSeco A-^i\i]o^ and substitute a pause for the last 
 half of the third and the sixth feet MTJVIV aeiSe, ded, X^aSeo) 
 A^tX.rj we have the form of the elegiac pentameter. Or 
 take the English distich which Coleridge translated from 
 Schiller, but unluckily forgot to mention Schiller s name in 
 connection with it r "In the hexameter rises the fountain s 
 silvery column, In the pentameter aye falling in melody 
 back." Here fill out in words the pauses of the second line, 
 " In the pentameter aye it is falling in melody downwards," 
 and you have a hexameter line, like the one before it. 
 Hephaestion had the grace to see this : he has given a correct 
 description of the so-called pentameter, which recognizes its 
 true relation to the hexameter. But there were writers who 
 gave it a different description, who made it consist of five 
 feet, whence the false name of pentameter -five feet, of which 
 the first two were either dactyls or spondees, the third always 
 a spondee, and the fourth and fifth anapaests. Such a de 
 scription answers certainly to the succession of long and short 
 syllables in this verse ; but it utterly fails to give an intelligible 
 conception of its rhythm. 
 
 But beside the ancient writers on metric, there were others 
 who treated of rhythmic, according to the distinction before 
 drawn between these terms. Unfortunately the works on 
 rhythmic seem to have been much less numerous than those 
 on metric, and the remains of them which have come down 
 
GREEK RHYTHM AND METRE. 85 
 
 to us are exceedingly scanty. In these, nearly everything 
 of any value has come directly or indirectly from a single 
 author, from Aristoxenus of Tarentum. This eminent phi 
 losopher was a pupil of Aristotle, and after the death of the 
 master aspired to lead the school of his followers, but Theo- 
 phrastus of Lesbos was preferred before him for this station. 
 He wrote on many subjects, but especially on music and the 
 arts connected with it. Among his works of this class was 
 one entitled *RvQ^uca aTOi^ela, l Rhythmic Elements, in 
 which he drew out for the first time in scientific form the 
 principles of rhythm which were embodied in the music and 
 poetry of his countrymen. If this work were preserved in its 
 entireness, we should doubtless have a systematic and tole 
 rably adequate conception of the ancient rhythms. But we 
 have in fact only fragments one large fragment and a num 
 ber of small ones. The large fragment was found in a Vatican 
 manuscript containing works or parts of works on musical 
 subjects. It was first published by Morelli at Venice in 1785, 
 and recently by Westphal in his Fragmcnte und Lelirscitze der 
 Gricchischen Rhythmiker (Leipzig, 1861). As printed by the 
 latter, it amounts to about 270 lines : it formed the beginning 
 of the second book in the work of Aristoxenus. Of the 
 smaller fragments, several come from a little tract, r jrpo\a(jb- 
 ftavbfjieva e/9 Trjv pvO/JLitcrjv 7rtcrT?7yLt7/i/, by Michael Psellus, a 
 Byzantine writer of the tenth century ; and others from an 
 anonymous manuscript in the Royal Library at Paris. Next 
 in importance to these remains of Aristoxenus, but next at a 
 great interval, is a work of Aristides Quintilianus, who seems 
 to have been a thinker of the revived Pythagorean school, 
 and probably lived in the century before Constantine. His 
 work may be called an Encyclopaedia of the Musical Arts. In 
 the first of its three books he treats of Harmonic, Rhythmic, 
 and Metric ; in the second he discusses the influence of mu 
 sical art upon the soul ; in the third he sets forth the numeral 
 relations which subsist among the tones of the musical scale 
 the numbers which express the ratios of the vibrations by 
 which the tones are produced and then proceeds, after the 
 Pythagorean fashion, to show the cosmical significance of 
 these numbers. The rhythmic, it will be seen, forms only 
 
86 GREEK RHYTHM AND METRE. 
 
 part of one book, though the second book contains an impor 
 tant passage on the different ifios or character (pathetic, ani 
 mating, terrifying, etc.) of different rhythms. But even if we 
 include this passage, the rhythmic of Aristides, as printed in 
 Westphal s collection just referred to, contains less than four 
 hundred lines : it is curt and meagre, and altogether insuffi 
 cient to give any satisfactory conception of the subject. In 
 parts of this brief sketch, Aristides has relied on Aristoxenus, 
 or perhaps on a writer who drew his materials from Aristox 
 enus, so that what we find here represents the Aristoxenian 
 theory of rhythm ; but in other parts he evidently follows a 
 different, and far inferior, authority. This mixture of different 
 authorities and views adds greatly to the difficulties which 
 attend the interpretation of Aristides. Besides Aristoxenus 
 and Aristides, Westphal s collection contains a few pages of 
 fragments drawn from other sources, mostly anonymous. 
 The student of ancient rhythmic is not oppressed by the 
 extent of his authorities. It would be no extraordinary feat 
 for him to commit tD memory every line of the texts which 
 he has to work upon. It is only of late that the importance 
 of studying these texts has come to be recognized. It has 
 been the prevailing impression of scholars that, in order to 
 understand the rhythm of Greek poetry, it was enough for us 
 to examine the extant remains of Greek poetry, being aided 
 in the work by the old writers on metric, and guided at the 
 same time by our own sense of rhythm. But it is coming to 
 be understood that there is great uncertainty and hazard in 
 thus applying our own notions of rhythm to the poetry of the 
 Greeks. The sense of rhythm is indeed common to all men. 
 There are many forms and successions which all men would 
 accept as rhythmical, many others which all would reject as 
 unrhythmical : but there are some which are not so definitely 
 marked, about which different men might differ as to their 
 rhythmical character/ And among forms acknowledged to be 
 rhythmical, some men would prefer one, some another ; the 
 choice depending very much on fashion and education. Where 
 a particular verse admits of different rhythmic constructions, 
 we cannot assume it as certain that the one which seems pre 
 ferable to us would have seemed so to the Greeks. There 
 
GREEK RHYTHM AND METRE. 87 
 
 are many particulars in which we can never know the actual 
 rhythraus of Greek verses without an authentic statement 
 from the Greeks themselves. I need not insist further on this 
 point. It will become more apparent as we proceed. The 
 first scholar who pointed out the importance of carefully 
 studying the remains of Greek writers on rhythm was Bockh, 
 in his Essay on the Metres of Pindar. In this study, how 
 ever, he did not himself proceed to any great extent. 
 
 It is only about twelve years since two young scholars- in 
 Tubingen, August Rossbach and Rudolf Westphal, set them 
 selves in earnest to the work. A volume, entitled Griechische 
 Rhythmik) appeared in 1854, written by Rossbach, but pre 
 senting the results of their united study. It was succeeded in 
 1-856 by a volume bearing the names of both scholars, the 
 Griechische Mctrik, in which the doctrines of rhythmic de 
 veloped in the former work were applied to the treatment of 
 Greek metres. It appeared, however, that in the interval of 
 two years which separated these works the views of the writers 
 had undergone changes on various points, and some of con 
 siderable consequence a circumstance not to be wondered 
 at, when we consider the newness of the subject and the great 
 difficulties which attend it. These volumes attracted much 
 attention, and were generally regarded as containing new and 
 important truth though how much, few would undertake to 
 decide. There were not many critics whose studies had fitted 
 them to pass a comprehensive and independent judgment on 
 works of this kind. But it is proper to except one or two 
 articles by H. Weil of Besan^on, who is known for an elabo 
 rate book on Latin accent, published by him and Benloew. 
 The articles to which I refer appeared in Jahn s JahrbilcJicr, 
 and contain valuable contributions to a knowledge of Greek 
 rhythmic. In 1861, Westphal came out again with a volume 
 entitled Die Fragment e und die Lehrsatze der Griechischen 
 Rhythmikcr. It professes to be a supplement to the Greek 
 Rhythmic of A. Rossbach. It is dedicated by the author to 
 his former associate Rossbach, who seems to have withdrawn 
 from these studies. The dedication speaks of numerous 
 changes of opinion as having been the result of further research, 
 and apologizes for the appearance which the new book has of 
 
88 GREEK RHYTHM AND METRE. 
 
 being in no small degree a polemic against the old one. The 
 work opens with an Introduction, on the sources from which 
 a knowledge of Greek rhythmic is to be derived, the manu 
 scripts in which we find them, the books in which they have 
 been printed, and the like. Then follow the texts themselves, 
 occupying some fifty loosely-printed pages. But by far the 
 largest part of the book consists of the Commentary on the 
 Doctrines of the ancient Rhythmic Writers. This Com 
 mentary follows, not the order of the texts to be elucidated, 
 but the natural arrangement of the subject : beginning with 
 the general idea of rhythm ; proceeding thence to arsis and 
 thesis in general, thence to kinds of feet, thence to extent of 
 feet, and so on ; giving under each head the dicta of the 
 rhythmic writers, and discussing, comparing, and illustrating 
 their statements. Only a few weeks after Westphal s book 
 was given to the public, there appeared another on the same 
 subject, and on somewhat the same plan, by Julius Caesar, a 
 scholar of long-established reputation. Its name is Die 
 Grundzllgc der GriecJiiscJicn RJiytJnnik, im Ansclihiss an Arts- 
 tides Quintilianus, crldutcrt von Julius Ccesar. The Introduc 
 tion to this work contains a very thorough investigation of all 
 that is or can be known concerning Aristides, with a view 
 especially to determine the time when he lived and wrote. 
 Then comes the text of Aristides that part of it which re 
 lates to rhythmic with careful statements of various readings. 
 And then a long Commentary, in which particular points of 
 rhythmic are taken up, one after another, and the statements 
 of Aristides are discussed in comparison with those of other 
 authorities. In these discussions Caesar is often led to notice 
 and criticise the views expressed by Rossbach and Westphal 
 in their RJiythmik und Mctrik ; and his criticism, though in 
 tended to be respectful and courteous, is somewhat irritating 
 in its tone. He assumes that they have been so biassed by 
 their own preconceived notions of rhythm as to be unfitted for 
 a fair understanding and appreciation of the texts on which they 
 build. This tone is especially marked in the Preface and the 
 Appendix ; these were written after the appearance of West 
 phal s Fragincnte und Lclirsiitze, and the last is taken up 
 with a discussion of the various points in which they differ. 
 
GREEK RHYTHM AND METRE. 89 
 
 One might suppose from Caesar s tone that there was a funda 
 mental difference between his views and those of Westphal. 
 But upon looking more closely it would be found that on 
 most of the main points they are agreed, and that the matters 
 about which they disagree are generally of minor conse 
 quence. It is not strange perhaps that Westphal should have 
 felt himself somewhat aggrieved by Caesar s criticisms ; but 
 the way in which he shows his resentment is not altogether to 
 be approved. He brought out last year (1863) a long-prom 
 ised work, a part of the same series with the Rhythmik und 
 Metrik, viz. Hannonik und Mclopoie der Gricchcn. To this 
 he has prefixed a long preface of fifty pages, much of it taken 
 up with one or two points on which he and Caesar are at issue. 
 And it is somewhat curious, as a manifestation of feeling, that 
 he everywhere suppresses the name of Caesar while combat 
 ing his views, although he is thus obliged to use some rather 
 inconvenient circumlocutions. Just at the end, however, he 
 does bring out the name, as if he had been reserving it to that 
 place for a grand final explosion. To rebut this attack, Suse- 
 mihl, who seems to have been a pupil of Caesar s, appears in 
 Jahn s Jahrbucher for Dec. 1863, with an open letter to his 
 master, in which he says that Westphal has carried on his 
 polemic in such a way that Caesar s self-respect will hardly 
 allow him to reply, and he (Susemihl) therefore takes up the 
 cudgels for him. But if he is a partisan, he is a fair-minded 
 one ; he concedes so much to his opponent that Caesar finds 
 it necessary to append a lengthy note, protesting against the 
 admissions of his friend, and fighting against his champion 
 hardly less than against his enemy. 
 
 Before taking up the principles of rhythmic, it may be well 
 to notice some objections urged against this study, which 
 Westphal states and answers in the Introduction to his Frag- 
 mente nnd LcJirsatze. The first is, that the doctrine of the 
 ancient rhythmists refers not so much to the rhythm of poetry 
 as to that of music. Westphal s reply runs thus : " This error, 
 which has arisen from the relation of our modern poetry to 
 music, appears to have been shared even by Hermann, and to 
 this day many philologians seem not to be wholly free from it. 
 The relation of the poet to the musician in classic Greek was 
 
9 GREEK RHYTHM AND METRE. 
 
 wholly different from what it is with us. It is true that the 
 ancients also had an instrumental music separate from poetry ; 
 but while this in modern times has been coming more and 
 more to be the crown of musical art, it was confined in anti 
 quity to the kitharistic and aulctic nomes ; but the centre of 
 gravity lay in vocal music with instrumental accompaniment, 
 in melodized poetry. Was then, we ask, the rhythm which 
 the ancient poet gave to his compositions different from the 
 rhythm of the song ? In our time this is certainly the case. 
 Our dramas are either designed to be declaimed throughout 
 and this is true of all which make pretensions to any high 
 poetic merit or the drama takes the form of an opera, ii> 
 which the music predominates with such unlimited ascend 
 ency that the text, with rare exceptions, is insignificant as 
 poetry, and even the metrical form is indifferent, for the com 
 poser in general arranges his rhythms (bars in music) without 
 reference to the number of verse-feet, and in the religious 
 opera frequently makes use of an unmetrical prose text. Such, 
 too, is the procedure of a musician when he melodizes a lyric 
 poem which he finds ready to his hand, a poem written with 
 out reference to musical composition. Very different was the 
 case in classic Greece. Excepting the Epos and a few other 
 species, every poem was expressly intended, in whole or in 
 part, for musical performance. To write a lyric poem for 
 mere reading or declamation was, with few exceptions, an 
 unknown effort of poetic art, and every drama (as Aris 
 totle says) contains the /u-eXwSta as an essential element, as 
 fieyicTTov t]^vorfj.dTwv. Not only choral songs and monodies, 
 but also parts of the dialogue were sung, and even when the 
 iambic trimeter of tragedy was spoken, it was delivered in a 
 melodramatic way, i. c. with instrumental accompaniment. 
 To this we must add that poet and composer were united in 
 one person. In the great lyric and dramatic authors of 
 Greece we are wont to see poets merely, but in antiquity 
 they were no less esteemed as the Coryphaei of music. When 
 Aristoxenus, the great musical art-theorist, protests against 
 the overloading of music with affected ornament, a style 
 which had been introduced by Philoxenus and Timotheus, 
 and which threatened a general corruption of taste, he refers 
 
GREEK RHYTHM AND METRE. 91 
 
 to the representatives of the good classical style as models for 
 imitation, naming as such Pindar and Pratinas : Whoever 
 (he says) has in his youth earnestly and zealously studied 
 the fjie\7) and /cpou/^ara, the melodies and instrumentations of 
 these masters, will remain ever after secure from many aber 
 rations, even if he should apply himself to the complex and 
 ornate TTOIKL^TJ /AOVCTIKIJ of Philoxenus (Plut. Mus. 31). If 
 Aristoxenus would show that the noble simplicity of classical 
 music was conscious and intentional, and by no means founded 
 in a defective knowledge of the resources of art (ov 8t ciyvoiav 
 aXXa 8ia irpoalpeaiv), he refers to the compositions of /Eschylus 
 and Phrynichus, who were well acquainted with the chromatic 
 treatment of the keys, but never applied it in their tragedies 
 (Plut. Mus. 20). So, too, Sophocles was a composer : Aris 
 toxenus calls him the first Athenian who introduced the 
 Lydian mode in i$ia acr/jLaia, i. <?. in monodies and threnes, 
 after the fashion of the dithyrambic poets (Vit. Sophocl. fin.): 
 and of later tragic poets we know that they were active as 
 musical artists as of Agathon, who took up the chromatic 
 keys, which his predecessors had declined to use. The lyric 
 and dramatic poets of the classical time were therefore the 
 composers of their own yu-eX?;. Why now do these poets 
 apply to the metrical form of their choral songs and dramatic 
 monodies a degree of care so extraordinarily great ? Why do 
 they constantly appear in this field as original artists, never 
 once repeating a strophic form, whether used by themselves 
 or by any of their predecessors ? For reading or declamation 
 their ^e\ij were not intended, but for musical performance. 
 Why then we repeat the question have they taken so much 
 pains with the metrical form of the Xeft?? The answer can be 
 none but this : the rhythm given by the metres of the words 
 was the same which appeared in the musical delivery, the 
 same which the audience were to hear in the performance of 
 the poem. 
 
 " It is true then that the writers on rhythm have especially in 
 view the rhythm of the song, but this is identical with the 
 rhythm of the metre, as manifested by the words. In a time like 
 ours, when poetry and music are tw r o independent arts, it may 
 happen that a composer like Beethoven, in melodizing three 
 
92 GREEK RHYTHM AND METRE. 
 
 similar verses of five feet, should make three bars out of the 
 first, two out of the second, and again three out of the third 
 (we refer to the song Einsam wandelt dcin Frcund im 
 Fruhhngsgarteii). Beethoven takes a song which a poet be 
 fore him had composed in a traditional metre, simply for 
 reading or recitation, without any reference to melodizing, 
 and in the treatment of this metre he proceeds with perfect 
 freedom. Only in this respect does the composer attach him 
 self to the poet, that the strong part in every musical bar 
 coincides with an accented syllable in the poem, though it is 
 not true conversely that every accented syllable of the poem 
 appears in the melody as the strong part of a bar ; in refer 
 ence to the weak part of his bars the composer pays no atten 
 tion to the poet. So stands the case with modern rhythmic. 
 But in classical antiquity, where the poet himself was always 
 composer, where he worked out his artfully constructed 
 metres only with a view to musical performance, every metri 
 cal arsis is also an arsis in the melody, every metrical thesis a 
 thesis in the melody, and as many feet as belong to the verse, 
 so many bars belong to the musical period. We repeat, then : 
 The tradition of the writers on rhythm relates indeed to the 
 rhythm of the melody, but the rhythm of the melody is iden 
 tical with the rhythm of the text. Hence it is that the feet 
 and kinds of feet of which these writers speak take their 
 names from the feet of the metre (70^0? &a/cTv\i/c6v, la^/Si/cov, 
 iraiwviKov, rpo^alo^ 1X0709, etc.) ; and moreover, when the 
 rhythmic writers wish to illustrate the form of a rhythm or a 
 rhythmic series, they always take their examples from metric. 
 Never does it appear that any rhythmic form of which they 
 speak finds no application to poetry : even the pceon epibatns, 
 which had its principal application in mere instrumental mu 
 sic, seems, in the earlier time at least, to have been used also 
 as a foot in poetry." 
 
 I have made this long quotation from Westphal, because 
 the views contained in it, if not altogether novel, are very 
 justly and forcibly expressed, and are important to be recog 
 nized and understood. A second objection brought against 
 this study is that the system of the ancient rhythmists had a 
 theoretical rather than a practical character, that it consisted 
 
GREEK RHYTHM AND METRE. 93 
 
 more of speculations than of facts, that it was not so much a 
 description of actual usage as a series of ideal principles and 
 categories. " If this were so, the rhythm which Aristoxenus 
 talks about might not be identical with that of the ancient 
 poets : Aristoxenus would stand on much the same level with 
 Hermann and Apel." In reply to this objection, Westphal 
 traces the procedure of Aristoxenus from point to point, 
 and shows that it is purely empirical ; "he states facts, 
 one after another, and states them simply as facts ; he seeks 
 appropriate definitions for current expressions of musical 
 art ; and beyond this he aims only to bring the facts designa 
 ted by those expressions into an intelligible order, and to show 
 that they are justified by some principle r analogy." 
 
 But there remains a third objection ; that the writers on 
 rhythm lived and wrote when art was in its decline, and may 
 therefore be supposed to represent, not the true classic usage, 
 but that of degenerate times and inferior artists. In regard 
 to this, I quote again at some length from Westphal : 
 " Aristoxenus stands on the border of the classic period : his 
 father might have seen Socrates, Epaminondas, and other 
 men of classic Greece ; but he himself belongs to a later gen 
 eration. He resided in Corinth at the time when the younger 
 Dionysius lived there as an exile ; of professed musical ar 
 tists he was particularly acquainted only with Telestes, the 
 dithyrambic composer, whom he met in his wanderings 
 through Italy. Afterwards he came into connection with 
 Aristotle, and after his death aspired to be his successor in 
 the Lyceum. This in truth was no longer the time of classic 
 life : it was a period when the creative spirit in rhythmic had 
 died out, when men like Chseremon and Theodectes held the 
 first rank in tragedy. But amid this corruption of ancient 
 art, Aristoxenus took a very peculiar place. Trained in the 
 conservative school of the Pythagoreans, he had early im 
 bibed a predilection for the norms of ancient art, and this 
 partiality for the old showed itself through his whole life, in 
 an opposition to the tendencies of contemporary art. He 
 appears in his works as an admirer of the art-style represented 
 by Pratinas, Pindar, Simonides, Phrynichus, and /Eschylus ; 
 in this style he recognizes the proper r/0o? (art-character) 
 
94 GREEK RHYTHM AND METRE. 
 
 (Plut. Mus. 31 and 20); even Euripides and Sophocles are not 
 named by him ; while against the dithyrambic composers 
 Timotheus and Telestes, and their modern style, he com 
 mences an exasperated strife. The good old time that was 
 for him the time when art reached its greatest height, when 
 choral melody was in its bloom, when musical art had its 
 genuine r/9os and truly served for TratSet a : but now was the 
 time of cr/crjviier) povcritcij, of exaggerated and affected stage- 
 airs and concert- solos. TJicn the iroiKikia pvO/JLiKij had a 
 meaning, the artists were (f>i\6ppv0^oi : but now all rjOos was 
 lost in the rhythms of the /ceK\a<TjjLsvafjLe\r). His stand-point is 
 most clearly exhibited in the work called crv^^tKra CTV^TTOTIKCL 
 (Ath. 14. 632). In these colloquies, which he held with friends 
 and pupils on matters of musical art and afterwards gave to the 
 public, he begins by referring to the people of Posiclonia in 
 Italy, who among Tyrrhenian and Latin neighbors had grad 
 ually become barbarians, forgetting their Hellenic customs 
 and language, and even their very name ; but on one clay of 
 every year they held a feast in commemoration of the old 
 Greek time, now passed and gone, and broke up their assem 
 bly with tears and lamentations. So will we also do, says 
 Aristoxenus : while the theatre is sinking more and more 
 into barbarism, and musical art is suing only for the favor of 
 the multitude and hastening on to its ruin, in our own little 
 circle we will remember the ancient /JLOVCTIK/^ And so he con 
 versed with his scholars on the OrjXvvofjievr] fjLovo-iKy] of the con 
 temporary theatre (Themist. Or. 33, p. 364), and to the tasteless 
 compositions of Philoxenus and Timotheus opposed the 
 norms of classic art as represented by Pindar and Pratinas. 
 . . . From this oppositional stand-point, which Aristoxenus 
 holds toward the musical art of his own time, like the battle 
 waged by Aristophanes against the monodies of Euripides 
 and the new dithyramb -writers of Athens, it follows, of course, 
 that the rhythmical doctrines of Aristoxenus are drawn from 
 the norms which are fundamental to the classical rhythmic 
 of Pindar, Simonides, ./Eschylus ; they are derived by ab 
 straction from the compositions of those great masters. In a 
 word, the rhythmical doctrines which Aristoxenus presents 
 to us are the same that were followed by the classical poets 
 
GREEK RHYTHM AND METRE. 95 
 
 of Greece. Yet, on the other hand, it would not be correct 
 to assume that the doctrines of Aristoxenus apply only to 
 the times of^Eschylus and Pindar, and are inapplicable to the 
 rhythmic of later time.s, as that of Euripides. Widely as the 
 rhythmopceia of yEschylus differs from that of Euripides, and 
 the rhythmopceia of Pindar from that of Philoxcnus, in 
 particular forms and combinations, the fundamental principles 
 of rhythmic as regards kind of feet, extent of feet, division 
 of feet, composition of feet, change of feet, etc. are the same 
 for both periods of musical art ; they remained without 
 change from the time of Alcman and Stesichorus to that of 
 the Romans. Now it is just these fundamental principles 
 which we learn from the rhythmic writers ; into the detailed de 
 scription of art-forms in rhythmopceia (rhythmic composition) 
 they have not entered. And thus no well-informed scholar 
 would think of constructing a complete metric from the doc 
 trines of the rhythmists : for this purpose we must have re 
 course to the old writers on metric, and above all to the old 
 poets themselves ; but no system of metric can have a sure 
 foundation unless it bases itself on those fundamental doc 
 trines of the rhythmists. Unhappily, of these fundamental 
 doctrines we are far from knowing all, for only a very scanty 
 portion of the rhythmical literature of the ancients has been 
 preserved to As ; but what we have received is absolutely in 
 valuable, and sheds clear light upon the darkest points." 
 
 We proceed now to consider some elementary facts and 
 principles of rhythmic as set forth in these remains of 
 .ancient rhythmic. And we begin with arsis and thesis. As 
 rhythm consists in a regular succession of proportional 
 times, it is necessary that the successive times should be so 
 marked off and distinguished that their proportionality and 
 regularity shall be made clearly perceptible. This is accom 
 plished by a greater intensity of action falling upon particular 
 moments, and distinguishing them from the intermediate mo 
 ments of weaker action. It is by alternate intension and re- 
 mission of effort that rhythm is made obvious to our senses. 
 The portion of time thus marked off by an intension and a re 
 mission of effort is a rhythmic foot. It divides itself, of 
 course, into two parts, one of them being the time of inten- 
 
96 GREEK RHYTHM AND METRE. 
 
 sion, the other that of remission. To these two parts of the 
 foot were given the names of arsis and thesis. The names, as 
 we have already seen, connect themselves with the dance : the 
 6ecns was the setting down of the foot, and the apcns was the 
 raising of it up. They are appropriate also to the practice of 
 beating time, whether with the foot or with the hand : the 
 Oecris corresponded to the downward beat, the apcrt? to the 
 upward. But the downward movement was stronger than 
 the upward, the force of gravity being an addition to the for 
 mer, and a subtraction from the latter. Hence, when these 
 times were applied to musical and metrical feet, the name 
 thesis was used for the stronger part of the foot, that which 
 was distinguished by the intension of voice or instrument, and 
 the name arsis for the weaker part, that which showed a re 
 mission of voice or instrument. In dancing to the sound of 
 music, there was a Oecns of the dancer s foot on the accented 
 part, and an apais on the unaccented part, of the musical foot 
 or bar. In beating time to poetry, when sung or recited, 
 there was a Oka-is, or downward beat, on the accented part, an 
 dpcris, or upward beat, on the unaccented part, of the metri 
 cal foot. This is unquestionably the proper use of the terms 
 arsis and thesis : it is the only use among the extant Greek 
 writers on rhythmic and metric, down to a very late period. 
 But it is remarkable that the Latin writers on metric show a 
 very different use. They employ arsis, or clatio, for the first 
 part of a foot, and thesis, or positio, for the last part, without 
 reference to the rhythmic accent. In an iambic foot the 
 short syllable is arsis, the long thesis ; in a trochaic foot, the 
 long syllable is arsis, the short thesis. The terms thus used are 
 mere designations of place, and have no rhythmical significance. 
 The singular accordance of the Latin writers in this misuse 
 makes it probable that they have derived it from a common 
 source, which would have to be placed as early as the second ~ 
 century after Christ. This "source would seem to have been 
 some Greek author on metric : I say some Greek author, for 
 we find the same usage in late Byzantine writers, who 
 would hardly have taken it from a Latin source. Wherever 
 and whenever it arose, it was the result of ignorance and mis 
 apprehension. A writer who had heard or read that in the 
 
GREEK RHYTHM AND METRE. 97 
 
 common dramatic trimeter the first part of each foot was call 
 ed arsis and the last part thesis, may have inferred that this 
 was the case in all kinds of verse. Or he may have been 
 misled by the circumstance that the writers on rhythmic, 
 where they had occasion to speak of arsis and thesis together, 
 always coupled them in this order arsis and thesis. But what 
 ever the explanation, the mistake was a most unfortunate one. 
 For the consequence is that the statements which we find in 
 Latin sources as to arsis and thesis are almost wholly unin- 
 structive ; they do not show us where the rhythmic accent fell. 
 And even when they speak of the ictus or beat, the case is no 
 better; for these Latin metricians who, it must be remem 
 bered, were not practical musicians seem in their scanning 
 of poetry to have beat time in the same way, raising the hand 
 on the first part of the foot and lowering it on the second- 
 thus, in anna vir unique, raising the hand for ar- and lower 
 ing it for -ma vi-. But the confusion in regard to arsis and 
 thesis has become still greater in modern times. As the Latin 
 metricians misunderstood the use of these terms in Greek me 
 tric, so Bentley, the first modern scholar who really applied 
 himself to metrical studies, misunderstood the use of the 
 Latin metricians : he took arsis for the accented part 
 of the foot and thesis for the unaccented, thus exactly 
 reversing the original and proper use.. In this he has been 
 followed by all scholars from his time to ours. Even Ross- 
 bach, Westphal, and Caesar, while protesting against the pre 
 valent use of these terms, have felt themselves obliged to con 
 form to it. But Westphal, in his Fragment e und Lehrsatze, 
 has ventured to break loose from the bonds of custom, and 
 return to primitive usage. In this, it seems to me, he has 
 done well, and the example will doubtless be followed by 
 other writers. But as our scholars are familiar with the Bent- 
 leian use, of arsis for the accented part of the foot and thesis 
 for the unaccented, so that statements in accordance with this 
 use would be most readily understood, I shall adhere to it in 
 the sequel of these remarks. 
 
 The foot, then, consists of the arsis or accented part, and 
 the thesis or unaccented part. The arsis, however, does not 
 always precede the thesis, as in the bars of our modern music. 
 3 
 
98 GREEK RHYTHM AND METRE. 
 
 It would have been^impler and better, perhaps, if the foot had 
 been constituted like the bar, so as always to commence 
 with an accent : an unaccented syllable at the beginning of a 
 verse could then have been treated as an anacrusis, introduc 
 tory to the proper rhythm. But the ancients regarded the 
 proper rhythm, the verse foot, as commencing with the com 
 mencement of the verse, and hence constituted feet of thesis 
 followed by arsis, as in iambic and anapaestic verses, not less 
 than feet of arsis followed by thesis, as in trochaic and dac 
 tylic verses. The two components of the foot, the arsis and 
 thesis, may have different ratios to each other. Thus, they 
 may have a ratio of equality, as in the dactyl, the anapaest, 
 the spondee, in which arsis and thesis divide the foot into 
 equal halves ; or they may have a ratio of two to one a 
 diplasic ratio, as the ancients called it as in the trochee and 
 the iambics, where the long arsis has twice the length of the 
 short thesis. It is obvious that these two ratios correspond to 
 the two varieties of time in modern music : the equal ratio 
 answers to our common time, and the diplasic ratio answers to 
 our triple time. But ancient rhythm has another ratio, which 
 is not recognized in the theory and the notation of modern 
 music. Though much less frequent than either of the others, it 
 is still by no means rare. I refer to the hemiolic ratio, as 
 the ancients call it, in which arsis and thesis are as ij to I, or 
 as 3 to 2. Thus, in the cretic foot, an arsis consisting of a 
 long and short syllable is followed by a thesis consisting of a 
 long ; in the first paeon, an equivalent of the cretic, an arsis 
 consisting of a long and short is followed by a thesis consist 
 ing of two shorts. The hemiolic ratio appears in cretic and 
 paeonic, as the equal ratio does in dactyl, anapaest, and spon 
 dee, and the diplasic ratio in iambus and trochee. Westphal 
 asserts that in modern music the bar is sometimes divided 
 into five equal parts, as in the hemiolic ratio ; and he gives as 
 example of this rhythm the German popular song of Prince 
 Eugene, the music of which he writes with the fractional pre 
 fix f . Beside these three ratios of arsis and thesis the equal, 
 diplasic, and hemiolic Aristoxenus mentions two others : the 
 triplasic, in which the two parts of the foot are as 3 to I , and 
 the epitritic, in which they are as 3 to 4. He intimates, how- 
 
GREEK RHYTHM AND METRE. 99 
 
 ever, that they differ widely from the preceding, not being used 
 continuously in rhythmic composition. Their real character 
 has been ingeniously divined and convincingly explained by 
 Rossbach. I shall pass by the explanation for the present, 
 and return to it further on. All other ratios between arsis 
 and thesis, aside from those which have been mentioned, 
 Aristoxenus sets aside as unrhythmical : they are too com 
 plex to be appreciated by our senses and enjoyed by our 
 feelings. 
 
 We proceed now to describe the rhythmic feet, according 
 to the definitions of Aristoxenus. The shortest which he ad 
 mits consists of three short times. The pyrrhic, of two short 
 times, is unsuited for rhythmical purposes ; the arses and theses 
 would succeed each other with too much rapidity for distinct 
 impression and agreeable effect. This reason would apply to 
 a verse made up of pyrrhics in immediate succession. Whether 
 Aristoxenus would disallow the occasional use of a single 
 pyrrhic- as the ^Eolian poets appear to use it in the opening 
 of a logacedic verse is a point on which we have 110 express 
 testimony. Laying this out of account, we have : 
 
 First. The foot of 3 short times, trochee, iambus, or tri 
 brach. The ratio is diplasic, an arsis of 2 times, a thesis of 
 
 3 | N 
 i time. It corresponds to our | time : thus, 090. 
 
 8 Inn .i. 1 
 
 Second. The foot of 4 short times, dactyl, anapaest, spondee, 
 proceleusmatic. The ratio is equal, an arsis of 2 times, a 
 
 thesis of 2. It corresponds to our f time : thus, 0000. 
 
 Third. The foot of 5 short times, cretic, first paeon, fourth 
 paeon. The ratio is hemiolic, an arsis of 3 times, a thesis 
 
 of 2. It may be represented by -| time: thus, 40099 
 
 Fourth. The foot of 6 short times, ionic a majorc, ionic a 
 minore, etc. The ratio is diplasic, an arsis of 4 times, a 
 
 thesis of 2. It corresponds to our -J time : thus, t 
 
 But 6 short times admit of another rhythmic arrangement, 
 corresponding to our -| time : thus, 000009 . To this -| 
 we give the name of compound time, as it is made up of two 
 
100 GREEK RHYTHM AND METRE. 
 
 |-,.one of which has a stronger accent than the other. This 
 stronger accent dominates the whole bat, and makes it a 
 unit, though a compound unit. In like manner the cor 
 responding Greek foot is called by Aristoxenus a compound 
 foot ; it is made up of two trochees, iambi, or tribrachs : i. e. 
 it is a trochaic or iambic dipody. The ratio is equal, with 
 one trochee or iambus for arsis, and one for thesis. Observe 
 that the ratio of the whole compound foot is equal, though 
 each of the two feet from which it is compounded has a di 
 plasic ratio. 
 
 A foot of 7 short times would be unrhythmical, or at least 
 unfitted for continuous composition. Dividing the 7 times 
 between arsis and thesis, we should have 6 to I, or 5 to 2, or 
 4 to 3, none of them ratios which Aristoxenus recognizes as 
 fitted for continuous rhythmic composition. -We come, then, 
 
 Fifth. To the foot of 8 short times. This corresponds to 
 our -| time, a variety of compound time made up of two J- : 
 
 thus, 00000000. Here, then, we have another com 
 pound foot, a dactylic or anapaestic dipody. The ratio is 
 equal, with one dactyl or anapaest as arsis, and one as thesis ; 
 and here the simple feet themselves have the same ratio. 
 
 Sixth, The foot of 9 short times. This corresponds to our 
 f time, a variety of compound time made up of three f : 
 
 9 | M M N 
 thus, 000000000. It is a compound foot, a trochaic or 
 
 iambic tripody ; the ratio is diplasic, with two trochees or 
 iambi for arsis and one for thesis. And the single feet them- 
 . selves of which the compound foot is made up have a similar 
 diplasic ratio. 
 
 Seventh. The foot of 10 short times. It may be represented 
 
 with our notation as ^ time : thus, 000000000 0. It 
 
 is a compound foot, a cretic or pseonic dipody ; the ratio is 
 equal, with one paeon as arsis and one as thesis ; while the 
 single feet themselves have the hemiolic ratio. 
 
 A foot of 1 1 times would be unrhythmical, for the same 
 reason as one of 7 times. You cannot divide the times 
 between arsis and thesis so as to have a ratio that is rhyth 
 mical. 
 
GREEK RHYTHM AND METRE. IOI 
 
 Eighth. The foot of 12 short times. This may be repre- 
 
 13 I M N i (s ! 
 sented with our notation by - 1 /- : thus, pp 
 
 This is the trochaic or iambic dimeter, a compound foot with 
 equal ratio, one trochaic or iambic dipody for thesis, and one 
 for arsis. But the 12 shorts may also take a form which we 
 
 should write as : thus, * & 0900090 * 9. This is an 
 
 ionic dipody, with equal ratio, one ionic foot for arsis and 
 one for thesis. But yet again, the foot might be written 
 
 9 p p. This is a dactylic or anapaestic 
 
 tripody, with diplasic ratio, two dactyls or anapaests for arsis, 
 and one for thesis. 
 
 A foot of 13 short times is inadmissible, as not allowing 
 any rhythmical ratio between arsis and thesis. One of 14 
 short times might be divided in an equal ratio, as a com 
 pound foot with 7 times for arsis and 7 for thesis ; but the 
 simple feet of 7 times each would themselves be unrhythmical, 
 and the compound foot must therefore be unrhythmical also. 
 We have, then, 
 
 Ninth. The foot of 15 short times. This may be written 
 
 13 I is I I M i N I 
 thus : p p.p p p pppppppp . It is a cretic or paeonic 
 
 tripody, a compound foot with diplasic ratio, two cretics or 
 paeons for arsis, and one for thesis. But again we may write 
 
 15 I h I M M M h . . i . , . 
 
 p*p*ppppfpppp0p . I his is a trochaic or iambic 
 
 pentapody with hemiolic ratio, three trochees or iambi for 
 arsis and two for thesis. 
 
 Tenth. The foot of 16 short times. This may be written 
 
 8 | | | | | | I I 
 
 ppppppppppppptpp. It is a dactylic or anapaestic 
 
 tetrapody, with equal ratio, having a dactylic or anapaestic 
 dipody for arsis and another for thesis. According to Aris- 
 toxenus, it is the longest foot with equal ratio, the longest 
 foot in which arsis and thesis are equal to each other. 
 
 A foot of 17 or of 19 times would allow no rhythmical ratio 
 between arsis and thesis. But we have 
 
 Eleventh. The foot of 18 short times. We may write it 
 
102 GREEK RHYTHM AND METRE-. 
 
 18 I h | hi hi hi hi h 
 
 000000000000000000. It is a trochaic or iam 
 bic trimeter, with diplasic ratio, having two trochaic or iambic 
 dipodies for arsis, and one for thesis. We cannot divide it in 
 equal ratio, making it consist of two tripodies, one for arsis 
 and one for thesis. For this would be contrary to the rule of 
 Aristoxenus just mentioned. There is, however, another 
 combination possible, which we may write after this fashion : 
 
 000000000000. It is an ionic tripody, 
 
 03000 
 
 with diplasic ratio, two ionic feet for arsis and one for thesis. 
 According to Aristoxenus, the foot of 18 times is the longest 
 foot with diplasic ratio, the longest in which arsis and thesis 
 are as 2 to I. But the hemiolic ratio may extend, according 
 to his express assertion, to the length of 25 short times. This 
 gives us two additional feet, viz. : 
 
 Twelfth. The foot of 20 short times. We may write it 
 
 -^00000000000000000000. It is a dactylic or 
 
 anapaestic pentapody, in hemiolic ratio, with three dactyls or 
 anapaests for arsis, and two for thesis. 
 
 Thirteenth. The foot of 25 short times. We may write it 
 
 35 I N \ ! hi | hi I N ; ! hi 
 
 0000000000000000000000000. It is a cretic 
 
 or paeonic pentapody, in hemiolic ratio, with three cretics or 
 paeons for arsis, and two for thesis. 
 
 Of the feet just described, by far the greater part are com 
 pound. The foot of 3 times is simple, so that of 4 times, 
 that of 5 times, and the first of 6 times. But the second of 
 6 times and all that follow it are compound, and are made up 
 of two or more simple feet, each having its proper accent, 
 though all under the power of one dominant and unify 
 ing accent. These compound feet are precisely analogous in 
 rhythm to the varieties of compound time in our music, to 
 i> f > >b I and *-/-. The only differences are, I. that the an 
 cients allowed ratios such as -|, -y~, ^-, etc., which are not defi 
 nitely recognized in modern theory, and are not found, at least 
 with such notation, in our written music ; and, 2. that the an 
 cients gave to these compound feet a much greater extension 
 than they have in modern praxis an extension reaching in the 
 
GREEK RHYTHM AND METRE. 103 
 
 equal ratio to |-, in the diplasic to J-y 8 -, and in the hemiolic to the 
 seemingly enormous extent of 5 / 1 . This great extent, however 
 strange it might seem to us, in the musical bar, has evidently 
 nothing impossible ; there would be nothing impossible in a 
 .bar of 16, 18, 20, or even 25 quavers under the dominion of one 
 principal accent. And w r e are assured by the express testimony 
 of Aristoxenus, a most competent witness, that ancient rhythm 
 made use of feet in which 16, 18, 20, 25 short times were thus 
 united under the dominion of one principal accent. 
 
 I ought to say that the foot of 8 short times w ? as not always 
 a compound foot. It was sometimes a spondee the <j7ro^Seto? 
 {jiei&v, or greater spondee, which I have already referred to 
 in which each long was equivalent to Tour shorts. In like 
 manner there was a foot of three long syllables, in which each 
 long had this same dimension of four shorts : this foot, when 
 it had trochaic rhythm, was called trocJumis scmautus, when it 
 had iambic rhythm, it -was called orthius. The trochceus 
 scmantus and the orthius were feet of 12 short times, but 
 still they were regarded as simple feet. In this connection we 
 must notice also the pceou cpibatus, or pxon of five long syl 
 lables, with hemiolic ratio, three longs for arsis and two for 
 thesis. The/<^;/ cpibatus was a foot of 10 short times, but it 
 was a simple foot. 
 
 As regards the beating of the compound feet, we are told 
 that, if the ratio was equal, the compound foot received two 
 beats, one of which w r as doubtless given to the arsis and one 
 to the thesis. If the ratio was diplasic, the compound foot 
 received three beats, two of which were doubtless given to 
 the arsis and one to the thesis. If the ratio was hemiolic, we 
 should expect five beats for the compound foot, three for 
 arsis and two for thesis ; but, singularly enough, we have it 
 on express and undeniable testimony that these hemiolic feet 
 received four beats instead of five. Simple feet of six short 
 times or less received only two beats, whatever the ratio, 
 whether equal, diplasic, or hemiolic ; the single trochee, dactyl, 
 or cretic had two beats, one for the arsis and one for the thesis. 
 
 The doctrine of the extended or compound feet is the most 
 valuable piece of instruction which we gain from the remnants 
 of the old Greek rhythmic. Let us look at some of the con- 
 
1 04 GREEK RHYTHM AND ME TRE. 
 
 elusions which either flow from it directly, or may be in 
 ferred from it when combined with other statements and 
 notices. We learn that the common iambic trimeter of the 
 dramatic dialogue was a single foot, that it had one accent 
 which was sensibly stronger than any other in it and made it a 
 rhythmic unity. This is a fact which could never have been 
 guessed at from the descriptions of the writers on metric. 
 They represent the verse as consisting of three dipodies, but 
 they say nothing which would suggest the idea that these 
 three dipodies are of unequal weight, or connected as princi 
 pal and subordinate. But now we learn that one was stronger 
 than either of the others. It is probable that the first dipody 
 had the strongest accent, the second a weaker accent, and 
 the third the weakest of all. But it is a question what part 
 of the dipody received the accent, whether it fell upon the 
 first or second iambus of the dipody. It was assumed by 
 Bentley, without any attempt at proof, that it was the first 
 iambus which was distinguished by the accent ; and the as 
 sumption has been tacitly accepted by all succeeding writers. 
 But Westphal denies the correctness of the assumption. He 
 maintains, not only that it is unproved, but that it is in con 
 flict with the express teaching of the ancient authorities. He 
 concludes with undoubting confidence that the accent fell on 
 the second iambus of the dipody : thus, co re/cva KO&JJLOV rov 
 Trakai vea Tpo&rf. But Weil, reviewing his argument, shows 
 that it rests on statements of Latin metricians, and that the 
 misuse of these writers in reference to arsis and thesis (which 
 we have before spoken of) deprives their statements of all 
 authority. At the same time, he points out an expression of 
 Aristides which implies pretty clearly (though nobody had 
 noticed it before) that the iambic dipody was accented on the 
 first foot. We are allowed, therefore, to go back to the 
 rendering of the trimeter which since Bentley has been the 
 usual one, and which is certainly far more consonant with our 
 modern notions of rhythm, and to read to reKva KdS/^ov rov 
 7rd\ai vea rpo(>/]. 
 
 But if the six feet of the dramatic trimeter form but one 
 compound foot, the same cannot be true as to the six feet of 
 the epic hexameter. For the six dactyls or spondees mak 
 
GREEK RHYTHM AND METRE. 105 
 
 24 short times which would have to be divided either in 
 equal ratio, as 12 and 12, or in diplasic, as 16 and 8. But 
 the longest foot with equal ratio has only 16 short times ; 
 the longest with diplasic ratio only 18. The epic hexameter, 
 then, is not under the dominion of one superior accent : it 
 must consist of at least two parts, the principal accents of 
 which are equal to each other. It appears certain, in fact, 
 that it was made up of two compound feet, of 12 short times 
 each, with diplasic ratio, having two dactyls for arsis and one 
 for thesis. Which of the three dactyls in each division of the 
 verse had the principal accent it is not so easy to determine. 
 Westphal finds reason to believe that it was the third : he 
 
 would give the opening line of the yEneid thus : Anna 
 
 virumqite cano Trojce qui primus ab oris. But the conclusion 
 seems to be far from certain. 
 
 The dactylic tetrameter is doubtless a single foot of 16 
 short times, the longest which the equal ratio admits of; and 
 the dactylic octameter consists of two such feet. So too the 
 anapaestic dimeter is a single foot of 16 times, and the 
 anapaestic tetrameter consists of two such feet. In like man 
 ner the iambic and trochaic tetrameters consist of two feet of 
 12 times each, with equal ratio. But what shall we say of the 
 trochaic tetrameter catalectic, of such a verse as TroXXa /nev 
 <yap ere 6a\d(rar}s, TroXXa 8 e /e ^epaov KCLKVL ? If the first half of 
 the line has its 12 short times, the second or catalectic part 
 would seem to have but II : but Aristoxenus, as we have 
 seen, rejects the foot of II shorts as being unrhythmical. 
 
 The true answer doubtless is that the pause at the end of 
 the line makes up for the deficient syllable, and supplies the 
 one time which is wanting. The iambic tetrameter is also 
 often catalectic, as in o> (HePicJoTarov Kpeas cro^&eo? ye Trpovvoi/crco 
 or a) TTCLGIV dvOpMTTOLs (pavels /jbeytcTTov ox^eX^/m. Here the 
 second compound foot, fiejiorrov Go^eX^a, or cro^eo? <ye 
 r jrpovvoY](Tw ) appears to consist of n times, instead of 12. 
 Shall we adopt here the same explanation as before, and as 
 sume that the deficiency was made up by a pause at the end 
 of the verse ? The more probable explanation for this case 
 is, that the last iambus has lost, not its second syllable, but 
 its first ; that the short syllable of the last iambus is sup- 
 
106 GREEK RHYTHM AND METRE. 
 
 pressed, and its time made up by prolonging the long sylla 
 ble of the preceding iambus. In the line just quoted, the 77 
 of voT) would be equal to three short times, and vorjcra), having 
 six short times, would be equivalent to an iambic dipody. 
 We can now understand, in connection with this example, 
 what Aristoxenus means when he speaks of a foot with tripla- 
 sic ratio i. e. a foot in which arsis and thesis are as 3 to I : 
 such feet, he says, are used occasionally, though not in con 
 tinuous composition. The foot vorj, just before the close of 
 the catalectic verse, answers perfectly to this description. 
 Its thesis is of one short time, its arsis of three. Its peculiar 
 ratio arises from a special cause, its arsis being prolonged so 
 as to occupy the time of a short thesis suppressed in the next 
 foot. It has therefore an occasional character, differing in 
 this respect from the ordinary iambus with diplasic ratio. 
 
 It is well known that in the iambic trimeter, and in other 
 iambic and trochaic verses, feet of 4 short times, spondees, 
 dactyls, anapaests, are to some extent intermixed with the 
 proper three-timed feet of these rhythms. How docs the an 
 cient rhythmic regard these apparently unrhythmical inter 
 mixtures ? It partly acknowledges their unrhythmical char 
 acter. Aristoxenus says there are three varieties of times 
 the eppvOfjioi or rhythmical, which have distinct and obvious 
 ratios to each other; the appvO^oi, or unrhythmical, which 
 have no intelligible or appreciable ratios ; and an intermediate 
 class, the /au^/xoetSe??, or rhythm- like, the ratios of which are 
 not indeed unintelligible, like the last, but are also not dis 
 tinct and obvious, like the first. There is no doubt that by 
 the last term, the pv0/j.oei,8ei$, he meant to designate such 
 times as belong to spondees in iambic or trochaic verse. A 
 term more frequently applied to them is a\oyoi, irrational, 
 without ratios, i. c. without integral ratios. The ordinary long 
 syllable is equal in time to two shorts : the ratio here is 2. A 
 moment ago, we saw in the penult of a catalectic tetrameter 
 a syllable which contained three short times : the ratio here is 
 3. We have seen that there were certain spondees in which 
 each long syllable was equivalent to 4 shorts : the ratio here 
 is 4. These longs of two, three, or four times are all rational : 
 their ratios to the short syllable taken as the unit of measure- 
 
GREEK RHYTHM AND METRE. 107 
 
 ment are all expressed by integers. But if we should rind that, 
 in a particular foot, a certain long syllable had the time of i-J- 
 shorts, such a long syllable would be irrational ; the number 
 which expresses its ratio to the unit of measurement is not an 
 integer, but a fraction. Now this is the case, according to 
 explicit testimony, with one long syllable in the spondees of 
 iambic and trochaic verses. The long arsis of those spondees 
 has the usual length, that of two shorts, but the long thesis, 
 we are told, was intermediate between one and two shorts ; 
 in other words, it bore to the unit of measurement the 
 fractional ratio of i^ to I. It was an irrational long; and 
 the foot to which it belonged was irrational also, the whole 
 length of the foot being expressed by a fractional designation, 
 viz. 3-J short times. The arsis and thesis of this irrational 
 foot bear to each other the ratio of 4 to 3. And there can 
 be no question that Aristoxenus refers to this foot, when he 
 speaks of the epitritic ratio (4 to 3) in the same terms that he 
 uses of the triplasic, as a ratio that does indeed occur, but 
 not in foot after foot through a continuous rhythmic series. 
 These irrational feet of the iambic and trochaic verses form 
 one of the most remarkable features of ancient rhythmic. It 
 is something which, to the same extent at least, does not 
 occur in modern art. It is analogous to a ritardando move 
 ment, by which certain barsor rather, certain parts of cer 
 tain bars receive a length greater than the ruling tempo of 
 the passage to which they belong. But the definiteness in 
 the amount of retardation and in the places of its occurrence, 
 and particularly the great frequency of its use, often three 
 times in one trimeter verse, are circumstances which have 
 no parallel in our music. 
 
 In regard to the feet of which we have been speaking, 
 their irrational character rests upon positive testimony. That 
 there were other cases of irrationality is probable enough 
 in itself, but they can only be determined by processes of 
 inference which are more or less uncertain. In the clactylo- 
 epitritic verses of Pindar, where dactyls are followed by 
 epitrites, each epitrite consisting of a trochee and spondee, 
 Westphal considers the trochees as equal in time to the 
 dactyls and spondees with which they are associated. In 
 
IOS GREEK RHYTHM AND METRE. 
 
 that case, they would contain irrational times ; the long of 
 the trochee would be |- and its short f of the short time which 
 serves as the unit of measurement ; while together they would 
 make l -f or 4 short times, which is the normal length of the 
 dactyls and spondees. But on the other hand, in logacedic 
 verses, which show dactyls mixed with trochees, Westphal 
 regards the trochees as having rational times, and the dac 
 tyls irrational. To the long of the logacedic dactyl he gives 
 the length off, to its first short f, to its last short the ra 
 tional length I : the dactyl consists then of -*-, f , I , making in 
 all 3 times, which is the normal length of the trochees. To 
 this measurement of the logacedic dactyl Caesar strenuously 
 objects : he makes the logacedic dactyl equal to a trochee, 
 but divides it into i.V, f , f- the long syllable i-J- instead of 
 two, and each short f instead of I. This difference of opinion 
 has been the theme of some pretty acrimonious discussion ; 
 but I cannot enter here into the merits of the quarrel. 
 
 I have already pointed out a long syllable of three times as 
 occurring in the penult of catalectic trimeters. I may now 
 say that the ancient rhythmic recognizes also a long syllable 
 of four times (this, too, we have had occasion to notice) ; and 
 even a long syllable of five times, equal to a whole cretic or 
 paeonic foot. It is often matter of doubt whether we are to 
 recognize such a lengthened long, or whether \ve are to as 
 sume a pause of one, two, three, or four times. Such vacant 
 times, Kevol %p6voL, are fully recognized in the theory of an 
 cient rhythmic. Thus, in the so-called elegiac pentameter, 
 aicrxyvT] Se </xXot? r)p,erepois eyevov, where the one syllable Xot? 
 takes the place of a dactyl, we might question whether A,o? 
 was actually prolonged to a length of 4 times, or whether it 
 had its usual length, but was followed by a pause of two 
 short times. In this case we have ancient testimony show 
 ing that the syllable was not prolonged, but that a pause was 
 made after it. But if times are wanting in the middle of a 
 word, as often happens, we may safely assume that they were 
 made up, not by a pause, which could hardly be tolerated in 
 such a place, but by the prolongation of a long syllable to 
 three, four, or five times. The recognition of deficient times, 
 which were made up in one way or the other, by a pause or 
 
GREEK RHYTHM AND METRE. 1 09 
 
 by the prolongation of a long syllabic, has put a new face on 
 ancient metres. It is perhaps the most important peculiarity 
 in the work of Rossbach and Westphal on Greek metric. In 
 the little metric at the end of my Greek grammar I have 
 adopted it from them, with the name of syncope, which they 
 had given it. Its effect in giving simplicity and unity to 
 many things which before appeared unconnected and un 
 meaning may be shown by a brief illustration. In looking 
 over the index of metres to President Woolsey s Electra, I 
 find five different names given to what I should regard as 
 syncopated forms of the iambic trimeter : I. Antispast and 
 iambic penthemimeris ; 2. Two iambic penthemimeres ; 3. 
 Iambic penthemimeris and iambic tripody ; 4. Iambic dime 
 ter hypercatalectic ; 5. Iambic dipody and ithyphallic. By 
 looking over the indexes to other plays, I have no doubt 
 that I could have increased this number. It is evident that 
 these names, beside their number, fail most of them to sug 
 gest any clear idea of a rhythmical unity. But if in all these 
 various and strange-looking lines we can truly recognize vari 
 ous forms of the iambic trimeter, we have certainly made a 
 great and most satisfactory advance in our understanding of 
 their rhythmus. 
 
VI. 
 
 ON THE NATURE AND THEORY OF THE GREEK 
 
 ACCENT. 
 
 1869. 
 
 EVERY Greek word of two or more syllables had one 
 syllable which was sounded on a higher key than the 
 rest of the word : thus, \v in \e\vKoitu, ice in \e\vieevai, /co? in 
 XeTuwc 9. Eor a long time, the Greeks in writing their lan 
 guage made no attempt to distinguish the syllable which was 
 thus sounded on a higher key : they aimed to represent the 
 substance of their sounds, the different articulations, but not 
 their relative pitch. It was not until the development of 
 grammatical study, in the Alexandrian period, that the gram 
 marian Aristophanes of Byzantium, about 200 years before 
 our era, invented a sign for this purpose. Over the vowel 
 which was sounded on a higher key he placed a wedge-like 
 mark, sloping downward to the left, which was called ^ o^ela 
 7rpo$M$ia, sharp accent, acute accent. But it often hap 
 pened, in the utterance of a long vowel or diphthong, that 
 the higher key with which the word began was not maintained 
 to the end ; that, after pronouncing the first part on a higher 
 key, the voice dropped down to a lower, and on this pro 
 nounced the last part of the long sound. For such cases 
 Aristophanes introduced a compound sign ; representing the 
 higher key as before, he added, to represent the lower key, a 
 similar mark, but sloping downward to the right. The roof 
 formed by the joining of the two marks was rounded ofT in 
 writing, and the whole sign was called 77 Trepio-Tray/uievT) Trpo^wBla, 
 twisted accent/ circumflex accent. The ordinaty lower 
 key was not generally represented by any distinctive mark : 
 if the vowel of a syllable had no mark of higher pitch written 
 above it, this was a sufficient indication of its lower pitch. 
 And, indeed, there was nothing in this lower pitch that called 
 for designation. The essential fact to be recognized and 
 
GREEK A CCENT. 1 1 1 
 
 made evident in the writing was not that some syllables were 
 lower than the rest : it was that some were higher than the 
 rest; or rather, that one syllable in each word was made con 
 spicuous and important above all others by the higher key on 
 which it was sounded. Yet there were two cases in which 
 the lower pitch was represented in writing. One of these 
 has just been noticed when the lower pitch followed the 
 higher in the same syllable, in the same long vowel-sound. 
 The circumflex accent used for such a syllable consisted of 
 two marks, a first representing the higher, and a second 
 representing the lower pitch. The other case relates to 
 oxytone words, where the higher pitch comes at the end of 
 the word and belongs to the last syllable : ayaOos, arpar^yo^. 
 If such a word is followed by other words in immediate gram 
 matical connection, the higher pitch of its last syllable changes 
 to a lower one, as in dyaOos o-Tparrjybs eyevero. Here now, 
 on the last syllable of an oxytone word, when in the connec 
 tion of discourse its higher pitch changes to a lo\ver, the 
 lower pitch is represented in writing, and represented in the 
 same way as in the latter part of the circumflex accent : that 
 is, by a mark sloping downward to the right, and called 77 
 {Bapela Trpo^wSia, heavy accent, grave accent. Aside, 
 then, from these two cases, the ordinary lower pitch is always 
 left without designation. 
 
 In this description of the Greek accent, it has been taken 
 for granted that there was an actual regular difference of pitch 
 between different syllables of a word, and that the proper use 
 of the written accent was to represent that difference. The 
 correctness of this assumption is implied in the very names 
 of the accents. The words o|fu9, sharp, and ftapvs, heavy, 
 are the ordinary words used in Greek music for what we in 
 our music call high and low. They are not used to de 
 note difference in stress, or strength of utterance. We might 
 find it natural that ofu?, sharp, should be applied to a sylla 
 ble which was pronounced with marked stress ; but it would 
 be strange if {3apv$, heavy, was used of syllables pronounced 
 without stress, the weaker or lighter syllables of a word. 
 The term TrposwSla itself, as well as the Latin acccntus, which 
 is used to translate it, comes from a root which means to 
 
1 1 2 GREEK A CCENT. 
 
 sing ; and in O^ltiriing the name, the ancient grammarians, 
 both Greek and Latin, tell us that it is a singing of the syllable. 
 There is a remarkable passage in a work of Dionysius of 
 Halicarnassus (de Comp. Verb., 11), in which he compares 
 the melody of speech with that of song. The melody of 
 spoken language, he tells us, is measured by a single interval, 
 the so-called fifth, or, as he explains it immediately after, the 
 interval of three tones and a semitone. He says that when 
 the voice rises to the acute, it does not go higher than this 
 interval ; and when it falls off again to the grave, it does not 
 sink lower than this interval. He adds that the two intona 
 tions, the acute and the grave, may be combined in the same 
 syllable, and that such syllables are said to be circumflexed. 
 It is perfectly evident that he is here speaking of the accent, 
 that he describes it as a difference of pitch, and that he makes 
 this difference about the same as the musical interval of a fifth, 
 that is (as he himself says), three tones and a semitone. All 
 this is pretty clear ; but it is made still clearer by the con 
 trasted description of music, which comes directly after it. 
 Music, whether vocal or instrumental, uses a number of inter 
 vals, and does not confine itself to the fifth alone (these are 
 his own words); but employs, for purposes of melody, 
 first the octave (this being first named, as the most important 
 interval), and the fifth, and the fourth, and the full tone, and 
 the semitone, and, as some think, even the quarter-tone (the 
 chromatic diesis), so as to be distinctly perceived. Further, 
 he says, music claims the right to subordinate the words to 
 the tune, instead of having the tune subordinate to the words ; 
 by which he means that it is the right of the musician to sing 
 the words to any tune he pleases, without reference to the 
 natural tune (if we may call it so) which by the accent they 
 have in spoken utterance. That this is his meaning becomes 
 abundantly evident from a subjoined illustration. He quotes 
 two or three lines from a chorus of Euripides, and points out 
 at some length how the natural tune or accent of the words 
 was wholly disregarded in the music to which they were sung. 
 Thus, the first word, criya, is sung upon a monotone, both 
 syllables on the same pitch, though the spoken word had two 
 tones, the acute and the grave, and indeed both of them com- 
 
GREEK ACC. 
 
 bined in the circumflex accent of its 
 word, apflv^rjs, which had the acute on the second syllable, 
 the music gave both second and third syllables the same pitch, 
 though in a spoken word the higher pitch of the acute accent 
 was never maintained through two successive syllables. In 
 TiOere, the natural tune of the word, a high sound followed 
 by two low ones, was completely reversed by the music, the 
 first syllable being sung to a lower note and the last two syl 
 lables to a higher note. In /cruTretre, the circumflex of the 
 second syllable, with its combination of acute and grave, wa? 
 lost in the music, that syllable being sung upon a single note 
 And in aTTOTrpofldre, a word of five syllables with acute accent 
 on the middle one, the higher intonation, which belonged to 
 the Trpo, was in the music transferred to the (3d. These are 
 the illustrations and explanations given by Dionysius himself, 
 whose authority on such a subject must be very high from his 
 intelligence and learning, as also from his date, in the first 
 century before our era. This passage alone, if there were 
 nothing else to confirm it, would leave no doubt as to the 
 melodic character of the Greek accent. 
 
 But is it not possible that this elevation of pitch, which 
 characterized the accented syllable, was accompanied by an 
 increased effort of the vocal organs, by a greater stress of 
 pronunciation, such as marks the accented syllable in our own 
 language and in other languages of modern Europe ? It is 
 possible, certainly ; but there is scarcely any evidence to 
 prove it true. In all that the ancient grammarians and other 
 ancient writers have left us on the Greek accent and the 
 aggregate is far from inconsiderable there seems to be no 
 statement or expression which implies that the accented sylla 
 ble was pronounced with more force than the rest of the 
 word. No such implication can be found in the remarkable 
 passage just referred to, where Dionysius speaks of the mel 
 ody of spoken language. We might perhaps expect to find 
 it rather in a following section, where his subject is the rhythm 
 of spoken language. But nothing of the kind appears there. 
 In speaking of rhythm, he refers to quantity of syllables, to 
 the succession of longs and shorts ; and he remarks that- com 
 mon speech, the utterance of prose, does no violence to these, 
 
1 1 4 GREEK A CCENT. 
 
 but keeps the syllables long and short according to their 
 received or natural quantity. But music, which disregards 
 the natural tune of words, disregards likewise their natural 
 rhythm ; it changes (so he says) the length of the syllables, 
 increasing or diminishing their quantity, and occasionally even 
 reversing the natural proportions ; for (he adds), instead of 
 making its own musical times subordinate to the natural 
 quantity of the syllables, it makes the quantity of the sylla 
 bles subordinate to the times of the music. In all this, there 
 is no hint that any one syllable of a word was regularly dis 
 tinguished from the others by its more forcible enunciation. 
 Gottling, who believes that the accented syllable was actually 
 pronounced with greater stress, can only refer in support of 
 his belief to the modern Phavorinus. The silence of the an 
 cient authorities may not prove that there was absolutely no 
 difference in stress between accented and unaccented syllables ; 
 but it certainly warrants the conclusion that the difference, if 
 there was any, cannot have been great or striking : it must 
 have been far slighter than in English or in modern Greek. 
 The grand fact about the accented syllabic, to the mind of the 
 ancients, was its higher pitch ; its greater stress, if it had any, 
 was either not noticed by them, or was felt to be compara 
 tively unimportant. 
 
 The same conclusion that the stress of voice on the ac 
 cented syllable was little, if at all, greater than on other syllables 
 may be supported by probabilities resting on other grounds. 
 It is the natural effect of a decided stress- accerft to weaken- 
 the following syllables of the word, and especially the one 
 which immediately follows the accent, so that the vowel of 
 that syllable is apt to be shortened or to be omitted alto 
 gether. In our own language, this tendency may be seen in 
 the short c of mystery as compared with the long sound in 
 mysterious, and in the suppressed c of every, wond(e)rous. 
 In Greek such changes are confined to a few words, as TLTTTC 
 in Homer for rl Trore, r]\6ov for earlier i}\v6ov. They are 
 perhaps hardly more numerous than the cases where an ac 
 cented vowel has disappeared: cases like dvyarpes in Homer 
 for Ovyarepes, (Bfjv for effrjv, etc. So far from being disposed to 
 shorten the vowel which follows the accented syllable, the 
 
GREEK A CCENT. 1 1 5 
 
 Greek shows rather a predilection for such forms as av 
 Ti6ri[jii J \v6r]croi(j9ov. Latin proper names like Dentatus, 
 Modest us, Salernum, the Greeks were perfectly able to pro 
 nounce with their Latin accent ; there was nothing in their 
 own system which forbade it : yet we often find such words 
 accented on the first syllable, devraros, MoSeo-ro?, ZuX&pvov 
 showing that an accented antepenult followed by a long penult 
 was a combination agreeable to the Greek ear and regarded 
 with a kind of preference. 
 
 Another consideration which goes to show that the Greek 
 accent was not accompanied by any very decided stress of 
 voice is found in the structure of Greek verse. In this, the 
 word-accent is wholly disregarded, the ictus of the verse being 
 quite as likely to fall on an unaccented as on an accented syl 
 lable. In the heroic hexameter, for example, we know that 
 there was an ictus, or verse-accent, that is, a special stress of 
 voice, on the first syllable of each dactyl or spondee. But if 
 we look at the first seven lines of the Iliad, out of the forty- 
 two cases of ictus which they present, only sixteen are found 
 on syllables which have the written accent. Now it would 
 seem to us unnatural in our own language to take the words, 
 " regarded with admiration and uncommon esteem," and read 
 them as a hexameter, " regarded with admiration and uncom 
 mon esteem;" or to take the line, <( tis as moonlight unto 
 sunlight or as water unto wine," and read it as an iambic 
 tetrameter catalectic, " tis as moonlight unto sunlight or as 
 water unto wine." But we must suppose that the Greeks did 
 something very much like this, if we assume that with them, 
 as with us, there was a decided stress of voice laid on the 
 accented syllable. It is true that there is something hazard 
 ous in such reasoning. The Greeks, in the construction of 
 their verse, may have treated their language with more free 
 dom than we allow in the treatment of ours. Different lan 
 guages, or, rather, the people who use them, differ widely in 
 this respect. Thus the German poet has greater liberty than 
 the French in departing from the established forms and idioms 
 of prose speech. We have seen that in Greek music the 
 natural tune of the word, the differences of pitch depending 
 on the accent, were not observed, but were freely superseded 
 
1 1 6 GREEK A CCENT. 
 
 by other combinations of tones at the pleasure of the com 
 poser. If, then, the Greeks in their music were willing to 
 substitute other tones for those given by the accent of the 
 spoken language, it is conceivable that in their verse also they 
 may have been willing to substitute other stresses for those of 
 ordinary speech, to lay the emphasis on other syllables, ac 
 cording to the rhythmical arrangement of the poet. Yet we 
 cannot but regard it as highly improbable that a stress-accent, 
 if it were as decided as ours, should be wholly neglected and 
 superseded in the composition of verse. And in this view we 
 are confirmed by the modern Greek, which, having a decided 
 stress-accent, makes it, as we do, the basis or determining 
 element of its verse-system. 
 
 Taking all these considerations together, we hold it all but 
 certain that the Greeks did not lay any marked or forcible 
 stress on the accented syllable. It is even doubtful whether 
 they laid any stress upon it, more than on other syllables of 
 the word. It is certain, however, that in the history of their 
 language they have either adopted a stress-accent, or have 
 given strength to a weak one existing from the first. In the 
 modern accent, the leading element is stress, and difference of 
 pitch, if it is not wanting, has at any rate ceased to be promi 
 nent and uniform. The distinction between acute and circum 
 flex has been wholly obliterated. The forms S^Xwcrat, to 
 make manifest, and ^Xwcrat, might make manifest, are 
 undistinguishable in the modern pronunciation. With this 
 change in the character of the accent was connected, as we 
 just saw, the adoption of accent as the basis of versification. 
 When verses began to be constructed on this basis, they were 
 called crri xpi nrokiriKOi, political or popular verses, in 
 contrast with the old quantitative verses, which continued to 
 be written, as a kind of literary exercise, long after the pro 
 nunciation on which they were founded had ceased to be 
 heard. Now these political verses were composed as far back 
 as the eleventh century, and probably much earlier. When 
 ever they began to be made, we may be sure that the Greek 
 accent had already changed its character, and had come to 
 exhibit a decided stress. But the change, we may presume, 
 was very slow, and may have been going on for centuries be- 
 
GREEK A CCENT. 1 1 7 
 
 fore the stress element was strong enough to express itself in 
 verse composition. It is quite supposable, therefore, that a 
 weak stress may have been heard on the accented syllable, as 
 regular accompaniment, even in the time of Herodian, the 
 principal authority on accent, if not yet earlier in the time of 
 Aristophanes of Byzantium, the inventor of the accentual 
 signs. If this were so, we might find in it an explanation of 
 the fact already noticed, that the last syllable of an oxytone 
 word -has an accent written upon it, even when its tone 
 changes from high to low. In j3acri\evs eyevero, the grave 
 accent on Xeu? shows (so we are told) that it was pronounced 
 on a low tone. But why then should it have any mark over 
 it, more than the other syllables of the same word (3acri\evs ? 
 Or why should pacriXevs, which in this case has no high tone, 
 which is all pronounced on a low tone, have any mark of 
 accent over it, more than an enclitic word, more than eariv 
 (for instance) in ^dXewov ea-riv, where the verb appears with 
 out accent? To such questions it might be replied, not 
 wholly without plausibility, that though ftacriKew in the case 
 supposed had no elevated pitch, no accent properly so called, 
 its last syllable was yet distinguished from the rest by a 
 slightly superior stress, and was therefore allowed to have a 
 distinctive mark over it ; while the enclitic ecmv was written 
 without any such mark, because it had neither elevated pitch 
 nor superior stress on either of its two syllables. This, I 
 say, would be a possible solution of the difficulty. But there 
 is another solution, of which I am now to speak, and which 
 brings up a question of much interest for the theory of the 
 Greek accent. 
 
 Did the Greek accent distinguish only two tones, a high 
 and a low ? Or was there some middle tone, having a regu 
 lar place in the system, some intermediate between the two 
 extremes ? It must be confessed that the evidence given on 
 this point by the ancient writers is not so distinct as could be 
 wished. , In general, they speak only of two tones, high and 
 low, or (as they term them) acute and grave. Still, they clo 
 not explicitly assert that all grave tones were equally low. 
 It may be that they thought it important only to distinguish 
 the high tone, as dominating the whole word, and that dif- 
 
1 1 8 GREEK A CCENT. 
 
 ferences between the lower tones seemed to them of too little 
 practical consequence to require mention. Yet we do find in 
 ancient writings indications of a middle tone. Thus Aristotle 
 (Rhet., iii. I. 4) speaks of three "tones, acute, grave, and 
 middle ; " though it is possible that by " middle " here he in 
 tends the circumflex, which, combining as it does both acute 
 and grave, might be regarded as having an intermediate char 
 acter. But the Greek grammarian Tyrannic of Amisus, as 
 quoted by Varro, enumerates four accents, grave, middle, 
 acute, and circumflex. Tyrannic, indeed, may have been 
 speaking of the Latin accent ; but Varro refers to other 
 writers as recognizing a middle accent, Glaucus of Samos, 
 Hermocrates of lasos, and the Peripatetics Theophrastus and 
 Athenodorus, some, if not all, of whom must have referred to 
 the Greek. And the grammarian Servius says : "It must be 
 understood that this doctrine of a middle accent is no inven 
 tion of recent times, but belongs to all who before the time of 
 Varro and Tyrannic have left anything on accent ; since the 
 majority of these, and the most distinguished writers, have 
 made mention of this middle accent, all of whom Varro refers 
 to as his authorities." This language of Servius, doubtless, 
 overstates the case. The number of writers who expressly 
 recognized a middle accent cannot have been so great as here 
 represented. But a reason may be found for this in a remark 
 of Servius himself, that "the middle accent, which is a sort of 
 border between the two others, resembles the grave more than 
 it does the acute, and is therefore reckoned with the lower 
 rather than with the higher tone." 
 
 The evidence of a middle tone which has come to us from 
 ancient writers, deficient as it is in definiteness and con 
 sistency, seems on the whole sufficient to warrant the inquiry, 
 whether the phenomena themselves of the Greek accent fur 
 nish any indications of such a tone. G. Hermann, in his 
 essay De cmendanda ratione Grammatics Grcecce (p. 66), 
 suggested that the grave accent, where it was written for the 
 acute on the last syllable of an oxytone word, was the sign 
 of a middle tone, intermediate between the acute and the un 
 marked grave. On this point Buttmann also, in his AnsfilJir- 
 liche Grammatik, takes substantially the same ground. More 
 
GREEK A CCENT. 1 1 9 
 
 recently, G. Curtius, reviewing Bopp s Accentuationssystem 
 in Jahn s Jahrbucher (1855, vol. 71), expresses the opinion 
 that the grave accent, where it forms the second part of the 
 circumflex, represents not the ordinary low tone of the word, 
 but an intermediate tone. This view seems to have been sug 
 gested to his mind by the Sanskrit system of accentuation : 
 here, the udatta (or elevated tone), corresponding to the 
 Greek acute, is regularly followed in the next syllable by the 
 svarita, which certainly differed from the anudatta (unele- 
 vated, depressed tone) of the other syllables. The Indian 
 grammarians describe the svarita as a combination of the 
 udatta and anudatta, similar to the Greek circumflex ; and 
 this must clearly have been its character when used as an 
 independent accent. But where it is the mere follower of an 
 z/dV?//,. and especially where it belongs to a short syllable, the 
 statement that it was a compound accent, a circumflex, seems 
 far from probable ; we could much more easily believe it to 
 have been an intermediate tone. Curtius inti-mates that he 
 would not confine his recognition of a middle tone in Greek 
 to the last part of the circumflex accent ; yet he has not 
 actually given further development to the theory. But a 
 recent writer, Franz Misteli, in an able article contributed to 
 the 1 7th volume of Kuhn s Zeitschrift, lias taken it up and 
 carried it to a much greater extent. He holds that the acute 
 accent in Greek, as in Sanskrit (?), was regularly followed by 
 a middle tone, this middle tone being either written as the last 
 part of a circumflex, or being merely understood on the syl 
 lable which comes after the acute. If the acute stands on the 
 last syllable of a word, where there is no room for a rm ddle 
 tone after it, the acute itself loses its high pitch and becomes 
 a middle tone, represented by the so-called grave accent. 
 Only at the end of a sentence, or before an enclitic, does the 
 acute under such circumstances retain its high pitch, and the 
 word appear as an oxytone. This theory of a middle tone 
 Misteli applies with much ingenuity to account for the general 
 laws of Greek accentuation. In showing how it may be made 
 to answer this purpose, I shall not confine myself to his state 
 ments, but shall take the liberty to depart from them in vari 
 ous particulars, and shall introduce some views (especially 
 
1 20 GREEK A CCENT. 
 
 those on Latin accent) which do not appear in his exhibition 
 of the subject. 
 
 The general laws here referred to are the four following : 
 I. The acute cannot stand on any syllable before the ante 
 penult. 2. The antepenult, if accented at all, must have the 
 acute ; but it cannot be accented at all, if the ultima is long : 
 thus. civOpwTTos, but avOpwirov. 3. The penult, if accented, 
 must have the acute, when the ultima is long (has a long vowel- 
 sound) : thus avOpooTrov, Toiavrr). 4- A long penult (one 
 which has a long vowel sound), if accented at all, must have 
 the circumflex when the ultima is short : thus TOIOVTOS. 
 
 The Greek accent is confined to the last three syllables of a 
 word. But in Sanskrit there is no such restriction. The ac 
 cent may go back to any syllable, however far removed from 
 the end of the word. In dbubodhishdmahi^ l we wished to 
 know, it stands at the beginning of a seven-syllable word. 
 The same freedom, we may presume, belonged to the primi 
 tive Indo-European language. There must have been some 
 time, therefore, in the history of the Greek language, whether 
 before or after it became distinctively Greek, when a change 
 took place in this respect some time when all accents stand 
 ing before an antepenult were carried forward and thrown 
 upon one of the last three syllables. If we ask for the cause 
 of such a change, none could be imagined more natural or 
 probable than a fondness for some particular succession of 
 tones at the end of a word. If the earlier accentuation had 
 a threefold distinction of tones, a high tone, middle tone, and 
 low tone, we can easily conceive that this succession, these 
 three tones in the order of their height, should have been 
 found an agreeable cadence for the close of a word. It is a 
 cadence which appears in Sanskrit in very many words ; it 
 may have been common in the Indo-European language prior 
 to the separation of its branches. What we should have to 
 suppose in regard to the ancestors of the Greeks would be 
 that at some time a special taste or fondness for this cadence 
 developed itself among them, with a dislike for any cadence 
 in which the high and middle tones were followed by more 
 than one low tone, so that, in order to secure what they liked 
 and avoid what they did not, like, they came at length to 
 
GREEK A CCENT. 1 2 1 
 
 change the accent of words, to shift it from an earlier to a 
 later syllable. Changes in an opposite direction, from a later 
 syllable to an earlier, were not made, even to secure this 
 favorite cadence ; or, if such a change took place in particular 
 cases, it did not become the general law. One thing further 
 we must suppose, to account for the Greek accent : that in this 
 cadence the Greeks preferred that the final low tone of the 
 word should be a short one : they did not like to have it 
 maintained through a long syllable. Our hypothesis, then, 
 may be stated in a single sentence that the early Greeks 
 changed the older accent of words so as to secure this ca 
 dencehigh tone, middle tone, short low tone wherever it 
 could be secured without throwing back the accent. This 
 single hypothesis will be found sufficient to account for the 
 four general laws already given. Thus : 
 
 1. "The accent cannot stand on any syllable before the 
 antepenult." In e Xe/Trero, was left, the accent (it can hardly 
 be doubted) was originally placed on the augment, as it 
 regularly is in Sanskrit : it fell, therefore, on the syllable be 
 fore the antepenult. The middle tone would then fall on the 
 antepenult Xet, and the remaining two syllables ire and TO 
 would have the low tone. But the preferred cadence allowed 
 only one syllable, and that a short one, for the low tone at the 
 end of the word. It was necessary, therefore, to place the 
 high tone or accent on the antepenult \ei, leaving Tre for the 
 middle tone and TO for the low tone. 
 
 2. " The antepenult, if accented at all, must have the 
 acute." Of course, if accented at all, it must have either the 
 acute or the circumflex. Suppose, then, that in the word 
 eXetrreTo the circumflex accent was placed on the antepenult 
 Xei : that syllable, from the nature of the circumflex, would 
 have the high tone on the first part and the middle tone on 
 the last ; and thus again as before there would be two sylla 
 bles, Tre and TO, for the low tone. The circumflex on the 
 antepenult would, therefore, be incompatible with the cadence 
 required. But the rule asserts also " that the antepenult can 
 not have any accent, even the acute, if the ultima is long." 
 For suppose that the first person eXetTro^v could have the 
 acute, the high tone, on the antepenult Xet : the middle tone 
 
1 2 2 GREEK A CCENT. 
 
 would fall on TTO, and the low tone on /A^V, a long syllable. 
 But the cadence required was " high tone, middle tone, short 
 low tone." Hence the high tone or acute must be placed on 
 the penult TTO, and the final long syllable ^v must be divided, 
 between the other two tones, its first half being sounded with 
 the middle tone, and its last half (which, of course, has the 
 quantity of a short syllable) being sounded with the low tone. 
 
 3. "The penult, if accented, must have the acute when 
 the ultima has a long vowel." For in the feminine Toiavrr), 
 such, suppose that the circumflex could stand on the penult 
 av. This syllabic, then, by the nature of the circumflex, 
 would have both high tone and middle tone, and the low tone 
 would fall on T??, a long syllable, which is inconsistent with 
 the required cadence. To secure this, the high tone, or acute 
 accent, must be placed on the penult av, and the final long TIJ 
 must be divided between the other two tones, the first short 
 time contained in it being sounded with the middle tone, and 
 the last short time with the low. 
 
 4- "A penult with long vowel-sound, if accented at all, 
 must have the circumflex when the ultima has a short vowel." 
 For suppose that the masculine roioOro?, such, had the acute 
 or high tone on the penult ov ; the middle tone would then 
 fall on the last syllable TO?, and the final low tone would be 
 excluded. In many words, as in ^070?, speech, this was 
 something unavoidable : the high tone falling onXo, the short 
 penult, must fill that syllable, the middle tone must fill the last 
 short syllable 709, and there is no room left for the closing low 
 tone. But with a penult long by nature, as in TOLOVTOS there 
 was no such necessity. It was enough to divide the long ov 
 between the high tone and middle tone in other words, to 
 sound it with the circumflex accent : the ultima TO? was then 
 left for the short low tone, and the desired cadence was thus 
 obtained. 
 
 We see, then, that these four general rules of Greek accent, 
 which have the appearance of being unconnected, arbitrary, 
 and capricious, are all of them direct corollaries from a single 
 hypothesis, all of them necessary results from the extension 
 of a single cadence high tone, middle tone, short lo\v tone 
 at the end of words. But what shall we say of polysyllables 
 
GREEK A CCENT. 1 2 3 
 
 like ^aXe-Tro?, harsh, with acute on the last syllabic, or 
 like ^aXe-Trw?, harshly, with circumflex on the last syllable, 
 or like XeXv/zez/o?, * having been loosed, with acute on the pe 
 nult before a short ultima ? In these words there is room for 
 the favorite cadence, but they do not have it : they close 
 either with the high tone itself (as in %aXe7ro?), or with the 
 high tone followed by the middle (as in ^aXevrco?, XeX j/LteVo?). 
 How can this be accounted for? By the last clause in our 
 hypothesis : " The Greeks changed the older accent of their 
 words, so as to secure the favorite cadence, wherever this 
 could be secured without throwing back the accent ;" We hold 
 it to be true, as a general fact, that words like those just given 
 were thus accented in the primitive period i.e. either on the 
 last syllable, or on the penult with short ultima following 
 and that they did not assume the threefold cadence, because 
 the tendency to this was not strong enough to produce a re 
 traction of the accent from a later to an earlier syllable. By 
 this we do not mean that such a retraction never occurred. 
 It may have taken place in numerous instances, but it never 
 became the general law. And thus the Greek has many 
 oxytones, which of course end with a high tone ; many pe- 
 rispomena, which end with a middle tone ; and not a few pa- 
 roxytones with short ultima, which likewise end with a middle 
 tone : and these, not in short words only, when the full 
 cadence was impossible, but in very many longer words, 
 where there was room enough and to spare for the succession 
 of high tone, middle tone, and short low tone. 
 
 But there was one branch of the Greek people the /Eoli- 
 ans of Asia Minor which went farther than the rest in the 
 fondness for this cadence. The ^Eolians did not hesitate to 
 retract the accent for the purpose of securing it. Thus Xe\u//,e- 
 z/o<? was changed by them to XeXuyLte^o?, ^aXevro? to ^aXevro?, 
 ^aXeTTw? to ^aXcTTO)? ^aXe-Tro)? they could not say, for that 
 would make a long low tone. If there was not room for the 
 complete succession of three tones, they secured as much of 
 it as they could, saying (for instance) cro^o? with high tone 
 and middle tone, instead of the common Greek 0-0^09, wise, 
 an oxytone word. Hence in the /Eolic dialect the only oxy 
 tones were monosyllables, and even these were oxytone only 
 
1 24 GREEK A CCENT. 
 
 when the vowel was short, as the articles rov, TO, ra. If the 
 vowel was long, there was room for a middle tone after the 
 high tone : the monosyllable was then pronounced with cir 
 cumflex accent, as r-fjv, rot)?, instead of the common Greek 
 TTfv, TOI;?. It was only in prepositions and conjunctions that the 
 ^Eolic, in agreement with the other Greek dialects, admitted 
 oxytone words of greater extent than one short syllable. 
 
 We observe now that a similar hypothesis may be used to 
 explain the peculiarities of Latin accentuation. The Latin, 
 beside the acute accent, or high tone, has a circumflex, which 
 it uses just where the^Eolic dialect would use the circumflex 
 that is, on all monosyllables with long vowel (except only ne 
 with the imperative), and on all penults with long vowel fol 
 lowed by short ultima. On all other accented syllables the 
 acute was used : viz., on antepenults followed by short penult, 
 on long penults followed by long ultima, on all penults long 
 only by position, on short penults of dissyllables, and on all 
 monosyllables with short vowel, enclitics of course excepted. 
 It is evident that in changing the primitive accent the Latin 
 has not confined itself to one direction, from earlier to later 
 syllables : like the ALolic Greek, it has freely moved the ac 
 cent backward, from later syllables to earlier, in order to 
 secure the desired cadence. But the cadence required to 
 account for the Latin accent is in one respect different from 
 that which served to explain the Greek. The Greek would 
 not allow the middle tone to be followed by a long low tone : 
 the Latin would not allow the low tone to be preceded by a 
 long middle tone, a middle tone extending over the whole of 
 a long syllable, whether long by nature or by position. Hence 
 for the Latin the cadence becomes "high tone, short middle 
 tone, low tone." For example, in Icgcre, Icgcres, Icgcrct, the 
 low tone fell on the last syllable, re, res, ret, without reference 
 to its quantity, whether long or short ; the middle tone on the 
 short penult^, and the high tone on the antepenult le. This 
 example is enough to show that, by the necessity of such a 
 cadence, the accent could never go farther back than the 
 antepenult ; that the antepenult, if accented at all, must take 
 the acute ; and that the antepenult could only be accented 
 when the penult was short. In such forms as gaudere Q^ gait- 
 
GREEK A CCENT, 1 2 5 
 
 dcrct, the acute must stand on the long penult de ; but if the 
 desired cadence is to be obtained, it must admit the middle 
 tone to the latter half of the long vowel, for the short ultima 
 has only room enough for the low tone : the word, therefore, 
 could only have the circumflex (the combination of high and 
 middle tone) on the penult. But in forms like gaitdcrcs, where 
 both penult and ultima have long vowels, the Latin preferred 
 to divide the long ultima res between the middle and low 
 tones, leaving only the high tone for the penult dc : the word, 
 then, has the acute on the penult. So too have forms like 
 legendus, where the penult has a short vowel, and is only long 
 by position. Here the circumflex is impossible : the short 
 vowel in gen cannot be divided between two tones, the high 
 and middle ; gen must have the high tone, dus the middle, 
 and the low tone is excluded by the necessity of the case. In 
 legcndi, the cadence could be made complete by dividing the 
 long di between the middle and low tones ; but in legendus 
 it is necessarily incomplete, as much as in legit or legunt, where 
 both vowels are short, or in the Greek \oyo$, having only the 
 high and middle tones. 
 
 It is not necessary to go further into details to show that all 
 features of the Latin accentuation may be accounted for by the 
 assumed tendency to close all words with the succession of 
 " high tone, middle tone, low tone," or so much of it as pos 
 sible, consistently with the one restriction that the low tone 
 must never have before it a middle tone which occupied the 
 whole of a long syllable. But if we compare the Greek ac 
 centuation with the Latin, and both with that freer system of 
 primitive Indo-European speech which is best represented to 
 us by the Sanskrit, we may naturally conclude that the first 
 step in the series of changes which gave the accentual systems 
 of the Greeks and Romans their peculiar character was caused 
 by a simple distaste for a succession of low-toned syllables at 
 the end of a word, I repeat, the beginning of a special Greek 
 and Roman accentuation would seem to have sprung from the 
 mere unwillingness to hear more than one low-tone syllable 
 at the end of a word. This unwillingness, carried into prac 
 tical effect, would confine the accent (that is, the high tone) 
 to the last three syllables of the word. But it would not 
 
1 20 GREEK A CCENT. 
 
 cause any retraction of the accent ; it would allow such forms 
 as ^aXeTTo?, %<zXe7r>9, XeXu/u-eVo? ; for these have not even one 
 low tone, still less have they a succession of low tones at the 
 end. And further, this simple distaste or unwillingness, as it 
 implies no restriction on the long or short quantity of the single 
 tones, would allow such forms as eXe/Tro/^y, avOpwirov, in 
 Greek, and such forms zsgauderet, legend us, in Latin. This 
 first step we may naturally suppose to have been made in the 
 Graeco-Latin or Graeco-Italican period that is, while the 
 common ancestors of these peoples spoke a language, differing 
 indeed from the original Indo-European, but not yet divided 
 into branches having a distinct character as Greek, and Italican 
 or Latin. But the next step must have been taken after this 
 division, as it is different in the two branches. In the Greek, 
 it springs from a distaste for a long low tone, .a low tone 
 stretching through a whole long syllable, following the high 
 and middle tones at the close of a word. This would require 
 eXenro/Arjv instead of eXetTro^z 1 , dvOpaiTrov instead of civOp^Trov. 
 The corresponding step in the Latin springs from a distaste 
 for a long middle tone, a middle tone stretching through a 
 whole long syllable, between the high and low tones at the 
 close of a word. This would require gandvrcs instead ofgau- 
 dcrcs, legvndus instead of legendus. In Greek the effect of 
 this second step was to make the cadence, "high tone, mid 
 dle tone, short low tone," the prevailing one for words which 
 were long enough to admit of it ; though there still remained 
 a large number of words, represented by ^ aXeTro?, ^aXe-Trw?, 
 XeXu/Ltez o?, which did not have it. Now one section of the 
 Greek race, the /Eolic of Asia, went further than this : they 
 took a third step, probably much later than the second ; they 
 threw back the accent of these words, so as to make the already 
 prevailing cadence universal, so far at least as the length of 
 the word would allow. Whether the whole Italican race, in all 
 its branches, Umbrian, Oscan, Sabine, etc., took a similar 
 third step, we are unable to say. It is certain that one branch 
 did so, the Latins ; they threw back the accent, so that the 
 cadence already prevailing should be made universal, as far as 
 the length of the word would allow. 
 
 Perhaps I should leave a false impression, if I were to close 
 
GREEK A CCENT. 1 2 / 
 
 without calling attention in one word to the hypothetical char 
 acter of what has been said here about a middle tone. The 
 existence of a middle tone in Greek and Latin has a good deal 
 of positive ancient testimony in its favor. But that a high 
 tone, when it did not come at the end of a word, was regularly 
 followed by a middle tone, is a proposition which, however 
 supported by Sanskrit analogies, has no direct evidence in the 
 statements of, the ancient writers. And of course, if there 
 were no doubt of its truth, still the use here made of it to ac 
 count for the ante-historic changes and the earliest historical 
 appearances of Greek and Latin accentuation, would be 
 purely hypothetical. At the same time, it may be said witli 
 justice, that the hypothesis is so natural in itself, it is so 
 readily suggested by known facts, and it offers so simple and 
 perfect an explanation for a variety of seemingly unconnected 
 and capricious phenomena, that one can hardly help believing 
 that it has a foundation in truth. 
 
 To some persons it may seem hard to believe that the ordi 
 nary utterance of discourse and conversation should have had 
 so much of musical intonation : that this threefold distinction 
 of tone should have found place in it as a recognized and con 
 stant element. But in the Chinese, and the languages cog 
 nate to it, as spoken at the present day, we find the musical 
 element playing a much larger and more important part. In 
 some of the popular dialects of China, a large proportion of 
 the syllables which make up the language are pronounced 
 with seven or eight intonations : thus, as a short abrupt mono 
 tone (compare the English preposition to in its ordinary short 
 pronunciation) ; or as a prolonged monotone (compare the 
 English numeral tivo) ; or with mixed falling tone (like the 
 Greek circumflexed TOI)) ; or with mixed rising tone (like the 
 English two at the end of a question: " two ? ") \ or with 
 .similar intonations duplicated on a lower key. Thus the same 
 syllabic may be pronounced in seven or eight different ways, 
 having each their special and widely diverse meanings. Com 
 pared with such complexity of musical intonation, that which 
 we have hypothetically ascribcd to the early speakers of Latin, 
 Greek, and Sanskrit, and the yet earlier speakers of the undi 
 vided Indo-European speech, is a very simple and easy matter. 
 
VII* 
 
 ON THE BYZANTINE GREEK PRONUNCIATION 
 OF THE TENTH CENTURY, AS ILLUSTRA 
 TED BY A MANUSCRIPT IN THE BODLEIAN 
 LIBRARY. 
 
 1870. 
 
 IN the second part of Mr. Alexander J. Ellis s work on 
 " Early English Pronunciation," mention is made (pp. 5 J 6- 
 527) of a document which seems to me of considerable inte 
 rest in reference to the history of Greek pronunciation. It 
 consists of a few manuscript leaves, written apparently by an 
 Anglo-Saxon hand, not far from a thousand years ago. On 
 these leaves are given passages from the Greek text of the 
 Septuagint, written not in Greek, but in Anglo-Saxon char 
 acters. They are Anglo-Saxon transliterations of the Greek 
 Septuagint, in which it seems to have been the object of the 
 transliterator to represent, at least approximately, in Anglo- 
 Saxon letters the current pronunciation of the Greek words. 
 These transliterations were noticed as long ago as 1705, by 
 the famous Anglo-Saxon scholar Hickes, in the preface to his 
 Lingnarnin Veterum Scptciitrionaliuin Tlicsaurns (pp. xi 
 xiii). The codex in which they are found is of a composite 
 and fragmentary character. There is a brief account of it in 
 the second volume of Hickes Thesaurus, with a description 
 of its contents, given by Humphrey Wanley. Mr. Ellis de 
 scribes it more at length, on the authority of Mr. G. Waring 
 of Oxford. He speaks of it as a small quarto volume, con 
 taining several unconnected pieces of great age and value. 
 Thus, in folios 1-8, we have part of the treatise entitled De 
 Conjugationibus distinguendis , by the grammarian Eutychius ; 
 in folios 10-18, an Anglo-Saxon homily on the finding of the 
 cross by St. Helena, the mother of the Emperor Constantine ; 
 in folios 20-22, a Lunar and Paschal Calendar for the years 
 817-832 ; in folio 23, Pauca dc Alcjisuris ct Pondcribus, a few 
 
GREEK PR O NUN CIA TION. 1 29 
 
 statements on weights and measures ; and in folios 37 to the 
 end of the volume, the first book of Ovid s Ars Amatoria. 
 It is a curious fact that three of these pieces the Eutychius, 
 the Pauca de Mcnsuris, and the Ovid s Ars Amatoria have 
 a number of Welsh glosses, renderings of Latin words and 
 phrases into the Welsh of that time. Zeuss, who makes 
 much use of these in his Grammatica Celtica, refers them to 
 the latter part of the eighth or the earlier part of the ninth 
 century. He pronounces them the oldest monuments of the 
 Welsh language ; and regrets that they are too scanty to show 
 us the system of the language as it then was. From similar 
 material, but more abundant, he has been able to reconstruct 
 the Irish of the ninth century ; for the Welsh, there is no suf 
 ficient material before the twelfth. This, however, is aside 
 from our present subject. The only parts of the manuscript 
 which concern us now are folio 19 and folios 24-36. These 
 contain extracts from the Septuagint, with the corresponding 
 passages from the Itala (or old Italian, ante-Vulgate, version). 
 And, of these, we may leave out of consideration folios 
 24-28, where the Greek text is given in Greek characters, 
 with many inaccuracies, which show that the copyist had but 
 little acquaintance with the Greek language. There remain 
 then folio 19 and folios 28 to 36, in which the Greek text is 
 given in Anglo-Saxon characters. Unfortunately this trans 
 literated text has never been published as a whole. But 
 Hickes in his preface before referred to has given specimens 
 of it, which belong to the 1st and 22d chapters of Genesis, 
 the 42 d Psalm, and the 4th, 5th, and 55th chapters of Isaiah : 
 a few verses from each, in all 25 verses. These same speci 
 mens are reprinted by Ellis, in a corrected form, having been 
 carefully collated with the manuscript by Mr. Waring. In 
 our remarks on Greek pronunciation as indicated by the 
 manuscript, we are confined of necessity to these 25 verses ; 
 but the evidence they give is probably as distinct on most 
 points as we should be able to draw from an examination of 
 the whole text. 
 
 But, before taking up particular points of Greek pronuncia 
 tion, it may be well to notice one or two questions which 
 naturally suggest themselves. What is the age of this manu- 
 9 
 
1 30 GREEK PR ON UN CIA TION. 
 
 script ? On this point I do not find that Hickes expresses 
 any judgment. Mr. Waring thinks that it was written in the 
 last half of the tenth century. His opinion does not seem to 
 be founded on the forms of the letters, though these would 
 probably give ground for a near guess to one expert in Anglo- 
 Saxon diplomacy, but on external or historical reasons. It is 
 not likely that an Anglo-Saxon, even of the better educated 
 clergy, would have known or cared much about the Greek 
 pronunciation of his time, unless his interest had been awak 
 ened and his information extended by living communication 
 with Greek persons.- Now Mr. Waring observes that Eadgith, 
 an Anglo-Saxon princess, married Otho I. of Germany in 
 930, and her son Otho II. married Theophania, a Greek prin 
 cess, in 972. He supposes that " at the court of Otho a 
 constant connection was kept up with the Anglo-Saxons and 
 the Greeks, and thus a means was opened for the priests of 
 the former to receive some tincture of Hellenic letters. We 
 shall therefore hardly be wrong," he says, " in referring such 
 transcriptions to the latter part of the tenth century. Want 
 of opportunity is against an earlier date, and the confu 
 sion and ruin occasioned by the Danish invasion in the early 
 part of the eleventh century, the close connection of Canute 
 with Rome, and the subsequent Norman influence through 
 Edward the Confessor, render a later date almost impossi 
 ble." To these historical reasonings, which do not seem to 
 me very decisive, he adds "the agreement of the Saxon 
 homily in the same book with the language of the tenth 
 centthry ; " and we may the more readily accept his conclu 
 sion, as most of the extant Anglo-Saxon manuscripts belong 
 to that century. 
 
 Is it certain that the scribe intended his transliteration to 
 be a phonetic one, that he aimed to represent the sounds 
 rather than the letters of the Greek text ? On this point 
 there seems to be no reasonable question. The single fact 
 that he represents OL uniformly by the letter y may be con 
 sidered decisive ; no reason can be imagined for this except 
 a desire to indicate the pronunciation. Neither the Greek 
 letters themselves, nor the o& diphthong which the Romans 
 used for them, could suggest a y as the symbol. It must be 
 
GREEK PR ON UN CIA TION. 1 3 1 
 
 admitted, indeed, that our writer is not uninfluenced by the 
 Roman mode of writing Greek words. Occasional variations 
 from his normal method can be traced to this influence. 
 Thus, in general, he takes no notice of the rough breathing, 
 which undoubtedly was no longer pronounced in his time ; 
 we may presume, therefore, that in the very few cases (six 
 only out of seventy-nine in the verses before us) where he 
 writes it by //, he was affected by the ordinary Roman tran 
 scription. For ai he usually writes a simple c (pronounced 
 nearly as our e in they) ; in a few instances (eleven only out 
 of eighty-eight), he writes CB, as in the ordinary Roman tran 
 scription. In this case the variation was the more excusable, 
 because the writer was probably accustomed in Latin texts 
 to sound ce exactly as he did e. So as to the consonant (/>, 
 which he generally represents by// if in two instances out of 
 fifteen he represents it as the Romans did, by//*, he certainly 
 did not think of ph as having any sound different from f. 
 Other variations and inconsistencies might be pointed out ; 
 but with one remarkable exception, of which we shall speak 
 further on, the cases of deviation from the rule bear only a 
 very small proportion to those of conformity with it. We 
 are thus able to make out in all points what may be called 
 the system of the writer, his normal method. We see clearly 
 what letters he thought best fitted to represent the several 
 vowels, diphthongs, and consonants of the Greek. 
 
 We proceed now to examine in detail this system of nota 
 tion, and to point out some conclusions which may be drawn 
 from it as to the current pronunciation of Greek in the, time 
 of the writer. The vowel a he represents, as we should ex 
 pect, by a ; e, by e ; i, by i ; o and w, by o. There is no 
 reason to suppose that the Greek pronunciation of these 
 vowels has altered since the earliest times, except that the 
 distinction of long and short quantity has ceased to be ob 
 served. Whether the distinction was still observed in the 
 time of our writer can hardly be determined. It is true that 
 the Anglo-Saxon scribes often mark the long sounds of their 
 own vowels by an accentual stroke ; but in this practice they 
 were so far from uniform that we can lay no stress on the 
 general omission of the stroke in this manuscript. Only four 
 
1 3 2 GREEK PR ON UN CIA TION. 
 
 instances of it occur in the 25 verses given. In three of these 
 it stands over o, in one over 77 ; in all the four it coincides 
 with a Greek accent. That the Anglo-Saxon vowels a, e, /, o 
 had the same sounds with the a, e, i, o in ancient and modern 
 Greek is sufficiently proved by other evidence, though the 
 confirmation which we find here is not unwelcome. The 
 Greek v our writer represents by y. He invariably distin 
 guishes it from i : he never uses y for i, and he never uses i 
 for v. This is unquestionable evidence that the two vowels 
 differed in pronunciation. Had L and v sounded alike, as 
 they do in modern Greek utterance, our scribe would have 
 confounded them in writing, as he confounds e and at, as he 
 confounds v and 01. Now, without reference to this manuscript, 
 there is strong reason to believe that the Anglo-Saxon y had 
 the sound of French u or German u. And the general 
 opinion of scholars has recognized this as the prevailing an 
 cient pronunciation of the Greek v. But here comes inde 
 pendent testimony that Greek v and Anglo-Saxon y were 
 sounded like each other, and both differently from the Greek i. 
 In what other sound different from i can they have agreed, 
 but in that which is most probable on separate grounds for 
 each of them, the sound of French n or German u? 
 
 The only simple vowel yet to be mentioned is 77. And in 
 reference to this we find a strange vacillation on the part of 
 the writer, a vacillation which has no parallel in his treatment 
 of other letters. He sometimes uses c and sometimes i, both 
 with nearly equal frequency. In some of the fragments e pre 
 vails, in others i but when all the instances are counted, there 
 is little difference in the sum. In TTCLCT^ 7-779 7779, Gen. i. 26, the 
 77 of the first word is given as c, that of the second and third 
 as i. In eSityrjcrev i) tyv%ij, Ps. xlii. 2, the case is exactly re 
 versed ; the 77 of the first word is given as i, that of the second 
 and third as e. Mr.Ellis s statement, here, has not his usual 
 accuracy. " For 77," he says, " we have most generally /, but 
 in about 50 instances e." If he had counted, he would have 
 found 55 instances of z, but 62 of e. The e s have a majority, 
 though only a small one, over the z s. Mr. Ellis suggests 
 that there was some confusion in the mind of the scribe, 
 "perhaps arising from the Latin transcriptions of 77, with 
 
GREEK PR O NUN CIA TION. 1 3 3 
 
 which he was necessarily more familiar." But this is contrary 
 to the analogy of the manuscript. The scribe was familiar 
 with os as a Latin transcription of CH, but in these verses he 
 scarcely uses it all. He was familiar with ce as a Latin tran 
 scription of at, but the instances in which he uses it are only 
 one-eighth of the whole number. In the other seven-eighths 
 he uses e, though he doubtless regarded ce as expressing the 
 same sound. If 77 had sounded to his ear like Anglo-Saxon i, 
 he would have represented it in general by that letter, and 
 only by an occasional slip have fallen into the e, which for 
 him expressed a different sound. On the other hand, if 77 
 had sounded to him like c, he would have represented it in 
 general by that letter, and only by an occasional slip have 
 fallen into the i. That he vacillates as he does between the 
 two is a sufficient proof that both were unsatisfactory ; the 
 Greek 77 did not seem to him like either his c or his i. Now 
 the ancient Greek 77 was a longer e, an open sound which must 
 .have been essentially the same as that of Anglo-Saxon e. 
 And the modern Greek 77 is not different in sound from L, and 
 of course not different from Anglo-Saxon i. It appears, then, 
 that in this manuscript we have caught the letter in a state of 
 transition ; it was on its way from the ancient to the modern 
 sound ; it had become closer than the first, but was not yet 
 so close as the second. It cannot have differed very widely 
 from the final sound of English they, pray, convey, etc., which 
 is certainly closer than Anglo-Saxon e, and has in fact a van 
 ishing sound like Anglo-Saxon i. The difficulty which puz 
 zled our writer may be illustrated by taking the three English 
 words ell, ail, eel. The Anglo-Saxon e was like the vowel 
 sound of ell the Anglo-Saxon i like the vowel sound of 
 eel. For the vowel sound of ail he had no equivalent in his 
 language : how was he to represent it ? If he writes an c, the 
 word will sound ell, not ail ; if he writes an /, it will sound eel, 
 not ail. No wonder that he vacillates between the two, un 
 satisfied with either. I suspect the 77, as he formed it, was 
 a little closer than our a in ail ; if not, I think he would more 
 generally have used his e. But that the 77 was then less close 
 than our e in eel that is, than Modern Greek 77 seems to me 
 proved beyond a doubt. 
 
1 34 GREEK PR ON UN CIA TION. 
 
 If now we pass on to the diphthongs, we shall find, as we 
 might expect, that a large part of them have lost the diph 
 thongal character. Beside those written with iota subscript, 
 which are represented here in the same way as simple a, rj, &>, 
 we have ov represented by //, ai represented by <?, and eu 
 represented by i. In all these cases, the change from the 
 compound diphthongal utterance to the simple sound had 
 begun to prevail either before, or not long after the Chris 
 tian era. Another very interesting change of this kind ap 
 pears in 01, which our scribe, as already stated, uniformly 
 represents by y. He does not distinguish it from f, and 
 doubtless heard it pronounced with the same sound, that of 
 French n or German u. The fact thatot was for a long time 
 sounded like v, while as yet they were both distinguished 
 from i, has already attracted the notice of scholars. On this 
 subject I may quote some statements made by G. Curtiusin the 
 Erlauterungen to his Grammar (p. 21). He tells us that 
 Liscovius, in a work on the pronunciation of Greek, pub 
 lished in 1825, cited from old grammarians a number of 
 orthographic rules which imply this identity of 01 and v. 
 Thus in the Erotemata of Basil, written in the fourth century 
 after Christ, it is said : Trdcra Xet? UTTO T^ KV <Tv\\a/3fjs 
 ap^opevri &ia rev v tyiXov ypdfarai irKrjv TOV Kol\ov. It is 
 strange that the rule should omit to mention /coi^da}, tcoivos, 
 and several others with initial /cot ; but it says, plainly enough, 
 that the first syllable of xolXov has the sound of KV. Similar 
 rules are found in the Epimcrismi ascribed to Herodian, 
 and in the Greek Etymologica. Thus in the Etymologicum 
 Magnum (p. 289, II) : ra et? vt; airavra Bia TOV v ^tXoi) jpd- 
 <j)TaL 7r\r/v TOV 7Tpoi. The word Trpol%, then, must have 
 sounded as if written rrpv^. If the modern pronuncia 
 tion had prevailed at that time, the writer should have added 
 words in i%, as 0pi%, Ki Xit;, fido-ri^, TrepStf, rerrtf, 
 , and a multitude of others ; also words in 17^, as 
 uX.& jrr)!;, fjLvpjur)^, vdp&rj!;, 7r?jX?;f, and many more ; for 
 the modern Greek pronounces all these as he would if they 
 were spelled with v in the last syllable. Other facts of this 
 kind arc given by R. F. A. Schmidt in his Beit rage zur 
 GescJiichte dcr Grammatik (p. 73 ff.) ; and he derives from 
 
GREEK PR ON UN CIA TION. 1 3 5 
 
 them a new and beyond all doubt a correct explanation of 
 the name v tyi\ov as well as e -v/aAoi . The old name of the 
 letter was v, i.e. the long sound of the vowel with circumflex 
 accent. But when 01 came to be sounded in the same way, 
 the adjective tyiKov, bare, simple, was often added : thus v 
 fyiXov, simple u, written with one letter, in distinction from 
 v 8i(j)Soy<yov, written with two letters, that is, 01. As to e, it 
 was first named e!, just like the Greek word for if, only 
 given with circumflex accent ; afterward it was named e, the 
 short vowel-sound itself serving as a name for the letter. But 
 when at, was sounded in the same way, the word ^L\OV was 
 often added : e -fyiKov, the simple, one-letter e, in distinction 
 from e SlijjQoyyov, the two-letter e that is, ai. 
 
 How (H should get the sound of v it is not difficult to un 
 derstand. First, the close i may have caused the o before it 
 to assume the closer sound of u. This would give the diph 
 thong uij which might naturally pass into the intermediate v. 
 If, instead of sounding the u and i each in its own position, 
 the vocal organs take the position for u and in that position 
 try to utter i, the result will be a simple v. It is a curious 
 fact that this pronunciation of oi appeared among the 
 Boeotians several centuries before the Christian era, as in 
 ru? aXXu? for TO? a\Xo9, FVtcta for olfc : a. What adds to the 
 wonder is, that the changes in at and et should have been 
 likewise anticipated among the Boeotians : as in ypdffreaSij, 
 Boeotic for ypdfaa&ai ; IJM, Bceotic for el^i. The Boeotians, 
 backward as their Athenian neighbors thought them, v;ere 
 certainly, as regards pronunciation, in advance of their 
 age. 
 
 We have yet to speak of the diphthongs av and ev. Our scribe 
 represents them by au and cu. In modern Greek they are 
 sounded as af and cf before surds (i. c. smooth and rough 
 mutes, also cr, f, ifr), but av and ev before sonants (i. e. all 
 other consonants and all vowels). m It seems clear that this 
 writer did not hear them as af and cf before surds ; for in 
 that case he would have used/" in writing them. To suppose 
 that he would uniformly give up a phonetic representation 
 for the sake of conforming to ordinary Latin transcription is 
 contrary to the analogy of his procedure in other cases. Nor 
 
I 36 GREEK PR O NUN CIA TION. 
 
 is it probable that he heard them as av and cv before sonants. 
 The u with which he writes them is the same letter that he 
 uses for ov ; it is most likely that he meant it to express the 
 same sound. This, however, is not entirely certain. The An 
 glo-Saxons, while they had a w (or what we write as such), had 
 no v y the Anglo-Saxon n is always a vowel. But, in writing 
 Latin, they used the same letter u both for u and for v. 
 We cannot, therefore, be quite sure that our scribe would 
 not have written au and en for av and ev t as he was wont to 
 do in writing Latin words like grains and levis. But if he 
 did so, if his au and eu were meant for av and ev t we may be 
 sure that the Greek /3 did not sound to him as v, which is its 
 present pronunciation. The modern Greek confounds av be 
 fore sonants with a/3 ; but this writer distinguishes them, giv 
 ing the former as av, the latter as ab / he cannot have sounded 
 both of them as av. 
 
 As to the rough breathing, we have already said that our 
 writer leaves it unrepresented in all but a very few cases, where 
 we may presume that he was influenced by the ordinary Latin 
 transcription. Undoubtedly it had ceased to be heard in pro 
 nunciation : and the Roman k, we know, has suffered the 
 same fate in the modern descendants of the Latin. 
 
 The consonants will furnish little subject for remark. We 
 have already observed that </> is pretty constantly represented 
 by/". It is remarkable that for a long time the Romans never 
 represented ^ by f. At first they used/ for this purpose, as 
 in Poino-s for $olvi% , purpura for Troptyvpa, etc. But after 
 ward they begin to express the Greek aspirates ; and thence 
 forward that is, from about the time of Cicero they used 
 ph for <j). It is not until late in the period of the empire that 
 we begin to find / for </>. This fact shows that the classical 
 pronunciation of <> must have been more than slightly differ 
 ent from that of// it must have been broadly distinguished 
 from the/, and nearer to Latin/. It was, in fact, what the 
 ancients describe it as being, a/ followed by an // distinctly 
 audible. But from this, its true aspirate sound, it passed into 
 the spirant// and that change had probably become fixed 
 some centuries before the date of our Anglo-Saxon writer. 
 The other aspirates must have undergone a similar change at 
 
GREEK PR ON UN CIA TION. 1 3 / 
 
 about the same time. They are represented here by th for 6, 
 and ch for %. It is surprising that we nowhere find the simple 
 character which is used in Anglo-Saxon writing for the th 
 sound of think, throw ; but there can be no real question as 
 to the sound represented. 
 
 It is worth noticing that, wherever two rough mutes suc 
 ceed each other, the first is always represented here as becom 
 ing smooth. Thus the word l%0va)v occurs twice, and both 
 times is spelt with cth, not chth ; the participle \6i<f>3-ev occurs 
 three times, and in each instance is spelt with pth, not fth. 
 Can it be that the long current pronunciation of the word 
 diphthong as diptlioug was founded on the usage of the Greeks 
 who served as teachers of their language at the revival of 
 learning? 
 
 In modern Greek pronunciation a smooth mute (TT, T, K), 
 when it follows a nasal, is vocalized and becomes sonant, 
 through the influence of the sonant before it. Hence /ZTT is 
 sounded as inb, vr as nd, y/c as ng. No such change is indi 
 cated in our scribe s transliterations : he writes ampelon (not 
 afnbelon) for a/7reXcoz>, panton (not pandon) for Trdvrayv, prose- 
 nencon (not prosencngoii) for Trpoo-eveytcov. 
 
 The modern Greek pronunciation has no middle mutes 
 except after a nasal. Everywhere else the /3, 7, S have ceased 
 to be mutes and have passed into spirants, with sonant utter 
 ance. From being the sonants of TT, , r, they have become 
 the sonants of <, %, 0, in their present pronunciation. The 
 writing of our scribe, who represents them by b,g, and ^/re 
 spectively, affords no clear indication of these spirant sounds. 
 It is true that the Anglo-Saxon g must often have had a weak 
 sound, not very different from our consonant y ; but there is 
 no reason to suppose that b and d had any other than their 
 present English sounds. But for the spirant sound of 8 the 
 sonant th in otJicr the Anglo-Saxon had a simple character, 
 which we should expect to find here, if that was really heard 
 by the writer. Still, as he has not used the corresponding 
 character for the surd th, we can hardly lay very much stress 
 upfon this fact. 
 
 In all that has been said thus far, I have spoken of the 
 scribe who wrote this transliterated text as being of Anglo- 
 
1 3 8 GREEK PR ON UN CIA TION. 
 
 Saxon race. But this cannot be regarded as certain. The 
 codex which contains it is composed of half a dozen pieces ; 
 and we have seen that three of them contain Welsh glosses. 
 Though there is one piece of Anglo-Saxon in the book, it is 
 evident that a Welsh hand or hands have had a large share in 
 its making up ; and it is altogether possible that this trans 
 literation may have been made by such a hand. This possi 
 bility derives some support from the fact that not one of the 
 special characters used by the Anglo-Saxons the w, the two 
 signs for th surd and sonant, the compound & is found in 
 this text. And none of them occurs in the Welsh glosses of 
 this volume, as printed by Zeuss at the end of his Gram- 
 matica Celtic a (with perhaps one exception, on p. 1087). It 
 must be said, however, on the other side, that the y, which 
 occurs so often in the transliteration, is not to be found in 
 these glosses. If it was a Welshman who made the trans 
 literation, we must suppose that, having in his own language 
 no sound that corresponded to the Greek v, he fell back upon 
 the familiar Latin equivalent y, which he used also for ot, be 
 cause that had the same sound as v. But it would still be 
 true that, if v and OL had a common sound, different from i, 
 that common sound could hardly have varied much from 
 German ii, which appears on independent grounds to have 
 been the ancient power of the v. And it might be shown in 
 detail, if there were time for it, that the other conclusions 
 which we have drawn from the manuscript would require little 
 modification, if we suppose it to have been written by a 
 Welshman rather than a Saxon. 
 
 The object of Mr. Ellis in giving a specimen of this trans 
 literated text, and commenting at some length upon it, is 
 mainly to throw light, on the current pronunciation of the 
 Anglo-Saxon in the tenth century. The light thus obtained, 
 if it does not discover anything absolutely new, gives a wel 
 come confirmation to views which were already probable on 
 other grounds. But Mr. Ellis is led to say something on the 
 pronunciation of Greek in ancient times. Without entering 
 into any discussion of the subject, he gives utterance to a 
 general conclusion, in the folio \ving terms : " We may never 
 be able to recover the pronunciation, or appreciate the quan- 
 
GREEK PRONUNCIA TION. 1 39 
 
 titative rhythm, of the Athenian tragedians or of the Homeric 
 rhapsodists, but we can read as Plutarch and as Lucian, and 
 we should be satisfied with that privilege, remembering that 
 if we pronounced these later authors otherwise than as the 
 modern Greeks, we should certainly pronounce wrongly. It 
 would indeed be just about as absurd to read Lucian with the 
 pronunciation of Aristophanes, as to read Tennyson with the 
 pronunciation of Chaucer." This is a kind of obiter dictum in 
 Mr. Ellis s book, for which, perhaps, he should not be held to 
 a very strict account. But we must be allowed to express 
 our surprise at hearing it from a scholar of so much candor 
 and judgment. He is right, indeed, in assuming that Greek 
 pronunciation changed in the five centuries between Aristo 
 phanes and Lucian. No one would deny this except those 
 who, like Professor Ross of Halle, maintain the antediluvi 
 an antiquity of the modern Greek sounds. But can 
 we suppose that Greek pronunciation has undergone no 
 changes in the seventeen centuries between Lucian and our 
 own day ? Have not the external and political conditions 
 been at least as unfavorable to continued uniformity of pro 
 nunciation in the .seventeen centuries since Plutarch and Lu 
 cian as they were in the five centuries before them ? But 
 Mr. Ellis asserts that " if we pronounced Plutarch and Lucian 
 otherwise than as the modern Greeks, we should certainly 
 pronounce wrongly." How any man can say this who looks 
 at the manuscript we have been discussing, and who believes 
 in it, is beyond my comprehension. Mr. Ellis recognizes in 
 the manuscript an attempt to represent the current Byzantine 
 pronunciation of the tenth century. But, so regarded, it 
 shows unequivocally that in the tenth century, seven or eight 
 hundred years after Plutarch and Lucian, the current pronun 
 ciation was still in many important points essentially different 
 from that of the present day. Unquestionably the most 
 characteristic feature of the present pronunciation is its iota- 
 cism. Like i are sounded the two vowels T] and v, and the 
 four diphthongs, et, 77, 01, vi : that is, there are six written 
 forms beside i which have the same sound with it. Now how 
 does it appear in this manuscript ? Here only one of these 
 six forms the diphthong et is as yet fully identified with t. 
 
1 40 GREEK PR ON UN CIA TION. 
 
 Three others, the vowel v and the diphthongs 01 and vi, are uni 
 formly distinguished from it, while the remaining two, T? and 77, 
 though clearly on the way toward an t-sound, have not yet 
 reached that goal. But the manuscript gives similar testi 
 mony in regard to other prominent features of the modern 
 Greek pronunciation : so as to the sounds of af and ef, for the 
 diphthongs av, euwhen followed by surds ; and probably also 
 the sounds av and cv for the same diphthongs when followed 
 by sonants. So, too, as to the medial sounds for the smooth 
 mutes TT, K, r, where they follow a nasal. I say nothing as to 
 the spirant sounds of the middle mutes, for in regard to these 
 the testimony of the manuscript can hardly be regarded as 
 decisive. But leaving these out of the account, the differences 
 indicated between the pronunciation of the tenth century and 
 that of the nineteenth arc extensive and important. How 
 then can it be said (as Mr. Ellis in effect says) that there is no 
 material difference between the pronunciation of the second 
 century and that of the nineteenth ? 
 
VIII. 
 ROSS ON ITALICANS AND GREEKS. 
 
 1858. 
 
 I PROPOSE to call the attention of the Society* for a short 
 time to a pamphlet of some ninety pages, recently pub 
 lished by Professor Ross, of Halle, and entitled " Italiker und 
 Grakcn. Sprachen die Romer Sanskrit odcr GriccJiiscJi ? " 
 Greeks and Italicans. Did the Romans talk Sanskrit or 
 Greek? In letters to a friend, by Ludwig Ross, Halle, 1858. 
 This rather quaint title is followed by an equally quaint dedi 
 cation, "to the Greeks, as being, in race and language, 
 parents of the Romans." The theory here indicated as to the 
 relation between Greeks and Romans is more distinctly stated 
 in the Preface a lengthy but lively document, addressed to 
 Professor Keil, of Pforte. The Preface, however, does more 
 than announce the author s views of this subject ; it explains 
 the reasons which have led him to the present publication of 
 his views. The immediate occasion for preparing the essay 
 came, as we learn, from the recent appearance of Mommsen s 
 History of Rome. The general merit of that extraordinary 
 work is fully and even warmly recognized by Professor Ross. 
 But he objects strongly to the opening chapters, complaining 
 in severe terms of the "rash, confident, and overbearing 
 manner" in which Mommscn dogmatizes " on the ante-his 
 torical period and the primitive history of Rome, on the 
 sources of our knowledge concerning them, on the ethno 
 graphic relations of the Latins and other populations of mid 
 dle and southern Italy, on the early importance of the 
 Etruscans, and all other matters connected with these " a 
 manner which (as he goes on to say) "makes it evident 
 enough that he has not thought it worth while to give these 
 subjects any serious investigation." The most important of 
 
 * That is, of the American Oriental Society, at its meeting in November, 1858. 
 
142 IT ALT CANS AND GREEKS. 
 
 Mommsen s errors, and the one specially combated in this 
 work, is found in a passage quoted by our author, asserting 
 that "philological researches teach us to distinguish three 
 primitive Italian stocks, the Japygian, the Etruscan, and the 
 Italican, as we may call it, the last of which divides itself into 
 two main branches, the Latin idiom, and that to which belong 
 the dialects of the Umbrians, Marsians, Volscians, and 
 Samnites." Professor Ross, on the other hand, maintains 
 (we give his own words) that "in all middle and southern 
 Italy, up to Ktruria and Umbria, so far as our knowledge ex 
 tends, but one great stock of languages the Greek pre 
 vailed, in different dialects ; so that Latins and Volscians, 
 Sabincs and Oscans, Messapians and Japygians, spoke noth 
 ing but corrupt Greek, and that admixtures of Greek are 
 found even in the Etruscan." And again, he asserts, in an 
 other place, that " the Latin, Sabine, Oscan, and, in general, 
 all the idioms of lower Italy, the Japygian and Messapian in 
 cluded, are nothing but Greek, corrupted in speaking, and at 
 last written with differently formed alphabets ; and that all 
 these idioms, and all forms of words developed in them, ran 
 together at last into the Latin as sole heir to all of them." 
 So certain and obvious docs this character of the Italican 
 idioms appear to Professor Ross, that from the failure of Bopp 
 and his followers to recognize it he does not hesitate to infer 
 the worthlessness of their methods. The passage in which 
 Indo-European philology is weighed in a balance and found 
 wanting is sufficiently curious to deserve extraction : 
 
 " You will ask, indeed," he says to his friend, Professor Keil, " in 
 what relation I stand to the Sanskrit, as I have placed upon the title 
 of these letters the question whether the Latin was Greek or Sanskrit, 
 while in the following pages the. Sanskrit is hardly mentioned. My an 
 swer is perfectly simple, I do not understand a word of Sanskrit. But 
 I comfort myself with the reflection that Mommscn and Curtius, whose 
 statements have called out this tract, understand hardly any more of it, 
 and have themselves been led by others into the path of error. The 
 thorough learning of a language requires far too much time for a classical 
 philologist to get it up in his spare hours : and with the mere skeleton of 
 a language there is very little gained. How Sanskrit sounded in actual 
 living utterance is probably pretty much unknown to our students 
 of Sanskrit ; and without the living sound a language is a stiff corpse, 
 which you may recognize, dissect, and understand, but can hardly set in 
 
ITALIC A NS AND GREEKS. 1 43 
 
 motion or action. Of Sanskrit studies in general I have a pretty low 
 opinion, for I do not see that, since they have come to flourish in Ger 
 many, and are represented in almost all our universities, they have pro 
 duced any important result, least of all any of a positive historical char 
 acter, unless it be the word Indo-Germanic a designation hardly sus 
 tained even yet by sufficient grounds with which so much speculation 
 of all sorts (Wesen und Unwesen, ( business and bother ) has been car 
 ried on, and which after all says nothing more than that the European 
 nations and their languages have their furthest roots in Asia a fact 
 known ever since the famous tower-building at Babel, only expressed in 
 a different way. As all comparative philology, in the many volumes it 
 has sent into the world, has never yet, to my knowledge, demonstrated 
 in extenso, as I have done in these pages, that the Latin is only a mix 
 ture of different Greek dialects, written in other letters, and afterwards 
 raised into a literary language I am fully justified in thinking of it thus 
 disrespectfully. I do not mean to deny that now and then a Greek 
 or Latin form or inflexion may properly be associated and compared 
 with the Sanskrit, but the same thing can be done with other languages." 
 
 By other languages which, according to Professor Ross, 
 throw as much light as Sanskrit on the forms of Greek 
 and Latin, he refers, as it appears, especially to the Egyptian 
 and the Phoenician. He suggests an attempt to point out the 
 Egyptian and Phoenician words in Greek and Latin, as a de 
 sirable antidote to the Sanskrit. And to this end he makes 
 a ^beginning himself with a long list of words, in which, 
 among others, vav<s navis, and TTOU? pes are traced to the 
 Egyptian. That the same words are found in Sanskrit (ndu, 
 pad), and in Anglo-Saxon (naca,fot), he would probably ex 
 plain, if his dislike of Indo-European philology would per 
 mit him to take notice of the fact, by referring to the con 
 quests of Sesostris. These he regards as the historical basis 
 for the Greek mythus of Dionysus conquering India (a 
 mythus, by the way, of which we find no trace before 
 the time of Alexander), and in both the Dionysiac myth and 
 the supposed history of Sesostris finds reason to believe that 
 the language and religion of India may have been largely in 
 fluenced by those of Egypt; so that "the multiform 
 and grotesque divinities of the Indians " (as he is pleased to 
 term them, with an evident unconsciousness of the immense 
 difference between the earlier and later mythologies of India), 
 " instead of being older than the Egyptian, are perhaps after 
 
144 ITALICANS AND GREEKS. 
 
 all only a disfigured copy of the latter." How this may be, 
 we will not stop to inquire ; we will rather imitate the 
 judicious forbearance of Professor Ross, who only throws out 
 the suggestion, without entering on the proof. 
 
 But we must be allowed to point out some misconceptions 
 which appear in the passage just extracted. In the first place, 
 he misconceives the significance of the term Indo-Germanic 
 or Indo-European (which Bopp with good reason prefers to 
 use) so strangely misconceives it as to furnish ground for a 
 suspicion that his frank and manly confession in reference to 
 his knowledge of Sanskrit should be construed as extending to 
 Indo-European philology. The term, he assures us, signifies 
 nothing more than that " the nations and languages of Europe 
 have their furthest roots in Asia." Now this in one view is 
 saying too little : in another, too much. Too much : for it 
 scatters the roots of Greek and Latin through the continent of 
 Asia ; so that, in following the indications of this statement, 
 we might look for them quite as much in Chinese and Man- 
 chu as in Zend and Sanskrit : indeed, Professor Ross, as we 
 have seen, finds them still more in Phoenician and Egyptian. 
 But this is not only not the meaning of the term Indo-Euro 
 pean : it renders the term absolutely unmeaning and absurd. 
 Again, it says too little : for it fails to express the fact that cer 
 tain languages of Asia have, not only the same ultimate roots 
 as the Greek and Latin, but also, what is vastly more impor 
 tant as proving identity of origin, the same grammar to a 
 great extent, the same systems of formation and inflexion. 
 We cannot help regretting that Professor Ross should have re 
 solved to learn no Sanskrit, because his other pursuits would 
 not allow him to learn all. He says, indeed, that " with the 
 mere skeleton of a language very little is gained." But the pre 
 sent case, we think, furnishes clear proof to the" contrary. If 
 Professor Ross would get some friend, who knows the Sanskrit 
 alphabet, to interline the paradigms of Bopp s Sanskrit Gram 
 mar with European equivalents for the oriental characters, it 
 would be easy for him to read them all over in less than two 
 hours. Among the nouns he would find such forms as Fern. 
 vdk, speech, in which he could not help recognizing the Latin 
 vox / Gen. vdcas, Lat, vocis ; Dat. vace t Lat. voci ; Ace. 
 
ITALICANS AND GREEKS. 145 
 
 wicam, Lat. vocem ; Nom. PL vacas, Lat. voces, etc. 
 Amonsf the verbs he would find such forms as Aorist adixam, 
 
 o 
 
 Greek eSeiga (originally eSeigav) ; 2d person adixcts, Gr. 
 eSetfe? ; 3d person adixat> Gr. ISetfe (originally eSeifer) : 
 Dual, 2d person adixatam, Gr. eBei^arov ; 3d person adix- 
 atduij Gr. eSei^drrjv : Plural, 1st person adixama, Gr. eSei^a/jiev ; 
 2d person adixata, Gr. eei%are ; 3d person adixan, Gr. 
 e$ei%av. We are greatly mistaken, if, after seeing such phe 
 nomena as these, which even the most cursory view of the 
 Sanskrit paradigms would force upon his notice, he could re 
 gard them as adequately represented by saying that " the 
 languages and nations of Europe have their furthest roots in 
 Asia." And if he could be persuaded further to spend two 
 hours or two days or two months on the languages of Egypt 
 and Palestine, with a view to find in their grammars any fair 
 offset or antidote to these Sanskrit parallels, we are greatly 
 mistaken if after such an attempt he could still assert that the 
 same comparison of Greek and Latin forms can be made with 
 these languages as with the Sanskrit. We believe that a few 
 hours of grammatical study applied in this way to the Sanskrit 
 and other oriental languages, though he might only become 
 acquainted with their skeletons, would effect a great change 
 in his views of comparative philology, and that thus very much 
 would be gained, if not for the cause of science, at least for 
 the enlightenment and reputation of Professor Ross. 
 
 But again, he misconceives the views of Indo-European 
 philologists as to the relation between Greek and Latin on the 
 one hand, and Sanskrit on the other. He evidently regards 
 them as deriving the Greek and Latin from the Sanskrit as 
 making this the parent language, and those its descendants. 
 This is implied in the interrogatory of his title-page, Did the 
 Romans talk Greek or Sanskrit?" According to him they 
 talked Greek that is, a language formed from the Greek : 
 and of course he assumes that according to his opponents 
 they talked Sanskrit a language formed from the Sanskrit. 
 But no philologist of reputation maintains, so far as we know, 
 that the languages of Europe are derived from the Sanskrit. 
 The Indo-European languages have many things in common 
 things which it would be absurd to attribute to accidental 
 10 
 
ITALICANS AND GREEKS. 
 
 coincidence ; things which can only be explained as inherited 
 from a common parent. But it is agreed on all hands, 
 that neither the Sanskrit nor any other extant language can 
 claim to be that common parent. The oldest of the Indo- 
 European dialects are only descendants of the primitive 
 mother. All of them have departed more or less widely 
 from the original type, modifying, omitting, and adding each 
 in its own way. The common language from which all have 
 sprung is to be reproduced, if at all, only by comparing all 
 its descendants. And if in such a comparison we derive es 
 pecial advantage from the Sanskrit more than from any sis 
 ter dialect, more than from the Greek, which in this respect 
 yields only to the Sanskrit the reason is, partly, that the 
 early language of India is preserved to us in monuments which 
 are at once very ancient and very copious ; and partly, that the 
 Aryan people of India, having wandered less w r iclcly from the 
 early home of their race, and being perhaps endowed by nature 
 with less mobility of character and habit, seem to have pre 
 served the primitive roots and forms of their language with 
 more tenacity than their kinsmen of the West. The original 
 Indo-European language must have had an extraordinary 
 abundance of undiscriminated synonyms, both in roots and in 
 formative syllables. Among such synonyms, the languages of 
 the West, as a general thing, take their choice, retaining each 
 of them one root or form, and suffering the rest to drop. Or, 
 if they retain more than one, it is usually by establishing some 
 distinction of sense between them, so that they are no longer 
 completely synonymous. The Sanskrit, on the other hand, 
 certainly stands nearer in this respect to the primitive condi 
 tion. It presents- especially as found in the Vedas a simi 
 lar copiousness of synonymous words and forms : a character 
 istic which of course renders it peculiarly valuable for the pur 
 poses of linguistic comparison. 
 
 But let us proceed now from the preface to the sequel of 
 Professor Ross s pamphlet. It will be observed that he 
 thinks of Latin as holding to Greek much the same relation 
 that French does to Latin. As French, though containing 
 some non-Latin elements, is yet substantially Latin, only al 
 tered by time and corrupted by popular use : so Latin itself 
 
ITALICANS AND GREEKS. 147 
 
 is in the same sense substantially Greek, only subjected to 
 similar modifying and degrading influences. This he endeav 
 ors to prove by an extended lexical comparison of the two 
 languages, Greek and Latin, embracing a large number of 
 words, and conducted with an openly avowed and evidently 
 sincere contempt for exact philological criticism. Of gram 
 matical comparison, between the systems of inflexion and for 
 mation in the two languages, we find but little, though in a 
 passing remark he acknowledges its importance. His strength 
 is in numbers the number of words which can be pointed 
 out and put together as really or apparently identical in Greek 
 and Latin. "A twentieth," he says, " at the utmost, a tenth, 
 of my comparisons, I am willing to surrender as still uncer 
 tain. Even then, if you take into account formations from 
 radical \vords, derivatives and compounds, more than half the 
 entire stock of words in Latin is shown to be Greek, and a 
 key furnished for deciphering the rest. Quod erat dcmon- 
 strandiuii^\\ concludes with triumphant emphasis. Now if 
 Professor Ross had not been so averse to a little knowledge 
 
 o 
 
 of Sanskrit, a cursory examination of Bopp s Glossariitm 
 Sanscritum would have shown him that the Latin words cited 
 in that work, as more or less probably connected with words 
 found in Sanskrit, make a list little less numerous than that 
 of his index a list which Bopp could easily have doubled, 
 if his philological principles had been of the same easy and 
 accommodating character as our author s. And yet, were it 
 even quadrupled, Bopp would not have drawn the inference, 
 which according to Professor Ross would be inevitable, that 
 the Latin language w r as derived from the Sanskrit. He would 
 only have seen in it additional proof of a fact, too clearly 
 proved before to need confirmation, that the Latin and the 
 Sanskrit are both descendants of a common parent. Profes 
 sor Ross must of course be a\vare that two languages may 
 have a close and pervading lexical resemblance, without 
 either of them being derived from the other. For French 
 words we have to a great extent similar words in Latin : and 
 the French language is in fact derived from the Latin. But 
 again, for French words we have to a great extent similar 
 words in Spanish : yet the French language is not derived 
 
148 ITALIC ANS AND GREEKS. 
 
 from the Spanish. The Scottish bane is not of necessity de 
 rived from the English bone : in fact, there is no reasonable 
 doubt that both have arisen independently from the Anglo- 
 Saxon ban. The Latin gcnu is not of necessity derived from 
 the Greek yovv : it is at least equally supposable that both 
 may have arisen independently from a third form say ganu 
 which we find, though with a degradation of the initial gut 
 tural, in Sanskrit y^//?/. If numerous resemblances exist be 
 tween the Greek and Latin, there are three possible supposi 
 tions to account for the fact. We may suppose, I. that Lat 
 in is derived from Greek ; 2. that Greek is derived from 
 Latin ; or, 3. that Greek and Latin are separately derived 
 from a common source. Professor Ross adopts the first of 
 these suppositions as the true one : to prove it true, he was 
 bound to show, not only that the resemblances are numerous 
 for so they arc between French and Spanish but that they 
 are of such a kind as to require the first supposition rather 
 than the other two. It is a little surprising that he should 
 have overlooked this obvious and capital necessity of his ar 
 gument. He has contented himself with showing that Greek 
 and Latin words are related in form, assuming in general, 
 without attempt at proof, that the Latin form was made out 
 of the Greek by some euphonic alteration or corruption. 
 Yet in many of these cases it is improbable, to say the least, 
 that the Latin form should have had such an origin ; and in 
 not a few, it is clear enough that the Latin stands nearer than 
 the Greek to what must have been the primitive form. We 
 will give one example from the multitude which offer them 
 selves to our hand. Professor Ross says in express terms, 
 " The usual origin of the Latin j is directly (gcradczii) from 
 the Greek f." As one instance he gives us jngum, yoke 
 (pronounced yuginri), from the Greek wyov ; and at the same 
 time refers, apparently with approbation, to a passage in 
 Plato s Cratylus, in which gvyov is derived from Svo and ayaiyi i, 
 as that which binds two animals for drawing. Probably, 
 however, he would lay but little stress on this etymology, 
 which is evidently conjectural, and is far from being support 
 ed by analogy. But what shall we say as to the main point : 
 arisen out of %vyov ? The thing, it must be owned, 
 
ITALIC ANS AND GREEKS. 1 49 
 
 has very much the appearance of a phonetic miracle. How 
 a ^tf-sound should transform itself to ya, it is not easy to im 
 agine. We assume here that the Greek f had the ordinary 
 sound of English z, for the real sound cannot have been far 
 remote from this, and Professor Ross is fully satisfied that this 
 and nothing else was the real sound. The opposite change 
 from ya to sa would not be difficult to understand ; for the 
 ya-sound is* closely connected with the palatals cha,ja, which 
 contain a sibilant element and readily pass into sibilants. It 
 is moreover attested by indisputable cases : thus the Latin 
 fuguni y which preserves its initial sound in Spanish yugo, be 
 comes giogo in Italian, with a palatal, and joug in French, 
 with a sibilant at the beginning. Can any cases equally in 
 disputable be adduced for the change of za to ya? for if it 
 can be shown to have taken place, we must believe it possi 
 ble, even though we cannot understand the rationale. Our 
 author presents no such cases, beyond two or three similar 
 parallels between Greek and Latin words. Of these we need 
 not notice Latin jungo referred to Greek Qvyvv/jM, Latin jugo 
 to Greek fixyoo), Latin conjnx to Greek ervfvjf, which present 
 the same root as juginn &yov, and .of course only repeat the 
 same question. The connection of Latin jus, juris, broth, 
 with Greek ^69, * broth, cop6v, pure wine, is made pretty 
 doubtful by the difference of form in one case (fw/^o?), and 
 the difference of meaning in the other (copov) : if admitted, 
 we must still ask which was the primitive sound,/ or s. The 
 comparison of Latin jejunus, fasting, with Greek e#&>z/o?, 
 well-girded scilicet, because a fasting man can more easily 
 clasp his girdle if made by an opponent of Professor Ross, 
 would have seemed a broad burlesque of his view. Latin 
 major is compared with Greek jjiei^wv : but here it is evident 
 that j has not arisen from f, nor vice versa, but bothy and f 
 from a combination <yi, gi, or gj \ fjuei^cov is for peyiwv (com 
 pare e%3-p6s, ej$tW), and major is for magior, magjor (com 
 pare the adverb magis, which retains the guttural). There 
 remains Jupiter, Greek Zevs Trarrjp. But here it is plain that 
 j and f are not derived one from the other ; but both from a 
 common St, dj y which appear in the Greek Genitive AIDS, 
 ) and in the Italican DiuvciJovi, to say nothing of the 
 
150 ITALIC ANS AND GREEKS. 
 
 Sanskrit dyaus, Gen. divas, heaven, which corresponds 
 with wonderful exactness to the Greek Zeu?, AI(F}Q^. It thus 
 appears that the Greek and Latin parallels by no means re 
 quire us to assume, notwithstanding the phonetic difficulty, a 
 change from ?to/. If other languages gave no testimony on 
 the subject, or if, with Professor Ross, we closed our ears to 
 their testimony, it would still be probable that the initial f in 
 wyov,is of later origin than the/ vijugum. But the moment 
 we turn to look at other languages, it becomes impossible to 
 doubt. Even our author would probably shrink from main 
 taining that the Sanskrit yuga, the Gothic juk(a), the Lithu 
 anian junga-s, and the Slavonic igo, all show in their initial 
 sound a corruption of the Greek f. 
 
 A remarkable error which Professor Ross has made in this 
 connection illustrates the danger of indulging too far in an 
 ignorant contempt for Indo-European philology. Though he 
 regards the / of Jupiter as a corruption of the f in Zevs, he 
 justly recognizes thejofyovts as having arisen from dj ; and 
 in like manner identifies, no doubt with correctness, the divine 
 names A IWVTJ in Greek and Juno in Latin. But unhappily it 
 occurs to him to go a step further in the same direction, and 
 identify the Latin juvcnis, youth, with the Greek Sioyevij*;, 
 akin to Zeus, which occurs so often in Homer as an epithet 
 of princes. Now here is at once a difference of form and a 
 difference of meaning, either one of which might have de 
 terred a less adventurous and uncompromising philologist. 
 Apart from the initial sounds, there is a variation of ending 
 the Greek stem being Sioyeves, not Stoyev, while the Latin 
 stem isjuven (seen in Gen. Plur.jttvtn-um), or, by a frequent 
 assumption of i,jnveni (hence Nom. Sing, juvcni-s). Pro 
 fessor Ross, however, regards it as weak-minded to attach 
 importance to such variations of ending or inflexion. Then 
 the change of 7 to v might occasion some uneasiness to weaker 
 consciences. True, we ourselves have guard and ward, guile 
 and wile, guaranty and warrant} . But in these, iu was the 
 original sound, which first strengthened itself by a prefixed 
 guttural, but afterwards fell away : thus ward y gward y g(ti)ard. 
 A change in the opposite direction is not so lightly to be as 
 sumed. Yet if it were altogether common, we should feel 
 
ITALIC ANS AND GREEKS. I 5 I 
 
 some surprise at finding it just here. For the Latin has the 
 root gen in gigno, genus ; and developes it, not less widely 
 than the Greek, in a long series of derivatives and compounds ; 
 it might be expected, therefore, in this word as well as in 
 others, to preserve the well-known root in its pervading and 
 familiar form. And, again, the connection of meaning be 
 tween * one of Jove s kin on the one hand, and a young 
 man on the other, is hardly close enough to tempt an ordi 
 nary philologist to encounter the difficulties of form just 
 enumerated. Indeed, it would seem as if Professor Ross 
 himself were not entirely at ease about it. For he adds, in a 
 tone of confidence which sounds a little hollow, "Forjuvenis, 
 I think I have given the correct explanation." Now here 
 again the cognate languages supply us with testimony, which 
 of course Professor Ross rules out of court, but which, if ad 
 mitted, must carry overwhelming and decisive weight. The 
 L.a.tmjuvenis reappears with the same meaning and with little 
 difference of form in nearly all the Indo-European languages. 
 The Sanskrit stem yuvan may be pronounced identical with 
 juvcn in the Gen. Plur. juven-um : for the short and 6 of 
 other languages are represented in Sanskrit by short a, that 
 language having but three short vowels, d, (, u. The San 
 skrit-stem has moreover in part of its inflexion the contract 
 formyjiJi, which is completely identical with the root of the 
 Latin comparative junior, as also with the Slavonic jnn t 
 youth. The Lithuanian has jauna-s ; the Gothic, juggs 
 (=jung-s)\ and even the Welsh presents us icuanc, youth, 
 iaiiy younger. We have here a word which must have be 
 longed to the primitive Indo-European language, since we 
 find it in unmistakable identity from the banks of the Ganges 
 to those of the Severn. The Greek only has suffered it to 
 drop, supplying its place by another word not less early in its 
 origin and not less extended in its diffusion, the adjective 
 ueo?, or rather Z^FO?, new. A somewhat similar case of omis 
 sion is presented by the Latin in the words for son and 
 daughter. The prevailing Indo-European designation for 
 son comes from a root su, to beget, found in Sanskrit. 
 Hence Sanskrit sunu and siita, son, Greek via? for o-f-to?, 
 Gothic sunu-Sy Lithuanian suna-s, Slavonic syn. That for 
 
152 ITALIC ANS AND GREEKS. 
 
 daughter comes, it should seem, from the root duk, to 
 milk. Hence Sanskrit duhitar, Greek QvyaTyp, Gothic 
 daulitar, Lithuanian dukte(r), Gen. duktere(s}> Russian docj, 
 Irish dear. These, now, the Latin has discarded, substituting 
 for \\\cmjili7is and Jilia, which have been generally regarded 
 as derivatives from the root seen in Greek c/uXo?, * dear. This 
 derivation, it must be confessed, is by no means certain ; 
 though it seems to be more probable than the one proposed 
 by Professor Ross, who takes filins and Jilia from the Greek 
 (frvXrj, as if for (frvXios <f>v\ia, with the meaning one of the 
 family, male or female. But Professor Ross goes further, and 
 derives the Greek via? itself from the same cpuXto?. In this 
 instance he departs from his usual practice : for he has to 
 admit here that the Latin filius is far nearer to the supposed 
 original form, shows vastly less corruption, than the cognate 
 Greek vlo$. Probably the admission seems to him harmless, 
 so long as the original form is seen to be genuine and unques 
 tionable Greek. But the manner in which he makes u/os out 
 of (f)v\Los is too characteristic to be omitted. First, the aspi 
 rate <f> passed into the spiritns aspcr into h. For this change 
 he adduces no parallel instance from the Greek, but finds it in 
 the Latin herb a for Greek fopfiij, feed, a comparison already 
 made by others, but of doubtful correctness. He finds it 
 also, and no doubt correctly, in the Spanish Jiablar from 
 fabulari, liicrro from fcrrum, hi jo fromjfi/ius. Thus," he 
 says, " the Greek made i/Xto? from <pv\io$ : then \ was elided, 
 and the i which arose in its place was absorbed by the pre 
 ceding v. Hence [v?o? or] with change of accent, uto?, to 
 distinguish the word from i/eto? [ belonging to swine ]. The 
 Italian," he continues, "retained the old full form films = 
 <j>v\tos ; but, as if to leave no doubt as to the identity of the 
 word, the elided Greek form returns again letter for letter in 
 the Spanish hijo = v/o?. Can all this," he concludes with 
 natural exultation, "be made out and explained from the 
 Sanskrit ? " Alas, no we are forced to confess, at the risk 
 of degrading the Sanskrit still further in his eyes no, it can 
 not. But just as little can it be made out from Greek and 
 Latin. It rests mainly on the Spanish. For the change of 
 (j> to h no Greek example is given ; for the disappearance of A 
 
ITALICANS AND GREEKS. I 53 
 
 between two vowels, no Greek or Latin example/ But the 
 first occurs in Latin, and both are frequent in Spanish ; and, 
 therefore, we need not hesitate to recognize them, both at 
 once, in a single Greek word. 
 
 We have said that it is the general practice of Professor 
 Ross, in his comparisons of words, to give priority to the 
 Greek form, and to represent the Latin as made from that by 
 additions, retrenchments, and other alterations. It would 
 even seem as if he regarded such words as sex, sedes y 
 somnus, etc., as made from ef, eSo?, VTTVOS, by prefixing an s / 
 and words like vimtm, volvo, etc., as made from ol^o?, etXww, 
 by prefixing a v. For though in his lists of these words he 
 puts the Latin form first, and thus appears to give it the pri 
 ority, yet he tells us further on, that "the Italican idioms 
 sometimes, though not often, prefix to the Greek vowels 
 other consonants than s or v" which w r ould seem to imply 
 that, after all, the s and v were really nothing but prefixes 
 added to the corresponding Greek words. We are willing, 
 however, to regard his expression in this case as an unob 
 served inaccuracy of language. For an opinion so perverse 
 ought not to be fixed on any scholar without the clearest 
 evidence. But there are cases beside that of wo? from (^uXto?, 
 in which he distinctly acknowledges corruptions in the Greek 
 form which do not appear in the Latin. Thus he admits that 
 the Greek has dropped an ;/ in the common verb e aw, to 
 permit, and the Homeric adjective ei/5, good (whence the ad 
 verb e, well ). These words he identifies with Latin sino and 
 bonus. Whether he considers the s and b as added by the 
 Latin or dropped by the Greek, does not distinctly appear. 
 Nor in regard to the ;/ do we clearly understand why he 
 speaks of it more than once as having been restored to these 
 words by the Latin. Can he suppose that the Latin received 
 the words in the forms seao and beiis, or sio and bous, and 
 reinserted the n which had before fallen out ? How came 
 they to put in the precise letter which had been lost, rather 
 than some other one ? Or is this only a further instance of inac 
 curacy in expression, similar to that which charity compelled 
 us to assume a moment since ? As to the identity of law and 
 sino, of eifc and bonus, we will allow it to rest on Professor 
 
1 54 ITALIC ANS AND GREEKS. 
 
 Ross s authority. We can more confidently indorse his com 
 parison in another case where he admits the corruption of the 
 Greek form : we refer to the genitives >yeveos in Greek and 
 generis in Latin, from the nominatives yevos, genus. Yet even 
 here he does not escape from his besetting inaccuracy of ex 
 pression : for he describes this as one of the cases in which 
 " the Latin inserts a liquid, /, ;/, or r y between two vowels 
 which liquid," he continues, "must therefore have existed 
 originally in the Greek form." Why, if it existed in the ori 
 ginal Greek form, the Latin should have been compelled to in 
 sert it, docs not appear. But perhaps our author only means 
 that the Latin has it there and the Greek has not though at 
 an earlier period, as he thinks, the Greek also had it. But in 
 this point again we cannot fully concur with him. For we 
 think it certain that the Greek yeveos never at any time had 
 an r in it : that it has arisen, not from 76^6^09, but from 
 7eyec709. In proof of this we shall not appeal to Sanskrit 
 neuters in -as, Gen. -asas, which correspond to Greek nouns 
 in -09, -eos 1 : as manas, Gen. manasas=- Greek yu-eVo?, Gen. 
 yLte^eo? (for yueyeero?). Professor Ross is proof against proof 
 of this kind, however convincing to others. But there 
 is no need of it here. We have only to mention three 
 facts which our author himself would not think of dispu 
 ting : i. The Greek Nom. 7^09 has s, which is most natu 
 rally explained as belonging to the stem, since neuters of the 
 third declension have no case-sign in the Nom. Sing.; 2. The 
 Greek, while it retains p between two vowels, has in a multi 
 tude of cases dropped cr standing in that position : as riSeai,, 
 TiQr), for TiQeacu ; 3. The Latin, though it does not drop s 
 between two vowels, has in a multitude of cases changed it 
 to r: as in Lares, earlier Lascs (Carmen Fratrum Arvalium) ; 
 crant, Greek rjcrav (r). 
 
 Our author s eagerness to derive every Latin form directly 
 from a Greek one, coupled with his fixed resolution to know 
 nothing of Sanskrit, has betrayed him into many palpable 
 errors. Even where the words which he compares are really 
 connected, he often misconceives the relation between them. 
 Thus he brings the Latin bibo directly from the Greek irivo) t 
 assuming that both TT and v are changed to b. A change 
 
ITALICANS AND GREEKS. I 5 5 
 
 of TT to b involves no particular difficulty : but the change of v 
 to b is not so easily admitted, and even our author calls it 
 vercinselty isolated. He seems not to have observed that 
 bibo has the appearance of a reduplicated form, analogous to 
 sisto. In the latter verb, sta, the stem-syllable, by redupli 
 cation becomes sista, which then allows its final a to be treated 
 as a mere connecting vowel, and thus conforms to the third 
 conjugation. The stem of TTLVW, TTO (as we find it in TreTrw/ca, 
 TrofoY), if treated in the same manner, would give TTLITW or 
 bibo. This derivation of bibo, which would appear, even 
 without reference to the Sanskrit, a plausible hypothesis, is 
 converted into absolute certainty when we turn to that lan 
 guage : for we find there that the root/, to drink, follows 
 the third or reduplicated conjugation, though instead of the 
 regular pipdmi, it makes Present pivdmi, and in the Vedic 
 dialect pibdmi, thus approaching very closely to the Latin 
 form. 
 
 Again, our author in many cases overlooks a certain or prob 
 able connection between the two languages, substituting for it 
 a superficial and illusory comparison. Thus the Latin cere 
 brum is set down as a corruption of the Greek ey/ce(j)a\ov, 
 brain ; though we must say, in justice to Professor Ross, 
 that he adds an expression of doubt. But surely, if he had 
 not been so intent on making every Latin word to be an 
 actual Greek word, only more or less corrupt, he would have 
 connected cerebrum with /cdpa, head, even if he did not ex 
 plain it with Bopp as a compound of cere, Kapa, Sanskrit 
 girds, with fero, (frepco, Sanskrit b/iri, meaning head-borne, 
 carried in the head equivalent to ey/ctyaXov, though wholly 
 distinct in etymology. 
 
 The Latin luna, l moon, is repeatedly identified with the 
 Greek creX?^. Now we have already noticed the relation of 
 Latin sex and Greek ef, serpo and epTrco, salio and a\Xo//.a6, to 
 which fifty more might be added of the same kind. We see 
 how tenaciously the Latin preserves the initial s, and how 
 readily the Greek sacrifices it, unless connected with another 
 consonant, as in (TKCLTTTW, c-reXXw, etc. If a is retained in avv 
 and its compounds, the reason is obvious in the earlier form 
 fw, where we find it connected with K. That (j^kr\vt] itself has 
 
I 56 ITALICANS AND GREEKS. 
 
 preserved the initial o-, is owing probably to a consonant F, 
 which must originally have followed it ; for it coincides in 
 root with Sanskrit srar, another form of sur, l to shine. 
 Yet along with a-\tfvr), we have the personal, or rather the 
 divine name r E\evrj (moon-goddess), which is probably the 
 same word, but with the ordinary Greek mutation of initial 
 a to h. If, therefore, luna had been the Greek form, 
 and o-\ijvrj the Latin, the connection of the two might have 
 appeared more plausible ; as it is, we must regard it as in a 
 high degree improbable. But, if unconnected with each 
 other, neither of the two is without connections in the other 
 language. Luna is for Inc-na (compare liuiicn for luc-men), 
 connected with lux and In ceo in Latin, and with Aeu/co?, 
 AeiWw in Greek. And o-eX^vrj, on the other hand, is con 
 nected with c-eXa?, brightness, which is clearly akin to 
 the Latin sol. These obvious relations our author has over 
 looked, in consequence of his determination to connect every 
 Latin word directly with a Greek word of the same meaning. 
 The Latin opus is represented as identical with the Greek 
 em)?, word, the exact correspondence of form find inflexion 
 being regarded as decisive proof of identity, notwithstanding 
 a pretty wide difference of meaning. Why our author should 
 lay so much stress here on coincidence of form is not quite 
 easy to understand : for with the almost unlimited range 
 which he allows to mutations of form, such coincidences might 
 easily arise by accident in words originally different. And in 
 fact he himself says elsewhere, in opposing some etymologies 
 of Mommsen, that Mike-sounding is not necessarily of like 
 sort" (gleichlatitendes nicJit nothwendig gleichartig ist) : and 
 proceeds to illustrate the maxim by giving three different ex 
 planations to the sul in consul, cxsul, and prcesul : prcesul 
 prce-silicns, leading in the dance ; ex sul=ex solo, driven 
 out from his native soil ; while consul is only a corruption of 
 the Greek crvfjL/3ov\o<$ ! %vv and con are one preposition, 
 and /3ciA,o?, by a simple change of {3 to s, comes very near to 
 sul. But in speaking of opus and ezro?, he says that " where 
 letters and inflexion both coincide, we cannot refuse to 
 acknowledge the identity ; " and he mentions Latin sidus, 
 1 constellation/ and Greek etSo9, form, figure, as another in- 
 
ITALICANS AND GREEKS. 157 
 
 stance of the same kind. Unluckily the supposed coinci 
 dence does not exist in either case. EZSo? is not for cret8o? 7 
 but for ^etSo9 : the original F is abundantly proved by the ap 
 pearances of the Homeric verse, and by the close relation to 
 elSov, I saw/ where Professor Ross justly recognizes it, com 
 paring Latin video. So, too, evro? is really FETTO^ : no word in 
 Homer presents more unquestionable traces of the initial F, to 
 say nothing of inscriptions which show us the letter itself. 
 But FeTro? cannot properly be said to coincide in form with 
 opus, unless we are willing to say that Xe-rro?, peel, or mvro?, 
 glen, coincides with opus : and even in meaning these are 
 not much further away than- 677-09. We will venture here to 
 offer a suggestion. A Sanskrit root like dp, Professor Ross, 
 we fear, would reject with scorn. After deriving mas, male/ 
 from Greek /jt,dj(\o$, wanton, by omission of % and change 
 of \ to r (though, by the way, masculus shows the Latin stem 
 to be, not mar, but mas), he says to his correspondent, " if 
 that won t do, make me another mas : but, mind you, he 
 must be Greek (aber wohlgemerkt , griechisch muss er seiii}" 
 Now we wonder that he did not find his Greek opus in /COTTO?, 
 which has the meaning toil, fatigue. The decapitation of 
 the word (if I may so term the suppression of initial K) he 
 could defend by Latin apcr, which he identifies with Greek 
 /caTrpos. And the neuter ending us (36 decl.) instead of mas 
 culine 09 (2d decl.), if it gave him any trouble, could be sup 
 ported by Latin corpus, carports, body, which he derives 
 from Greek /copras, Kopfiov, a block cut from a tree. And 
 yet it is almost a pity to say anything against the eiro^-opus 
 etymology : for thereby hangs an interesting piece of literary 
 history, which we give as we find it. "According to these 
 analogies, opus is nothing but eVo?, and we gain the conclu 
 sion, certainly not unimportant for Greek literary history and 
 the question as to the antiquity of Greek literature, that, 
 in the earliest period, 67709 had passed from the meaning 
 word to the meaning work of poetry, and hence work 
 in general : for only thus could it come to the Italican opus. 
 As, therefore, in the last centuries opera, work, has passed 
 into the opera of music and poetry, so in opus the opposite 
 change must have taken place from three to. four thousand 
 
158 ITALIC ANS AND GREEK$. 
 
 years ago. - But we must add that eVo?, though unconnected 
 with opus, has its kindred in Latin. The same root reir ap 
 pears in the aorist elvro^. Homer ereiirov for erereTrov ; and, 
 with a regular variation of vowel, in the noun oSjr, that is, 
 FOTT-S, voice. But no one who observes how often the 
 Greek has substituted a labial sound for a primitive guttural 
 can hesitate to identify FOTT-? with Latin vox, voc-is, whence 
 the verb voco, vocarc. This unquestionable connection is over 
 looked by our author, who derives voco from Greek fiodco, to 
 cry. He must certainly have forgotten the Latin boo, boare, 
 which corresponds exactly in form and in meaning to Greek 
 /3oda). Both these have a common derivation, either directly 
 from /SoO?, bos ; or, more probably, from a root /3u=Sanskrit 
 gu, to sound, from which /3ov$ bos are also derivatives : in 
 any case they are radically distinct from voco and elirov. 
 
 We will mention further an interesting, not to say amusing, 
 derivation of populus. Professor Ross considers it as a 
 diminutive of TTOTTO?, a word which occurs only in the Homeric 
 exclamation a) TTOTTQI. This has been regarded generally by 
 recent grammarians as a mere interjection ; but our author 
 contends, not without force, from its connection with co, and 
 its appearance only at the commencement of addresses, that 
 it must have been a real vocative. But he discards the ex 
 planation of ancient grammarians, who make it mean Sal/nove?, 
 divinities, and adopts out of his own head the meaning 
 men, people. It thus serves him a double purpose, not 
 only accounting forflofluhis, but also for ////// j, pupa, pupnlus, 
 a. boy or girl, a mannikin or doll. At the same time, he 
 does not altogether discredit the statement of Plutarch that 
 the Dryopes used TTOTTOL for Saijjioves, divinities ; and that of 
 a grammarian in Bekker s Anecdota, that the Scythians used 
 it for cvyd\/4a,Ta, images. He suggests in explanation that, 
 as the oldest idols of the Dryopes were certainly pretty rude 
 and inartificial, people may have given them the name of 
 inannikins ; and that the Greeks may have ridiculed the 
 Scythian images by the same disparaging epithet. And if 
 Zeus in Homer commences a speech to the divine assembly 
 on Olympus with o> ironroi, our author looks upon the address, 
 *O folks (ok leittchdi) as a flattering expression of familiar 
 
IT^LLICANS AND GREEKS. 159 
 
 confidence (vertraulich und schmeichelnd). He fails to ob 
 serve that c5 TTOTTOL is used by Homer even in addresses to a 
 single individual, showing that it must already have lost its 
 meaning, if it ever had any, as a plural substantive, and sunk 
 down into a mere exclamation. Nor does it strike us as very 
 probable that the Dryopes would call their sacred images 
 puppets, or that the Italicans would designate the collective 
 nationality as a little man. Considering these improbabili 
 ties, as well as the total absence of direct proof for the mean 
 ing men given to TTOTTOI, we shall adhere for the present to 
 the etymology which makes populus a reduplicated form of the 
 stem which appears in TrXeco?, full, 7rA?}$o9, multitude, and 
 perhaps in 7ro?U9, city. 
 
 It must be evident already that Professor Ross is one of 
 those resolute philologists who are not easily turned aside 
 from their purpose by vowels and consonants. He laughs 
 good-naturedly at his weaker brethren, who trouble them 
 selves about anlaut, inlaut, ablaut, umlaut, and all the other 
 lauts. For himself, he has no idea of paying too much 
 respect to letters and syllables ; he has in fact the lowest pos 
 sible opinion of their steadiness. This is a result of his exten 
 sive and valuable travels in lands and islands where Greek is 
 spoken. The endless variations of the local dialects have im 
 pressed him deeply with the extreme mutability of spoken 
 sounds, and have thus given him what he regards as a peculiar 
 qualification for his present undertaking. He commences his 
 Preface with a long list of corrupt forms, brought together 
 out of the various and widely separated patois of the modern 
 Greek : and he has no difficulty in admitting such corruptions, 
 any or all, as means for effecting the proposed transmutation 
 of Greek into Latin. In particular the modern Greek /capa- 
 ^>X6? ? bald, for the ancient (f>a\aKpds, has convinced him that 
 the order in which the letters of a word may stand is pretty 
 much a matter of indifference, and that so long as you have 
 all of them in one aggregate, or most of them, or others more 
 or less resembling most of them, you need not concern your 
 self about the particular succession. The transposition ol 
 vowels and liquids metathesis is an ordinary and familiar 
 phenomenon of language. But our author s transpositions 
 
160 ITALICANS AND GREEKS. 
 
 take a far .wider range. Thus vcrto is derived from rpeTra) by 
 spelling the root backwards, ipeir pert, and by change of/ 
 to , ff/ A The Latin word is thus made to illustrate its own 
 meaning by a complete turn-about : and if the Sanskrit has 
 the same form vart. Present vartumi, it may perhaps be 
 ascribed to a certain original perversity in that mischief-mak 
 ing language. The adjective ra^i;?, swift, by a similar pro 
 cess is converted into ^<ZTU?, which readily furnishes citns : 
 though the Latin word, if we could suppose it endowed with 
 consciousness, might feel some surprise at this sudden disrup 
 tion from do, to move, stir, with which it has been so long 
 and so comfortably connected, and the total revolution which 
 it is compelled to undergo. The Latin famulus is derived by 
 transposition of//, and \, and by interchange of rough mutes, 
 from Greek $aXo.yuo?, chamber. That a valet-de-chambre 
 should be called simply and shortly chamber, is perhaps not 
 incredible ; yet we must confess a preference, on the score 
 both of form and meaning, for the derivation from facio 
 famulus for fac-mnlus, with the same ending and the same 
 omission of a guttural as in stimulus for stig-mulus, which 
 Professor Ross refers, probably with good reason, to the root 
 of Greek o-Ti fw, to stick. In some cases we should be much 
 at a loss if our author were not kind enough to explain his 
 processes. In 8njrao>, to thirst, he assures us (we quote his 
 words) that " the IT falls out, the a and S are transposed and 
 thus we get o-iSday, Latin st tio." In alcr&dvofjiat,, to perceive, 
 the Future ala-S-ijaropai gives evidence of a Present aio-Qew, 
 which is supposed to be for ava&eco : and this, by transferring 
 <T from the middle to the beginning of the word, gives cravSeay, 
 from which any one can easily find his way to Latin sentio. 
 In other cases the explanation itself requires to be explained. 
 For the Greek lyvs, fish, he assumes a change of %$ to i/r, 
 giving f\|rf9 (which has remained, he thinks, in the common 
 word o^rov, fish, or flesh eaten with bread ), and then he 
 adds: "From n/w? the Italicans by transposition formed 
 their Disci s." We have tried to spell the word ityvs in every 
 possible order of the letters : but we have not yet succeeded 
 in making out the form piscis. Manage as we will, there is 
 always a K=C wanting. May it not be that the Italican se- 
 
ITALICANS AND GR 
 
 cured his piscis by an ingenious combinatioi 
 with the supposed n/ru?, getting his / from one and his c from 
 the other ? And were the Goths then obliged to go through 
 with the same process, which certainly is a little complex, 
 to form their fisk(a)s, which corresponds exactly to the Latin 
 piscis ? 
 
 We will only add two or three cases to illustrate further the 
 startling originality and the bold defiance of difficulties which 
 characterize Professor Ross as an etymologist. The Latin 
 (minis, river, he derives from Greek de\\a, blast ( stream 
 of air, and hence stream of water ) ; and in like manner 
 omnis, all, from Greek doXXrj?, collected together ; in 
 both he assumes, without other examples, the rather remarka 
 ble euphonic change of XA into mn. Amo, to love, he derives 
 from Greek dyajrw, through a supposed intermediate form 
 ayTTo), which most philologists would regard as still inconve 
 niently and discouragingly remote from amo. Auric io, to 
 clothe, he derives not from am andjacio, to throw around, 
 but from Greek afifaevvvfii,. The prefix am is doubtless con 
 nected with dfj,<f) : but few philologists would venture to de 
 rive icio from evvvpi (or rea-vv^t,) ; even our author only 
 states the fact, without adding an explanation. Of course he 
 could give one, if called upon : did he think it too easy and 
 obvious to require insertion ? He does better for us, when 
 he derives duco, to lead, from Greek yyeofjiat, : for he gives 
 us the Etruscan lucumo, derived from Greek qye/jLcoV) as a key 
 to unlock the enigma. The /, he tells us, takes the place of a 
 spiritus aspcr : but / and d are often interchanged (as lacrima 
 = SaKpv/jLa) ; and the vowel-change is supported by luna from 
 creXtfvT) : and thus -all is clear. 
 
 It must not be supposed from what we have said that Pro 
 fessor Ross recognizes no such thing as fixity or permanence 
 in language. Whatever else may change, one thing has re 
 mained constant from the beginning, secure "amid the wreck 
 of letters and the crush of words;" and that is the modern 
 Greek pronunciation of vowels, consonants, and diphthongs. 
 " I consider it," he* says, " as being from the time of Inachus, 
 or whatever else may be more ancient than Inachus, the only 
 correct pronunciation." All honor to courageous faith ! It is 
 1 1 
 
1 62 ITALIC A NS AND GREEKS. 
 
 something, as the world goes, to believe in the historic reality 
 of Inachus, a millennium earlier than the earliest known mon 
 uments of the Greek language. But it is much more to be- 
 lieve that the Greeks, when they first adopted the Phoenician 
 alphabet, added a v only to express a sound already ex 
 pressed by i ; and that, having afterwards applied the T; to ex 
 press the long quantity of this same sound, they were not yet 
 satisfied, but took up ei, rji, ot, ut, as four additional represen 
 tatives of a sound which had three representatives already : 
 and then that the old Greek grammarians should have made 
 the enormous blunder of describing this sound, when repre 
 sented by 77, as a prolongation of e : with much more of the 
 same kind, which is inseparable from a belief in the primitive 
 antiquity of the modern Greek pronunciation. It is not 
 surprising, perhaps, that early prepossession and patriotic en 
 thusiasm should lead the modern Greek to credit such mar 
 vels : though in a distinguished member of our own Society, 
 who has written on the ancient pronunciation of his language, 
 we see that such influences are not always proof against in 
 vestigation and reason. But we should hardly have expected 
 that Professor Ross s travels in Greece, extensive and useful 
 as they have been, would so far transform him into a native 
 Greek. 
 
 We have seen that our author, though always disposed to 
 regard the Latin word as the corruption of an extant Greek 
 word, is yet obliged to acknowledge that in some cases the 
 extant Greek word is the corruption of a form more faithfully 
 preserved in Latin. Nor is there any good reason why he 
 should not have done this much more frequently. For he 
 holds that the separation of the Greek and Italican races oc 
 curred at a very remote period, ages before the time at which 
 the Greek language first becomes known to us. He sup 
 poses that the Greek colonization of Italy may have com 
 menced some two thousand years before the Christian era, 
 and perhaps far earlier. And though he regards it as having 
 been in progress for several centuries, he intimates that 
 nearly all the words which passed from Greece to Italy must 
 have been carried over before the Greek itself was reduced to 
 writing. And the Greeks, he tells us, had the art of writing, 
 
ITALICANS AND GREEKS. 163 
 
 and used it, for fifteen centuries before Christ. Professor 
 Ross, it appears, believes in Cadmus as well as Inachus, and 
 perhaps in tradition generally : he thinks it very hard that 
 Daedalus, for instance, should be stripped of his individual 
 historic personality, and made to be only what his name 
 means, a * cunning artificer/ merely because he was unfortu 
 nate enough to live a few centuries earlier than some other 
 men. Possibly a respect for tradition may have led him to 
 assign a date so early for the Italican migrations. For he 
 might naturally think that a body of tradition, which he ac 
 cepts as reliable, would have included these migrations, if 
 they had not been in great part anterior to its beginning. 
 Let us then take the year 2000 B.C. as a convenient middle 
 term between the extremes of time suggested by his lan 
 guage : his theory would be, that two thousand years before 
 Christ, be the same more or less, the Greek language was 
 carried into Italy. But what, we may ask, was the Greek 
 language of the year 2000 B.C. ? It might be hard to tell ; 
 but it would be easy to see that it was something widely dif 
 ferent from the language of Homer. For if we carry Homer 
 back to the year 900, and he cannot with probability be car 
 ried further, he is still separated from 2000 B.C. by a broad 
 gulf of eleven centuries. What was English eleven centuries 
 ago ? Something quite as hard for us to understand and 
 learn as modern German. Who from the English of the 
 present day, or the French or German of the present day, 
 could without other helps form any tolerable conception of 
 those languages as they were eleven centuries ago ? Our 
 author, it is true, appears to regard the Greek of 2000 B.C. 
 as not very different, either lexically or grammatically, from 
 the literary language. The division of dialects he conceives of 
 as already established at that early period. But what assu 
 rance can he give that even the grand distinctions which sepa 
 rate Ionic and Doric the rj for long a among vowels, 
 and the a for a primitive r among consonants were developed 
 eleven centuries before the time of Homer ? Yet he goes so 
 far as to point out Laconian forms in the Italican idioms, and 
 asserts that the Sabines were mainly Laconians. The philo 
 logical reason for this statement, which has no support in tra- 
 
1 64 ITALIC ANS AND GREEKS. 
 
 dition, must be that some Sabine words are thought to re 
 semble the peculiar forms of the Laconian dialect more closely 
 than they do the other dialects of Greek. But many Latin 
 words have a closer resemblance to the Sanskrit forms than 
 to those of any Greek dialect : as scptcm, Sanskrit saptan, 
 Greek eTrrd ; jiigum, Sanskrit yugam, Greek vydv ; jeciir, 
 Sanskrit yakrit, Greek rjjrap ; corvns, Sanskrit kdravas, 
 Greek icopaj; ; and a hundred others. Shall we say then that 
 the Latins were mainly Hindus ? 
 
 But the comparative philologist can prove conclusively that 
 the Greek of 2000 B.C., assuming that the Latin was derived 
 from it, must have differed immensely from any known Greek 
 dialect of the historic time. For in cases where the Latin 
 agrees with the prevailing type of the Indo-European lan 
 guages and the Greek departs from it, this departure, it is 
 plain, could not have belonged to that supposed Greek lan 
 guage which was imported into Latiurn. To assert that the 
 original type was first abandoned in Greece and afterwards 
 restored in Italy (as our author appears to do in one or two 
 places, if his language is to be strictly construed) is a gratui 
 tous assumption, and an improbable one. We can only bring 
 forward here a few prominent facts to illustrate the mode of 
 reasoning just indicated, and for convenience we will give 
 the name " Grseco-Italican " to that primitive Greek (if Greek 
 it was) from which the Italican idioms are supposed to have 
 sprung, and which our author imagines to have been the 
 spoken language of Greece about 2000 B.C. We say then 
 that this Grseco-Italican must have preserved the initial s 
 before a vowel, which all the Greek dialects have changed 
 into // / thus it must have had sex not ef, scrpo not epTrco, 
 sistdmi not tcm7/u, and many others. So too it must have 
 preserved the medial s between vowels, in many forms where 
 all the Greek dialects omitted it, and where the Latin has 
 changed it to r; thus nitsos, daughter in law/ not vvo$, 
 Latin nurus ; vises, poison, not 169, Latin virus ; gencsos, 
 of birth, not ryeveos, yevovs, Latin generis. It must have 
 preserved the final m, which all the Greek dialects have 
 changed to v, or (after a] have dropped altogether: thus 
 6vom y egg, not wbv, Latin ovum; pateram, father/ not 
 
ITALIC ANS AND GREEKS. 165 
 
 Trarepa, Latin patrem ; esiem, I might be, not efyv, Old 
 Lat. stem; esam, I \vas, not rja or rjv, Latin cram for esain. 
 It must have preserved the final t, which all the Greek dia 
 lects have rejected, but which the Latin retains either as t or 
 d: thus esiet, he might be, not eirj, Old Lat. sict ; csant, 
 they were, not fjaav, Latin erant for csant ; alyot, other, 
 not aXXo, Latin aliud. It must have preserved the conso 
 nant y, which, after producing a multitude of euphonic 
 changes, vanished from all the dialects of Greek, but which 
 the Latin retains or changes to i thus yugom, yoke, not 
 v<y6v, Latin jugum f jepar, or rather jekar, liver, not rjTrap, 
 Latin jccur / salyomai, to leap, not aXko^ai, Latin salio. 
 In the inflexion of nouns, the Graeco-Italican must have had 
 a case which in all the Greek dialects has vanished from case- 
 inflexion, the ablative in /, Latin d: thus dolot, from art, 1 
 Oscan dolnd, Old Latin dolod=dolo ; 1121 ctot from night, 
 Latin noct-e, earlier noctcd. In the comparative degree of 
 adjectives, it must have had the form ions. Gen. ionsos, or 
 iyonSy Gen. iyonsos, like Sanskrit iyatfis, Gen. iyasas. From 
 this ending the Greek in all its dialects omitted the s, leaving 
 lov (Nom. uov, iov], Gen. LOVOS ; while the Latin omitted the 
 n t leaving ios t iosis, which afterwards became tor, ioris. 
 Thus to Sanskrit mahfyansas, greater (Nom. Plur.), cor 
 responded a Graeco-Italican magi(y)onses, Greek fjuetfyves, 
 Latin ma(g)joses, which is expressly handed down as the 
 earlier form of major cs. In the verb it must have preserved 
 the old infinitive in tuiii, tu, which remains in Latin as in 
 Sanskrit, but which in all the Greek dialects appears only as 
 a verbal noun, without the proper character of an infinitive. 
 It must have preserved likewise the old perfect participle 
 passive in tos, which remains in Latin as in Sanskrit, but 
 which in all the Greek dialects appears only as an adjective, 
 without the proper character of a participle. 
 
 This enumeration of a few leading points may suffice to 
 show that the supposed Grseco-Italican must have differed 
 very widely, not only from the known dialects of Greek, but 
 also from the common or unitary Greek language, out of 
 which these dialects sprang, and to which we may ascribe 
 such features as belong alike to all of them. Now we cannot, 
 
1 66 ITALICANS AND GREEKS. 
 
 with any approach to certainty, assign the period at which 
 this unitary Greek began to undergo the changes which re 
 sulted in the formation of its various dialects. But one thing 
 we can say with entire certainty, in view of such evidence as 
 that just presented. We can say that the period here referred 
 to, however remote it may have been, was long subsequent 
 to another to the period when the assumed Graeco-Italican 
 began to undergo the changes which terminated at last in the 
 formation of the Greek with its dialects on the one hand, and on 
 the other in the formation of the Italican idioms, Latin, Oscan, 
 Umbrian, and the rest. Our Graeco-Italican is thus driven 
 far back into the darkness of an ante-historic past. When 
 our author asserts that it was spoken in Greece and carried 
 over by sea from Greece to Italy, he asserts that which may 
 be true, though he is very far from having proved it so. 
 But when he puts it later than the formation of the Greek 
 dialects, and makes it in fact a mixture of those dialects, he 
 maintains that which is certainly and clemonstrably untrue. 
 The real question for philological study at the present time is, 
 whether there was any such Gra^co-Italican, distinct from 
 the primitive Indo-European language ; in other words, 
 whether there was any special relation between Greek and 
 Latin, more than between either of them and the Armenian, 
 for instance, or the Gothic. It has been, and still is, the pre 
 vailing opinion that there is a special relation between Greek 
 and Latin ; that both have sprung from a common Gra^co- 
 Italican, distinct from, and of course posterior to, the primi 
 tive Indo-European. This view is set forth in the works 
 both of Mommsen and of Curtius. But there are respectable 
 philologists who maintain the opposite opinion. 
 
 Before concluding, we ought, perhaps, to offer a word of 
 apology. We have, perhaps, fallen below the dignity of this 
 occasion, in giving an extended notice of a work so slight 
 and unimportant a work proceeding, it is true, from an ac 
 complished classical scholar, and a. justly esteemed traveller 
 in classic lands, but unworthy alike of his position and his 
 reputation. Possibly some little mixture of the light or the 
 amusing may be found to season not disagreeably the or 
 dinary gravity of our assemblies. We will not confess here 
 
ITALIC ANS AND GREEKS. 1 67 
 
 we are ashamed to own it even to ourselves a certain secret 
 satisfaction in finding that Germany before which we hide 
 our diminished heads, acknowledging her to be first without 
 second in philological studies can send out from the high 
 places of her universities specimens of fantastic absurdity 
 scarcely equalled on this side of the Atlantic. A more legi 
 timate pleasure may be derived from a book which allows us 
 to see the immense difference between the present and the 
 past between the Indo-European philology, with all the de 
 ficiencies and uncertainties that cleave to it, and the unscien 
 tific and unsatisfactory etymologizing that preceded it. If 
 at times we grow weary of the twilight in which we still move, 
 and sigh with despairing hope for the perfect day, it may be 
 well that a Professor Ross should come, and open for us a 
 glimpse into the darkness visible, the realm of chaos and old 
 night, which we have left behind us, as we trust, finally and 
 forever. 
 
IX. 
 
 ON INDO-EUROPEAN ASPIRATE? MUTES. 
 1862. 
 
 IN order to understand fully the relations of the Indo-Euro 
 pean languages, we need to know the forms of that primi 
 tive speech from which they have all set out on their diver 
 gent progress. Hence it must be the object of comparative 
 philology to reconstruct the original Indo-European language 
 as it was spoken by the common ancestors of Hindu and Per 
 sian, Greek and Goth and Gael. That this is an object not 
 to be easily or soon attained, is obvious enough : it is rather 
 an ideal to be aimed at than a result which we can expect to 
 see realized. In any such attempt at reconstruction, the first 
 point must be to determine the phonetic system of the primi 
 tive language ; what elements of sound were heard and utter 
 ed by our forefathers in that early time when as yet Sanskrit 
 and Greek, German and Celtic, had no separate existence. 
 It may seem at first view almost chimerical to think of re 
 calling the fugitive sounds of a period so remote from our 
 own ; but it must be remembered that their echoes have been 
 sounding in Asia and Europe down to the present day. By 
 comparing the phonetic systems of the different Indo-Euro 
 pean languages, we may ascertain in many cases with a high 
 degree of probability the phonetic elements which belonged 
 to their common mother. I propose to give a specimen of 
 such inquiries, by discussing the aspirate mutes of the original 
 Indo-European language. Did that language possess aspirate 
 mutes ; and if so, what were they, and how were they sound 
 ed ? First, then, let us see what letters of this class we find 
 in the various languages, especially the ancient languages, of 
 our family. 
 
 The Sanskrit has the greatest abundance of aspirate mutes. 
 In this language and in its modern descendants, every unas- 
 
IND O-EUR OPE AN ASP IRA TES. 1 69 
 
 pirate mute, whether surd or sonant, has its corresponding 
 aspirate, made by pronouncing after it the sound of h ; at 
 least, such is the modern pronunciation. Thus, of surd mutes, 
 along with k t we have kh ; with t, th ; with p, ph. And in 
 like manner of sonants ; along with g, gh ; with d, dJi ; with 
 b, bli. We have also aspirates for the so-called " cerebrals," 
 lingual sounds peculiar to India, and borrowed probably by 
 the Sanskrit-speaking people from languages which they found 
 in that peninsula sounds produced by turning the tip of the 
 tongue far back into the mouth : thus, without aspiration, t, 
 d; with aspiration, th, dJi. Even the palatals c, j (as in our 
 church, judge], which are not exclusively mutes, but contain 
 a fricative element, have their corresponding aspirates, ch,jh. 
 These aspirates, as sounded in the living languages of India 
 and in the traditional Indian pronunciation of the Sanskrit, 
 have a compound character ; they are made up of the unas- 
 pirate mute with an //-sound distinctly audible after it. It is 
 conceivable, however, that these letters may have changed 
 their sound in the course of ages, and that the living 
 pronunciation of the old Sanskrit may have been something 
 quite different. This, then, is a question which requires to 
 be considered : whether there is reason for believing that the 
 ancient pronunciation of these letters was materially different 
 from the modern. Now if they were not originally com 
 pound aspirates of the kind just described, they must in all 
 probability have been simple spirants, like our / and v and 
 our two sounds of th (surd in thin, sonant in this). In that 
 case, the aspirate corresponding to k w r ould have sounded as 
 German ch (in machcti) or as modern Greek % ; the aspirate 
 of g, as 7 in modern Greek Xctyo? ; the aspirates of t and d, as 
 the two sounds just described of English th ; the aspirates of 
 / and b, as / and v respectively. It might be urged in favor 
 of these sounds that, in the Sanskrit alphabet, the aspirate 
 mutes are represented by single letters a procedure which is 
 perfectly natural if they were in reality simple sounds, like/, 
 v, and the rest, but is somewhat surprising if they were com 
 pounds containing a distinctly audible h. But this argument 
 is much more than balanced by others on the opposite side. 
 Thus, it would be difficult to find simple spirants which should 
 
1 70 IND 0-E UR OPE AN ASPIRA TES. 
 
 answer to the cerebrals, /, d ; to pronounce anything like our 
 th with the tip of the tongue bent back to the roof of the 
 mouth seems almost a physical impossibility. It would be 
 perhaps even more difficult to imagine any spirants which 
 should answer to the palatals c t j. Besides, if the letter which 
 is now pronounced pit was anciently pronounced f, we must 
 recognize a change of sound from /to///. But this seems a 
 very improbable change, and cannot be supported, so far as I 
 know r , by any certain example. And if it is hard to suppose 
 that such a change should have occurred in one case, how 
 much harder to assume it for the whole series of mutes, surd 
 and sonant ; to admit that v has passed into bh t and til (as in 
 this) into dli (d-Ji) t and so throughout. We should have to 
 suppose, too, that the change took place over the whole ex 
 tent of northern India, in Bengali, Hindui, Marathi, and other 
 languages, varying widely from each other, but none of them 
 preserving any of the supposed original spirant mutes. But 
 there is another circumstance which stro ngly supports the tra 
 ditional pronunciation, by making it probable that these letters 
 contained an audible //-sound ; and this is that even in the 
 Sanskrit we find them occasionally changing into //. This is 
 true particularly of those which, as we shall see further on, 
 are the most important of the whole series, the aspirates of 
 b, d, and^-. Thus the common root han, to kill, shows in 
 many of its derivative forms that it was originally gJian. The 
 word hansa, goose, is proved by the Greek yi^v and the 
 German gans to have been originally ghansa. The original 
 ending did of the imperative, 2d sing., which corresponds to 
 Greek Qi in crrfjdi, IcrQi, is preserved in many Sanskrit verbs, 
 but in many others it becomes /// . The pronoun mahyam, 
 to me, is proved by the analogy of tnbhyam, and also by 
 the analogy of those case-endings which contain the /V/-souncl, 
 to have been originally mabhyam ; though in this instance 
 the change to // might seem to have been much more ancient, 
 and to have preceded the separation of the languages, since we 
 find it also in the Latin miJii as compared with tibi. These 
 changes of the aspirate mutes to // have been carried to a 
 much greater extent in the Prakrit dialects, which supplanted 
 the old Sanskrit in the mouths of the people some centuries 
 
IND 0-E UR OPE AN ASPIRA TES. 1 7 1 
 
 before the Christian era. In these the surd aspirates, as kh 
 and th, have also been reduced in many cases to a mere /i. 
 Now it is true that in some languages a spirant sound f, for 
 instance, or cJi has passed into h ; but a movement of that 
 kind so general as we have seen it in the languages of India 
 is a thing unexampled and almost incredible. But if the In 
 dian aspirates were what tradition shows them, compounds of 
 mutes with a distinctly audible /i, then the change we have 
 described is a perfectly intelligible and natural phenomenon. 
 It is merely slurring and sinking one element in the phonetic 
 compound. The h was perhaps uttered with a degree of 
 force which threw into the shade, and at length totally eclipsed, 
 the mute sound that preceded it. We are brought to the 
 conclusion, then, that the letters in question could not have 
 been simple spirants, but must have sounded in ancient, very 
 much as they do in modern India, like the unaspirate mutes 
 followed by a clearly pronounced Ji. 
 
 The argument here drawn from the change of aspirate mutes 
 to h seems unfavorable to certain views which have been re 
 cently expressed by Lepsius in a valuable memoir on the 
 sounds of the Arabic language and the mode of representing 
 them in occidental characters. Speaking of the Sanskrit as 
 pirates, he recognizes their compound character, as differing 
 from the unaspirate mutes only by an added breathing : but 
 this added breathing he holds to be nothing more than be 
 longs regularly to the/, t, k of the English and the German. 
 He declares, indeed, that English, French, and German have 
 no tenues, but only aspirates and mediae ; their /, /, k are 
 not tenues, but aspirates. And he says that he cannnot see 
 how t, for instance, can be more fully aspirated than it is in 
 the German tail, cable, or than, clew. On the other hand, 
 he tells us that the Hungarians have true tenues, but no aspi 
 rates. The Greek and Sanskrit had both tenues and aspi 
 rates : they had one/, for instance, pronounced in the Hunga 
 rian manner, and another/ in the English. Now it is true 
 that in forcible enunciation we speak our/, t, k with a sharp 
 explosive utterance which approaches perhaps to the charac 
 ter of an aspiration. At the same time, it seems almost ab 
 surd to suppose that an h could arise from these letters, pro- 
 
1/2 INDO-EUROPEAN ASPIRATES. 
 
 nounced as we pronounce them. If the slight trace of as 
 piration which we sometimes give to these letters should 
 become constant, and then should be intensified so as to bear 
 a measurable proportion to the proper mute sound, the latter 
 might in time fall away and leave only the h. But the sub 
 stitution of this sound for the original mute could only come 
 about by some such process. It ma}* be considered certain 
 that, before the Sanskrit aspirates passed into h, they had 
 assumed sounds in which the aspirate element, the h, was 
 vastly more prominent than anybody can reasonably suppose 
 it to be in our English /> /, k. But if they had such sounds 
 
 o J " j 
 
 when the change to Ji took place, it would be difficult to say 
 when they did not have them. For some changes of this 
 sort are to be found even in the earliest Sanskrit ; and in 
 mahyam for mabJiyam (Lat. iniJii for inibi) we have a case 
 which perhaps goes back beyond the formation of the San 
 skrit itself. 
 
 If now we turn to the ancient Bactrian, the language of the 
 Zend-Avesta, we find aspirates, both surd and sonant, as in 
 the Sanskrit. Their use, however, seems to depend in a great 
 degree on euphonic conditions, so that in kindred roots they 
 do not correspond with regularity to the aspirates of the Sans 
 krit, the Zend often showing an aspirate mute where the 
 Sanskrit has the unaspirate, and vice versa. Whether they 
 were pronounced with a distinct h sound is a point which it 
 might be hard to determine in reference to an ancient lan 
 guage which has left no direct descendants, and which is so 
 imperfectly represented by documents. 
 
 Coming next to the Greek, we find the three aspirate mutes 
 (/>, 6, %. In the present pronunciation of the language these 
 are spirant letters, with the sounds of/, English tJi surd, and 
 German ch. There is reason to believe that this pronuncia 
 tion goes back at least to the time of the later Roman Em 
 perors. In many inscriptions of that period, the Greek (/> is 
 represented by a Romany/ which never was the case in 
 earlier times. The Latin grammarian Priscian, about 500 
 A.D., tells us. that the sound then expressed by / was origi 
 nally signified by p with an aspiration (that is, by pJi). This 
 original ph y he adds, was retained in the Latin writing of 
 
IND O-EUR OPE AN ASP IRA TES. I / 3 
 
 Greek words, as Orpheus, but changed to/ in the writing of 
 Latin words, as fama. He thus implies clearly that he re 
 garded the /// of Orpheus and the f of fama as substantially 
 identical. In fact, he says in another place that the lips are 
 more firmly fixed in pronouncing ph than in pronouncing f t 
 and this is the only difference between / and ph. Here the 
 expression " this is the only difference " seems naturally to im 
 ply that he regards the difference as not considerable. Now 
 the Latin f was unquestionably a spirant ; and if the Greek (f> 
 in Priscian s time was likewise a spirant, we can hardly doubt 
 that at that period the 6 and the % had the same character, 
 though we have not the advantage of being able to compare 
 them with any similar sounds in the Latin. 
 
 But there is a variety of considerations which go to show 
 that these letters were not originally sounded as spirants, but 
 like the Sanskrit surd aspirates, as compounds of /, /, k with 
 a distinctly audible h. In the first place, I may mention the 
 fact that they are never doubled, but that TT</> is used for (fxj), 
 rO for 66, /e%for ^%. The spirant sounds are easily doubled. 
 Thus Saf-fo offers perhaps less difficulty in pronunciation than 
 Sap-fo. But if (/> was sounded as/-//, then it must obviously 
 have been almost impossible to double the aspiration. The 
 organs closing on the first /-sound would open with the 
 second, which would then be followed by its aspiration (Sap- 
 p-ho) ; and this would give the sound of 7n/>, not of 00. In the 
 second place, these letters have actually arisen in many cases 
 from the combination of a smooth mute with a rough breath 
 ing. Thus a0 ov, from which, has arisen out of air ov ; 
 e (/>oo?, approach, has arisen" out of eVoSo?. There can be no 
 doubt whatever that these were once sounded ap liou, ep- 
 liodos : the only question would be one of time when did 
 they change from ap lion to af ou, from cp-Jwdos to cfodos? 
 In the third place, two of these letters were originally written 
 as compounds of a smooth mute with a following h thus, 
 as TIH, and % as KH a mode of writing abundantly attested 
 by extant inscriptions. When letters are first used to write 
 a language, they are used according to phonetic principles : 
 the presumption certainly is that the IT and K and H had the 
 same sound in these combinations that they had elsewhere. 
 
1 74 IND O-E UR OPE AN ASPIRA TES. 
 
 As to 0, the grammarians tell us that this also was once 
 expressed, like the other aspirates, by TH. But this state 
 ment is not borne out by the evidence of inscriptions, and is 
 not probable in itself. The truth appears to be that, as the 
 primitive Semitic alphabet had two signs for /, one for the 
 ordinary lingual and the other for an intensive variety of it, 
 the Greeks from their first adoption of Phoenician characters 
 employed the second to express their own aspirate /. And it 
 was probably the fact of their having already a simple charac 
 ter for this aspirate that led them afterwards to invent simple 
 characters for the other two. Should any one wish to infer 
 from this adoption of simple characters that the sounds repre 
 sented were likewise simple, we need only point to the double 
 consonants f and *fy to show the hazardousness of such an 
 inference. In the fourth place, we must attach some weight to 
 the well-attested difference between the Greek (/> and the Latin 
 f. Thus Cicero, it is said, in pleading for Fundanius, ridi 
 culed a Greek witness on the other side, because he could 
 not even pronounce the name (Fundanius) of the accused 
 party. This perhaps might not necessarily imply that there 
 was anything more than a slight difference between the 
 Latin letter and the Greek. But it is a more significant cir 
 cumstance, that for centuries the Greek d> is never represented 
 by a Latin// if the difference between the two sounds was a 
 slight one, it is strange that, with the constant occasion to ex 
 press Greek w r orcls in Roman writing, the familiar/ should not 
 occasionally be used for a sound approaching nearly to it. 
 And in the fifth place, the preceding argument is confirmed 
 by a similar fact in relation to the Coptic. The alphabet of 
 this language, like the Roman, has what is believed to be a 
 spirant/.- but this letter is never used in transcribing the 
 Greek proper names Philippos, Philotera, and the like. The 
 Greek cf> itself is borrowed by the Coptic for this purpose. 
 
 These arguments in their collective weight seem to be deci 
 sive, and to make it certain that the Greek aspirates were 
 once sounded as the surd aspirates of the Sanskrit. It is true 
 that this conclusion involves one pretty serious difficulty a 
 difficulty presented by the familiar combinations $0 and yQ. 
 It is well known that, before 6, the Greek not only allows </> 
 
IND O-EUR OPE AN ASPIRA TES. 1 7 5 
 
 and % to stand unchanged (eypd^Oijv, erv^^i*), but even 
 changes TT and ft to </), K and 7 to ^ (eVe/i^^z/, e\e-%0r}i>). 
 These combinations, $0, ^0, with the spirant pronunciation, 
 are natural and easy; but to pronounce/-// -/-/ , giving each 
 of the four letters its proper sound, appears well-nigh impos 
 sible. It would seem that euphony must require us in 
 such cases to give up the first aspiration, to say irQ (p-t-/i) in 
 stead of<f)0, and K0 (k-t-Ji) instead of %0. This, in fact, is the 
 prescription of Sanskrit euphony. I must confess my inability 
 to offer any satisfactory solution for this difficulty. We might 
 suppose that the earliest Greek had TT#, K0, and that these 
 were changed by assimilation into (j>0, %0, after the aspirates 
 had assumed their present spirant sounds : but this would 
 carry the spirant sounds so far back that they would precede 
 the time of transcription into Coptic and Latin, and we should 
 be unable to account for the peculiarities just noticed in that 
 transcription. However this may be, it can scarcely be 
 doubted that there was a time when the Greek aspirates dif 
 fered from the smooth mutes only by the addition of an h- 
 souncl. They were thus, as before remarked, identical in 
 sound with the surd aspirates of the Sanskrit. But it is a 
 curious and important fact that, in roots and words common 
 to both languages, it is almost always the sonant aspirates of 
 the Sanskrit, rarely the surd, that correspond to the Greek (, 
 0, %. Thus Greek (frvco answers to Sanskrit root bJin ; Greek 
 cf)ep(jL> to Skt. root bhar; Greek Ti 0?7/u, root $e, to Skt. root 
 dJid j Greek ov0ap to Skt. udhar, our udder y Greek e\a%vs, 
 small, to Skt. laghus, Might; Greek e^t?, serpent, Skt. 
 ahis for agliis. The Sanskrit roots and stems which have surd 
 aspirates, if they reappear in Greek, appear there almost al 
 ways with tenues instead of aspirates : thus Skt. st/iri, Greek 
 crra in ICTT^/U, not a0a ; Skt. prithus, Greek TrXaru?, broad/ 
 not TrAatff? ; Skt. root splnir, to tremble, Greek crTraipco, 
 ao-Tralpw, not <r&aipw. It should be said also that, among 
 the Sanskrit roots and stems which have these Jetters, there 
 is a very large proportion which have not been satisfactorily 
 identified either in the Greek or in other languages of the 
 Indo-European family. 
 
 Passing on now to the Latin, we seem to lose sight almost 
 
1 76 IND O-E UR OPE AN ASP IRA TES. 
 
 completely of the class of letters under discussion. The only 
 Latin mute of this species is the labial/. It appears to be 
 certain that, from the time of our earliest notices, this letter 
 was no compound of a smooth mute and rough breathing, but 
 a simple spirant, not differing essentially from our English/". 
 It is true that, from the expressions of Quintilian and Priscian 
 in relation to it, Corsscn has inferred, with some appearance 
 of probability, that it was pronounced with a very loose and 
 imperfect contact between the upper teeth and lower lip, so 
 as to have a particularly rough or aspirate sound. But this 
 belongs to the special coloring of the sound, not to its general 
 nature, and is unimportant in relation to our present inquiry. 
 It concerns us more to observe how it corresponds to the 
 Greek and Sanskrit aspirates in words which belong to the 
 Latin in common with those languages. At the beginning 
 of words, the Greek (/>, Sanskrit bh, appears in Latin as /. 
 thus Skt. root b/iar, Gr. fyepoi, Lat./rr^/ Skt. bhratar, Gr. 
 (bpanjp, Lat. f rater. But a Latin initial/ corresponds also in 
 a number of cases to Gr. 9, Skt. dli : thus Gr. @vpa, Skt. 
 dvaraui for dhvdram, Lat. fores Gr. 6rfp, Lat. fcra. But 
 the Greek initial ^, Skt. gh or /t, is usually represented by 
 Latin Ji : thus Gr. ^ei/j-cov, winter, Skt. ]iimas, snow, Lat. 
 hicms. In the middle of a word, on the other hand, the 
 aspirates of the Greek and Sanskrit usually appear in Latin 
 as unaspirated sonants ; that is, as b, rf, g. Thus Gr. a/jLcjxo, 
 both, Skt. nbJunt, Lat. ambo ; Gr. 6p0o$, straight, steep, 
 Skt. -urdhvas, Lat. arduns ; Gr. Xer^w, to lick, Skt. root lih 
 for Ugh, Lat. lingo. It is instructive to look at the words for 
 reel in these languages. The Greek has epvOpos, the Sanskrit 
 rudhiras, showing a primitive rndh or pvO. In Latin, the 
 lingual aspirate passed into a labial, which appears -either as 
 an unaspirated sonant, in rnbci\ or as an aspirate surd, in 
 rufjis. 
 
 If in Latin we find the aspirate mutes reduced to a mini 
 mum in the single spirant/, we are met by a still more com 
 plete deficiency when we turn to the languages of North 
 eastern Europe, to those of the Lctto-Slavic class. The /it 
 self is wanting in the Lithuanian and in the Old Slavonic, and 
 in both, the aspirates of the Greek and Sanskrit appear regu- 
 
INDO-EUR OPE AN ASPIRA TES. 1 7 / 
 
 larly as nnaspirated sonants, as b, d, g. Thus Gr. <j>vco, Lat. 
 fui, Skt. root bhu, Lith. inf. biiti, to be, O. S. byti ; Gr. 
 (frpanjp, L,at. f rater, Skt. b/irdtar, O. S. bratru, Lith. broils, 
 brotcrelis ; Gr. epv6pd<s, red, Skt. rndJiiras, Lith. randimas, 
 1 red, O. S. rudycti, to redden ; Gr. 0/^X97, mist, cloud, 
 Skt. mcgha, O. S. migla, Lith. migld. 
 
 In the Celtic languages, we are struck at first view with the 
 great abundance of aspirate mutes. For each of the three 
 tenues, /, /, , and for each of the three medials, , ^ g, we 
 have corresponding aspirates, which in the Irish are written 
 ///, t/i, ch, bh, etc. These are sounded, not as compound 
 aspirates, but as simple spirants. The proper spirant sounds 
 of the th and dh, though corrupted and lost in the modern 
 Irish, are preserved in the Welsh. We find, however, on a 
 little examination, that the use of these aspirates depends on 
 euphonic conditions and principles peculiar to the Celtic lan 
 guages, and that they do not correspond, save rarely and ex 
 ceptionally, to the aspirates of the Greek and Sanskrit. 
 These latter, indeed, are represented in the Celtic languages 
 precisely as they are in the Letto-Slavic, by b, d, g. Thus, 
 for Gr. (frva), Skt. root bhu, we have Ir. bi j for Greek 
 (frparrjp, Skt. bhrdtar, O. Ir. brdtkir, brother ; for Gr. 
 epv6po<s, Skt. rudhiras, O.W. rliud, red ; for Gr. ^et/^co^, 
 Skt. iiimaSy O. \r.gaim, Q.VJ.gauam, winter. 
 
 Turning now in the last place to the Germanic languages, 
 we find much the same state of things as in the Celtic. The 
 Mceso-Gothic, the oldest specimen of this class, has spirant 
 mutes, but those only surd. Thus it has the labial f it has 
 the lingual th surd (as in thin), represented by a single char 
 acter. For the palatal spirant ch, it shows the simple breath 
 ing h. In the further development of the Germanic languages, 
 we find added to these their corresponding sonants, the sounds 
 of v, th in this, and gh. But even the aspirates of the 
 Mceso-Gothic, like those of the Celtic, are of secondary 
 origin ; they do not connect themselves with those of the 
 Greek and Sanskrit. The Germanic languages stand here on 
 the same footing with the Celtic and the Letto-Slavic. They 
 show unaspirated sonants, b, d, g, in place of the Greek and 
 Sanskrit aspirates. Thus in our be and brother, compared 
 
 12 
 
I?8 INDO-EUROPEAN ASPIRATES. 
 
 with cj)va> and (frparrjp ; in ;Y<^, compared with epvOpds ; and in 
 goose, Germ, gans, compared with %??^, %^o9 (apparently for 
 
 After looking thus at the different branches of our family, 
 
 we are prepared to take up the question whether there were 
 
 aspirates in the primitive Indo-European language. If they 
 
 were found only in the Sanskrit and the Zend, we might sup 
 
 pose that they had sprung up in the Aryan or Indo-Persian 
 
 branch after the separation of the other branches. But the 
 
 fact that they appear in Greek, and that the same roots and 
 
 words which have them in Greek have them also in Sanskrit, 
 
 compels us to carry them back to a more ancient period, 
 
 to times when the mother-tongue of Greek and Sanskrit was 
 
 yet undivided. Possibly, however, it might be imagined that 
 
 the languages of Northern and Western Europe, the Letto- 
 
 Slavic, the Germanic, and the Celtic, which present only 
 
 medials in place of the Greek aspirates, were separated from 
 
 the common stock at a yet earlier period, while the ancestors 
 
 of the Greeks and Hindus continued to live and speak 
 
 together ; and that it was then, after the separation of those 
 
 former branches, that this class of letters began to develop 
 
 itself. But even this hypothesis will be found untenable. 
 
 For our Germanic languages, while they have lost the primi 
 
 tive aspirates, afford yet very curious indications of their 
 
 primitive existence. If we take the Greek words Ovpa, Si/o, 
 
 and rpels, and look for their English equivalents, for Ovpa we 
 
 have door, for Svo, tiuo, for rpels, tlircc. If, prior to the 
 
 separation of the Germanic languages, the word door began 
 
 with d, we ought in English to have toor (as for primitive Bvo 
 
 we have two) ; if in the same period it began with t, we ought 
 
 in English to have tJioor (as for primitive rpet? we have three). 
 
 So with the scries Gr. yfiv, Eng. goose, Gr. 761/09, Eng. kin, 
 
 Gr. Kva)v, Eng. Jiound : if the ante-Germanic goose began 
 
 with^-, we should have koosc in Eng. (like kiti) ; if it began 
 
 with/ , we should have hoosc (like hound}. But these series 
 
 are only examples of a general law of consonant relations 
 
 (Grimm s Lautverschiebung). We see then that the Ger 
 
 manic languages represent the aspirate mutes of the Greek in 
 
 a different manner from its tenues and mediae ; from which it 
 
IND O-EUR OPE AN ASPIRA TES. 1 79 
 
 follows that the aspirate mutes were already distinct from the 
 tenues and mediae before the Germanic branch came to 
 be separate from the parent-stock. And this is equivalent to 
 saying that there existed in the Indo-European language, 
 prior to the separation of its branches, a class of mutes dis 
 tinguished from the unaspirated surds and sonants, a class 
 represented in later times by the surd aspirates of the Greek 
 and the sonant aspirates of the Sanskrit. 
 
 It now remains for us to ask how this peculiar class of 
 mutes was pronounced in that primitive language. Had they 
 simple spirant sounds, like our/and v and our two /// s, or were 
 they, like the Greek and Sanskrit aspirates, compounds of 
 tenues and mediae with an audible //-sound ? That they must 
 have had one or other of these two characters can hardly be 
 doubted, when we see what sounds have proceeded from them 
 in the linguistic history of our family. Nor can there be 
 much doubt that they had the character last-mentioned, that 
 they were in reality compound aspirates. If the spirant 
 sounds were the earliest, it would be strange that we should 
 not find them either in the Greek or in the Sanskrit, which 
 present to us the most ancient specimens of Indo-European 
 speech. And besides, if the .spirant sounds were the earliest, 
 we should have to suppose both in Greek and Sanskrit such 
 changes as that from/ to/-//, from v to b-h, etc. a supposition 
 which has been already characterized as forced and improba 
 ble. 
 
 But there is still another question to be considered. We 
 have seen that the Greek aspirates were surd, made up of a 
 smooth mute and rough breathing. Yet we have seen that 
 in corresponding words and forms they answer to the sonant 
 aspirates of the Sanskrit, made up of a middle mute and 
 rough breathing. Which now was the original pronuncia 
 tion ? Did the primitive Indo-European give to the root of our 
 verb to bear the sound of phar as in Greek (f>epo), or of bhar 
 as in Skt. bhardmif Now the foregoing survey has shown 
 us that in nearly all the great branches of our family the 
 primitive aspirates are represented by sonant letters. Thus 
 the labial aspirate of the original language appears in San 
 skrit as bk, in Zend as bh or b, in Letto-Slavic, Germanic, 
 
1 80 IND O-EUR OPE AN ASPIRA TES. 
 
 Celtic, and partly in Latin, as b : only in Greek, and partly 
 also in Latin, does it appear as a surd, ph or/. Here, then, 
 we find strong reason for believing that the Indo-European 
 aspirates were originally sonant, bh, d/t, gh. And accord 
 ingly this opinion has been generally adopted by the most 
 eminent philologists : by Curtius in vol. ii. of Kuhn s Zeit- 
 schrift(i853), by Boppin the second edition of his Comparative 
 Grammar (1856), and by Schleicher in his recently published 
 Compendium of Comparative Grammar (1861). 
 
 Only one voice, so far as I know, has been raised on the 
 other side. Kuhn, in a late number of his Zeitschrift (vol. xi., 
 1862, p. 302 ff.) reviewing Schleicher s work, expresses his dis 
 sent from the commonly-received view, and his belief that the 
 primitive aspirates were surd, as in Greek. In Sanskrit, he 
 thinks, they have been generally weakened into sonants, 
 while the surd aspirates of that language are mostly of later 
 origin : though in some cases he holds these to be remnants 
 of primitive usage. His reasons he only indicates in a brief 
 sketch, promising to exhibit them more fully at some future 
 time. They are in part directed against a view which I sus 
 pect that no one entertains : namely, that the Sanskrit surd 
 aspirates were not only produced after the sonant, but were 
 actually developed out of them in the same words and forms. 
 This may, not improbably, have occurred in some exceptional 
 and anomalous cases ; but it is quite certain that the mass of 
 surd aspirates did not originate in this way. Let us look then 
 at his other reasons, which may be reduced to three. First, 
 if the primitive aspirates were sonant, then in Greek they 
 must have changed to surd. But this is a change from weaker 
 to stronger sounds, and is therefore unnatural and improba 
 ble. There is doubtless force in this argument. The change 
 from sonant to surd is certainly less probable than a change 
 in the opposite direction. And yet there are abundant and 
 striking examples of the less probable mutation. It is enough 
 to mention that in the Germanic languages, by the first Laut- 
 versckiebzing, all medials have passed into tenues (tivo for Svo, 
 etc.) ; and that the modern Armenian pronounces all the 
 medials of the ancient language as tenues, making up for this 
 by pronouncing all the old tenues as medials (thus the ancient 
 
IND 0-E UR OPE AN ASPIRA TES. 1 8 1 
 
 Tigranes is the modern Dikraif). Second, whenever new, un 
 original aspirates have arisen in the Indo-European lan 
 guages, they have always been surd aspirates, never sonant, 
 though sonant aspirates and mediae have often been developed 
 from them in the progress of time. The remark has refer 
 ence especially to the Germanic and the Celtic languages. Of 
 the first I believe it to be strictly true ; but not altogether so 
 of -the second. The sonant aspirates of the Irish and Welsh 
 have not been developed out of surd aspirates, but out of 
 mediae. It is not even quite certain, in case of the Irish, that 
 they are of later development ; for though the earliest monu 
 ments have only pk, th, ch, the aspirates of p, /, c, yet there 
 are not wanting indications that the b, d, g had already un 
 dergone a change analogous to the aspiration of the /, /, c. 
 Still, I think it more probable that in the Celtic, as in the 
 Germanic, aspiration began with tenues, and that medial as 
 pirates were of later growth. And the presumption thence 
 arising of a similar progress in early Indo-European times is 
 not without its force. Third, he says that if the surd as 
 pirates have arisen from the sonant, it is very strange that 
 in a number of cases the Sanskrit ph y th, kh correspond to 
 Greek (/>, 6, ^, and this partly in forms which, like verb-end 
 ings, go back to the beginnings of linguistic formation. Of 
 Greek and Sanskrit words thus related, the number yet 
 pointed out is not large, and in part of them there may be 
 no real etymological connection. In some, we may suppose 
 that, by an exceptional and anomalous change, the Sanskrit 
 bh, etc., has passed into ph, etc. In others, it may be that 
 both languages have independently changed the tenuis into a 
 surd aspirate, p into ph, etc. : for it could easily be proved 
 that such changes have been frequent in both languages. As 
 regards the verb-endings referred to by Kuhn, it is curious to 
 compare the Greek and Sanskrit forms. The Sanskrit has 
 the sonant dh in several endings : in the 2d person sing, 
 impcr. active, dhi corresponding to Greek 61 ; in the first 
 person plural of the middle, mahc, mahi (originally madhe, 
 mad/ii], corresponding to Greek neOa ; and in the 2d person 
 plural middle, dlive, dhvain, corresponding to Greek aSe. 
 Here the relation of the aspirates is regular and constant. 
 
1 82 INDO-EUROPEAN ASPIRA TES. 
 
 But it is not so where the Sanskrit has the surd th. To be 
 sure, thdm in the 2d person dual of the middle answers to 
 crOov in the Greek, though for crOov, o-0rjv, of the 3d person 
 dual the Sanskrit has tarn. So tha in the second person 
 singular perfect active may correspond to Gr. 6a in olaOa, 
 fjaOa, e<f)7]cr9a, K\aioLcrOa. But for thas in the 2d person 
 singular of the middle there is no corresponding aspirate 
 form in Greek, and as little for thas in the 2d person dual of 
 the active, and for tha in the 2d person plural of the active. 
 So that in verb-endings the surd aspirates of the Sanskrit do 
 not appear to stand in any definite and satisfactory relation 
 to those of the Greek. 
 
 But to my own mind these arguments, giving them all the 
 weight which they can possibly claim, seem insufficient to 
 balance that general consent of the Indo-European languages, 
 by which, in all except the Greek and partly the Latin, the 
 original aspirates are represented by sonant consonants. 
 The only way, as it seems to me, to weaken the overwhelm 
 ing weight of this fact, would be to show that the sonancy 
 may perhaps have arisen from the aspiration ; that there is a 
 tendency in such an aspiration to change a preceding tennis 
 into a medial, so that we might expect an original/// to pro 
 duce a b rather than a/, an original th to produce a d rather 
 than a t. Now there are some facts which might lend a color 
 to this view. In the Irish language there are certain influ 
 ences which change an original /, /, c into ph y ///, ch : aspi 
 rating influences, we may therefore call them. But in the 
 kindred Welsh, p, t, c, under the same conditions, pass into 
 b, d, g the aspirating influences have here changed the 
 surds into sonants. That the influences in question tend to 
 aspiration even in the Welsh is evident from the fact that 
 they change an original Welsh b into v ( = bh), an original 
 d into dd (=dh), an original g into a spirant gh- sound, which 
 has disappeared from the language. A very curious parallel 
 to these phenomena of Celtic euphony is presented on Ro 
 manic soil, among the descendants of the Latin, in the dia 
 lect of Central Sardinia. Here, too, we find that the same 
 euphonic conditions which aspirate an initial b, d, g change 
 an initial/, /, c into the sonant b, d, g. Thus boc, ox, pre- 
 
IND O-E UR OPE AN ASP IRA TES. 1 8 3 
 
 ceded by the indef. art. unu, gives unit oe, an ox, the aspi 
 rated b disappearing in pronunciation; butfloveru, pauper, 
 with the same article, nnu boveru, a. pauper. If these 
 analogies would authorize us to assume that there is a pho 
 netic tendency in an aspirated/, /, k to pass into b, d, g, we 
 might suppose with plausibility that the original aspirates, 
 which have produced b, d, g in so many classes of languages, 
 were not bh t dli, gJi, but///, th y k/i, as Kuhn supposes. 
 
 I confess, however, that these analogies do not seem to me 
 as yet definite and extensive enough to be fully relied on. In 
 the present state of the question, I regard it as probable that 
 the primitive sounds under discussion were sonant rather 
 than surd. 
 
X. 
 
 ON THE FORMATION OF INDO-EUROPEAN 
 FUTURES. 
 
 1859- 
 
 I PROPOSE to occupy a short time in describing the dif 
 ferent methods used in different languages of the Indo- 
 European family for making a future tense. The contrivances 
 resorted to for this purpose are for the most part intelligible 
 to us ; and in their variety we may find a striking illustration 
 of that abundance of resources which the makers of language 
 (and that is only another term for the users of language) have 
 always had at their disposal. The facts which I shall have to 
 present are generally known ; but it may not be uninteresting 
 or unsuggestivc to sec them brought together and exhibited 
 in one view. I shall not, of course, undertake in this paper 
 to give all the expressions which arc used in any one of the 
 languages mentioned to convey the idea of futurity. Such 
 an attempt would be almost without limit. We shall consider 
 merely those modifications of the verb by inflection, compo 
 sition, or use of auxiliaries, which are most constantly and 
 regularly employed to refer the action expressed by the root 
 of the verb to a time posterior to the present of the speaker. 
 
 Let us, at starting, cast a glance outside of our own class 
 of languages, at those of the Semitic family. Here, at first 
 view, a remarkable prominence appears to be given by the 
 grammatical organism to the idea of future time. For, while 
 the system of verb-inflection embraces only two tenses, one 
 of these, and the one most extensively used, goes by the 
 name of " future." But this first appearance is hardly borne 
 out by a closer scrutiny, directed to the actual use of the 
 tense-forms; for this so-called future does service also as a 
 present, and even, under particular conditions, as a past tense. 
 And, in fact, Ewald in his Hebrew and Arabic grammars de- 
 
INDO-EUROPEAN FUTURES. 185 
 
 nies it the name of a future. According to him, the proper 
 office of the two Semitic tenses is to express, not past and 
 future, but finished and unfinished action. Thus an action 
 thought of as finished, at whatever time, past, present, or fu 
 ture, is expressed by the Semitic perfect : and, in like man 
 ner, an action thought of as unfinished, or thought of without 
 any idea of completion, is expressed by the Semitic future, 
 to whatever time it may belong, whether future, present, or 
 past. This theory, which goes far to account for the seem 
 ingly irregular and capricious use of the tenses in the Hebrew 
 and its sister-languages, has found many adherents : it ap 
 pears even in the recent editions of Gesenius s Grammar, pub 
 lished by Rodiger. In the Hebrew Grammar of Dr. Nord- 
 heimer, we find a view essentially the same as this, but with 
 an ingenious and interesting modification of form, which in a 
 measure reconciles it with the old established view. In this, 
 the Semitic perfect and future are made to represent, not 
 strictly past and future time, but prior and posterior action 
 action prior and posterior to some assumed stand-point or 
 dividing position this assumed stand-point, or dividing posi 
 tion between prior and posterior, being often coincident with 
 the immediate present time, or instant of speaking, but very 
 often ilso some different epoch suggested in the context. 
 The main peculiarities in the Semitic tense-system, according 
 to his view of it, would be : first, the freedom with which 
 this dividing-point is allowed to shift its place, and to take up 
 any required position ; and, second, the fact that every action 
 is thought of as preceding or following, never as contempo 
 raneous with, the moment of time thus fixed upon. It is not 
 my purpose, however, to enter upon any discussion with re 
 ference to these points of Semitic grammar, but only to draw 
 attention to the fact which is equally true, whatever theory we 
 may adopt in accounting for it that the present and future of 
 our Indo-European languages are to a great extent confound 
 ed, that is, represented without distinction of form, in Semi 
 tic expression. And I do it with the purpose of adding that 
 the same confusion, the same want of formal distinction be 
 tween present and future, is to be found also within the do 
 main of Indo-European speech ; and that; too, at a point very 
 
1 86 IND 0-E UR OPE AN F UTURES. 
 
 near home to every one of us. If we look at our own mother- 
 tongue, or, more properly, the mother of our mother-tongue, 
 the Anglo-Saxon, we shall find a corresponding deficiency. As 
 in Hebrew, so in Anglo-Saxon, the same form of the verb is 
 regularly and ordinarily employed both as a present and as a fu 
 ture. The words which appear in the Anglo-Saxon New Testa 
 ment as aversion of John xvi. 2, might be rendered into mod 
 ern English, either " They will put you out of the synagogues, 
 and the time will come, that whosoever shall slay you, will 
 think that he doeth God service," or, if we look only at the 
 forms of expression, without reference to the actual circum 
 stances of the case, " They put you out of the synagogues, 
 and the time cometh, that whosoever slaycth you, thinketh 
 that he cloeth God service." Nor is the Mceso-Gothic, the 
 most ancient and most perfect representative of the Germanic 
 class of languages, better off in this particular than the 
 Anglo-Saxon. If we turn to the translation of the Bible 
 made by Bishop Ulfilas in the fourth century for his Mceso- 
 Gothic countrymen, we shall find that the inflected forms of 
 the verb which arc used for the present, and which are shown 
 by a comparison of other Indo-European languages to belong 
 properly to that tense, are also in constant use for the expres 
 sion of the future. The same thing is true likewise of the old 
 Norse ; and it appears, indeed, that this want of a special 
 form for the future in distinction from the present belongs to 
 the Germanic languages generally, in all the earliest forms 
 with which we are acquainted. Perhaps it was even charac 
 teristic of primitive Germanism. By this I do not mean that 
 the common Indo-European stock had not developed a dis 
 tinct form for the future at the time when the progenitors of 
 the Germanic people began to have a separate subsistence 
 and dwelling-place. I mean, rather, that a distinct form for 
 this tense supposing it to have been, as I believe it was, al 
 ready in existence may perhaps have been lost by the primi 
 tive Germans at a very remote period, and thus the confusion 
 of present and future may have been among the earliest of 
 those deviations from the common type which at length im 
 pressed a peculiar character on the Germanic idioms. The 
 modern languages of this class have endeavored to supply the 
 
INDO-EUROPEAN FUTURES. 1 87 
 
 want of a future by the use of auxiliaries : of these attempts 
 we shall speak further on. We observe now, that a similar 
 deficiency exists in one of the Celtic languages, the Welsh ; 
 and not only so in the older forms of that language, as the 
 laws of Hywel Dda (Howel the Good), which belong to the 
 ninth century, but also in the current idiom of the present 
 day. In one verb only do we find a distinction made between 
 the present and the future in the verb to be. Sum, es, 
 est y etc., are expressed by tvyv y wyt, yw y or ydwyv, ydwyt, 
 ydyw : but ero, cris, crit are expressed by byddaf, byddit, 
 bydd, etc. The root bydd, employed for the future, is only 
 a modified form of bu, which is also found in Welsh, and, like 
 our English be, is identical in origin with Sanskrit bhu, Greek 
 <u-w, and Latin/?/-/. The added aspirate of bydd is perhaps 
 to be compared with the Greek 6 in forms such as 7r6\d9-a> = 
 TreXa-o), to draw near/ /uw0-&) = Lat. minu-o, to diminish/ 
 and many others. This formative 9 is now generally identi 
 fied with the root dhd, Gr. 6e in T/^/U, Eng. do. Germ. thnn. 
 If the dental in bydd is to be explained in this way, we should 
 have for the primary meaning of byddaf I do be/ I do or 
 make the being/ a present form of expression, which, how 
 ever, as it implies that the being is not already done, that the 
 existence of something is not already brought to pass, may 
 naturally enough be employed as a future. The Welsh forms 
 of the verb to be are much used in periphrastic expressions 
 for continued action, which are even more prominent in 
 Welsh than in English ; and in these, of course, it is able to 
 discriminate the future from the present : he will be loving 
 is expressed by bydd yn cant, he is loving by yw yn earn: 
 though he loves and he will love are both expressed by one 
 and the same form, car. W r e cannot, however, suppose in this 
 case what we saw to be at least not improbable for the Ger 
 manic idioms that the want of a special future form was an 
 original and distinctive feature in the class of languages, the 
 Celtic, of which the Welsh is a member. For in the Irish, the 
 Gaelic of Ireland, which in many respects has preserved the 
 primitive characteristics of Celtic speech better than the Welsh, 
 we find a regular formation for the future, one which goes back 
 evidently to the earliest periods of the language, and bears a 
 
1 88 INDO-EUROPEAN FUTURES. 
 
 remarkable analogy to the Latin future in bo. Thus from the 
 root mol (infinitive molain, to praise ) come the present, mo- 
 laim, molais, molaidh, l laudo, landas, landat ; and the future, 
 mo If ad (with f), violfair, molfaidh, laudabo, laudabis, lau- 
 dabit : while the Welsh has molaf, molit> uiaivl, for both 
 tenses. And yet the Welsh has a single femnant which 
 shows that it anciently participated in the Irish formation of 
 the future. Forms like moliff (with final f) are occasionally 
 met with, in the third person singular only, and always in a 
 future sense. 
 
 But while the Germanic idioms*, as a class, with on branch 
 of the Celtic, resemble the Semitic tongues in confounding 
 future and present, the great body of the Indo-European 
 languages have distinct forms for the expression of the fu 
 ture. In examining these forms, which present considera 
 ble variety, we notice first the fact that some of them have 
 been derived from a subjunctive or potential. The future is 
 uncertain ; it is dependent on conditions, either now exist 
 ing or to arise hereafter. We need not be surprised, there 
 fore, to see it expressed under the form of contingency ex 
 pressed as that which may or might be. A rudiment of 
 this practice is seen in the Greek language, where the sub 
 junctive, especially the aorist subjunctive, with or without the 
 contingent particle av or KV, is sometimes found in cases 
 where we should expect a future indicative. Thus in the first 
 book of the Iliad, Nestor, speaking of the heroes of an earlier 
 time, says : ov yap TTCO -rolovs tSov dvepa? ou8e o)<w/m, for never 
 yet saw I such men, nor (may I sec, am I likely to sec, i.e.] 
 shall I see hereafter. This usage, however, is nearly con 
 fined to the Epic poetry. But that which we see in Greek 
 occurring in this occasional and irregular way is matter of 
 regular formation in Latin. It is a fact now generally recog 
 nized, that the future indicative of the third and fourth con 
 jugations, and, of course, the future indicative of nearly all 
 primitive verbs, was originally a subjunctive that it corres 
 ponds to the potential in Sanskrit and the optative in Greek. 
 The characteristic of this mode in Sanskrit is yd or i : in Greek 
 it is iri or i (Solrjv, Solpev). In Latin it was the same as in Greek, 
 ic or i. These two forms are seen in the Old Latin siciii, and 
 
INDO-EUROPEAN FUTURES. 1 89 
 
 its later equivalent sim* The vowel i, used as mode-sign of 
 the subjunctive, is seen also in velim, nolim, malim, duim 
 dein, and perhaps one or two other old forms. It is implied 
 likewise in the present subjunctive of the first conjugation, 
 as amcmus, laudenms, if these forms are really, what they are 
 generally regarded as being, contracts made from amaimus, 
 laudaimus. This is the contraction which has occurred in 
 the future indicative of the third and fourth conjugations, and 
 has made legemits, luenms, out of legaimus and luaimus 
 forms which correspond perfectly to the Greek optatives 
 /,- Xvoipev (or more closely to the Doric forms, \eyoipes, 
 It is probable, too, that in the future of the verb 
 sum, in eris, erit y erimus, the i is in fact this same sign of the 
 contingent mode, attached to the syllable cr, originally es, the 
 stem of the verb. Now it is a curious fact that the Armenian 
 language, which belongs to the Iranian branch of the Indo- 
 European family, and is therefore quite remote from the Latin, 
 forms its future tense in a manner closely analogous to that 
 just described. The Armenian future has for its characteristic 
 a letter which is pronounced as ts : thus, from the stem da, to 
 give, comes dats, I shall give/ daises, .thou wilt give, 
 datse, he will give, etc. At first view, we should think of 
 this Armenian ts as connected with the Greek a of the fu 
 ture, and should compare dats, datses, datse with Scoaa), S&aeis, 
 Secret. But Bopp, in the second edition of his Comparative 
 Grammar, has shown that the ts which plays an important 
 part in the Armenian system of inflection has arisen every 
 where from an original Indo-European y consonant, by a 
 phonetic change, an assibilation, analogous to that which ap 
 pears in English joke from Latin jocns, English June from 
 Latin Junius. Thus dats, datses, datse correspond, not to 
 3a><7o>, t>cocret9, Soxret, but to Soiijv, 80/779, Sot?;, Sanskrit 
 deydsam, deyds, dcydt ; that is, they are potential or optative 
 forms, employed for the expression of the future. 
 
 We have just alluded to the common Greek future in <r, as 
 seen in a>o-o>, Swo-et?, Soocret. Identical with this is the ordinary 
 future of the Sanskrit: thus, from the stem dd, to give, we 
 have in the future ddsydmi, ddsyasi, ddsyati. In these forms, 
 comparative philology has long recognized the character of 
 
190 INDO-EUROPEAN FUTURES. 
 
 compound words, made by attaching to the stem da a future 
 of the verb to be. It is true that, in Sanskrit, the root as of 
 the substantive verb (Greek and Latin e?, Germanic is) has no 
 future existing as a separate form. It has a potential sydm, 
 syas, syat, for asyaui, etc., made by annexing the potential 
 characteristic^?/ to the radical as, which then loses its initial 
 a, as in the corresponding Old Latin sic in. Now in the ordi 
 nary Sanskrit future, we find the stem of the verb, whatever 
 it may be, followed by the forms sydmi, syasi, syati, which 
 very closely resemble the pgtential forms just given, differing 
 from them scarcely at all, except in having the full personal 
 endings ;///, si, ti, instead of the shortened ;//, s, t. It is 
 therefore natural to suppose that this sydmi, etc., was in its 
 origin identical with syaui / that it once existed as a separate 
 word, having at the outset a potential meaning ( may be, 
 might be ), but afterwards employed as a future ( shall be, 
 will be ), by a transition the same as that just pointed out in 
 Latin futures of the third conjugation ; and that in this use it 
 was compounded with other verb-stems to give them a future 
 meaning, and was thus retained in composition long after it 
 had ceased to exist in a separate state. The compound future 
 made in this way is seen not only in the Zend and the Greek, 
 but also in the Lithuanian, where we find, for example, dnsn, 
 I shall give, dusi, them wilt give, dns, he will give. It 
 will be observed that the proper future element in this forma 
 tion is not the letter s, but the potential ya which follows it, 
 and is very distinctly preserved in the Sanskrit syiimi, etc., 
 while in the common Greek aco, era?, crei it is nearly obliterated. 
 But in the Doric Greek, the endings of the future are crico 
 or erect), creet?, creet, which correspond well, according to the 
 phonetic relations of the two languages, to the Sanskrit end 
 ings. The Greek, on the other hand, has this superiority, 
 that it has retained the future of the stem as or e? in a sepa 
 rate form eao/uai, Epic eVcroO/iicu, from ecrcreoyu-at or ecrcricftat, 
 though with middle endings, instead of active. The Sanskrit 
 sometimes appends its future terminations directly to the 
 stem, but more generally inserts a short i as a connecting 
 vowel. In Greek this relation is reversed : occasionally we 
 find the connective, as in jj.a^6crofiai , from /xa^o/tat, to fight/ 
 
IND O-E UR OPE AN F UTURES. 1 9 1 
 
 but in general the ending is applied immediately to the stem. 
 In one case, however, the Greek almost invariably employs 
 a connecting e in the case of liquid verbs ; though here the 
 <r, standing between two vowels, has fallen away, as it does 
 so often in the Greek language. Thus from /*e z/r/>, to re 
 main, we have future IJLGVGCTM instead of jjievaw from which 
 comes jjLeveco, contracted fjievw, the usual form. 
 
 If now we return to the Latin, we shall find that the forma 
 tion just described, by which the future of the root as or es 
 appears as the final element in a compound verb form, is by 
 no means unknown to that language. It does not appear, 
 indeed, in the first future ; but in the future perfect it is the 
 established method of formation. Lcgcro, dixcro, docncro, are 
 nothing but the future of sum annexed to the stem of the per 
 fect active : they mean I shall be in the state of having read, 
 having said, having taught. Thus Icgcro in the active corres 
 ponds closely to Icctus cro in the passive, the difference being 
 that Icgcro is a true and proper compound, while in Icctus cro 
 the elements retain their separate form and construction. 
 
 We have not yet spoken of the Latin future in bo, which 
 appears regularly in the first and second conjugations, and 
 sometimes in the fourth amabo, docebo, audibo. This end 
 ing, too, like the one which we have been considering, con 
 tains the verb to be. It is not, however, derived like that 
 from the root as, but from another root no less widely diffused 
 through the Indo-European languages Sanskrit b/iu, Greek 
 <pv (<uft>), Latin /// (fui, f) c }> Germanic and Celtic bi, be. 
 It is true that the future ending bo shows the medial b, while 
 the separate fui, fore, etc. in Latin have the aspirate/"/ but 
 this circumstance involves no real difficulty. The primitive 
 sound was a medial aspirate, bh, as we find it in the Sanskrit. 
 In Latin, this bh either gives up its h and is reduced to the 
 medial b, or it fuses b and h in the homogeneous aspirate/: 
 in general, / is used for bh when it stands at the beginning of 
 a word, and b when it stands in the middle. Thus the root 
 bhar, to bear, gives fcro as a separate verb, but appears also 
 in the ending -bruin of candelabrum, candle-bearer, cere 
 brum, brain, lit. borne in the head a case which is 
 entirely parallel to the one in hand. It is matter of doubt 
 
192 INDO-EUROPEAN FUTURES. 
 
 whether bo, bis^M t ought to be regarded as present forms of 
 the root bhu used in the sense of futures, like what we have 
 seen before in the Welsh inolaf, etc., or whether they are 
 futures in form as well as in meaning 7-th at is, whether they 
 are not, like the Sanskrit syami, etc., potential forms applied 
 to designate futurity. In the latter case, bo, bis, bit might be 
 looked upon as contractions of bio, bits, biit, and this appears 
 to be the most probable explanation. 
 
 Closely analogous to this Latin future in bo, and resting 
 upon the same root, is the regular future formation of the 
 Irish, which we have had occasion to notice already. Here we 
 find the endings fad, fair, faidJi, famaoid, faidh, faid j in 
 which the derivation from the root bha may be regarded as 
 unquestionable. The endings in the ancient Irish, as given by 
 Zeuss (Gramm. Celt.), arc still more evidently connected with 
 the Latin forms ; thus nb (from bit), Lat. bo j fc (from y?), Lat. 
 bis j bid, fid, Lat. bit; bcm, fam, fcm, Lat. [limits; bid, fid, 
 Lat. bit is ; fct (for /<-;//), Lat. /;;////. That this formation was 
 not peculiar to the Irish, but belonged originally to all the 
 Celtic idioms, is proved by the fact likewise noticed before, 
 that a remnant of the same formation in the third person 
 singular is still found occurring occasionally in the Welsh. 
 
 These Latin and Celtic futures have been further illustrated 
 by Bopp from the Slavonic languages. In these the ordinary 
 future is made by combining a participle of the verb with the 
 future of the root /;;/, to be. Here, however, the two ele- 
 ments remain distinct, and do not coalesce into a single com 
 pound word, as in the Irish and Latin. Thus in the Car- 
 niolan we find bom igral, igrala, igralo, I will be playing, 
 he, she, or it that plays. The future of the root to be, 
 which occurs in these combinations, has a singular resem 
 blance to the Welsh future bydd, which we have considered 
 already. In the Old Slavonic it is bndn, which contains, if 
 Bopp is right, the radical of German thun and English do, so 
 that it would signify I do be, I do or make the being. 
 The Servian language stands distinguished from all its sister 
 idioms of the Slavonic class by the fact that it does not 
 confine this method to the substantive verb, but employs 
 it as a general mode for the formation of a future : thus, 
 
INDO-EUROPEAN 
 
 in the Servian, igra-dyit, ( I make a play 
 play. 
 
 The compound forms which we have thus far described 
 contain a future made from some root signifying to be, 
 which future is combined in one way or another with the dif 
 ferent verbs of the language. We might represent them all 
 loosely by such English expressions as he will be praising, 
 he will be giving, etc. But there is another method of pro 
 ceeding, in which the present of the substantive verb is used, 
 and which may be represented therefore by the English I 
 am about to praise, I am about to give, etc. Such are the 
 Latin daturus sum, laudaturus sum. And such is a form of 
 the future which appears in Sanskrit, though much less fre 
 quently than the one already analyzed. Thus the stem da, 
 to give, makes two futures : one ddsydmi, dasyasi, dasyati, 
 etc., corresponding, as we have seen, to the Greek Scoo-co, 
 &cret?, Secret ; the other datdsmi, ddtdsi, data, which an 
 swers to the Latin daturus sum, es, est. What is remarkable 
 in these Sanskrit forms is that data, except in this formation, 
 has nothing future in its meaning ; it is a common nomen 
 agcntis, and corresponds both in form and sense to the Latin 
 dator, giver. Indeed, it can scarcely be doubted that the 
 Latin daturus is in its origin only another form of dator, and 
 that the idea of futurity was attached to it by a sort of con 
 vention, as to the Sanskrit data. In the third person, singu 
 lar, dual, and plural, of this Sanskrit future, the verb sub 
 stantive is omitted, and we have only the singular, dual, and 
 plural of this nomen agoutis, so that in the dual, for instance, 
 we have only the expression two givers to signify they 
 two will give. It is also curious to observe that, although 
 data varies thus in the third person of the different numbers, 
 where it stands by itself, yet in the first and second persons, 
 where it is followed by the auxiliary, the form data, which 
 is the nominative singular, is used also for the other num 
 bers. It is as though in Latin we should say, not daturi 
 sumus, but daturus sumus. This shows that the Sanskrit ex 
 pression is not to be regarded, like the Latin, as a mere col 
 location of the substantive verb and the future participle ; 
 but that there is a real fusion, a quasi-composition in the case. 
 13 
 
194 INDO-EUROPEAN FUTURES. 
 
 Thus far it is the verb i to be which has figured in all the 
 futures, whether made by composition or by the use of sepa 
 rate auxiliaries, that have come under our notice. But other 
 verbs may be laid hold of by the language-making spirit, and 
 compelled to. perform a similar service. Among these is the 
 verb to have. The old Slavonic translation of the Evange 
 lists frequently uss the present of imam, to have, in con 
 nection with an infinitive, to form a future tense : thus, priiti 
 imaty syn == vcniet filius (lit. the son has to come ). The 
 idiom of our own language makes it easy for us to feel the 
 force of this periphrasis, which, as we use it, always looks for- 
 ^ward to the future, though combining with the proper idea of 
 tense the modal notion of obligation or necessity. A similar 
 periphrasis is occasionally found even in the Mceso-Gothic 
 of Ulfilas. But it is in the Romance languages that this 
 mode of <( futurizing" (if we may so call it) has shown itself on 
 the largest scale and with the greatest constancy. The 
 Italian, French, and Spanish future is nothing but a combi 
 nation of the infinitive with the present forms of the verb 
 haberc, to have. Thus loner ai, loucras, loucra, loneront, are 
 nothing but loner ( landarc] -f ai, as, a, out, the present 
 forms of the verb avoir. The origin of these forms, if any 
 doubt could be entertained in regard to it, is clearlv demon- 
 
 o J 
 
 strated by the usage of the Provei^al, in which there is often a 
 tmesis, or separation of the two elements, as in dar vos rial 
 (^jc i 1 cms en donncrai }, dir vos ai ( dicer c vobis Jiabco, I have 
 to say to you, jc Tons dirai }. The universality of this for 
 mation in the Romanic languages is a fine illustration of the 
 reality and power of a spracligeist, a mental attitude and 
 tendency common to all the speakers of a language, however 
 remote in place and disconnected in social relations, which 
 leads them to develop simultaneously and independently 
 similar modifications and new creations in their language. 
 No idea can be entertained, in this case, of a fashion spring 
 ing up in a particular locality, and propagated thence by in 
 tercourse and imitation. The Spanish future could not have 
 been borrowed from the Italian, nor the Italian from the 
 Spanish. The phenomenon is a natural, not an artificial one. 
 The Italian, French, and Spanish have alike grown out of a 
 
INDO-EUROPEAN FUTURES. 195 
 
 mixture of Germanism with a degenerate Latinism, and it 
 lies in the nature of the ingredients that, when thrown to 
 gether, they should precipitate among other things a future 
 such as this. 
 
 It is a little singular, however, that the Germanic languages 
 themselves, in their separate and independent development, 
 have not shown the same uniformity as to the creation of the 
 future tense which we have just seen in their influence upon 
 the corrupt Latin, which they altered and adopted in the 
 south of Europe. And yet, even here, the uniformity is too 
 great to be the result of accident. In seeking to supply that 
 ancient deficiency of special future forms which we have be 
 fore described, not one of these languages has adopted a com 
 pound inflexional form, like the Greek and Sanskrit future ; 
 not one has combined the verb" to be with a participle or a 
 periphrastic form, like the future in nearly all the Slavonic 
 languages ; not one has connected the verb to have with the 
 infinitive, like the future of the Romance languages. The 
 Germanic languages have taken up the verbs of volition and 
 necessity, and made their futures by the help of these auxili 
 aries. In low Dutch, it is the verb of necessity alone, the 
 shall, which is thus used to form the future. In Danish and 
 Swedish, both verbs are thus employed. And this is the case 
 in our classical English, which uses them not indiscriminately, 
 but according to a somewhat subtle and refined distinction. 
 In the English of Scotland and Ireland, and a large part of 
 the United States, it is the verb of volition alone, the will, 
 that discharges this office, the shall being never used except 
 with its primitive and proper notion of necessity. It is a 
 curious fact that the modern Greek has adopted a similar for 
 mation of the future, ^eXe*, va ypd(j)rj or $eXet va ypd-^y, 
 corresponding exactly to the he zuill write of our own 
 language. 
 
 Th ere is, however, one of the Germanic idioms which has 
 departed from the general analogy, and adopted a wholly differ 
 ent auxiliary. I refer to the modern German, and its use of 
 the verb ivcrdcn y to become, to express the future. What 
 makes this peculiarity the more striking is the fact that the 
 Althochdeutschy though standing to a great extent in the posi- 
 
I9 6 INDO-EUROPEAN FUTURES. 
 
 tion of the Mceso-Gothic and the Anglo-Saxon, and using its 
 present as a future, had begun to introduce the verbs of will 
 ing and shalling to their auxiliary function, like the other 
 languages of the same class. So that the HocJidcntsch lan 
 guage has actually thrown aside the general expedients of 
 its class, after it had, in part at least, adopted them. Bopp 
 congratulates his countrymen on this difference, as having 
 made a more felicitous choice than the rest of their family. 
 Wcrdcn is certainly not ill-adapted to the expression of the 
 future ; though its use as an auxiliary interferes somewhat 
 injuriously with its use as a separate verb. Yet the reason 
 chiefly urged by Bopp is a damaging one for his own argu 
 ment. " That which is now becoming," he says, " certainly 
 will be in the future." But, in the first place, this is not 
 true ; for the present evolution may be arrested. And next, 
 if it were true, the same thing might be said, with even 
 greater force, of the shall, the verb of necessity. 
 
 It remains only to add a word in reference to the historical 
 question, whether any of these formations can be referred to 
 the primitive period of Indo-European unity whether in the 
 common language, which afterwards, under different condi 
 tions of place, time, and circumstances, became Sanskrit, 
 Greek, German, and the rest, any of these forms were already 
 in use for the expression of the future or whether we are to 
 pronounce them all, as we must certainly pronounce many of 
 them, to be of later growth, developed in particular sections 
 of the Indo-European family, after these had separated them 
 selves from the common stock. In regard to one of them, 
 the future in s, there can be little doubt of its primitive an 
 tiquity. It occurs, as we have seen, with evident identity of 
 form and meaning, in Sanskrit, Greek, and Lithuanian. For 
 such an agreement in languages so remote from one another, 
 there can be but one satisfactory explanation. It is impossi 
 ble to regard it as the result of an accidental coincidence. 
 We are compelled to believe that the common formation was 
 derived by all the three from a common source ; and that can 
 only be the one common language of which they are all 
 descendants. This conclusion, if it needed any confirmation, 
 might receive it from an interesting fact, which seems to 
 
INDO-EUROPEAN FUTURES. 197 
 
 prove that the same mode of formation once existed also in 
 the Celtic. The Irish language presents us with an isolated 
 bhus, meaning * it will be, in which it appears safe to recog 
 nize with Bopp a future in s made from the root bu a form 
 strictly identical with Lithuanian bus, Greek (pvaei,., Zend 
 busyeiti, Sanskrit bhavishyati y and we may add Indo-European 
 bhusyati. But this conclusion as to the future in s seems to 
 draw after it, if not with necessity, yet with a good degree of 
 probability, a similar conclusion as to that formation of the 
 future which appears in the Latin third conjugation, and is to 
 be traced in the inflection of the Armenian verb. We have 
 seen that Sanskrit ddsydmi (Greek GO<J&>) compounds the 
 stem da with syaini or asyami y an old future of the root as y 
 and that this last was made by adding to that root the poten 
 tial ya, and was in fact a potential formation applied to desig 
 nate the future. If, then, the primitive Indo-European lan 
 guage made a future from the root as by applying to it the 
 potential ya, it is very likely to have formed other futures 
 from other roots by a similar process. It may have had these 
 two formations of the future existing side by side, and used 
 perhaps even in the same verbs. Thus the root dik, to 
 show, may have had the future dikayami y dikayasi, dikayati, 
 corresponding nearly to Latin dicam y dices, dicet, and, at the 
 same time, the future daiksydmi, daiksyasi, daiksyati, corre 
 sponding to the Greek BSL^M, Seifeis, Setfe*. 
 
 The only remaining formation which could be regarded as 
 having any plausible claim to a like primitive antiquity is the 
 Latin future in bo. Outside of the Latin, we find it only in 
 the Celtic languages ; and our opinion as to the period of its 
 development must depend a good deal on the relation which 
 we conceive to exist between the primitive Celtic and the 
 Latin. If these are as widely separated as the Greek and 
 Lithuanian, the fact that a form belonged alike to both of 
 them would go far to prove it an element of the common 
 Indo-European language. This would be a natural conclu 
 sion for those who hold, with Ebel and Lottner, that the 
 Celtic languages are connected by special affinities with those 
 of northern Europe, the Germanic and the Letto-Slavic : un 
 less, indeed, they should hold, as Lottner also does, that the 
 
198 INDO-EUROPEAN FUTURES. 
 
 Latin and the other Italic languages are in the same condi 
 tion, and stand nearer to Celtic, Germanic, and Letto-Slavic 
 than to the Greek. But there is a different view, set forth 
 recently by Schleicher (Beitrdge zur vergleichenden SpracJi- 
 forschung, vol. i.), which I am strongly inclined to believe 
 will be sustained by further study that the Celtic has a 
 special connection with the Grrcco-Latin languages, but more 
 particularly with the Latin and its sister languages of Italy. 
 And this formation of the future, which belongs both to the Italic 
 and the Celtic, and to these only, supplies him with one of 
 the main pillars for his theory. Another, similar to this, but 
 even stronger, he finds in the passive formed with r, which 
 belongs exclusively to these same languages. According to 
 this view, it would be natural to regard the future in bo as 
 having arisen on Italo-Celtic ground, rather than as imported 
 from the common home of the Indo-European family. 
 
XL 
 
 ON PASSIVE FORMATIONS. 
 1867. 
 
 I PROPOSE to occupy your time this evening for per 
 haps an hour with some observations on the formation of 
 the passive voice in different languages. The object will be 
 to show the nature and variety of the methods which have 
 been adopted by the makers of human language and it 
 must be remembered that all users of language are makers 
 of language to express the idea of the passive. 
 
 If knowledge is to follow the lead of charity, as St. Paul 
 would recommend, then knowledge must begin at home. 
 Before entering the Babel of outlandish languages, we will 
 linger awhile with our own kindly and intelligible Eng 
 lish. Yet we must confess that this good old mother-tongue, 
 with all its resources and faculties, has, strictly speaking, no 
 passive verb : I do not say " no passive," but "no passive 
 verb." . For such words as broken, struck, loved, praised, etc. 
 are not verbs : they are nouns, adjective nouns, attributive 
 words, appropriate to the object of an action, appropriate 
 to that which somebody breaks, strikes, loves, praises. 
 In themselves they are not predicate-words : they become 
 so only when they are connected with a verb. We 
 connect them with the verb to be, as he is praised, lie was 
 loved / and thus, if we do not strictly form a passive verb, 
 we at least supply the place of one ; we express under a 
 predicate-form the conception or relation of the passive. In 
 this respect we stand on the same footing as the Anglo- 
 Saxon. Our ancestors a thousand years ago were, like our 
 selves, without a passive verb, and they supplied the defi 
 ciency in the same way that we do. In one respect, how 
 ever, there is a difference. We have but one verb the 
 verb to be, in its various parts which we employ regularly 
 for this purpose. The Anglo-Saxon uses the same verb 
 
200 PASSIVE FORMATIONS. 
 
 very frequently ge gcliyrdon, that gccwcden wees, ye have 
 heard that it was said (was quothed, or quoted), etc. But fre 
 quently it uses another verb, one which has all but disap 
 peared from the present language. Every one will remem 
 ber the complaint of Fitz- James in Scott s Lady of the Lake : 
 "Woe worth the chase, woe worth the day, That cost thy 
 life, my gallant grey " that is to say Woe is the chase, or 
 rather, woe befalls it, woe betides it. This worth is a 
 remnant of the old Anglo-Saxon weorthan, to become, 
 which was often used, like bcbn, to be, for expressing the 
 passive. Thus " when his father saw him (the Prodigal Son), 
 he was stirred with mild-heartedness lie wcctrtJi mid 
 mildheortnisse astyrod ; " strictly, he became, he came to be 
 excited with compassion. Now these two verbs, to be and 
 to wort h (if I may give it so"), were similarly used in the older 
 forms of the High German, the Low German, and the Dutch, 
 the idioms most closely related to the Anglo-Saxon. But 
 in the modern forms of all these languages, the verb to be has 
 lost this use, and only the verb to worth is employed for ex 
 pressing the passive. 
 
 Thus, for he is loved, the German says, not cr ist gcliebt, 
 but cr wird geliebt. If we find in German such an expres 
 sion as cr ist geliebt) the participle has sunk into a mere adjec 
 tive, as though one should say " he is dear," " he is accepta 
 ble." These forms arc evidently quite different from the real 
 passive, in which we have the idea of an action proceeding 
 from an agent to an object : as, he is loved by his friends, 
 where the German can only say cr wird geliebt von scinen 
 Freimden. It is certainly a curious circumstance that out of 
 the two auxiliaries which were originally employed both by our 
 own language and by its nearest kindred, the English should 
 have adopted the one which all the rest have discarded at 
 least, in this use of it. But it is a fact which admits of a 
 plausible explanation. The English language has not been 
 left to an independent and spontaneous development. 
 Brought by the Norman conquest into contact with the 
 French, it shows many traces of the exotic influence. It is 
 not strange if, in a choice between two forms of expression al 
 ready existing, the result should be determined by the an- 
 
PASSIVE FORMA TIONS. 2O I 
 
 alogy of the intruding language. If the French had used 
 devenir, to become, or some word of similar meaning, in 
 the expression of the passive if in the present passive of 
 loner, to praise, they had said il dcvicnt loue instead of il 
 est lone it is probable that we ourselves should now be 
 saying he worth praised (German, cr wird gelobf) instead of 
 he is praised. But the French language, like its sisters of 
 Italy and Spain, makes the passive by usingf the verb to be. 
 And for this also there is an obvious historical reason. The 
 Latin, from which they are all derived, which stands to them 
 in much the same relation as Anglo-Saxon to English, forms 
 a large part of its passive in the same way. The method 
 which the Latin adopted in its complete tenses Perf. amatus 
 sinn or fui y Plup. amatus cram or fueram, etc., etc. this 
 method is extended in the Romance languages to all the 
 tenses, the incomplete as well as the complete. In this pro 
 cess the Latin participle in tus has lost the idea of time which 
 generally belonged to it in Latin ; it has changed from a per 
 fect participle passive to a passive participle pure and simple. 
 The Latin laudatus esl means (commonly, at least) he is the 
 object of a past act of praising, i.e. he has been praised ; 
 whereas the French il cst loue means he is the object of an 
 act of praising, i. c. he is praised. 
 
 If now we pass on to the Latin passive of the incomplete 
 tenses, we come for the first time to a true passive verb, an in 
 flexional form which is in itself a full predicate, and which 
 is made to take for its subject the real object of the action. 
 Its characteristic element is the r at or near the end of the 
 form, which appears in all modes and tenses : as amor, amabar, 
 amabor, amcr, amarcr, amari, etc. ; and in all numbers and 
 persons : as amor, amaris or -re, amatnr, amamnr, amantur. 
 This statement, however, suggests a very notable exception. 
 The second person plural amamini (amabamini, amemini] is 
 in glaring contrast with the analogy of the other forms. It is in 
 fact a very different thing, being simply a participle, the no 
 minative plural masculine of a lost participle. It is identical 
 with the Greek participles in /ze^o? which are so abundant in the 
 middle and passive voices : as, from \u-co, to loose, /Vuo /ze^o?, 
 , \vad[JLevos, \v6r)(ToiJievos, etc., etc. The Latin has 
 
202 PASSIVE FORMATIONS. 
 
 other traces of this participle, as for instance mfemina, which 
 appears to be the participle of a root fc, meaning to milk, 
 suckle. The familiar word alumnus is equivalent to alnminus> 
 a participle of the verb alo, meaning one who is nourished or 
 reared up. Apparently the Romans must once have used #;;*#- 
 mini estis for ye are loved ; but when the participle ceased to 
 be employed in other uses, the true character of amamini was 
 forgotten, it dropped the verb cstis, which would then appear 
 unnecessary ; and, though masculine in form, was applied to 
 a feminine subject not less readily (perhaps, if the sense of 
 the word is considered, more readily) than to a masculine. 
 
 The passive in r is one of the most remarkable features of 
 the Latin language. It is not, however, confined to the Latin. 
 It is found in the scanty remains of the Oscan and Umbrian, 
 and there can be no doubt that it was common to that whole 
 class of languages the Italian class, as it is called the an 
 cient idioms of Middle Italy. Outside of this class, it is found 
 in the Celtic languages, and in those only. The Welsh and 
 Irish, down even to the present day, have a passive in ;-, and 
 the forms of this passive, as they are found in the oldest 
 written monuments of the Irish language, dating back about a 
 thousand years, have a marvellous resemblance to those of the 
 Latin. In the Celtic, as in the Latin, the formation in r does 
 not extend to the perfect tense. In the Celtic, as in the 
 Latin, the formation in r is excluded from the second person 
 plural. Both in the perfect tense, and in the second person 
 plural, the Celtic, like the Latin, uses forms which appear to 
 have been originally participles. These surprising coincidences 
 in reference to the passive are the main prop of the hypothesis 
 recently propounded, according to which the Celtic languages 
 and the Italican stand in a special and close relation to each 
 other a closer relation than that which the Latin, for in 
 stance, bears to the Greek, or the Celtic to the German. 
 
 The question comes, now r what is the nature of this passive 
 in r what are its origin and its meaning? Before attempting 
 an answer to this question, let us notice some facts which are 
 not without their bearing on that answer. 
 
 The Italian language sometimes uses an active verb with 
 the reflexive pronoun si, where \ve should employ the pas- 
 
PASSIVE FORMATIONS. 203 
 
 sive. Thus the sentence by a good man virtue is not loved 
 on account of utility may be expressed in Italian, da un 
 uomo buono 11011 si aina la virtu per r iitile i. c., virtue does 
 not love itself (Lat. 11011 se amat virtus}. The Spanish has a 
 similar idiom : the sentence many lies are told may be ren 
 dered by sc dicen muchas mcntiras (Lat. sc dicunt mult a mcn- 
 dacid] i. e. many lies tell themselves, or, get themselves 
 told. Were this the only instance of the kind, one might 
 suppose that the Spanish people wished to relieve themselves 
 from apprehended blame by representing their little devia 
 tions from truth as self-produced. But the idiom is familiar 
 to these languages : it is, however, confined in them to the 
 3d person singular and plural. But there is one daughter of 
 the Latin, a forlorn and neglected child, which is said to 
 use this as its regular form for expressing the passive in all 
 persons and numbers. I refer to the language of the Wal- 
 lachians in Hungary and Turkey, the descendants of the 
 Romanized Dacians. For I am praised, these people say 
 jo me lander, i. e. I praise myself (Lat. ego me laudd] ; for 
 them art praised, tit te laudi, thou praisest thyself; for 
 he is praised, el se laudd, he praises himself, etc., etc. 
 
 The same phenomenon presents itself again in a language 
 which as regards place lies very near to the Wallachian, 
 though belonging to a different branch of the Indo-European 
 family the Old Slavonic, as it is called, or the Old Bulgarian, 
 the language of the Bulgarians a thousand years ago. The 
 Old Slavonic passive is made by using with the active verb 
 the word sen, which is plainly a reflexive pronoun, like the 
 Latin sc and the German sick. But while the Wallachian, as 
 we said, uses different pronouns for the different persons me 
 for the first, te for the second, and sc only for the third the 
 Old Slavonic uses sen for all persons of both numbers.* Sen 
 thus used does not really signify myself, thyself, himself, 
 etc., though we have to translate it by those pronouns: it 
 signifies merely self; it expresses the reflexive relation and 
 nothing further. 
 
 * Thus citun, I honor, citun sen, I am honored, lit. I honor myself; ci- 
 tesi, thou honorest, citesi sen ( thou honorest thyself), thou art honored; 
 cited, he honors, citeti sen ( he honors himself), he is honored. 
 
204 PASSIVE FORMATIONS. 
 
 If I am not mistaken, the reflexive passive here described 
 is found in all the modern languages of the Slavonic class, 
 the Russian, Polish, Bohemian, Servian, etc. 
 
 Nearly akin to the Slavonic idioms is the Lithuanian, a 
 language of scarcely any literary cultivation, but of much 
 interest and importance to the philologist. The Slavonic sen 
 is reduced in the Lithuanian to a mere s, which of course 
 does not form a word by itself, but is attached as a final 
 sound to the forms of the active verb.* 
 
 Now these forms in s, vczits, vczes, vcsas, and the rest, are 
 not passive the Lithuanian makes its passive by connecting 
 the verb to be (csmi, I am ) with a passive participle. 
 They have a reflexive sense, I bear myself, them bcarest thy 
 self, etc. they constitute a reflexive voice, like the middle 
 voice of the Greek. But I have brought them in here for the 
 sake of showing how the reflexive pronoun, reduced to a 
 single letter, may attach itself as a grammatical ending to the 
 inflected forms of an active verb. The bearing of this will be 
 obvious as we proceed. 
 
 The Scandinavian languages the Old Icelandic or Old 
 Norse, with its modern descendants, the Swedish and the 
 Danish form one section of that Teutonic family to which 
 belong the German and the English. But as regards the sub 
 ject now before us, these Scandinavian languages are strik 
 ingly distinguished from the other members of that family. 
 The Swedish and Danish have a passive in s, a passive made 
 by adding^ to the forms of the active. Thus, in Swedish, from 
 trycka, to press, or I press, comes the passive tryckas, 
 to be pressed, or I am pressed ; from pj-isadc, I praised, 
 the passive firtsades, I was praised, etc. Instead of this s, 
 the Old Icelandic has st. Thus skjota, to shoot, throw, 
 passive skjotast, to be thrown, to fall; Impf. skaut, he 
 shot, threw, passive skaut-st, he was thrown. But in the 
 oldest manuscripts we find sk instead of st : skjotask, to be 
 thrown, skantsk, he was thrown. And it may be regarded 
 as perfectly certain that this sk is only a shortened form of 
 
 * Thus, from vezu, I cai ry (Lat. ve/io), with added .c, comes the form vezu-s ; 
 from vezi, thou earnest, comes veze-s ; from vcza, he carries, veza-s and so 
 on, through the dual and plural. 
 
PASSIVE FORMATIONS. 205 
 
 the reflexive pronoun sik, the same as German sick. The 
 formation here described is in trie highest degree interesting, 
 because the process goes on, so to speak, under our own eyes. 
 We see the reflexive pronoun sik losing its separate character, 
 and, under the shortened form sk, attaching itself to the 
 active verb, which it converts into a passive : then, uncon 
 scious of its origin, passing for easier pronunciation into st ; 
 and finally, going into the yet easier s, which only a philolo 
 gist would think of as ever having been a separate word. 
 
 After this somewhat lengthy detour, which has taken us 
 the round of Europe, we come back to our Latin passive in 
 r. This is regarded now by the students of comparative 
 philology as a formation of precisely the same nature as the 
 one just considered : it is believed to have been formed by 
 adding the reflexive sc, or at least the s of it, to the forms of the 
 active. Thus landor is supposed to be for laudos, or lando sc, 
 I praise self (myself) ; laudaris for landasis, or laudas(i)se> 
 thou praisest self (thyself; ; laudatur for laudatus, or landa- 
 t(ii]se, he praises self (himself), etc., etc. The change of s 
 to r is one of the most familiar phenomena of the Latin 
 language. It occurs most frequently between vowels, as in 
 genus, generis, gcncri (for genesis, gcnesi] ; in cram, I was 
 (for esain], from the root cs, which appears in cs-sc, cs-to, cs-tis ; 
 and a multitude of other cases. Yet it is by no means un 
 known at the end of a word : we have both arbos 
 and arbor, tree, both honos and honor, honor, where the 
 form in s is unquestionably the older of the two. We can 
 therefore easily believe that laudor should be the later form 
 for an earlier landos, and the rest in the same way. It must 
 be confessed, indeed, that the case is not quite so clear a one 
 as this statement would seem to make it. For we have to 
 consider not the Latin only, but the other Italican languages, 
 the Oscan, Umbrian, etc., which have the passive in r ; and 
 not these only, but the Celtic languages, the Welsh, Irish, 
 etc., which have it likewise. Most of these other languages 
 do not show the tendency so common in Latin to exchange s 
 for r. Nor is there any reason to suppose that the parent 
 language from which the various Italican or Celtic idioms de 
 rived their origin had any general tendency to exchange s for 
 
206 PASSIVE FORMATIONS. 
 
 r. It is enough to point out this difficulty : I need not go 
 into the considerations by which its force perhaps may be in 
 some measure abated. It seems impossible to explain the 
 Latin passive in any other way ; to explain it in this way, as 
 a reflexive formation, is at once natural in itself and support 
 ed by strong analogies from various quarters. Philologists, 
 therefore, notwithstanding the difficulty referred to, have 
 generally adopted this explanation. 
 
 I -will only add that this way of accounting for the passive 
 throws a welcome light on the deponent verbs. Why, it may 
 be asked, should verbs of purely active meaning such as 
 conor, ( to endeavor, rcminiscor y to remember, ntor, to 
 use, orior, to rise, /;rcw, to pray, etc., etc. why should 
 they receive a passive ending ? The best answer would be 
 to deny the facts : these verbs have not a purely active mean 
 ing, and they have not received a passive ending. They are 
 verbs of reflexive meaning, and they have received, as it is 
 most proper that they should, a reflexive ending. Thus 
 conor means to exert one s self, reminiscor, to remind one s 
 self, utor, to serve one s self (with something, whence it is 
 followed by an instrumental ablative), orior, to raise one s 
 self, prccoi j to ask for one s self, etc. 
 
 From Latin to Greek is a natural transition. Here the 
 fact which strikes us first is that the Greek passive is to a 
 great extent borrowed from the middle or reflexive voice. If 
 there are any here who were brought up under the old regime in 
 Greek grammar and fed upon imaginary forms of TUTTTW, they 
 will perhaps be surprised by this statement. For they have 
 learned that 7V7TTo/j.ai and ervirro^v are true passive forms, 
 which the middle, to supply its own deficiencies, is compelled 
 to borrow 7 . But they must know that all this is changed now- 
 a-days. There has been a revolution, in which the passive 
 (fulfilling its own name) has been a sufferer, and the long de 
 frauded middle has come to its just rights. The passive now 
 has but two tenses of its own, its aorist audits future : for all 
 the rest, for present and imperfect, perfect and pluperfect 
 and future-perfect this last being the " paulo-post future " of 
 the old grammars for all these it must acknowledge its de 
 pendence upon the middle. In other words, it is now ac- 
 
PASSIVE FORMATIONS. 2O/ 
 
 kncwledged that such forms as present \ovopai, imperfect 
 eXovbpJiv, perfect \e\ovpai,, pluperfect eKeXovptjv, are primarily 
 reflexive forms, and that their passive use is secondary : that 
 in \ovopai, for example, the first meaning was I wash my 
 self, and the second I am washed, precisely as the Latin 
 favor, according to the explanation just given, must have 
 meant * I wash myself, before it meant I am washed. Now 
 the peculiarity of the Greek middle as compared with the 
 active its formal peculiarity lies in its personal endings ; it is 
 the personal-endings which distinguish the middle voice as 
 such. Thus, in the singular of the active, the original end 
 ings were pi, &i, ri: we see them in the verb to be - 1st 
 person dpi (for ecr-pi), I am, 2d ecr-crl (Horn.), * thou art, 
 3d ecr-W, he is. It is obvious that /-a, CTL, ri are pronouns, and 
 express the subject of the verb : thus, ecr-pl -- is me or I 
 am ; e<r-o-i is thec or thou art ; eV-Tt = is that or that 
 ,(one) is, he is. For pi, cri, TI, the middle has pai, aai, rat, ; 
 as in i crrapai, I o-racrai, iuraiai. Why, then, should an active 
 form become middle by simply changing pi, ai, ri to pai, crai, 
 rai? On this point there are two views, each of which h^s 
 its adherents. According to one view, pat,, aai,rai are mere 
 ly lengthened or strengthened forms of LJU, crt ? rt, and by their 
 greater length indicate the greater importance which belongs 
 to the grammatical subject in the reflexive verb. For, in the 
 expression / wash myself^ the first person, the I, is a very 
 important person, being at once subject and object of the ac 
 tion, its original and end. The importance of this person 
 may be expressed, not unfitly, by strengthening the syllable 
 which represents it. According to the other view, pai, crai, 
 rai are modified forms of pa-pi, ea-ai, ra-ri, in each of which 
 the person is expressed twice, first as object, then as subject. 
 Thus Xovopai, for \ovo-pa -pi, would mean wash-me-I, i.e. 
 I wash myself; \ove(cr)ai, for \ove-cra-cn,, \vash-thee-thou, 
 i.e. thou washest thyself; \oi e-rai, -for Xove-ra-n, ( wash- 
 him-he, i.e. he washes himself. I shall not discuss these 
 two views ; neither of them can be regarded as anything more 
 than a hypothesis, until it is shown that some similar explana 
 tion can be given for the other personal endings. 
 
 I remark, however, that a passive formed in the same way 
 
208 PASSIVE FORMATIONS, 
 
 once existed in our own Teutonic family of languages. 
 There is no trace of it, indeed, in the Anglo-Saxon, nor in 
 the Old Icelandic, nor in the Old High German : but if we 
 go back beyond all these to the earliest monument of 
 our family, to the Gothic Bible-translation made by Bishop 
 Ulfilas in the 4th century, we find it there. The Gothic 
 bairada, he is borne/ corresponds exactly in every respect 
 to the Greek Qeperai ; the Gothic bairanda, they are borne, 
 corresponds in the same way to the Greek fyipovrai. It is appa 
 rent that in the time of Ulfilas the Gothic was already beginning 
 to lose this formation. It is only found in the present, in 
 dicative and subjunctive : even here the first person singular 
 is gone, its place being supplied by the third, while in the 
 plural both the first and second persons have disappeared, 
 and the third does duty for them. It is not surprising, there 
 fore, that succeeding monuments of the Teutonic languages, 
 the earliest of which are some three centuries later, should 
 show no trace of this formation. If the Bible of Ulfilas, which 
 survives only in scanty fragments, had perished altogether, 
 there would be no evidence that this middle-passive was 
 for ages the common possession of Teutons and Grecians. 
 
 I have said that, while the Greek passive borrows most of 
 its tenses from the middle or reflexive voice it has two an 
 aorist and future which are peculiar to itself. But before 
 looking at these we will glance at the ancient language of 
 India. The Sanskrit passive, as compared with the Greek, 
 shows remarkable coincidences along with remarkable differ 
 ences. In both languages the perfect passive is borrowed 
 from the middle. But in the aorist and future, where the 
 Greek has a distinct form for the passive, the Sanskrit does 
 not generally distinguish it from the middle. And on the 
 other hand, in the present and imperfect, where the Greek 
 passive is always the same as the middle, the Sanskrit has a 
 distinct formation for the passive. This is made by adding to 
 the stem of the verb the syllable ya, which always takes the 
 accent. Thus, from tnd, to strike/ a root which appears in 
 the Latin verb tiuido, the Sanskrit makes tud-yd-tc, he is 
 struck/ tud-ya-ntc, they are struck. Now the syllabic ya 
 occurs in Sanskrit as the root of a separate verb, which means 
 
PASSIVE FORMATIONS. 209 
 
 to go : and it is believed that the formative ya of the pas 
 sive is no other than this root yd, to go. If this be so, 
 tudyate would mean he goes to a striking, he meets or in 
 curs a striking. A curious parallel as regards the meaning 
 is found in the familiar Latin veneo, to be sold, infin. venire, 
 undoubtedly for vennm co, vermin ire, to go to sale, and 
 hence to be sold. 
 
 This Sanskrit passive reappears in the Armenian, where 
 the syllable j/# is reduced to the simple vowel i. Probably 
 also it reappears in the Greek, where the second aorist pas 
 sive is formed by adding to the stem an e or 77, which may 
 easily be connected with the Sanskrit ya. Thus eVX^y//^, 
 from 7r\ij(7o-a), to beat, would mean I went to a beating, I 
 was beaten. As regards the Greek first aorist passive, the 
 aorist in OTJV, that is still an enigma : solutions have been pro 
 posed for it, but none that can be accepted as satisfactory. 
 
 We have now reviewed the modes of passive formation 
 which appear in the Indo-European class of languages. And 
 we have seen that one of the most common modes, perhaps 
 the most common of all, is the use of some form which was 
 originally and properly reflexive. For he is praised it has 
 been, and still is, a very common thing to say he praises him 
 self. If we inquire into the rationale of this expression, we 
 shall easily recognize in it the principle of qui facit per alium 
 facit per se. One who makes others praise him praises him 
 self. I praise myself, if I proceed and act in such a way as 
 to get praise from those around me. And this is tantamount 
 to saying that, if I am praised, I praise myself. What I wish 
 .to say is, that the reflexive which serves to express the pas 
 sive is a causal reflexive. The Latin laudor, the Swedish 
 prisas, the Greek eiraivovfjicn,, all mean I make men praise 
 me, I let men praise me, I get men to praise me, or something 
 of that kind. They are causal reflexives. In this view, it is a 
 significant circumstance that the Sanskrit, which has both a 
 causal and a passive, forms them both in very much the same 
 way : it adds ya to form the passive, and ay a to form the 
 causal. But if we go outside the bounds of the Indo-European 
 family, we shall find cases more striking and unequivocal of a 
 connection between causal and passive. The Magyar, or Hun- 
 H 
 
2 1 o PASSIVE FORMA TIONS. 
 
 garian, belongs, as all know, to that class of languages which is 
 variously called Tartaric or Altaic or Turanian ; the same class 
 to which belong the Finnish and the Turkish and the Manchu. 
 The Magyar forms both the causal and the passive by adding 
 the same syllable tat or at (tct or cf) to the root of the verb : 
 thus, from the root Idt, meaning to see, comes Idt-tat, which 
 is used both for the causal and for the passive. How, then, 
 are these two uses distinguished from each other ? By 
 the endings. The Magyar has endings which are applied 
 only to transitive verbs, such as to love, praise, strike, 
 etc., and other endings which are applied only to intransi 
 tive verbs, such as to sleep, dream, walk, fall, etc., verbs 
 in which the action is confined to the sphere of the agent or 
 subject, affecting no other object. If then Idttat, which is 
 made by adding the formative syllable tat to the root ldt t to 
 sec, is to be used as a causal, it takes the transitive endings : 
 Idttat ok, I cause to see, i. e. I cause a seeing of which (not I 
 myself, but) something else is the object. But if Idttat is to be 
 used as passive, it takes the intransitive endings : lattat o?n t 
 I am seen. i. c. I cause a seeing which has no other object 
 than me, I cause a seeing which affects myself only. In like 
 manner the Finnish, the Esthonian, and other languages 
 which stand nearest to the Magyar, show a close connection 
 between causal and passive, though they do not exhibit the 
 nature of that connection with the same beautiful distinctness. 
 But there is a language spoken by Indians of South America, 
 the Arauack language, which seems to be more transparent 
 in this respect than even the Magyar. It forms a reflexive 
 by adding niia to the root of the verb : thus, from assukussun, 
 to wash, comes assnkussunnua, to wash one s self. It 
 forms a causal by adding knttun to the stem : thus, from 
 assuktissuu, to wash, comes assukussukuttun y to cause to 
 wash. Now the passive is made by combining these two 
 signs, kuttnn for the causal and nna for the reflexive ; it adds 
 kuttunnua to the stem of the verb : dssukussu-kuttun-nua t to 
 cause to wash one s self, to cause a washing of which the 
 object is one s self/, c. to be washed. 
 
 It sometimes happens that an intransitive verb is used to 
 express the passive. Thus, instead of "he was killed in 
 
PASSIVE FORMA TIONS. 2 1 1 
 
 battle" we often say "he fell in battle." The Greek hardly 
 ever uses the passive aorist of Kreivw, to kill, but substitutes 
 for it the aorist active of the intransitive aTToOvr^a-Kci) : aTredave, 
 he died, instead of e/crdOrj, he was killed. Most languages, 
 doubtless, would show occasional instances of this kind ; but 
 in some languages the idiom assumes a more definite and 
 formal character. This is the case in Gothic, which, as we 
 have seen, had begun to lose the old middle-passive that 
 belonged to it in ccmmon with the Greek. It often uses in 
 place of that a peculiar series of intransitives, made by adding 
 na to the stem of the active verb. Thus, vei/ian, to conse 
 crate, but veihnan, to be consecrated ; aft air an, to tear 
 off, but aftaurnan, to be torn off; fraliusan (Germ, ver- 
 licreii)S to lose, but fralusnan, to be lost, to be forlorn. 
 In the Semitic languages (Hebrew, Arabic, Syriac, etc.), 
 passives are made to a great extent by merely changing the 
 vowels of the active verb : thus Arabic qatala, he killed/ 
 but qutila, he was killed. Passives such as these are 
 properly regarded as intransitives used to express the pas 
 sive. This view of their nature may perhaps account for the 
 singular fact that, in the Arabic, passives like qutila cannot be 
 used when the agent or doer of the action is to be expressed. 
 If I wish to say in Arabic the vizier was killed, I can use 
 the form qutila : qutila Tvazlru ; but if I wish to say the 
 vizier was killed by the sultan/ I can no longer use qutila I 
 must take some other form of expression. If qutila had the 
 nature of a true and complete passive, it could hardly be sub 
 ject to a restriction like this. 
 
 In English and the same thing is more or less true of 
 other languages there are some verbs which may be used 
 without any change of form both actively and passively. 
 Thus, "the water filled the ship," and "the ship filled with 
 water;" "the bad air sickened him," and "he sickened 
 with the bad air;" "the woman washes the calico," and 
 "the calico washes well," etc. In one auxiliary formation, 
 this double use as active or passive belongs to the established 
 idiom of our language. We say "he is building the house," 
 and "the house is building;" "while he was doing these 
 things," and " while these things were doing." It is in order 
 
212 PASSIVE FORMATIONS. 
 
 to avoid the ambiguity of this form that men have latterly 
 begun to say " the house is being built," " these things were 
 being done." These forms of the continued passive have 
 had to undergo much severe criticism, and it can hardly be 
 denied that they are somewhat illogical and perhaps yet more 
 unwieldy. Still, they have their convenience : probably 
 there is no man who, if the fear of the grammarians were 
 taken from him, would not occasionally use them. Like 
 other things which are homely and useful, they will be likely 
 to maintain themselves against all opposition, at least in the 
 language of conversation and extemporaneous speaking, 
 and may possibly gain a place at last even in polite litera 
 ture. 
 
 There are languages, however, in which this indifferent use 
 of the same form as active or passive is carried to a yet great 
 er extent. In the Tibetan language, all transitive verbs can 
 be used as passives ; and indeed, this is by far the most fre 
 quent use of them. It is not strange that in monosyllabic 
 languages, like the Chinese, which are of course destitute of 
 all inflection, the same verb should be employed at pleasure 
 as active or passive. The Chinese, however, does contrive to 
 make a sort of passive, by prefixing to the verb a certain 
 relative particle : thus, for it is seen the Chinese can say it 
 is what see ; for it is heard, he can say it is what hear, 
 and the like. 
 
 It might be supposed that when we have come to a language 
 which has no special form for the passive, but uses the same 
 form as active or passive indifferently, we have reached the 
 minimum point. But this is not the fact. There lies beyond 
 it an absolute zero. There are languages which not only 
 have no form of expression for the passive, but have not even 
 the idea of the passive. Such is the language of the Osage 
 Indians, in which the words " knock, and it shall be opened 
 to you " have to be translated by, knock and they will open 
 to you. So, too, the language of the Dacotah Indians, in 
 which the words "Jesus came to John and was baptized in 
 Jordan," are rendered Jesus came to John_and they baptized 
 him in Jordan. In these idioms, the agent or doer of the action 
 always appears in the nominative, and if the agent is unknown 
 
PASSIVE FORMATIONS. 
 
 213 
 
 or unthought of, the indefinite third person plural takes the 
 place of it : as they will open it, instead of " it will be 
 opened;" they baptized him, instead of" he was baptized." 
 
 On the other hand, there are languages which show a 
 marked predilection for the passive. This is true of the Ti 
 betan, just mentioned, which with transitive verbs regularly 
 employs the passive construction, except in certain limited 
 and comparatively unfrequent cases. Sometimes languages 
 which are nearly related to each other differ widely in this 
 respect. Thus, in Western Africa, the Mpongwe has a de 
 cided preference for the passive, while the kindred Bakele as 
 decidedly prefers the active, and the Benga employs both 
 active and passive with equal readiness. 
 
 We have noticed that there are passive formations which 
 are never used where the agent or doer of the action is to be 
 mentioned in the same connection. But as a general thing, 
 the agent is or can be introduced in some dependent con 
 struction along with the passive. In such cases, the relation 
 of the agent to the action is differently conceived and ex 
 pressed in different languages. In the largest number, per 
 haps, the agent is thought of as the source or starting-point 
 from which the action proceeds. Thus in Latin, discipulus 
 laudatur a magistro, the pupil is praised by the master, 
 literally from the master. So in the modern descendants of 
 the Latin : Italian dal maestro, French du maitre, etc. So in 
 German, von dcm Meister, and so in a multitude of other lan 
 guages. Some languages (as the Tibetan), which have an in 
 strumental case, use it for the agent with a passive verb. In 
 like manner we, in English, use the instrumental preposition 
 by ; though it must be observed that the instrumental use of 
 this preposition is a secondary one. Primarily it denotes 
 nearness or association : the instrument is that with which, in 
 presence of which, in connection with which, something 
 takes place. "The resolution was read by the chairman," 
 simply represents the chairman as near to the reading, and 
 closely connected with it. In Greek, the preposition VTTO, 
 under, is placed before the agent : he is thought of, appa 
 rently, as standing over his work, supervising it and having it 
 under his control Not a few languages put the agent in the 
 
2 1 4 PASSIVE FORMA TIONS. 
 
 genitive a pretty clear sign that the so-called passive is then 
 in reality a verbal noun. 
 
 A curious statement is made in regard to certain languages of 
 the Philippine islands. They have three different forms of 
 the passive verb, according as the subject of the passive is the 
 object of the action, or the instrument of the action, or the 
 place of the action. Thus a sentence like "the servant 
 searched for the book with the light in the chamber," can in 
 those languages be expressed passively in three different 
 ways, with three different forms of the passive verb, accord 
 ing as we take the book, the light, or the chamber, for the 
 subject-nominative. In all of them the agent (that is, the 
 servant) will be put in the genitive ; and this shows that the 
 three passive forms are in reality verbal nouns denoting the 
 object, the instrument, or the place of an action. The first 
 is analogous to the Latin alumnus* the object of nourishing, 
 the person nourished; the second to the Latin aratrum, the 
 plough, the instrument of ploughing ; the third to the Latin 
 auditorium, the place of hearing. The three passive senten 
 ces, then, in the Philippine language, would be exactly ren 
 dcrcd into English as follows : 1st. The servant s object of 
 search was the book with the light in the chamber ; 2d. The 
 servant s instrument of search foj the book was the light in 
 the chamber ; 3d. * The servant s place of search for the 
 book with the light was the chamber/ 
 
XII. 
 
 REMARKS ON THE USES OF THE LATIN SUB 
 JUNCTIVE. 
 
 1870. 
 
 IN all the uses of the subjunctive there is a common nega 
 tive element. It never expresses the conception of reality. 
 This conception, which is always present with the indicative, 
 is always absent from the subjunctive. But in the uses of 
 the subjunctive there are also positive elements ; there are 
 special conceptions which it is fitted to express. The first of 
 these probably in the order of time is that of wishing. There 
 is no doubt that si m, vclini, cdim, duim contain the formative 
 i or} a of the Sanskrit potential, the Greek optative, the Teu 
 tonic subjunctive. So also the present subjunctives \ncm t es, 
 et t etc., of the first conjugation, and possibly those in am, as, 
 at, etc., of the other conjugations ; though the latter are gen 
 erally supposed to correspond to the Greek subjunctive and 
 the Sanskrit Let-form of the Vedic language. As to the 
 other tenses of the Latin subjunctive, while the imperfect at 
 least is still involved in much obscurity, it can hardly be 
 doubted that they all contain this same /or ya. And of that 
 element the primitive function was probably optative ; it was 
 at first the expression of a wish. On this assumption, appa 
 rently, we can most easily and naturally explain the uses of 
 the forms which contain this element through our whole 
 family of languages. At this point, then, we begin with the 
 Latin Subjunctive. 
 
 i. The Latin subjunctive is used to express an action as de 
 sired or wished for. Thus in exhortation, request, command, 
 deprecation, prohibition, etc. With the idea of wishing is 
 naturally associated that of aim : from desire we proceed to 
 purpose, and then to effort for the attainment of our purpose. 
 Hence 
 
2 1 6 LA TIN SUBJUNCTIVE. 
 
 2. The Latin subjunctive is used to express an action as 
 aimed at or striven for (acted for). Fearing implies wish or 
 aim that something* may not take place ; and it is therefore ex 
 pressed by ne with the subjunctive : that is, expressed as ne 
 gative purpose. In the uses hitherto considered, there is a 
 conscious tendency toward an object which is both expected 
 and desired. But the idea of desire may fade out, leaving 
 only that of conscious tendency toward an expected object. 
 Hence 
 
 3. The Latin subjunctive is used to express an action as 
 looked toward and waited for. So in many dependent sen 
 tences beginning with until and as long as. This forward 
 looking of the subjunctive is illustrated by the fact that it is 
 commonly used with antcqnam and priusquam, while post- 
 quain regularly takes the indicative. But the tendency ex 
 pressed by the subjunctive is not necessarily a conscious (or 
 subjective) one, a movement of the mind ; it may be an un 
 conscious (or objective) tendency, a movement of circum 
 stances or causes, tending toward some result, that result 
 being conceived of not as real, but only as the object of a 
 tendency. Hence 
 
 4. The Latin subjunctive is used to express an action as a 
 result, as that toward which there is a tendency. The Latin 
 is apt to think of things characteristic or customary in this 
 way, to conceive them as what the circumstances, conditions, 
 or cases tend to produce. Non is cst qni hoc patiatur there 
 is no tendency in his case to such endurance, it is not his 
 habit or character to endure this. But the idea of tendency 
 which we have traced through all these uses may itself grow 
 dim and fade out, partially or wholly, leaving that of an ex 
 pectable or possible event. Hence 
 
 5. The Latin subjunctive may express an action as a pos 
 sible event. And by " possible" here I do not mean feasible, 
 practicable, that for whose production there exist adequate 
 powers. I mean simply liable to occur, that which has 
 more or less chance for existence. This is the potential sub 
 junctive. It presents the mode in its dimmest, most attenu 
 ated state, with the minimum of positive elements over and 
 above that negative one, that mere absence of a concep- 
 
LA TIN SUBJUNCTIVE. 2 1 / 
 
 tion of reality, which we noticed at the outset. In many cases, 
 indeed, it might be argued, not without plausibility, that the 
 positive elem.ents were wholly effaced, and that the subjunc 
 tive expressed the action simply as divested of\all conception 
 of reality. This is Baumlein s theory of the Greek optative ; 
 but I have never been able to satisfy myself that the uses of 
 the optative could be accounted for with that factor only. 
 
 We have thus endeavored to find a probable (or, at least, 
 a natural) genetic development for the uses of the Latin sub 
 junctive, starting from what seems to be the historic starting- 
 point, in the idea of wishing. If we look at the actual uses 
 as they lie before us, irrespective of origin or order of de 
 velopment, the prevailing notion which links them together 
 appears to be that of tendency. The subjunctive expresses 
 an action as that toward which there is a tendency, either in 
 the conscious desires, aims, expectations of some person, or 
 in unconscious bearings of circumstances known or unknown 
 to the speaker. We have in English a form of expression 
 in which this idea of tendency is very explicitly and sharply 
 presented ; it consists of the substantive verb followed by the 
 infinitive with to : as, they are to go, that is, they are toward 
 going. But this differs in two points at least from the Latin 
 subjunctive : it is confined in use to objective or outward 
 tendencies, and it contains (where the substantive verb is in 
 the indicative mode) a conception of reality. 
 
 It is often difficult to say, in regard to particular uses of the 
 mode, where they belong in this scale of development, to 
 which of these categories we ought to refer them. Thus, 
 the so-called subjunctive of deliberation, or of deliberative 
 questions. We might think of it as a potential subjunctive, 
 and so Professor Harkness gives it in his Grammar. There 
 may be no objection to this, considered as a practical ar 
 rangement. But it can hardly be doubted that the subjunc 
 tive here is in reality one of wish or aim. The real question 
 in such cases is what the person addressed would like or 
 would counsel : quid faciamus, what would you have us do, 
 what would you propose or advise that we should do ? A 
 visne y or something of that kind, may always be supplied. 
 Tofaciatis used without interrogation, in the sense may you 
 
2 1 8 LA TIN SUJ3 JUNCTIVE. 
 
 do, you are (by my desire) to do, corresponds the ques 
 tion quid faciamus, l what may we do, what are we (by 
 your desire) to do ? On the other hand, in the dependent 
 question, and in the so-called intermediate clauses of the 
 oratio obliqna, the subjunctive must be referred to the last 
 head, as expressing possible event. What in the oratio recta 
 is viewed as a reality becomes a possibility in the obliqua. 
 By the first speaker it is conceived as a reality ; to the second 
 speaker (the quoter), who sees it only as a conception of the 
 first, the character of reality falls away ; to him it is a mere 
 possibility. Thus the question " when did they come," di 
 rectly put, is equivalent to they came : tell me when : let 
 me know the time of what I view as a real coming. But the 
 indirect form, " my friend asked when they came," is equiva 
 lent to my friend asked the time of their coming a coming 
 which I have nothing to say about : however real to him, it 
 is only a may-be to me. Again, the subjunctive of cause is 
 probably to be explained from the idea of result. Qua 
 quum it a slut, the case being such that tfaese things are so, 
 as circumstances tend to these results, it follows, etc. The 
 most difficult use of the mode is that of the imperfect and 
 pluperfect subjunctive with quuui, when : as in quum Ccesar 
 vcnirct in Galliam. Though this construction is often used 
 where the two facts, expressed by the principal and the de 
 pendent sentence, are connected only in time, it may be con 
 jectured that this was not always so, that the subjunctive 
 with quum was at first used where the facts had a causal con 
 nection, and that the merely temporal use grew out of a 
 gradual and not strictly proper extension of that form. Not 
 that quum vcnirct became absolutely equivalent to quum 
 vcnit : the conception of reality was still wanting to the 
 former ; quum vcnirct was nearly equivalent to tcmpore ve- 
 niendi. But I think that it stands by itself among the uses 
 of the subjunctive; and that it is not to be explained by a 
 natural development of meaning in the mode, but rather by 
 the extension of a form beyond the sense which properly 
 belonged to it the form which was properly employed for a 
 causal connection was improperly extended to cases where 
 there was only a temporal connection. 
 
LA TIN SUB JUNCTIVE. 2 1 9 
 
 I will notice, further, what seems to me a very unfortunate 
 feature in Madvig s treatment of the Latin subjunctive. Not 
 content with the four tenses usually recognized in this mode, 
 he has given it two more, a future and a future perfect. His 
 future subjunctive is a periphrastic form composed of the 
 future participle with the present subjunctive of the verb to 
 be : amatiirus sim. His futur eperfect is a mere repetition of 
 the perfect ; amaverim is given twice over, once as perfect, 
 and again as future perfect. These tenses, however, he con 
 fines to the active voice : to the subjunctive passive he allows 
 neither perfect nor future perfect. But if amaverim is two 
 tenses in one form, why not amatus sim also ? If amaverim 
 has the two uses * I may now have loved and ( I may here 
 after have loved, the passive amatus sim is susceptible of the 
 two similar uses, I may now have been loved and I may 
 hereafter have been loved. If we recognize amaverim as a 
 future perfect in the active voice, there is no reason why we 
 should not recognize amatus sim as a future perfect in the 
 passive. That Madvig has not done so, that he has failed to 
 give a future perfect subjunctive in the passive, is probably 
 because he did not see what to do for a future subjunctive in 
 that voice. He could not find anything which would cor 
 respond to his future subjunctive active amatiirus sim. There 
 is no future participle of the passive voice : amandus sim 
 would convey an idea of necessity. But if we leave the 
 passive out of view, and look only to the active, the scheme 
 is misleading. It *is right in recognizing a double use of 
 amaverim, as a present and a future of complete action. But 
 it covers up the fact, which is of great importance in the 
 theory of the mode, that the form amcm is similarly double 
 in use : it means not only * I may now love, but also I may 
 hereafter love. If amaverim is repeated for a perfect and a 
 future perfect, amem ought equally to be repeated for a 
 present and a future. -The student of Madvig s Grammar 
 would naturally think that amem had only the present sense, 
 1 I may now love, and that the corresponding future I may 
 love hereafter must be expressed by amatiirus sim. But in 
 fact amatiirus sim has not precisely this meaning. It differs 
 from amem used in future sense just as amatiirus sum differs 
 
220 LA TIN SUBJUNCTIVE. 
 
 from the future indicative amabo. The truth is that 
 amaturus sum is a present indicative of prospective (or con 
 templated) action ; and in like manner, amaturus cram is an 
 imperfect indicative, and amaturus sim t amaturus cssem are 
 present and imperfect subjunctive, of prospective action. 
 We have a full series of these tense-forms for prospective 
 action. To take one of them, as Madvig does, and treat it, 
 without reference to the others, as belonging to the same 
 series with amem, arnarcm, amavcrim, etc., can hardly be a 
 justifiable procedure. 
 
XIII. 
 
 ON THE ORIGIN OF THE ENGLISH POSSESSIVE 
 
 CASE. 
 
 1866. 
 
 A FAVORITE poet declares, in a passage often quoted, 
 that " Error, wounded, writhes with pain, and dies 
 among his worshippers." The saying has its truth, doubtless, 
 when taken in reference to the grand results of world-history ; 
 but it gives no hint of the extreme tenacity of life which 
 error often shows, or of its frequent strange revivals after 
 seeming death. The champion of science cannot presume 
 that an error, once crushed, will never raise its head again : 
 too often he is forced to pe.rsonate the comic hero of whom it 
 was said that " thrice he slew the slain." 
 
 These remarks are suggested by a recent essay on the s of 
 the English possessive the s seen in horse s head, Mary s 
 book, children s dress, and the like. Its author s object is to 
 infuse new life into the old theory which viewed this s as a 
 remnant of the possessive pronoun his ; so that the examples 
 just given would be forms, abridged and altered in rapid 
 utterance, of what was originally horse HIS head, Mary HIS 
 book, children HIS dress. As the essay referred to appears in 
 a place so respectable as the published Transactions of the 
 Philological Society (London, 1864), as .it occupies nearly 
 ninety pages of that journal, and is written with evident labor 
 and considerable appearance of learning, it may be worth 
 while to spend some little time in reviewing its positions and 
 arguments. In one respect, as an instance of authorship in 
 advanced years, it may almost be reckoned among the curiosi 
 ties of literature. Its author "James Manning, Q.A.S., 
 Recorder of Oxford " describes himself as writing in the 
 eighty-third year of his age. It should seem that, after a life 
 of active labor in his profession as a lawyer, he has found a 
 pleasant and honorable recreation for his old age in philologi- 
 
222 ENGLISH POSSESSIVE CASE. 
 
 cal studies. That these studies have excited a keen interest 
 in his mind must be apparent to any reader of this essay. 
 He discusses his theme perhaps we should rather say, he 
 pleads his cause with all the warmth and earnestness of an 
 advocate. His materials of argument and illustration could 
 not have been collected without considerable expenditure of 
 time and effort. He has looked into the grammar of the 
 Anglo-Saxon, and even into that of the Mceso-Gothic. To 
 the grammar of the popular dialects of Germany he has given 
 a good deal of attention ; and still more to the Semi-Saxon 
 form of our own language, as represented in the two texts of 
 Layamon s Brut, and to the Old English which followed it. 
 It must be confessed, however, that his knowledge is far from 
 sufficient ; and that he has fallen into numerous misappre 
 hensions and errors, from which more thorough study would 
 have saved him. 
 
 Mr. Manning s thesis may be stated thus : The old inflec 
 tional genitive of the Anglo-Saxon was given up in the I3th 
 century ; it was almost wholly discarded at that period ; and 
 was replaced in most of its uses by the preposition of which 
 preposition, by the way, he strangely designates as " Scan 
 dinavian ; " forgetting that it is found, not only in the earliest 
 Anglo-Saxon, but also in the Mceso-Gothic af, and even in 
 the High German ab ; that it belongs, therefore, to the 
 whole Teutonic class of languages, and not specially to the 
 Scandinavian branch of it. He maintains (I say) that the 
 Anglo-Saxon genitive, discarded in the I3th century, was re 
 placed in most of its uses by the preposition of but that for 
 the possessive relation there sprung up at the same time a 
 different form of expression, a form in which the name of 
 the possessor was followed by the name of the thing pos 
 sessed, with the pronoun his interposed between them : thus, 
 horse his head. This interposed his he describes as sex-less and 
 number-less, denoting mere possession, without distinguish 
 ing either the gender or the number of the possessors ; so as 
 to be used with perfect facility and propriety in such expres 
 sions as Mary his book, children his dress. From this his the 
 h was naturally dropped in rapid utterance, leaving YJ/ and 
 in time the vowel of is was generally suppressed in pronun- 
 
ENGLISH POSSESSIVE CASE. 223 
 
 elation, leaving a mere s, which, since the i6th century, is 
 written with the apostrophe. 
 
 On the other hand, the prevailing, and (as we think) the 
 correct view of this subject may be stated thus. The old 
 inflectional genitive of the Anglo-Saxon has never been given 
 up ; it has never at any period ceased to be in use ; it has 
 come down by uninterrupted tradition from the earliest times 
 to the present day. But it has undergone extensive changes in 
 use and function. In the Gothic and so, doubtless, in the 
 primitive Teutonic s was the universal ending of the geni 
 tive singular, in all genders and for every kind of stem. The 
 Anglo-Saxon preserves this s in vowel-stems of the masculine 
 and neuter genders, but has lost it in all feminines and in all 
 ;/-stems : in fact, as regards the s of the genitive singular, the 
 Anglo-Saxon stands on much the same footing as the modern 
 German. But in the early English, we find the s extended 
 again to the feminines and the /^-sterns (these last losing their 
 n) ; so that s becomes once more the universal ending of the 
 genitive singular : although in some few words of common 
 occurrence such as fader, vwder, brother, ladi, saivle 
 ( soul ), etc. the genitive without s, which belonged to 
 -them in Anglo-Saxon, maintained itself, at least in occa 
 sional use, down to the time of Chaucer, or even later still. 
 The s- of the genitive, indeed, extended itself still more 
 widely, and passed into the plural, making, for instance, 
 men s, children s, as genitives of men, children. At the same 
 time, there is a change in position and in function : the geni 
 tive can no longer stand after the word on which it depends, 
 but must always come before it ; and it no longer expresses 
 the same variety of relations as before, but is now, in the 
 main, a possessive case. On this latter point we add a few 
 words of explanation. The Anglo-Saxon genitive, like the 
 Greek and Latin, expresses a great variety of relations 
 between one substantive and another in general, all the 
 relations which are, or may be, expressed by our of. The 
 English genitive, on the other hand, is called a possessive, 
 and properly so; for this is its leading use, and "a potiori 
 nomcn fit." But Mr. Manning is clearly wrong when he 
 contends that our case in s always carries a possessive mean- 
 
224 ENGLISH POSSESSIVE CASE. 
 
 ing. In the phrase (i in consequence of the prisoner s being 
 absent," he holds that the " prisoner " is thought of as a 
 possessor, and the " being absent" as a thing in his posses 
 sion. This is, evidently, forced and unnatural. The truth is, 
 that the English case in s has not only the possessive use of 
 the Anelo-Saxon genitive, but the other uses which stand 
 
 o o- 
 
 nearest to this. Thus it is constantly employed to denote 
 connection in family, or state, or society : as in John s 
 brother, Henry s neighbor, England s queen, the king s ene 
 mies in old English we find even the king s traitors. Mr. 
 Manning might, perhaps, argue that to say the king s ene 
 mies implies that "the king has enemies," and expresses, 
 therefore, a possessive relation. But the verb have is a word 
 of very general meaning, which can be used in a multitude of 
 cases where there is no possession, properly so called, and 
 sometimes even where our possessive case would be inad 
 missible. Thus, every apple has a half, but we cannot say 
 " every apple s half." Still further, our case in s is used to 
 express the subject of an action or an attribute : as in cow 
 ard s fear, God s love, the prisoner s being absent. But 
 relations which stand at a wider distance from the possessive 
 cannot be expressed in this way. Thus, the objective rela 
 tion : we do not say God s fear, but the fear of God ; not 
 the child s guardianship, but the guardianship of the cliild. 
 We do indeed say England s ruler, the child s guardian ; but 
 here it is political or social connection that is thought of, 
 and not the object of an action. In like manner, our case 
 in s cannot be used as a genitive partitive (not women s 
 loveliest, but loveliest of women] ; nor as a genitive of 
 material (not leather s girdle, but girdle of leather) ; nor as a 
 genitive of designation (not Italy s kingdom, but kingdom of 
 Italy}. 
 
 We have thus described the view commonly taken by 
 grammarians and philologists in reference to the English pos 
 sessive in s : it is, that this case is derived by uninterrupted 
 tradition, though not without important changes in use and 
 function, from the Anglo-Saxon genitive in s. Mr. Manning 
 admits that this is the established view 7 : but he considers it 
 to be only a long-received and deeply-rooted error, which 
 
ENGLISH POSSESSI VE CASE. 225 
 
 must be dislodged from the minds of men in order to make 
 room for the admission of the truth. He therefore assails it 
 with a strenuous polemic, bringing forward a great array of 
 objections, which he numbers from I to n. In these, how 
 ever, there is some repetition, and we may reduce them all 
 under a few heads. 
 
 First, he contends that, if the English possessive were the same 
 with the Anglo-Saxon genitive, it would be confined, as the lat 
 ter is, to masculines and neuters, and to a part only of these ; it 
 would be excluded from all feminines, from all //-stems, and 
 from all plurals. But surely, there is nothing more common 
 in the history of language than to find an inflection extended 
 beyond the class of words that originally had it, and applied 
 to other classes that originally excluded it. Thus the plural 
 of nouns in Spanish ends in s ; originally this s belonged only 
 to masculines and feminines : as libros, books, Lat. ace. 
 plur. libros ; mesas, tables, Lat. vicnsas ; virtudcs, vir 
 tues, Lat. virtutes but it has been extended to neuters also : 
 pomos, apples, Lat. poma / rcynos, kingdoms, Lat. rcgna. 
 The Proven9al dialect of the Romance has preserved the 
 Latin s of the nominative singular as in ans, year, Lat. 
 annus but has extended it to words which were without 
 it in the Latin as librcs, book, Lat. liber and even to 
 neuters as aurs, gold, Lat. aurum ; eels, sky, Lat. 
 cce linn cors, heart, Lat. cor; flums, river, Lat. jluincn. 
 The ancient Greek itself, in many adjectives of the vowel-de 
 clension, has extended the masculine form to the feminine : 
 thus, 770-1^09, quiet, which is properly masculine, is used 
 also as feminine. That this is an innovation of the Greek 
 appears from a comparison of the Sanskrit, Latin, and Gothic, 
 in which and, doubtless, in the Indo-European before 
 them the adjective tf-stcms always have a distinct form for 
 the feminine. 
 
 Secondly, it is objected that, if the English possessive were 
 the same with the Anglo-Saxon genitive, it would, of neces 
 sity, have all the same functions, and would be capable of 
 serving for all uses of the Anglo-Saxon genitive. But again, 
 it is a common thing in the history of language to find an in 
 flectional form undergoing some change of function, giving 
 15 
 
226 ENGLISH POSSESSIVE CASE. 
 
 up a sense which it once had, or taking on a sense which it 
 had not. An extreme instance of this kind is seen in the 
 Greek present infinitive passive (e. g. ypd^eo-Oai^ to be 
 written ), which, in the modern Greek, according to the 
 Grammar of Professor Sophocles, is used only after the verb 
 Oe\w, to form a continued future passive : thus 6e\ei 
 ypd(j)e(T0at,, it will be written, or, more exactly, it will be in 
 process of being written. Here the identity of form with 
 the ancient <ypd^>ea9ai is beyond all question; but to what a 
 minimum has shrunk the widely extended use of the old in 
 finitive ! The Latin perfect indicative has two uses, which 
 are often distinguished as definite and indefinite : thus, 
 posucrunt may mean either they have placed (equivalent to 
 the Greek perfect), or they placed (equivalent to the Greek 
 aorist). But in the Romance languages, the corresponding 
 form has lost the first of these uses, retaining only the last : 
 in Spanish, for instance, pusicron, which represents the Latin 
 posucrunt, means only they placed ; for they have placed, 
 the Spanish, like the English, uses the verb have with a pas 
 sive participle : thus, Jian pucsto. The Indo-European pres 
 ent tense, as we find it in the old Teutonic languages, has 
 taken on an additional use, and serves both as a present and 
 a future : thus the Gothic gibitJi means both * he gives and 
 he will give. In the modern Teutonic languages, this ad 
 ditional use is given up, and the tense has become, what it 
 was at first, a present only. And yet the t/i of our givetli, 
 or the s of our gives, is unquestionably the same ending as the 
 th of Gothic ^vY;////, the t of Latin dat, and the ti of Indo-Eu 
 ropean daddti. 
 
 Closely connected with the objection we are considering is 
 another which has been urged by Professor Goldstiicker in a 
 review of Mr. Manning s essay (see " The Reader ; " London, 
 Sept. 24, 1864). It is impossible, he maintains, that an in 
 flexional form, like the Anglo-Saxon genitive in s, should 
 maintain itself in use alongside of a prepositional expression 
 viz. of with the accusative which could be used instead 
 of it everywhere with precisely the same meaning. But to 
 this impossibility we may oppose a reality offered by the 
 Latin language. Here the old inflectional locative did 
 
ENGLISH POSSESSIVE CASE. 227 
 
 maintain itself in certain words, as a designation of the place 
 where, alongside of a prepositional form viz. in with the 
 ablative which could be used instead of it everywhere with 
 precisely the same meaning. Examples of this old locative 
 are Ronite, originally Romai, at Rome, Corinthi, originally 
 Corinthdi) at Corinth, rnri, in the country, domi, at 
 home, etc. It has become greatly restricted in use, being 
 nearly confined to names of towns, and among these to words 
 of the first or second declension. Other words used to desig 
 nate the place where are put in the ablative with in, as in pro- 
 vincia, in situ ; and even names of towns of the first or 
 second declension are susceptible of the same construction. 
 True, it is not idiomatic to say in Roma / but when an ap- 
 positive (as urbs) is added, in urbc Roma is the regular form. 
 Even in these points our English usage presents a curious 
 analogy. We do not say this is the house of Jack ; the form 
 is intelligible, but nobody uses it ; we say this is Jack s house : 
 but with an appositive added, this is the house of my cousin 
 Jack is a familiar form of expression. 
 
 Thirdly, the ending s of the English possessive is used in 
 various connections and constructions where, in Mr. Man 
 ning s opinion, it is impossible to explain.it as a genitive case- 
 inflection, or indeed in any way except as the pronoun his, 
 Thus in Ccesars crossing the Rubicon, he asserts, with much 
 confidence, that no inflected language could use its genitive 
 to translate Ccesar s ; that, instead of a genitive, it would be 
 necessary to use some expression with instrumental meaning. 
 But surely, there are scholars enough in his own city of 
 Oxford who could have told him that 77 Kaiaapo^ 8id/3ao-is 
 rov PovfMfcwvos is at once idiomatic Greek and an exact 
 rendering of Ccesar s crossing the Rubicon. The Ka/<rapo? 
 and Cczsar s in these expressions are not, indeed, true posses- 
 sives, as he holds in reference to the latter ; they are genitives 
 used to denote the subject of an action. Again, he finds a 
 more reasonable cause of difficulty in forms such as JoJui and 
 Walter s house, husband and wife s children : the latter, he 
 says, if wife s were a real genitive, would be like saying in 
 Latin vir et uxoris libcri, which, of course, would give a dif 
 ferent meaning, viz. the husband and the children of the 
 
228 ENGLISH POSSESSIVE CASE. 
 
 wife. But would this be further out of the way than vir et 
 nxor ejns liberi ( husband and wife, the children of that one ), 
 the Latin form which would correspond to Mr. Manning s 
 theory? He has himself suggested the true explanation, 
 when he refers to ordinals such as three and twentieth, where 
 the formative ending ctJi is applied to the copulative three and 
 twenty taken as one complex whole. He docs not deny that 
 three and twentieth is good English, though tres et viccsimus 
 would be very poor Latin. He refers also to similar forms of 
 expression in other languages, as half- und hoffnungslos in 
 German, literally help- and hopeless, i. c. helpless and 
 hopeless ; fcliz- y valerosamente in Spanish, literally fortu 
 nate- and valorously, i. c\ fortunately and valorously. Yet 
 again, he finds a difficulty in such expressions as King of 
 England s croiun : if our possessive were a real genitive, we 
 should have to say (he thinks) Kings crown of England, or 
 crown Kings of England. He will not admit that King of 
 England can be treated as one complex whole, receiving in 
 flection at the end ; though it is hard to see how this would 
 differ in principle from the formation just noticed in three and 
 twentieth. Can any one say that in three and twenty the two 
 numerals are more closely connected by and than the two 
 substantives in King of England are connected by of? On 
 the contrary, it would be more plausible to regard the prepo 
 sition as forming the closer connection. It is to be observed 
 that this formation is not readily used except when we are 
 accustomed to hear the combination the two substantives 
 connected by of as the recognized name of some well-under 
 stood object : thus we say the King of England s crown, but 
 hardly a bishop of the church s income, rather the income of a 
 bishop of the church ; hardly the men of property s influence, 
 rather the influence of the men of property. We observe also 
 that examples are not wholly wanting of other endings applied 
 in the same way: thus, from Church of England is formed, 
 at least in popular parlance, if not in the language of books, 
 the expression Church-of-Englandism, for which assuredly 
 no one would think of substituting Chnrchism of England. 
 There remains still another form of expressionseen in the 
 examples, he is a servant of viy brother s, I am going to try 
 
ENGLISH POSSESSIVE CASE. 229 
 
 that new horse of my neighbor s which Mr. Manning regards 
 as irreconcilable with the genitive origin of our possessive. 
 The preposition of (he says) requires an objective case, and, 
 according to his explanation, would have it even here ; a 
 servant of my brother s being equivalent to a servant of my 
 brother HIS, in which brother is an objective case, and his a 
 possessive agreeing with servant. It cannot be denied that 
 there is some difficulty in explaining this usage. Lowth 
 treated the genitive here as having a partitive construction, and 
 depending on a word understood : a servant of my brother s 
 he viewed as standing for a servant of [i. e. of the number of] 
 my brothers servants : the latter was the full logical form, 
 but the repetition in it of the word servant was unnecessary 
 and disagreeable, and was therefore dispensed with. To this 
 explanation of Lowth, Mr. Manning objects that a servant of 
 my brother s does not of necessity mean one of my brother s 
 servants : if my brother has but one servant, and I am aware 
 of the fact, I may still say with propriety " the person whom 
 you saw was a servant of my brother s." Or, to take a clearer 
 case, a man might say " that wife of my son s is always teas 
 ing me," without in the least accusing his son of bigamy (or 
 ofBrigham-y, if one may coin a Mormon name for a Mormon 
 institution). It is possible, however, to modify Lowth s ex 
 planation so as to escape the force of this objection. We 
 may regard the possessive, when thus used, as depending at 
 least in many cases not on a particular word repeated from 
 what precedes, but on a general indeterminate conception of 
 that which is possessed. In speaking of one particular thing, 
 I may say " it is my brother s house ; " of a second, " it is my 
 brother s servant ;" of a third, " it is my brother s heart ; " and 
 so on. Of course, there are a great multitude of things which 
 can be put after my brother s in such a sentence. Conceive 
 now the aggregate, the collective totality, of these things, and 
 you have the general, indeterminate conception of which we 
 are speaking. It is an element of some importance in the 
 theory of syntax. In the sentence "deal as thou wilt with 
 him and his," the subject of the possessive his can only be 
 this general conception : his belongings, the sum of all 
 those things of which each one can be called his* such is the 
 
230 ENGLISH POSSESSIVE CASE. 
 
 sense of the expression. If we say " all mine is thine," it is 
 plain that both possessives have this same indeterminate sub 
 ject. If, then, we say " all mine is my brother s," it is plain 
 that the possessive case brother s must depend on the same 
 indeterminate subject as the possessive pronoun mine. Now 
 it is easy and natural to apply the same explanation to such 
 forms as this soul of mine, thai wife of my brother s. There 
 is a greater difficulty in explaining that friend of hers, of ours, 
 of yours, of theirs, where we should expect of her, our, your, 
 their. It is known, however, that these forms hers, ours, 
 etc. are not very ancient ; and it can hardly be doubted 
 that, from being already accustomed to hear and to say it is 
 JoJui s, it is Jllctry s, it is my cousin s, it is tJic Prince s, and the 
 like, men were led by false analogy to say it is hers for the 
 earlier // is /ier, it is ours for / / is our and in like manner, 
 a friend of hers for a friend of her, a friend of ours for a 
 friend of our, etc. There are very few, probably, who will 
 think it more plausible to explain these forms as Mr. Manning 
 does that is, as abridged for the fuller and earlier forms, a 
 friend of her liis, a friend of our his, and so on. 
 
 Such are Mr. Manning s objections to the common view of 
 our possessive case. It need hardly be said that they arc not 
 decisive against it. They do not require its falsity. The ut 
 most we can say is that, if another view were strongly sup 
 ported by positive proof, these objections might be used, 
 with more or less effect, to clear the ground for its reception. 
 Positive proof for his own thesis Mr. Manning seeks to draw 
 from two different sources, the popular dialects of Germany, 
 and the early English writers. What he has collected on the 
 German dialects forms the most interesting part of his essay. 
 He shows, in fact, that a form of expression analogous to 
 John Smith /its book is widely current in them. Thus we find 
 in popular German dcs Vcttcrs sein Buck ( the father s his 
 book ) ; dcs Goldschmieds sein Junge ( the goldsmith s his 
 apprentice ) ; dein Aufwand iibertrijft den Aufivand des 
 Farsten seinen ( thy expenditure exceeds that of the Prince s 
 his ) ; Jcder hatte ein Pferd mitgcbracJit ( each had brought 
 a horse with him ), aber des einen seins war blind ( but the 
 one s his was blind ), des andcrn seins, laJim ( the other s 
 
ENGLISH POSSESSIVE CASE. 23 I 
 
 /its, lame ). Similar expressions are met with even in re 
 spectable literary works. Schmidt, in his Geschichte der 
 Dciitschcn, says des Alfonsus seine MdcJitc ( Alphonso s his 
 powers ). Gellert has dies Beiivort ist noch mahlerischer als 
 Homers seines ( this epithet is yet more picturesque than 
 Homer s his ). It will be observed that in all the instances 
 now given the name of the possessor is. put in the genitive. 
 But in Upper Germany the dative is used instead of this gen 
 itive : as in dem Vatcr scin Bitch ( to the father his book ) ; 
 dcm Goethe scin Gedicht ( to Goethe his poem ). And in the 
 popular language the name of the possessor can be put in 
 the same case as the thing possessed : as, fass Kurdchcn scin 
 Hiit c hen ( take little Conrad his little hat ), where Klirdchen 
 and Hiitchcn are both in the accusative ; in dem Wolfe seinem 
 Lcib ( in the wolf his body ), where Wolfe and Leib are both 
 in the dative. Mr. Manning speaks of these forms as very 
 old apparently, because he finds them in the brothers 
 Grimm s Kinder- und hausmdhrchen. I cannot see that he 
 traces them farther back than the i/th century. He gives us 
 no example of them from the Middle High German. And 
 there is one thing about them which makes strongly against 
 his hypothesis. When the possessor is -feminine or plural, 
 sein is not used, but ihr, her, or ihr, their, is used instead. 
 Thus, dcr Mutter ihr Klcid ( the mother s her gown ), der 
 Kinder ihr Spielzeug ( the children s their playthings ), den 
 Eltcrn Hire Sorgen ( to the parents their cares ), Frau Wolf 
 Hire Tochter ( Mrs. Wolf s her daughters ). According to 
 the same analogy, we should expect in English, not moiJier 
 his gown, the supposed original -of our mother s gown, but 
 mother her gown / not children his playthings, the supposed 
 original oi children s playthings, but children their playthings. 
 Mr. Manning feels the difficulty, and takes great pains to get 
 rid of it. He shows that, in the old German, scin, used as a 
 reflexive, could mean her and their, and that it some 
 times had this use even when not reflexive. This may be 
 true, but it makes against his cause : for if scin was capable 
 of meaning her and their, it is all the more remarkable 
 and significant that people never said dcr Mutter sein Kleid, 
 der Kinder sein Spielzeug, but always used the pronoun ihr 
 
232 ENGLISH POSSESSIVE CASE. 
 
 instead. Mr. Manning goes back to the Mceso-Gothic, where 
 sein is always reflexive, and shows that, as such, it can mean 
 her and their as well as his. He shows that the Latin 
 suns, which is likewise always reflexive, has the same range 
 of meaning ( his, her, their ). And from all these facts 
 he draws the conclusion that our English his could be used 
 with equal range of meaning that it could be used . for 
 her and their. But this does not follow. Suns and 
 scin are derived from a root sc, svc, sva, self, which denotes 
 personality without gender : sinis and sciu mean belonging 
 to self, of whatever gender or number that " self" may be. 
 Not so the Anglo-Saxon genitive /us. It is true that the root 
 hi, from which it comes, is found also in the feminine and the 
 plural ; but then it always shows a different suffix : belong 
 ing to a feminine hi is expressed by hi-rc (Gothic Jii-zos) ; 
 belonging to a plural/ / is expressed by /ti-ra (Gothic lii-ze, 
 Jii-zo} ; his can only mean belonging to a singular hi of the 
 masculine or neuter gender. Mr. Manning endeavors to 
 show 7 that the forms tie, his, Jiiin are pretty promiscuously 
 used in early English, for the feminine as well as for the mas 
 culine and neuter. This is true as to he, and for an obvious 
 reason: the accusative singular feminine of the Anglo-Saxon 
 pronoun was hi, and in early English this accusative (written 
 as hi or he) is sometimes used as a nominative, instead of the 
 regular heo (the nominative she is of later appearance). But 
 it is decidedly not true as to his and liim. Where these ap 
 pear to be feminine, it will be found almost always that they 
 refer to such words as wife, maiden, child, and others, which 
 are neuter in Anglo-Saxon and Old English. Mr. Manning- 
 has utterly failed to show that his was ever freely applied in 
 reference to feminine words. That it was ever freely used 
 for the plural, he has not even attempted to show. And, in 
 deed, he is evidently aware that his theory is open to objec 
 tion at this point; he sees that, for some at least, it will be 
 hard to believe that his was employed with distinct conscious 
 ness of its pronoun-character in such expressions as the 
 mother his gown, the children his playthings. He has no 
 trouble in believing it himself; but for those " of little faith" 
 who cannot do so, he suggests a way of getting over the dif- 
 
ENGLISH POSSESSIVE CASE. 233 
 
 ficulty. We may suppose (he says) that his was first em 
 ployed in this manner only after nouns of the singular num 
 ber and of the masculine or neuter gender ; that it was con 
 fined for a time to such expressions as the father Ins book, the 
 horse his Jicad, the land his ruler ; that in these it gradually 
 lost its force as a separate pronoun, restricted (like other pro 
 nouns) in gender and number, and came to be regarded as a 
 mere sign of the possessive relation ; and that in this condi 
 tion it began at length to be employed after feminities and 
 plurals. The process, thus hypothetically traced, is certainly 
 not inconceivable. Whether it was a historical reality, must 
 be determined by looking at the literary monuments of our 
 language. . We are thus referred to the usage of the Semi- 
 Saxon and the Early English writers. In these, if anywhere, 
 we must find the convincing proofs for the theory under con 
 sideration. 
 
 In this field, Mr. Manning s battle-horse is the later text of 
 Layamon s Brut. That interesting and invaluable relic of the 
 Semi-Saxon period which recites the mythic history of Bri 
 tain in a poem of some 32,000 short lines is preserved in two 
 manuscripts. The earlier manuscript was written about the 
 year 1200. It shows us the genitive case very much as it 
 stood in the Anglo-Saxon, in form, position, and function ; 
 with some restrictions in its use, but still to a great extent 
 the same thing. In the later manuscript, written (it is sup 
 posed) some fifty or sixty years after the first, we find con 
 siderable changes in this, as in many other features of the 
 language. It is true that the ending-^ of the genitive is not 
 often extended to words that did not have it in the Anglo- 
 Saxon. But in position, the case is no longer free to stand 
 either before or after the word on which it depends : with 
 only rare exceptions it stands first, as in modern English. In 
 function also it is restricted, for the most part, in the same 
 way as in modern English. In the earlier text, Mr. Manning 
 tells us that he has been able to discover only two instances of 
 the possessive pronoun his applied as a substitute for the 
 Anglo-Saxon inflected possessive genitive. Unfortunately he 
 has omitted to specify these two instances. Sir Frederic 
 Madden, the learned editor of Layamon, says that the geni- 
 
234 ENGLISH POSSESSIVE CASE. 
 
 tive expressed by his rarely occurs in the earlier text, but is 
 found in two places, which he names (vol. i., pp. 175, 279). 
 It is probable that these are the two instances spoken of by 
 Mr. Manning-. But any one who looks carefully at the two 
 passages will see .that in neither of them does his refer to the 
 substantive immediately before it ; so that they are not real 
 instances of the form we are discussing. But whether this 
 form does or does not occur in the earlier text, there is no 
 doubt that it is frequent in the later. Mr. Manning says even 
 that "nearly all " the Anglo-Saxon inflexional genitives of the 
 earlier manuscript become pronominal possessives (i.e. forms 
 written with his instead of -cs) in the later. This, however, 
 is a strange exaggeration unintentional, doubtless, but not 
 the less extraordinary. I have run over the first 9,000 lines, 
 or more than a quarter of the poem, with reference to this 
 point. From proper names of persons I have noted, in all, 
 forty-nine genitives. In twenty-three of them the genitive 
 is expressed by his : these are all masculine, and in about 
 half the name ends in s. In twenty-one instances (four of 
 them feminine names) we have the inflected genitive in s : in 
 none of these does the name itself end in s. But there are 
 five instances of names ending in s which are used without 
 change as genitives, the inflexional s being apparently fused 
 (as it often is now) in the final s of the name. 
 
 We see, then, that less than half the genitives from proper 
 names of persons are expressed by his. And it does not ap 
 pear that in the genitives expressed by his the possessive idea 
 is at all more distinct or emphatic than it is in the inflectional 
 genitives. We have " Eubrac his soncs " (sons), but we have 
 also "Argalrc sones ; " " Gorbonia his brother," but also 
 " Morgans brother ; " " Julius his men," but also " Cesar es 
 men;" " Cunagcs his hond " (hand), but also " Belyn^ 
 hond ; " " Jaines his temple," but also " Appol mcs temple " 
 (nom. Appoliii] ; and so on. In fact, we find " Albanac his 
 lond" (land) followed, eight lines after, by " Albanackrj folk ; " 
 and <c Belyn^ forth -fare " (departure, decease) followed, two 
 lines after, by " Belyn his cleathe." Is it not evident that 
 these arc mere capricious variations of orthography, without 
 difference of use, and apparently without difference of pro- 
 
ENGLISH POSSESSIVE CASE. 235 
 
 nunciation ? From, proper names of places or countries I 
 have noted fifteen genitives : two of these " Wales his 
 louerd " (lord), and " Leogris his lond " are made with his ; 
 while the remaining thirteen are written with inflective s, as 
 " Norwey^y king," " Lombard i^ lond," and so on. From 
 common nouns I have observed only two instances of the 
 genitive expressed by his: viz., "hem his mochelc mod" 
 (uncle s micklemood), and " man his froucre " (man s savior). 
 On the other hand, we have " hemes name " (uncle s name), 
 marine hond " (man s hand), and so on. Indeed, the gen 
 itives of common nouns, written with inflective s, which I 
 have noted, are about eighty in number: of these, many are 
 possessive in the strictest sense, and four-fifths perhaps nine- 
 tenths are susceptible of being rendered by our English pos 
 sessive. Again we ask whether it is not evident at least as 
 regards the first quarter of the poem that the genitive ex 
 pressed by his is only an occasional orthographic varia 
 tion of the old inflectional genitive a variation restricted, in 
 the main, to masculine names of persons, in which it would 
 be most natural for an unlearned scribe to substitute the pro 
 noun Jiis for the inflection -es, the two forms being equivalent 
 to his mind, and almost or altogether equivalent to his ear. 
 
 The conclusion thus drawn is strongly confirmed by the 
 evidence of the Ormulum. This poem, of about 20,000 short 
 lines, is preserved in a single manuscript, which is, not im 
 probably, the autograph of the author, the monk Ormin him 
 self. It belongs apparently to about the same age as the se 
 cond manuscript of Layamon. In the character of its noun- 
 inflection it has a decidedly later aspect than that manuscript ; 
 it approaches altogether more nearly to. the modern inflection 
 of the noun. But when all indications of age are taken into 
 the account, it will have to be considered, probably, as pretty 
 closely contemporaneous with the later manuscript of Laya- 
 man. The Ormulum is especially valuable from its regular and 
 careful phonetic spelling. It is a standing refutation of the 
 oft-repeated fallacy, that a cast-iron orthography a system 
 which follows tradition, without regard for present pronuncia 
 tion however inconvenient in other respects, is useful or in 
 dispensable for philological purposes. Here is a work, wholly 
 
236 ENGLISH POSSESSIVE CASE. 
 
 without literary merit, valuable chiefly to the philologist, and 
 to him chiefly from the fact that the scribe, departing freely 
 from traditional modes of spelling, undertook to make his spell 
 ing represent his pronunciation. In the Ormulum we find the 
 genitive restricted in position and function almost entirely as it 
 is in modern English. It is regularly made with the ending -ess, 
 the s being doubled merely as an orthographic sign, to show 
 that the c is short. Of the form with his (or hiss) as a substi 
 tute for the inflection-ending, not a single instance is found in 
 the Ormulum. Nor can it be claimed with any probability 
 that this genitive-ending -ess has come from the pronoun his 
 by suppressing the // ; for the difference between the vowels e 
 and 2, though in ordinary manuscripts no stress could be laid 
 upon it, is decisive here : when so careful a speller writes his 
 always with /, and the genitive-ending always with e, we may 
 be sure that he sounded the vowels differently in the pronoun 
 and the ending. But, indeed, there can be no reasonable 
 question that the possessive of the Ormulum is the descendant 
 and representative of the Anglo-Saxon genitive. Mr. Man 
 ning himself, quoting the line "till helless thesternesse " (to 
 hell s darkness), says that " we find here the old Anglo-Saxon 
 genitive." He was aware, probably, when saying this, that 
 the Anglo-Saxon word Jicl , being feminine, makes Jicllc (with 
 out s) in the genitive ; yet he could justly describe he! less as 
 an Anglo-Saxon genitive in this sense, that it shows an 
 Anglo-Saxon formation, extended in this instance a little be 
 yond the bounds which it had in that language. But if 
 this formation in the Ormulum is, by Mr. Manning s own ad 
 mission, identical with the Anglo-Saxon genitive, how can it 
 
 o o 
 
 be denied that our English possessive in s is identical with 
 the Anglo-Saxon genitive ? The -ess of the Ormulum has 
 very much the same range of use as our possessive s, being al 
 ready extended to feminities, to words of the //-declension, 
 and to plurals ; it takes the same uniform position before the 
 word on which it depends, and it is restricted substantially to 
 the same syntactical relations. And if the -ess of the Ormu 
 lum has a vowel-sound which is generally wanting to our pos 
 sessive, we find a perfect parallel to this in the ending of the 
 plural, which is -ess in the Oimulum, and is now generally re- 
 
ENGLISH POSSESSIVE CASE. 237 
 
 duced to a mere s. It is true that in writing the possessive we 
 use a sign the apostrophe which we do not use in writing 
 the plural. And it may be true that those who first used the 
 apostrophe did so because they believed the possessive s to be 
 only a remnant of the separate word his. But such a belief, 
 entertained at the close of the i6th century, has no more 
 weight of authority than when entertained now. It could 
 have rested then on no proofs which are not accessible now ; 
 and if the proofs now appear insufficient to support it, they 
 must have been insufficient then. False etymology is con 
 fined to no age. It appears in various particulars of our old 
 established orthography in spelling island (for instance) as 
 if it were connected with isle j in spelling sovereign as if it 
 were connected with reign and so on. If the apostrophized 
 s indicates a supposed connection with his, it may be only 
 another instance of the same kind. 
 
 Mr. Manning cites, from writers of the fourteenth and fol 
 lowing centuries, a series of passages showing possessive 
 cases made with his. He carries these citations even into the 
 seventeenth century. In the authorized version of the Bible, 
 as first printed, there were (he says) three cases of the kind ; 
 one of which " Holofernes his head " is to be seen, still 
 unchanged, in the Apocrypha. As for the other two " Asa 
 his heart " and " Mordecai his matters " he complains that 
 they have been " altered by some careless or earless printer" 
 into "Asa s heart" and " Mordecai s matters." But are 
 these three cases distinguished by any special /it s -ness from 
 the common run of possessives in the Bible ? and, if not, then 
 is it not, probably, mere carelessness, either of the printer or 
 the scribe, which gave them their peculiar form in the first 
 edition ? From Ben Jonson several instances are given, 
 like " Sejanus his fall," " Horace his judgment." But as this 
 is not the ordinary form of the possessive in Jonson s works, 
 and as Jonson, noticing this form in his English Grammar, 
 calls it (to Mr. Manning s great disgust) a " monstrous syn 
 tax," might it not be more charitable to ascribe those in 
 stances to " a careless or earless printer?" Mr. Manning 
 does not disdain even to gather up some few examples from 
 writers of the eighteenth century, Addison, Pope, and Sterne ; 
 
238 ENGLISH POSSESSIVE CASE. 
 
 as if he could acid anything in this way to the force of his 
 previous citations ; as if he were not rather weakening the 
 force of those citations, by showing that men could write as a 
 separate his what they undoubtedly pronounced as a mere 
 inflective s or es ! 
 
 Mr. Manning has given many examples of a possessive 
 written with Jtis : he might easily have given many more. 
 But we cannot see that he has proved, or even seriously at 
 tempted to prove, that this was the prevailing form of our 
 possessive at any period of time, or even in any single author. 
 If in a particular book he finds one, two, or three examples 
 of this form, he seems never to raise the question whether 
 these are specimens of a general practice, or exceptions to a 
 general practice ; though, plainly enough, the force of his 
 examples must depend largely on the answer to this question. 
 He notes the fact, as if it were important, that one of Chaucer s 
 Canterbury Tales is entitled the " Nonne Prest his Tale." 
 But he docs not notice that the same tale was just before de 
 signated as the " Nonne Presto Tale." In Wright s text of 
 the Canterbury Tales, which is regarded as the best hitherto 
 published, the possessive written with his is not only an ex 
 ception, but a rare exception, to the prevailing form. 
 
 My opportunities for studying old English texts have been 
 exceedingly limited, and I cannot therefore speak with posi- 
 tiveness as to their usage in this particular. I have examined, 
 however, with reference to this point, the series of extracts 
 given in Mr. Marsh s Origin and History of the English Lan 
 guage. Those extracts are somewhat extended ; they repre 
 sent the leading productions of English literature from the 
 close of the Anglo-Saxon period to the middle of the six 
 teenth century ; they are taken from the best texts which 
 had been printed when that work appeared. In these ex 
 tracts there are many examples of the possessive in s ; but 
 aside from the later text of Layamon, which has been dis 
 cussed already I have not been able to find a single instance 
 of a possessive written with his. I say "not a single in 
 stance ; " for as to the line in the Surtees Psalter " man his 
 claies ere als hai " -I do not regard it as an instance of the . 
 kind. Man here was undoubtedly uttered as an independent 
 
ENGLISH 
 
 word, followed by a pause ; exactly like homo in the Latin 
 original homo sic lit foenitm dies ejus and like man in the 
 older Wycliffite version " a man as hey his dayes." 
 
 We conclude, then, that the genitive in j, which belonged 
 to the earliest Anglo-Saxon, has never dropped out of the 
 language ; that it has never ceased to be in use. It has un 
 dergone great changes, both in the range of words that take 
 it, and in the range of uses for which it is employed ; but the 
 result of all these changes, the modern English possessive, is 
 connected by an unbroken historical tradition with the inflec 
 tional genitive of the Anglo-Saxon. This is the common be 
 lief of scholars, and it is likely to be sustained, rather than 
 shaken, by the new discussion which Mr. Manning has opened 
 in so spirited a manner. On the other hand, the possessive 
 made by his appears, with all the new light which he has 
 thrown upon it, to be only an occasional variation of the 
 English possessive. It is (or rather, it was) a variation, not 
 in the form itself, .as heard in living utterance, but in the way 
 of representing that form in writing. It arose, apparently, 
 from a mistaken and fanciful etymology, which seemed to 
 furnish a plausible explanation for the possessive sense ; and 
 the degree of currency which it gained is, at least partly, due 
 to the general inexactness and confusion of early English 
 orthography. 
 
XIV-. 
 ELLIS S EARLY ENGLISH PRONUNCIATION.* 
 
 1870. 
 
 THE Second Part, which is to complete this learned and 
 valuable work, was expected, according to the author s 
 statement, to have about the same extent as the First, and to 
 be ready for publication before the close of the year 1869. 
 It will investigate the pronunciation of Anglo-Saxon and Old 
 English prior to the fourteenth century, with that of the Teu 
 tonic and Scandinavian sources of the English ; it will dis 
 cuss the correspondence of orthography and pronunciation 
 from Anglo-Saxon times to the present day ; and it will con 
 tain a series of documents and illustrations relating to the 
 pronunciation of our language in the successive periods of 
 its history. Its appearance will be awaited with much in 
 terest ; yet it will probably be little more than an extended 
 supplement to the part now r before us. It is safe to assume 
 that this first part is much the more important of the two, 
 not only exhibiting the author s method, but presenting us 
 with the general views and opinions to which it has led him. 
 We shall be in little danger of doing him injustice if we 
 criticise what we already have, without waiting for that which 
 is yet to come. 
 
 It is saying little, to say that Mr. Ellis has surpassed all 
 predecessors in the same field. We believe that he is the 
 first who has really endeavored to collect everything which 
 can throw light on the history of English pronunciation, and 
 to treat the whole subject with scientific precision and. 
 
 * On Early English Pronunciation, with especial reference to Shakespeare and 
 Chaucer. By Alexander J. Ellis, F. R.S., etc. Parti. On the Pronunciation of 
 the XlVth, XVIth, XVIIth, and XVIIIth Centuries. London : Published for 
 the Philological Society, by Asher &Co., London and Berlin; and for the Early 
 English Text Society, by Trubner & Co., 60 Paternoster Row. 1869. Svo. 
 pp. viii, 416. 
 
EARL Y ENGLISH PR ON UN CIA TION. 24 1 
 
 thoroughness. In the collection of his material he has used 
 exemplary diligence, sparing no pains to make it complete 
 and exhaustive ; and in the discussion of it he has shown a 
 fairness of mind, a freedom from prejudice, a simple love 
 of truth, not less exemplary. He is always careful to present 
 the evidence on which his conclusions are founded, and to 
 distinguish conclusions which seem to him. only probable, in 
 greater or less degree, from those which he regards as cer 
 tain. He does not fail to recognize the uncertainties which 
 affect much of the evidence uncertainties arising either from 
 the nature of the subject or from the peculiarities of individual 
 witnesses. Nor does he keep back the evidence which seems 
 unfavorable to what he thinks the best-supported conclusions ; 
 but presents the whole case, making it possible for the reader 
 to form an independent judgment. 
 
 In the notation of spoken sounds, Mr. Ellis uses a compre 
 hensive system, \vhich he calls by the name palceotype (only 
 the old types being used in it), and sets it forth in a brief in 
 troduction. It is based on the Roman alphabet, and contains 
 no sign that cannot be found in the cases of an ordinary 
 printing-office. To secure the necessary variety, italics, 
 small capitals, and (in some instances) inverted letters 
 are used to denote sounds distinct from, though akin to, 
 those expressed by the corresponding Roman letters. As a 
 further means to the same end, the forms (h, j, id) are used, 
 without any consonant power, merely as diacritical signs, 
 modifying the sounds of the letters with which they are 
 connected ; while the forms (ll, J, w, and q) represent the 
 consonant sounds in /;ay, j/ea, way, \\ing. Long vowels are 
 represented by doubling the signs which stand for the cor 
 responding short vowels ; diphthongs, by writing their ele 
 mentary vowels in immediate succession ; successive vowels, 
 if they do not form a diphthong, are separated by a comma. 
 Words and sounds written in palaeotype, if mixed with or 
 dinary writing, are distinguished by enclosing them in marks 
 of parenthesis : thus, (Hee,zq) for Jiaying. Mr. Ellis is care 
 ful to explain that this mode of writing is not designed to 
 supersede the current orthography in popular use ; it is in 
 tended for scientific purposes, as a means of designating 
 16 
 
242 
 
 EARLY ENGLISH PRONUNCIATION. 
 
 conveniently and exactly the sounds heard in English and in 
 other ^languages.. In his tabular Key to Palaeotype he gives 
 more than two hundred and fifty distinct sonnds, with their 
 notation in his system. He also compares these signs of his 
 with the letter-forms devised by Mr. Melville Bell, and de 
 scribed in his " Visible Speech." The number of signs re 
 quired in treating of English pronunciation past and present 
 is, of course, much less than two hundred and fifty. Among 
 those which occur "often in the book we maybe allowed to 
 give here the most important, as it will be convenient to use 
 them in the statements and criticisms that we have to offer. 
 The short vowels (a, e, i, o, u) have the Italian sounds ; but 
 these are almost all different from the English short vowels 
 mpat, pet, pit, pot, put, which are represented in palseotype 
 by (ae, e, i, D, ?/). The long vowels (aa, ee, ii, oo, uu), with 
 Italian sounds, correspond more nearly to the English long 
 in par, pale, peel, pole, pool, which are represented in 
 palseotype by (aa, ee, ii, oo, uu), where (ce, oo) are closer 
 sounds than (ee, oo). The forms (y, ce) stand for the German 
 u, o. The English short and long a in want, war, are ex 
 pressed by the small capital (A, AA) ; the short u in but, by 
 the inverted (o) ; the diphthongs in height, house, by (si, 
 9u) ; the diphthongal u in pure, by (in). As to the conso 
 nants, we have already spoken of (ii, j, w, q) ; and need only 
 add (j) for the weak final r in fair, and (th, dh, sh, zh, tsh, 
 dzh) for the spirant sounds in thcw, tiion, sJioe, azure, chew:, Jeiv. 
 Mr. Ellis commences his inquiry with the pronunciation of 
 the sixteenth, seventeenth, and eighteenth centuries. From 
 the known present he goes back first of all to the recent past. 
 For this period he has the aid of a long series of orthoepical 
 writers ; and he begins by enumerating these in the order of 
 trme, giving full titles of works, and adding brief descriptions 
 and criticisms. The writers to whom he refers most fre 
 quently are: John Palsgrave (French Grammar), 1530; W. 
 Salcsbury, 1547, 1567; Sir Thomas Smith, 1568; John Hart, 
 1569; William Bullokar, 1580; Alexander Gill, 1619; Ben 
 Jonson (English Grammar), 1640; John Wallis (English 
 Grammar), 1653; Philip Wilkins (Philosophical Language^, 
 1668 ; Owen Price, 1668 ; C. Cooper, 1685 ; John Jones, i/oi ; 
 
EARL Y ENGLISH PR O NUN CIA TION. 243 
 
 an anonymous Expert Orthographist, 1704; James Buchanan 
 1766; Benjamin Franklin (Scheme for a New Alphabet), 
 1768. He manifests an especial preference for Salesbury and 
 Wallis. The latter, who was Professor of Mathematics at 
 Oxford, wrote a Grammatica Lingua Anglicance, with a 
 prefixed Tract at us grammatico-physicus de Loquela, in which 
 he gives an elaborate .description of the English sounds, with 
 the positions and movements of the vocal organs in produ 
 cing them. The former, a Welshman, educated at Oxford, 
 prepared an English Dictionary for the use of his country 
 men, with an interesting and valuable introduction, written in 
 Welsh, on English pronunciation. 
 
 From such materials Mr. Ellis endeavors to reconstruct the 
 prevailing pronunciation of our language during the three 
 centuries which preceded our own. He does not disguise 
 from himself the very great difficulties of the task. There 
 are few things harder than to understand the descriptions of 
 spoken sounds. Even when the writer is intelligent, it is no 
 very easy matter to reproduce the precise position of the 
 vocal organs, the precise utterance, which he meant to de 
 scribe. But most writers have been ignorant both of the 
 physical mechanism of speech and of the true relations of 
 sounds. The terms which they have used are very often 
 inexact or unmeaning. Who can be sure as to the force of 
 thick, thin, full, round, flat, hard, sdft, rough, smooth, coarse, 
 fine, sharp, dull, clear, obscure, and many similar, epithets, 
 which are so commonly and so fancifully applied to vowels 
 and consonants ? If the writer identifies a particular English 
 sound with one in some foreign language, as the French, a 
 variety of doubts at once suggest themselves. Are we sure 
 that he meant to -assert an absolute identity, or only a resem 
 blance, between the sounds compared ? Are we sure that 
 the French sound has not varied since the time in question ? 
 Are we sure that it was uniform at that time ? Are we sure 
 that the writer correctly apprehended the French sound of 
 which he speaks ? Misapprehension of foreign sounds is a 
 thing of constant occurrence. Mr. Ellis mentions a remark 
 able instance in a lecture by Professor Blackie, of Edinburgh, 
 on the pronunciation of Greek ; after saying that long a in 
 
244 EARL Y ENGLISH PR ON UN CIA TION. 
 
 Greek had the sound of Italian a in amarc, the lecturer added 
 immediately that " long a should always be pronounced like 
 English aw or au , as in cawl, maid, etc." ! Even experts 
 may differ as to the real character of the foreign sound. 
 Speaking of the French vowels before the nasal n in an, vin, 
 on, un, Mr. Ellis represents them, first as they seem to his 
 own ear, and then as they appear to Dr. Rapp, M. Feline, 
 M. Favarger, and Mr. Melville Bell ; and no two of these 
 gentlemen agree entirely with each other. Mr. Ellis further 
 tells us that he differs from Mr. Bell in his pronunciation of 
 several of the key-words which the latter has used to show 
 the exact phonetic value of his symbols, 
 
 While fully recognizing these difficulties which beset his 
 inquiry, our author has not allowed himself to be deterred 
 from pursuing it. Taking up in succession the English 
 vowels and consonants, he endeavors to ascertain, from a 
 detailed examination of his authorities, how each was sounded 
 in the sixteenth century, what changes (if any) it has under 
 gone since then, and at what time they occurred. Most 
 readers, we presume, will be surprised at the amount of 
 change which he finds in English pronunciation, and espe 
 cially in the vowel sounds, since the sixteenth century. Thus, 
 to commence with the short vowels, he holds that only 
 e and I were pronounced then as they now are that is, as 
 (e) and (z) ; as to a, 6, u, he believes that they were then 
 pronounced (a, o, u), the first two as in Italian, and that the 
 now prevailing sounds for them, viz. (ae, o, 3), as seen in the 
 words cat, cot, cut, came in during the seventeenth century. 
 This may be true as to 6 and u ; but we cannot help thinking, 
 for reasons which will appear presently, that our current 
 sound of a is older than he makes it that it belonged to the 
 pronunciation of the sixteenth century. 
 
 Next, as to the long vowels. According to Mr. Ellis, the e, 
 which in Chaucer s time (the fourteenth century) was always 
 (ee), began to take the sound (ii) during the fifteenth. In 
 the sixteenth, a practice arose of representing the latter sound 
 (ii) by doubling the vowel, as ee ; while the old sound (ee), 
 where it remained, was often distinguished by an added a, as ea. 
 Thus been, reed, greet \VQYQ pronounced as (biin, riid, griit) ; but 
 
EARL Y ENGLISH PR O NUN CIA TION. 24$ 
 
 bean, read, great, as (been, reed, greet). At length, however, 
 the new sound (ii) was extended to words which had for a time 
 retained the old one, the change being particularly rapid about 
 the end of the seventeenth century, so that early in the eigh 
 teenth ea had come to have in most cases the same sound as ce. 
 It is curious to compare the lists of words with ea y given by 
 orthoepists during the transition period, and to note the 
 progress of the change. We find here, as in many such 
 revolutions, that particular individuals carry out the innovat 
 ing tendency to an extent which is not finally sanctioned by 
 the prevailing usage ; some writers give break, great, in- 
 deavonr, deaf, etc., with the sound of (ii). Mr. Ellis, by the 
 way, speaks of (diif) as a pronunciation which he has never 
 been fortunate enough to hear. 
 
 The o has up to a certain point the same history as the e. 
 Before the close of the fifteenth century, it had in many 
 words passed from its proper sound of (oo) into the closer 
 (uu). And here also the practice arose in the sixteenth 
 century of representing the new sound (uu) by doubling the 
 letter, as oo ; while the old sound, where it remained, was 
 often distinguished by an added a, as oa. Thus moon, rood 
 were pronounced as (muun, ruud) ; but moan, road, as 
 (moon, rood). Here, however, the parallel ceases. The 
 movement had already spent its force before the Elizabethan 
 time. While ea in most words passed on from (ee) to (ii), oa 
 has never passed from (oo) to (uu) ; with few exceptions 
 such as move, prove , and (with shortened vowel) love, dove 
 the o has taken on the (uu) sound only in words where that 
 sound was indicated in the sixteenth century by the writing oo. 
 
 Long I and on, which in Chaucer s time were simple sounds, 
 the first being pronounced (ii) and the second (uu), had to a 
 great extent become diphthongs in the sixteenth century. 
 In Mr. Ellis s opinion, they were more clearly diphthongal 
 then than now, being sounded as (ei) and (on), where each 
 vowel must be understood as having its proper force distinctly 
 audible. In the present pronunciation the first element is 
 obscure ; the initial position of the organs is not maintained 
 long enough to give a fully characterized utterance ; hence, 
 orthoepists differ much as to the first vowel sound in English 
 
246 EARL Y ENGLISH PR ON UN CIA TION. 
 
 long I and ou. Mr. Ellis regards it as (9), and writes the 
 diphthongs as now pronounced (si, su). This pronunciation, 
 he thinks, came in during the seventeenth century, or perhaps 
 during the latter part of the sixteenth. 
 
 For ei and ai our author s results are particularly interest 
 ing. He shows that in the sixteenth century they were true 
 diphthongs, differing little, if at all, in their pronunciation, 
 which must have been much the same with that of our af 
 firmative aye. It is observed that in Shakespeare s minor 
 poems there is but one real instance in the words mane, 
 again of a rhyme between a and ai. The change by which 
 these combinations came to be sounded as simple vowels 
 usually as (ee], but ci in some words as (ii) is referred, like 
 so many other changes, to the seventeenth century. 
 
 As to long a in sale, came, fate, etc. Mr. Ellis s conclu 
 sions will cause greater surprise, and will perhaps meet with 
 less acceptance. He holds that in the sixteenth century such 
 words were pronounced (saal, kaam, faat), with the Italian 
 sound of a; and that, in the scventeeth, the pronunciation 
 changed to (saeael, kaiajm, fseaet), differing only in length of 
 sound from our Sal, Cam, fat. This sound, which he finds 
 first distinctly apparent in the description of Wallis (1653), 
 would seem to have been only transitional, as it gave place at 
 the close of that century to (ee), which has since passed into 
 the closer (ee] : thus, ($cc\, kmn, fat). What now is the reason 
 for believing that English long a had in Shakespeare s time 
 the Italian sound of a ? Palsgrave, in 1530, identifies the 
 English letter with the Italian ; but Palsgrave s ear, as Mr. 
 Ellis admits, was none of the most delicate. Hart, in 1569, 
 identifies the English letter with the German, Italian, French, 
 Spanish, and Welsh a ; but this, as Mr. Ellis says, is too wide 
 a comparison, and leaves us in doubt as to the real sound. 
 The witness really relied upon is Salesbury, who says, in 
 1547, that " a in English is the same sound as a in Welsh," 
 and represents the pronunciation of ale, pale, sale, etc., by 
 writing them for Welshmen aal, paal, sal, etc. This 
 testimony makes it pretty clear that English a had not then 
 the same sound as at present. If it had, Salesbury would al 
 most certainly have compared it with Welsh e ; and Palsgrave 
 
EARL Y ENGLISH PR ON UN CIA TION. 247 
 
 and Hart, with e in Italian, French, etc. No one now could 
 for a moment think of giving a, rather than e, in those lan 
 guages as the nearest equivalent for English long a. But as 
 proof of the sound (aa), this evidence is not equally convinc 
 ing. It is curious that in the following century Wallis, whose 
 testimony is regarded as clearest for the sound (seae), still 
 identifies the English and Welsh a. It appears that a fraction 
 of the Welsh people now give this sound (se) to their own a / 
 and such possibly may have been the pronunciation of Sales- 
 bury. It is, doubtless, more probable that his Welsh a was 
 an (a) ; but if so, he had no Welsh letter which would cor 
 rectly represent an (ae), and he may very naturally have re 
 garded a as the nearest Welsh equivalent. But what is most 
 important, we have distinct positive evidence from indepen 
 dent witnesses that the English a, at the beginning of the six 
 teenth century, differed from the French a, and approached 
 to the French e. In the fragment of a treatise on the pro 
 nunciation of French, by an unknown author, but with the 
 date 1528 (two years earlier than Palsgrave s book), is found 
 the following statement as to the French a and e: "A. 
 ought to be pronounced fro the botom of the stomak, and all 
 openly ; E. a lytell hyer in the throte, there properly where 
 the . englysshe man soundeth his a." Similarly, Gilles du 
 Guez, in his account of French pronunciation, which seems to 
 have been printed about 1532, says: "Ye shal pronounce 
 your [French] a as wyde open mouthed as ye can ; your 
 [French] e, as ye do in latyn , almost as brode as ye pro 
 nounce your a in englysshe." These passages do not prove 
 indeed that English a was identical with French e ; in fact, 
 the last of them excludes the idea of such an identity. But 
 they prove an approximation of the English a to the French e, 
 which is scarcely reconcilable with the sound (a) for the for 
 mer. If we should assume that English a, at the opening of 
 the sixteenth century, had, nearly or exactly, the sound of (ae), 
 we should account in the most natural manner for the expres 
 sions of these writers. A similar sound, at the end of this 
 century, is indicated by Peter Erondell, in his French Gram 
 mar (London, 1605). Distinguishing the French a from the 
 English, he represents the sound of English ale by the French 
 
248 EARL Y ENGLISH PR ON UN CIA TION. 
 
 writing eslwith silent s : that is, he finds the nearest equivalent 
 for English a in the French open e. And Mr. Ellis himself finds 
 the same sound, a = (aeae), clearly set forth in the description 
 of Wallis (1653). Are we not then warranted in concluding 
 that English a, as early as the year 1528, had varied from the 
 normal sound of (a, aa), and had assumed this sound of (ae) 
 for a, (cere) for a ? The change may then be referred with 
 much probability to the fifteenth century ; and may naturally 
 be regarded as the starting-point in that great revolution 
 which, since the days of Chaucer, has transformed the whole 
 vowel system of our language. 
 
 If we find ourselves obliged to dissent from Mr. Ellis on the 
 pronunciation of long a, we are equally unable to agree with 
 him on that of long //. He regards this vowel as having had, 
 through the sixteenth century and much of the seventeenth, 
 the same sound as the French u. The conclusion is not, in 
 deed, as incredible as it may seem at first view. Our English 
 u is nearly confined to words which have come to us either 
 from the French itself or from the Latin after our language 
 had fallen under French influence. The Anglo-Saxon and the 
 oldest English had a long u, but for five centuries it has been 
 represented by ou or oiv. The n in Chaucer s time, and 
 ever since, belongs to the Romance part of the language. 
 The sound of French u (y, yy), as represented in palaeo- 
 type is certainly a strange and difficult one for most speakers 
 of English. But Mr. Ellis assures us that it is common in 
 some of the English dialects at the present time. "In East 
 Anglia, in Devonshire, in Cumberland, as well as in Scotland, 
 (yy) and its related sounds are quite at home." We must ad 
 mit, then, as something quite possible, that this may have 
 been a current and prevailing sound of u in the sixteenth and 
 seventeenth centuries. It is simply a question of evidence. 
 Here again, as for long a, we find a number of orthoepists 
 asserting the identity of the English and French sounds. 
 Among these we may reckon Palsgrave (1530), who speaks 
 only of English cw, but appears to have meant the same 
 sound as that of u; also, Sir T. Smith (1568), Hart (1569), 
 Bullokar (1580), and lastly, Wallis (1653), whose testimony 
 on this head is perfectly distinct and positive. Yet in this 
 
EARL Y ENGLISH PR ON UN CIA TION. 249 
 
 testimony of Wallis we find the clearest proof of the unreli 
 able character of such identifications. For Bishop Wilkins, 
 in 1668, speaking of the French ?/, declares, not only that the 
 English do not use it, but that it is very hard for them to pro 
 nounce it. His language is: As for the?/ Gallicum, or 
 whistling ?/, though it cannot be denied to be a distinct sim 
 ple vowel, yet it is of so laborious and difficult pronunciation 
 to all those nations amongst whom it is not used (as to the 
 English) " etc. Let it be observed that Wallis and Wilkins 
 were contemporaries, that both were natives of Southern 
 England, that both were for some time fellow-collegians in 
 Oxford, that both must have mixed in the same society, and 
 that their books were separated by an interval of only fifteen 
 years. The discrepancy between their statements cannot be 
 accounted for by difference of time, place, education, associa 
 tion, and the like. One or the other must have been in error. 
 But a writer s statement that a foreign sound is strange to his 
 own people, and difficult for them to utter, is not in itself likely 
 to be erroneous ; and in this case it is confirmed by other evi 
 dence, all going to show that the pronunciation of it since 
 Wilkins s time has been what it now is. If then the sound of 
 French u was not used, and could hardly be pronounced, by 
 the English in 1668, it cannot possibly have been the current 
 sound of English u in 1653. Wallis s identification must have 
 been an error ; the native and foreign sounds which he com 
 pared were not really identical. And if a writer so intelligent 
 and careful as Wallis could fall into this error, we need not be 
 surprised to find it in Palsgrave, Smith, Hart, etc. Indeed, 
 the last-named writer seems to be at variance with himself, as 
 in an earlier treatise he identifies the vowel u with the word you. 
 He also describes it as a diphthong, composed of i and u ; it is 
 true, he applies the same description to the French u ; but that 
 may only show that he misapprehended the foreign sound. 
 
 But we have positive evidence that the French and English 
 sounds were not identical. Erondell (1605), to whose French 
 Grammar we have referred before, gives a careful description 
 of the way in which French u is to be uttered, and directs 
 the learner to pronounce musique y punir, subvenir, " not after 
 the English pronunciation, not as if written nniesiquc> pucnir, 
 
250 EARL Y ENGLISH PR ON UN CIA TION. 
 
 suevenir" with the English uc of sue, due, etc. Holyband 
 (" French Littleton," 1609) says : "You must take paine to 
 pronounce our v \i.e. French //] otherwise then in English ; 
 for we do think, when Englishmen do profer v [i.e. their own 
 u~\, they say, you." Here Mr. Ellis finds a distinct recognition 
 of our present sound of u. But we should hardly speak of 
 it as " the first distinct recognition." For Salesbury himself 
 (1547) indicates the same sound, when he says: " U vowel 
 answers to the power of the two Welsh letters u, w, and its 
 usual power is tiw, as shown in the following words, TRUE 
 truw verus, VERTUE vcrtuw probitas." It is true that the 
 Welsh uw, as heard in Duw, ( God, is not quite the same 
 with our long u (as in cue) ; but it is a pretty near approxima 
 tion to it, and the nearest which is possible in the Welsh lan 
 guage. The difference is that in the Welsh uw the two ele 
 ments of the diphthong are a little more distinct ; we can 
 hear the initial element as an ({], very short, but perfectly re 
 cognizable, before the closely following (?/). The English 
 diphthongal u of ate, pure, mute, etc., has the same elements, 
 but not with the same distinctness of utterance ; the initial 
 position of the organs is not held long enough to give a clear 
 ly characterized sound. In this respect it resembles the Eng 
 lish diphthongs i and ou ; and hence, like these, it has often 
 been regarded as a simple sound. Wallis speaks of it as 
 such, in distinction from the Welsh mv ; and when he identi 
 fies it with French u, he gives prominence to this fact, that 
 both are simple vowels. 
 
 We believe, then, that English u had in Salesbury s time- 
 that is, in the first half of the sixteenth century substantially 
 the same pronunciation as at present. In the loose identifi 
 cations with French //, made by some writers, we find no suf 
 ficient proof of the contrary ; the last and most distinct, by 
 Wallis, is refuted by the nearly contemporaneous state 
 ments of Wilkins. In tracing back the pronunciation of the 
 letter from the nineteenth century to the sixteenth, there is a 
 presumption in favor of the present sound, unless pretty 
 strong evidence can be found for a different one. We do find 
 in our authorities some evidence of this kind ; but it is liable 
 to grave suspicions, and is more than biilanced by positive 
 
EARL Y ENGLISH PR O NUN CIA TION. 2 5 1 
 
 indications of the present sound. Perhaps we have dwelt 
 longer than we ought both on this point and on the sixteenth- 
 century pronunciation of a. But they seemed to be among 
 the most interesting questions raised by our author in his en 
 deavor to determine the actual living utterance of Henry 
 VIII. and Elizabeth ; and they are well fitted to illustrate the 
 amount and character of the evidence on which he relies in 
 the discussion of these questions. We have only to add that 
 the testimonies which we have brought forward in opposition 
 to his views have all been derived from his own pages, where 
 they stand fairly presented and candidly considered. 
 
 We pass on to the consonants, the treatment of which by 
 our author will furnish much less occasion for remark. Of 
 the weak final r, as heard in car, care, which he represents 
 by (j), he finds no trace whatever in the sixteenth century. 
 Even Wallis and Wilkins are silent in regard to it. But Ben 
 Jonson (1640) implies its existence, when he speaks of the 
 letter as sounded in two ways, " firme in the beginning 
 of the words, and more liquid in the middle and ends." As 
 to its present weakness, the language of Mr. Ellis is very em 
 phatic ; he represents it as little more than a vanishing quan 
 tity, and indeed, as having vanished to a great extent from 
 English pronunciation. He says: "This second (.1) may 
 diphthongize with any preceding vowel. After (a, A, o), the 
 effect is rather to lengthen the preceding vowel than to pro 
 duce a distinct diphthong. Thus, farther and lord scarcely 
 differ from father, laud : that is, the diphthongs (aj, D.I) are 
 heard almost as the long vowels (aa, AA). That a distinction 
 is made by many, by more perhaps than are aware of it, is 
 certain ; but it is also certain that in the mouths of by far the 
 greater number of speakers in the south of England the ab 
 sorption, of the (a) is as complete as the absorption of the (1) 
 in talk, walk, psalm, where it has also left its mark on the 
 preceding vowel. When Dickens wrote Count Smorl Tork, 
 he meant Small Talk, and no ordinary reader would distin 
 guish between them" And again : " The diphthongs (ej, oj) 
 are very difficult to separate from each other and from (39). 
 But the slight raising of the point of the tongue will distin 
 guish the diphthongs from the vowel in the mouth of the 
 
252 EARLY ENGLISH PR ON UN CIA TION. 
 
 careful speaker, that is, one who trains his organs to do so. 
 No doubt the great majority of speakers do not make any dif 
 ference." It is fortunate for this much-abused letter that so 
 large a part of the English-speaking world is found in America, 
 where the first English settlers / brought this r in a less attenu 
 ated state, and where their descendants have been largely 
 reinforced by users of a yet stronger r from Ireland and Scot 
 land and the continent of Europe. Instead of losing the final 
 r, like our brethren in Southern England, we are more likely 
 to restore it to its ancient equivalency with the initial letter. 
 As to the combination wr in write, wreck, wrath, etc. 
 it appears from the testimony of Hart that the w was not en 
 tirely lost in the sixteenth century. It is Mr. Ellis s opinion 
 that wr was sounded, not only then, but from the earliest 
 Anglo-Saxon times, as (rw), or more exactly as (rw), a labial 
 ized r, the product of an effort to pronounce r and w at one 
 and the same time. The French roi he represents by (rwa) ; 
 and he holds that wrath, wreck, write were pronounced 
 (rwath, rzc ek, ra cit). This seems to us improbable for Old- 
 English, and still more for Anglo-Saxon. If wratli had been 
 pronounced thus, it would almost certainly have been written 
 rwatk. To English ears the French roi appears to begin 
 with the sound of r followed by that of w ; it is safe to say 
 that not one man in a hundred would think of it as beginning 
 with a w followed by an r. Even the words what, when, 
 white, etc., were by the Anglo-Saxons originally and gener 
 ally written with hw, not wh. Would they not have followed 
 the same analogy by writing rw instead of wr, if the sound 
 had been what our author supposes ? The copyists were ac 
 customed to spell very much according to their own ear and 
 taste ; would they not, sometimes at least, have used the 
 order rw f If hw has been changed to wh, it was probably 
 not from any doubt as to the real order of the elements, but 
 from the influence of the combinations ///, cJi, tJi, rli, con 
 stantly presented in Latin orthography. There could be no 
 such reason for adopting wr in preference to the seemingly 
 more natural order rw. And if we look at other Teu 
 tonic languages, we find everywhere the same succession. 
 Thus wrath appears in the Old Icelandic as vrcidhi, in 
 
EARL Y ENGLISH PR O NUN CIA TION. 253 
 
 Swedish and Danish as vrede ; wrong in Icelandic is vrangr, 
 Danish vrang, Swedish vrang. For wring the Gothic has 
 vriggan for wreak it has vrikan, vrakjan / for write it has 
 vrits (Gr. icepaia, point of a letter ). In view of these con 
 siderations, who can doubt that the Anglo-Saxon writing wr 
 represents the real order of the sounds, or what would ap 
 pear such to hearers generally ? If so, then (rzc ) cannot have 
 been the Anglo-Saxon pronunciation ; nor is there any 
 reason for supposing that the Old-English pronunciation dif 
 fered in this case from the Anglo-Saxon. The only argu 
 ment we can see for Mr. Ellis s (rw) is the difficulty of mak 
 ing a true English w audible before r at the beginning of a 
 word. We do not deny the difficulty ; nor do we undertake 
 to determine the precise sound of wry but we could easily 
 believe that w may have had in this case a somewhat stronger 
 sound than we are wont to give it a sound perhaps ap 
 proaching to the South German w (a v pronounced without 
 pressure of the teeth against the lips), which Mr. Ellis (p. 
 290) finds on British ground in the Aberdeen pronunciation 
 of write as (bhriit). 
 
 In this connection we have to confess some feeling of 
 doubt, if not of skepticism, as to our author s whole theory 
 of labialized consonants. He finds in French loi (he/a) a 
 labialized /, which he thinks existed once in English talk 
 being once sounded (talzc/k) "but it has died out:" yet 
 why not recognize it in ahvays ? So he finds a labialized k 
 in quell (kwel), a labialized t in twin, a labialized d in dwell, 
 etc. Here the qu, tzu, dw are in his view simple consonants, 
 a k, t, d, pronounced in a labial position. To us they still 
 appear as composite sounds. Compare high dwell the birds 
 with hide well t/te birds : in continuous utterance are they 
 not perfectly alike ? What Mr. Ellis says of twin and dwell, 
 " that the opening of the lips [from the rounded closure re 
 quired for a /] is really simultaneous with the release of the 
 (t, d) contact," we are unable to reconcile with the testimony 
 of the ear as to our own pronunciation and that which we 
 hear around us. 
 
 The wretched weakness of utterance which has changed 
 know to no (converting science into mere negation) was, ac- 
 
254 EARLY ENGLISH PR ON UN CIA TION. 
 
 cording to all the authorities, still unknown, or at least not 
 prevalent, in the sixteenth century; nor did it become uni 
 versal before the close of the seventeenth. So too it appears 
 that the gh in light, weigh, bough, etc., was heard, though 
 probably with only a feeble utterance, in the sixteenth centu 
 ry. As to the precise sound, there is an uncertainty ; per 
 haps, as in Scotland now, it may have had both sounds of the 
 German ch. It hardly survived the middle of the seventeenth 
 century. In most words it simply dropped away without any 
 further change ; but a preceding short i became long, as in 
 light, night, etc. In some cases, however, it was replaced by 
 another spirant sound, an/, as in laugh, cough, rough, and 
 in the vulgarisms oft, tJwft, for ought, thought. In sigh, 
 height, drought, it was replaced, at least in occasional use, by 
 the spirant th : the pronunciations sith, hcith, drouth are 
 mentioned by orthoepists at the beginning of the last centu 
 ry ; and drouth is still heard in our country, and has even been 
 adopted in an American Dictionary of the English language. 
 The suppression of the h, like that of the weak r, would 
 perhaps have become an accomplished fact if our language 
 had been wholly dependent on the people of Southern Eng 
 land. On this subject our author remarks : 
 
 " In England the use of the // (H) among the illiterate seems to de 
 pend upon emphatic utterance. Many persons when speaking quietly 
 will never introduce the (H), but when rendered nervous or excited, or 
 when desiring to speak particularly well, they abound in strong and un 
 usual aspirations. It is also singular how difficult it is for those accus 
 tomed to omit the h, to recover it, and how provokingly they sacrifice 
 themselves on the most undesired occasions by this social shibboleth. 
 In endeavoring to pronounce the fatal letter, they generally give them 
 selves great trouble, and consequently produce a harshness quite un 
 known to those who pronounce (H) naturally. An English author, S. 
 Hirst, writing an English Grammar in German, in which fifty quarto 
 pages are devoted to a minute account of the pronunciation of English, 
 actually bestows one hundred and sixty-seven quarto lines of German, 
 measuring about ninety feet, upon attempting to show that formerly"// 
 was not pronounced in English, and that it was altogether an orthoepistic 
 fancy to pronounce it, saying that almost all non-linguists would admit 
 that h was generally mute, or at most scarcely audible, and that linguists 
 who denied this in theory gave into the practice. The division of the 
 people is not exactly into linguists and non-linguists, but it must -be 
 
EARL Y ENGLISH PR ON UN CIA TION. 255 
 
 owned that very large masses of the people, even of those tolerably ed 
 ucated and dressed in silk and broadcloth, agree with the French, 
 Italians, Spaniards, and Greeks, in not pronouncing the letter H." 
 
 The sounds of sh in such words as sure, pressure, mission , 
 special, motion, and zJi in such as measure, vision, excision, 
 are unrecognized by the orthoepists of the sixteenth century, 
 and are not mentioned before the middle of the seventeenth. 
 Yet traces of these changes are pointed out in Shakespeare 
 and Rowley. Nor can it be doubted that, of the other 
 changes ascribed to the seventeenth century, quite a number 
 were already in progress during the sixteenth. Thus, if Hart 
 had not written, there would have been no evidence that the 
 pronunciation of ai as a simple vowel (ee) was known to the 
 sixteenth century : all other orthoepists of that age make it a 
 true diphthong; but Hart in 1569 gives it uniformly the 
 sound of (ee). Mr. Ellis regards the fifteenth and seven 
 teenth centuries as the grand periods of disturbance and revo 
 lution in the history of our language ; he thinks, with much 
 reason, that they owed this character to the political agitations, 
 the great civil contests, by which they were shaken. The 
 intermediate century, the sixteenth, he looks upon as a period 
 of comparative stability and repose, both in politics and in 
 pronunciation. Perhaps this is the true conception. Yet to 
 our mind the facts reported in this book suggest rather the 
 view that a great revolution in English pronunciation was 
 going forward from the opening of the fifteenth to the close 
 of the seventeenth century, and that it moved on with unin 
 terrupted and almost unremitted progress from the beginning 
 to the end of its course. 
 
 In the fourth chapter, which concludes this first part, Mr. 
 Ellis deals with the pronunciation of English during the four 
 teenth century. Here he finds his principal means of investi 
 gation in the poetry, and especially in the rhymes, of Chaucer 
 and Gower. He foresees that his procedure will be objected 
 to, on the ground that imperfect rhymes, which often occur in 
 modern verse, are likely to have been yet more common in 
 that earlier and ruder age. But he denies the force of the ob 
 jection, contending that the probabilities of the case point the 
 other way. A rhyme which is good to the eye, but not to 
 
256 EARL Y ENGLISH PR ONUNCIA TION. 
 
 the ear, may be tolerated by reading men ; to men who do 
 not read and such to a great extent were the public of 
 Chaucer and Gower it is no rhyme at all. Appealing main- 
 ly to the ear, these poets were actually less likely than later 
 rhymers to satisfy themselves with loose and inexact corre 
 spondences of sound. This d priori argument he fortifies 
 with remarkable success by an examination of their verses, as 
 represented by the best manuscripts. In Wright s edition of 
 the Canterbury Tales, founded on the Harleian MS. No. 
 7334, and containing 17,368 lines, he finds less than fifty 
 rhymes in which the spelling indicates a difference of pronun 
 ciation. Of these few exceptions most can be removed, 
 either by the readings of other manuscripts, or by conjectu 
 ral changes of a simple and natural character. The 33,000 
 verses of Gower s Confcssio Amantis furnish only nine in 
 stances of faulty rhymes, and for these it is shown that the 
 editor, not the author, is really responsible. The nature and 
 force of the evidence derived from the rhymes of these poets 
 is perhaps most strikingly seen in the case of the long i or y. 
 These letters appear in a multitude of rhymes like the follow 
 ing : wyse and justise, write and mcrite, vice and office, wyn 
 and famyn, side and Cnpide, lykc and ret or ike, wliilc and 
 Virgil e, Bible and possible, fyndc and Inde, I and enemy, 
 thcrby and mercy, sky and tritely, wJiy and almighty, by and 
 lady, etc. Of the second words in these pairs, the last three, 
 truely , almighty, and lady, come to us from the Anglo-Saxon. 
 They still have in the last syllable the same vowel sound, or 
 nearly the same, as in that language. It is not to be sup 
 posed that in the fourteenth century they should have taken 
 up a sound like our diphthongal i long, only to lose it again 
 by the sixteenth. Still less can we suppose this in reference 
 to justise, merite, office, famyn, Cupidc,rctorikc, Virgil c, pos 
 sible, Indc, enemy, mercy, which in Chaucer s time had just 
 come in from the French, retaining still their native French 
 accent, and could scarcely have undergone a change of vowel 
 sound so sudden and extreme. But these violent supposi 
 tions are only to be avoided by admitting that the first words 
 in the several rhyming pairs wysc, write, vice, wyn, side, 
 lyke, while, Bible, fyndc, sky, why, I, by, and thcrby still 
 
EA RL Y ENGLISH PR ON UN CIA TION. 257 
 
 retained in the fourteenth century their primitive sound of i 
 as (ii) or (//). By similar evidence less abundant, indeed, yet 
 sufficiently decisive it is shown that Chaucer s ou or ow, in 
 hons, hoiv, rtfoc ^ , about c, powrc, doun, broun, fouude, etc., 
 had not yet acquired the diphthongal sound which it bore in 
 the sixteenth century, but was generally pronounced as a sim 
 ple (uuj ; only where the on or ow corresponded to an Anglo- 
 Saxon ow, azv, did it have a diphthongal pronunciation, as 
 (oouj. And in like manner it is proved that a, c, o had the 
 same sounds as are generally given to those vowels in the 
 languages of continental Europe. 
 
 But the most important, and at the same time the most dif 
 ficult point in the language of Chaucer and Gower is the pro 
 nunciation of the final unaccented e. Mr. Ellis has greatly 
 enriched his work by taking into it, in a condensed form, the 
 masterly researches oirthis subject by Professor Child of Har 
 vard University, which were published in the Memoirs of the 
 American Academy, Vol. IX. Professor Child has proved 
 that the unaccented final e was generally sounded in the 
 poetry of Chaucer and Gower ; but also that it was frequently 
 silent. Of these exceptional cases, he is able to refer much 
 the greater part to certain general principles or habits ; but 
 others seem to depend on the mere caprice of the author, and 
 serve to indicate a varying usage and a progressive tendency 
 to suppress the letter. The parallel furnished in German 
 poetry by the frequent and capricious omission of the final 
 e is noticed by Professor Child ; though here the influences 
 of education and literature will doubtless save the letter from 
 total extinction, while the sensible habit of leaving it unwrit 
 ten where it is unpronounced will save much time and toil to 
 the philologists of the future. 
 
 Mr. Ellis proposes a uniform orthography for Chaucer, 
 which he would wish to apply also to other writers of the 
 fourteenth and fifteenth centuries. It is founded on the ortho 
 graphy of the Harleian MS., already referred to ; the usual 
 modes of spelling in that MS., the representations usually 
 given by it to the different sounds of the language as it then 
 was, are to be made universal, the occasional exceptions being 
 altered into conformity with the general rule. It is the spelling 
 17 
 
258 EARLY ENGLISH PR ON UN CIA TION. 
 
 which the copyist of that MS. would presumably have used if 
 he had been intent on a uniform orthography. A text of the 
 fourteenth or fifteenth century, printed according to this sys 
 tem, could be read without difficulty, and by a little practice 
 even fluently, with a pronunciation not widely different from 
 that of its own time. It seems to us desirable that some such 
 texts should be prepared and printed for school use. If a 
 historical study of the English language and literature is to be 
 made common in our educational institutions, it is important 
 to lessen, in some degree at least, the difficulties arising from 
 the endless and needless variations of Old-English spelling. 
 It would be possible, and perhaps expedient, in such books, 
 to give in the margin the actual spelling of one or two good 
 manuscripts, in the most important cases where it differed 
 from that used in the text. At the same time we think that, 
 if Chaucer is really to be popularized, it can only be done by 
 modernizing his orthography. The words which belong to 
 the modern language must appear in the spelling with which 
 all are familiar. But it is not less necessary that the text 
 should be so given that it can be read rhythmically. An un 
 certain and halting rhythm, which fails to fulfil its own pro 
 mises, and is continually leaving the reader in the lurch, is be 
 yond measure disagreeable, and even painful, to a rhythmic 
 ear ; plain prose is infinitely better than such a rhythm. To 
 cure this defect, the unaccented c must be supplied just as far 
 as is requisite to give each verse a proper rhythmic succession 
 of syllables and accents. We venture to add the opening 
 lines of the Prologue to the Canterbury Tales, as a specimen 
 of the arrangement which we have* in mind. Where the ac 
 cent of a word varies from that now usual, it is marked in the 
 text by an acute over the vowel of the accented syllable. A 
 double dot over an e which is now silent shows that it is to 
 be pronounced, and to make a separate syllable, in reading. 
 A small circle over an c which is now heard shows that it is 
 to be suppressed, or at least that it does not count as a sylla 
 ble in the verse. The double clot and the circle may also be 
 used for like purposes over other vowels. Where an e which 
 is not written even now has to be supplied in reading, it is 
 represented by a reversed apostrophe, as [ ]. 
 
EARL Y ENGLISH PR ON UN CIA TION. 2 5 9 
 
 " When that April with his shower s sweet 
 
 The drought of^vlarch hath pierced to the root, 
 
 And bathed every vein in such liquor, 
 
 Of which virtue engendered is the flower; 
 
 When Zephi rus eke with his sweet breath 
 
 Inspired hath in every holt and heath 
 
 The tender crop s, and the young sun 
 
 Hath in the Ram his half course y-run, 
 
 And small fowl s make melody, 
 
 That sleep all the night with open eye, 
 
 So pricketh them nature in their a corages : 
 
 Then long folk to go on pilgrimages, 
 
 And palmers for to seek strange strands 
 
 To b feme halwes, d kouthe in sundry lands ; 
 
 And specially, from every shire s end 
 
 Of Eng land, to Canterbury they wend, 
 
 The holy blissful martyr for to seek, 
 
 That them hath holpen when that they were sick." 
 
 "hearts. ** distant. c saints. d known. 
 
 In such a text we should have a genuine Chaucer, the ipsa- 
 verba, if not the ipsissima y of the poet himself. But it would 
 be free from the worst of those stumbling-blocks which now 
 vex and baffle the ordinary reader. The general aspect of 
 the words would be familiar to his eye ; the sense would be 
 almost or quite as easy to comprehend as that of Shakespeare ; 
 and the lines would yield without pains and puzzling a rhythm 
 fairly satisfactory to the ear. In thus giving to Chaucer a 
 modernized orthography, we should only be treating him as 
 we do Shakespeare ; no editor thinks of reproducing the 
 spelling of the old folios. If Shakespeare were accessible 
 only in that antiquated orthography, it is certain that the 
 number of his readers and the extent of his influence would 
 be seriously diminished. 
 
 Since the foregoing paragraphs were written, we have re 
 ceived " Part II." of the work under review. This, however, 
 does not finish the book. "On account of the unexpected 
 length " of the author s " investigations, the Societies for 
 which they are published have found it most convenient to 
 divide them into /<?;/; parts, instead of tivo, as previously con 
 templated." This second part contains a little more than two. 
 
2 6o EARL Y ENGLISH PR ON UN CIA TTON. 
 
 hundred pages, of which about two thirds are occupied with the 
 pronunciation of English prior to the fourteenth century, as 
 well as that of the Anglo-Saxon, Old Icelandic, and Gothic ; 
 while the remainder treats of the correspondence of orthogra 
 phy with pronunciation from Anglo-Saxon times to the present 
 day. A hasty glance has shown us that it is rich in curious and 
 instructive matter ; but we cannot yet undertake to criticise, or 
 even to describe, its contents. We will speak of only one 
 point which has chanced to attract our attention while turn 
 ing over its pages. In representing, as he does by palaeo- 
 typc, the original pronunciation of various specimen-texts of 
 early English, it seems to us that Mr. Kills has hardly been 
 as careful and exact as could have been desired in reference 
 to vowel quantities. That he has been satisfied with treating 
 this matter somewhat easily may be inferred from the fact 
 that he occasionally varies from himself in the course of one 
 or two pages. Thus asc, as, in selections from Dan Michel, 
 is given (aase) on page 412, but (asc) on page 413. In an 
 extract from Richard cle Hampolc, on page 414, the adverb 
 here is given both as (lleer) and (ller). In a paternoster, on 
 page 442, we find in forms of the verb do both (don) and 
 (miisdoon), where, too, one is surprised to see a long quantity 
 in the prefix mis / compare also (doon) on the next page. 
 In an ave and credo, on page 443, lauird, Morel, is repre 
 sented by (laavird), but laucrd, a variation of the same 
 word, by (laverclV On the same page, the adjective dcd, 
 dead/ appears both as ided) and as (deed); and the im 
 perative of the verb forgive appears as (forgiiv), but the 
 indicative as (forgiveth). In this last case, if any distinction 
 was to be made, we might have expected (forgiv) and (for- 
 giiveth) ; authority for this could be produced from the 
 Ormulum. In the proclamation of Henry III. (1258), the 
 ending -liche> the modern -ly ( appears as (-liitshe) on page 
 501, but as (-litshe) on page 503. 
 
 The English language has sometimes been spoken of as 
 having wholly lost the distinctions of vowel quantity which 
 belonged to the Anglo-Saxon; as if it stood related to the 
 mother language in this respect as the French is related to 
 the Latin, or the modern to the ancient Greek. It is true 
 
EARL Y ENGLISH PR ON UN CIA TION. 261 
 
 that, in comparing Anglo-Saxon with English, we find exten 
 sive changes of vowel quantity. Long vowels, where they 
 follow the accented syllable, have been shortened almost 
 uniformly. Thus the Anglo-Saxon -lice > early English -licke 
 (-liitshe), has become ly in adjectives and adverbs of modern 
 English. But in accented syllables, if we mistake not, a 
 careful comparison will show 7 that the vowel quantity of the 
 Anglo-Saxon has been preserved in a majority of the words 
 which have come to us from that source. And most of the 
 changes, whether from long to short or from short, to long, 
 can be referred to a small number of euphonic principles or 
 tendencies.* The effect of a weak r (before a consonant or at 
 the end of a word) to lengthen the vowel before it, as in far, 
 aware, horse, etc., is one of the most important cases. The 
 similar lengthening before Id and nd, as in child, find, etc., 
 is as old as the thirteenth century, being the general rule in 
 the Ormulum ; and it is a fair question whether child and wild 
 on page 483, and Jiundcs, hounds, on page 479, should not 
 have been given with a long vowel. Still more numerous 
 changes have been occasioned by the tendency to lengthen 
 the short vowel of an accented penult when separated by only 
 one consonant sound from the following unaccented vowel. 
 The long sounds in water, naked, evil, open, etc., owe their 
 origin to this tendency ; and so do those in name, make, law, 
 eat, hope, and a multitude of others, which originally ended 
 in an unaccented syllable. Now this tendency also is found 
 in the Ormulum, which carries it to a greater extent even than 
 modern English : thus, narrow), heavy, body, love, give, writ 
 ten, summer, etc., are found in the Ormulum with a long 
 vowel in the first of their two syllables. In such cases as 
 these, the modern English agrees with the Anglo-Saxon in 
 using a short vowel ; a fact which shows that the long vowel 
 could not have been universal in the thirteenth century, the 
 time of the Ormulum. We think that Mr. Ellis, who writes 
 a long vowel in some of these words such as give, written, 
 summer in other works of that century, would have done 
 better to give them short, as in Anglo-Saxon and mod 
 ern English. Still less can we approve of his writing in 
 * See the next following article. 
 
262 EARL Y ENGLISH PR ON UN CIA TION. 
 
 cases where it is at variance with the Anglo-Saxon, the 
 Ormulum, and the modern English, together : as in (wen) for 
 wen (page 412), A. S. wen, Orm. wettenn, to ween/ 
 (leev) for Icve (page 414), A. S. lifian, Orm. libbcnn, to 
 live ; (gret) for grete (page 479), A. S. great, Orm. greet, 
 great. It is not unlikely, however, that some of these 
 cases are mere errors of the press. 
 
 We hope that Mr. Ellis, in the parts yet to appear, may 
 regard this matter of vowel quantity as worthy of a closer at 
 tention than he seems thus far to have given it ; and may 
 thus add to the completeness and exactness of his admirable 
 work, 
 
XV. 
 
 ON ENGLISH VOWEL QUANTITY IN THE THIR 
 TEENTH CENTURY AND IN THE NINETEENTH. 
 
 1871. 
 
 IT is a well-known fact that the Modern Greek has lost the 
 system of vowel quantity which belonged to the ancient lan 
 guage : Kofjiij, hair, and K^^TI, village, are pronounced 
 alike by the Modern Greek ; nor is it otherwise with \vcreis, 
 thou wilt loose, and \6a-is, ( a loosing. In like manner, the 
 Romanic languages have lost the system of vowel quantity 
 which belonged to the ancient Latin. Thus the Italian and 
 Spanish mano, rosa, French main, rose, are alike in vowel 
 sound with Italian and Spanish vano, prosa, French vain, 
 prose, though the former come from Latin manus, rosa, and 
 the latter from vanus, prosa. It is an interesting question 
 whether our own language has had in this respect the same 
 experience as the Modern Greek and the Romanic tongues ; 
 whether we have wholly lost those distinctions of vowel quan 
 tity which undoubtedly belonged to the Anglo-Saxon. 
 
 It is sometimes said that we have no proper distinction of 
 long and short quantity in English : all our vowels are alike in 
 quantity ; they are all equally susceptible of prolongation and 
 abridgment ; or rather, any vowel rnay be so prolonged as to 
 exceed the ordinary quantity of any other : thus,_/z// may be 
 so prolonged in utterance as to take more time than is ordina 
 rily given to feel or file. But this must have been the case 
 also with Greek and Latin vowels : their absolute time was 
 not fixed, but variable. One speaker must have talked faster 
 or slower than another ; the same speaker must have talked 
 faster or slower at one time than at another ; even on the same 
 occasion one sentence or clause must have been uttered faster 
 or slower than another : and thus it could hardly fail to hap 
 pen that a short vowel would sometimes have a longer sound 
 
264 ENGLISH VOWEL QUANTITY. 
 
 than the ordinary or average quantity of a long vowel. But 
 the average quantity of a short vowel was less than the 
 average for a long one ; or, with the same general rapidity of 
 utterance, the quantity of long vowels exceeded the quantity 
 of short. I say simply, the quantity of long vowels exceeded, 
 was greater than, the quantity of short ; not that the former 
 was just double the latter. It may be presumed that in ordi 
 nary spoken utterance there was not any so exact relation ; 
 with the same tempo, the long might equal I y 2 of the short, 
 or 1 5/ of the short, or I3/^ of the short, as well as just 2 ; and 
 probably the average actual ratio was rather less than 2 to I ; 
 the average long would occupy rather less than twice the time 
 of the average short. The fixing of 2 to I as the precise nu 
 merical relation was probably the work of rhythmopceia, or 
 of rhythmopceia and melopceia together. When longs and 
 shorts were combined in rhythmic composition, and especially 
 when a musical accompaniment was added, the longs and 
 shorts must have a definite and fixed ratio to each other ; and 
 the ratio of two to one was the most simple, obvious and 
 convenient. 
 
 Now as regards our English syllables, it is certain that we 
 have nearly or quite lost the feeling of length by position. 
 By this I mean that we do not recognize a short vowel fol 
 lowed by two consonants as having any distinct relation to a 
 long vowel followed by one consonant. We can perceive, of 
 course, that there is more sound in fist than in fit, and more 
 sound in fight than \\\ fit ; but we do not recognize any 
 special relation of quantity between fist ai\<\fght. But in re 
 ference to vowel sounds, it should seem that our case is not 
 essentially different from that of the ancient Greeks and 
 Romans. There are certain sounds which, with the same 
 tempo, the same general rapidity of utterance, we recognize 
 as occupying more time than others ; we thus recognize the 
 former as long, the latter as short. It is true, the English 
 short vowels differ somewhat in quality of sound that is, in 
 position of the speech organs from the long vowels which 
 most nearly resemble them, the short being a little more open 
 than the corresponding long : there is a more open sound in 
 fill than mfcei, in/?/// than in fool, in fell than in fail, etc. 
 
ENGLISH VOWEL QUANTITY. 26$ 
 
 But this does not affect the relation of quantity. It is clear 
 that we have long vowel sounds in file and foul, in feel and 
 fool, in fail and foal, and in fall ; and that we have short 
 vowel sounds in fill and full, \\\ fell, in doll, in dally, in dull. 
 Pronounce the two series in succession : on the one hand, file, 
 foul, fed, fool, fail, foal, fall ; on the other, ///, /////, fell, 
 doll, dally, dull the difference of quantity is manifest and 
 unmistakable. 
 
 If then we have long and short vowels clearly distinguished 
 from each other, it becomes a matter of interest to inquire 
 whether the distinction is recent or ancient ; whether, and how 
 far, the vowels now sounded long or sounded short were 
 sounded long or short respectively in early periods of the 
 language. The question here raised is not whether our vowels 
 have kept the same sounds, but whether they have kept the 
 same quantities. It matters not that long a has passed from its 
 original sound (of all] to that heard mfamc; long e, from the 
 sound in prey to that in key / long /, from the sound in pique 
 to that in pike / long o, from the sound in bone to that in 
 boon / long u, from the sound in prude to that in proud. 
 These changes have taken place to a very great extent, but 
 they do not affect the question now before us : the old long 
 vowel remains long still. So if the short vowels I, #, u (in 
 fill, full, fell) have become a little more open ; if, on the 
 other hand, a has generally become closer (as in dally}, ap 
 proaching the sound of <5 (in dell) ; if the old 6- has disap 
 peared altogether, passing into a sound more open than that of 
 a in fall, and often described as its corresponding short (thus 
 fall, folly) ; if ft, and sometimes other short vowels, have 
 sunk into the obscure and undistinguished sound heard in 
 dull, done these changes do not concern us here, as the old 
 short vowel still remains short. 
 
 In the inquiry whether, and how far, we can trace back the 
 present long and short quantities of English vowels, we find 
 our best guide and help in a single (and singular) production 
 of early English literature. I refer to the Ormulum, so called 
 from the name of its author, the monk Orm, or Ormin, who 
 -wrote in the eastern part of England, some time in the thir 
 teenth century, or fully six hundred years ago. A series of 
 
2 66 ENGLISH VOWEL QUANTITY. + 
 
 metrical homilies on the successive daily lessons of the church 
 service, its interest is philological much more than literary. 
 In the only manuscript from which it is known to us not 
 improbably the only manuscript of it that was ever written 
 we find on the^part of the writer a careful and systematic 
 regard to vowel quantity. He has a spelling of his own, to 
 which he adheres with much consistency ; and in this spelling 
 the most peculiar and conspicuous feature is the doubling of 
 every consonant which follows a short vowel. There was a 
 tendency to this in the general English orthography even at 
 that early period ; what is remarkable in this author is that 
 he consciously carried out this tendency as a uniform and uni 
 versal rule. Thus he writes it with a double /, if with a 
 double/", hundred with a double ;/ and a double d, lastctli 
 with a double s and a double sign for th. So much is he at 
 tached to this spelling, in spite of its strange and whimsical 
 appearance, that in a preliminary address (seemingly very 
 little needed) to the future copyist he insists upon a careful 
 conformity to it : 
 
 Annd tatt he lokc wcl tatt he 
 
 An boc staff write twi^css, 
 E^w/uer thccr itt uppo tliiss boc 
 
 Iss writcnn o thatt wise. 
 Lokc he well thatt Jict write swa, 
 
 Forr he nc ma\\ nohht clless 
 Onn EnngJissJi write/in riJiJit te word 
 
 Thatt wife he wcl to sothc. 
 
 Which may be modernized thus, in the same measure, all but 
 the unaccented ending of the even lines : 
 
 And that he look [full] well that he 
 
 A letter write twice [over], 
 Wherever it upon this book 
 
 Is written on that wise. 
 Look he well that he write it so 
 
 For he (nc) may not else 
 In English write aright the word, 
 
 That wit he well to sooth. 
 
 Thus then the spelling of the Ormulum enables us to say, 
 in the case of every vowel followed by a consonant in the 
 
ENGLISH VOWEL QUANTITY. 267 
 
 same word, whether the author pronounced it or, at least, 
 supposed that he pronounced it long or short. In order 
 now to arrive at some general conclusions as to the persis 
 tence of vowel quantity in English, I have looked through the 
 vocabulary of the Ormulum, as presented with much fulness 
 and clearness in the second volume of White s edition (Oxford, 
 1852), and have noted the words which survive in the English 
 of to-day, including some few which, though lost out of the 
 common language, are still generally known from their use in 
 literary works of the modern period. I propose to state the 
 general results of this comparison between the quantities of 
 words as indicated in the Ormulum, and the quantities of the 
 same words as heard in modern English. It should be said 
 that the vocabulary of the Ormulum is not large. The work 
 is nearly as long as the Odyssey ; but the number of different 
 words used in it is much smaller than in the Greek poem. 
 There is a great sameness, a wearisome sameness, in the 
 contents of the book : the leading facts, principles, and pre 
 cepts* of the Christian system are repeated over and over 
 again with little variety of expression. Still there are words 
 enough to give a fair idea of the relation between the vowel 
 
 o o 
 
 quantities of Saxon English six hundred years ago and those 
 of Saxon English at the present time. I say " Saxon Eng 
 lish," because few Latin words (aside from those already 
 taken into the Anglo-Saxon), and fewer French words, are to 
 be found in the Ormulum. Its language is as purely Teu 
 tonic as the modern German. 
 
 The most general statement suggested by the extended 
 comparison of which I have been speaking is this : that in the 
 great majority of cases the vowels which had a long sound six 
 hundred years ago are long now ; those which had a short 
 sound then are short now And -if the exceptions are pretty 
 numerous if in a good many cases the long sound of the 
 Ormulum corresponds to a modern short, or the short sound 
 of the Ormulum to a modern long most of these exceptional 
 cases depend on a few obvious conditions, on clearly marked 
 euphonic influences and tendencies, so that cases of capri 
 cious variation, of variation without apparent principle, are 
 comparatively few. It will probably be most instructive, if 
 
268 ENGLISH VOWEL QUANTITY. 
 
 we consider first (and indeed, chiefly) these euphonic condi 
 tions and tendencies which have led to differences of quantity 
 between the language of six hundred years ago and that of 
 to-day. 
 
 In the first place, then, let us look at the cases where the 
 loss of a consonant sound has occasioned the lengthening of a 
 short vowel before it : as in alms, O. allmcss ; buy, O. biggcnn, 
 and the like. Such changes can hardly be considered as vio 
 lations of the old system of quantities. If there is here an 
 alteration, an increase, of vowel quantity, it only takes the 
 place of a consonant quantity withdrawn. It is simply that 
 the time before occupied by a vowel and consonant is now 
 occupied by the vowel prolonged. This is sometimes de 
 scribed as an absorption of the consonant by the vowel, 
 sometimes as a vocalization of the consonant. Neither of 
 these expressions gives a distinct idea of the nature of the 
 change. Of course, in every instance of the kind there must 
 have been a time of transition, when the consonant was be 
 ginning to be omitted, when the very speakers who omitted 
 it were perfectly aware of its existence, and perhaps generally 
 pronounced it, but occasionally let it drop, with a lengthening 
 of the preceding vowel. Now this consciousness of a conso 
 nant with a claim to be pronounced is an important element 
 in the phenomenon. The speaker who does not really pro 
 nounce it does not feel that he can omit it altogether ; he 
 does not feel that he is altogether omitting it. To his own 
 feeling, he gives it a kind of recognition. He perhaps brings 
 the organs of speech into some sort of approach toward the 
 position required for pronouncing the consonant, so that the 
 preceding vowel passes into a sound more or less modified, 
 which does duty for the consonant. If this modification con 
 tinues to be made, then the resulting long vowel-sound will 
 not be a mere simple prolongation of the preceding short, but 
 something different, perhaps a diphthong. Yet it may very 
 well happen that, in this ^z/mz-pronunciation of the consonant, 
 the approach made by the organs to the position for that con 
 sonant will grow more and more slight, and the sound pro 
 duced will differ less and less from a mere continuation of the 
 preceding vowel ; until finally and perhaps very soon it 
 
ENGLISH VOWEL QUANTITY. 269 
 
 :omes to be just that and nothing else, and the consonant is 
 replaced, as its claim for utterance is felt to be satisfied, by a 
 simple addition of quantity to the preceding vowel. But, 
 whatever may be thought as to the rationale of the process, 
 it is one of which we find numerous instances in comparing 
 the Semi-Saxon of the Ormulum with modern English. Thus 
 where /, followed by another consonant, has been suppressed 
 in utterance, though still retained in writing, the short vowel 
 before it has become long in 
 
 O. alluiess (alms), A. S. Cdmesse, Lat. clcemosyna. 
 calif (calf ), A. S. cealf. 
 follc (folk, people), A. S.folc. 
 hallf (half, behalf}, A. S. hcalf. 
 illc (each, every, Sc. ilk), A. S. illc, clc, ylc. 
 sallfe (salve, ointment), A. S. scalf. 
 sallme (psalm), A. S. scalm, Lat. psa Imu s. 
 
 In should (O. sJwllde), would (O. wollde and woldc), we 
 have a short vowel sound ; but we may see from the OIL that 
 the vowel was first lengthened (sJioii d, luou d, with ou as in 
 youtlt) ; though afterward it became short again, by a new 
 and independent change, similar to that by which good and 
 stood have received their present short pronunciation. The 
 same change has occurred also in could, which never had an 
 I actually sounded. The Ormulum, like the Anglo-Saxon, 
 has catlic, in which the til became d, and the vowel was after 
 ward shortened. As people were accustomed to write a 
 silent / in should, would, and regarded could as a word of 
 similar character, they put a silent / into that also. There 
 would have been more propriety in the insertion of a silent ;/ ; 
 for this letter belongs to the root as seen in can. It is a fea 
 ture of the Anglo-Saxon in its earliest known form that it 
 drops 11 before th or s, and lengthens the preceding vowel : as 
 gos (goose] for gans, sodh (sooth) for santh (which means 
 being, existing, and is identical with Lat. -sens, in praesens, 
 -scntis) ; and so cudhc for cunt he (could). 
 
 Again, where g has been suppressed in utterance, the short 
 vowel before it has become long in 
 
 O. biggenn (to buy), abiggcnn (to aby, pay for), A. S. byegan, dbycgan. 
 leggcnn (to lay), A. S. Iccgan. 
 scggenn (to say), A. S. secgaii. 
 
270 ENGLISH VOWEL QUANTITY. 
 
 But in most cases of this kind the g appears in the Ormulum 
 softened into the consonant /-sound (:), which after the short 
 vowel is written double (]]) : thus 
 
 O. daft (day }, pi. datfess, daftess, A. S. dr>g, pi. dagas. 
 driftc (dry}, A. S. drygc. 
 eftlenn (to ail), A. S. eg/an. 
 either r (cither), A. S. agdher. 
 faftcrr (fair), A. S.fXger. 
 fteftl( flail), Gcrm.flcgel, *L%.\.. flagellum. 
 
 geftncnn (to gain),ga$/icnn (gain}, O. N. gagn (advantage). 
 geftnlikc (aptly, cf. ungainly), O. N. gcgn (apt, clever), A. S. un- 
 
 gagne (of no effect). 
 
 laft (lay), from lin (to lie), A. S. %% from licgan. 
 left (impv. //), /^Jrtfe toW), from ^ggemi, A. S. /^r, /^^, from 
 
 ma ^ (lie ;;/^y), A. S. ;;///. 
 
 wmj,^ (w^y, maid), A. S. imrg (femina, virgo). 
 
 maftdcfin (maiden), A. S. mfigden. 
 
 naftlcnn (to nail), A. S. niiglian. 
 
 rcftii (rain), A. S. n^//. 
 
 soft (impv. say), seftt/i (sait/i), seftdc (said), from seggcnn, A. S. 
 
 scge, s eg d It, s<!gde, from sccgan. 
 inn-sc^lcss (seals), A. S. sigcl, inscgel. , 
 tiveftoin (twain), A. S. twcgcn. 
 twiftcss (twice), also twiyss, A. S. /w/><7 (/ ?). 
 thriftess (thrice), also thriyss, A. S. thriga (I ?). 
 2X ^ ( // (waggon, wain), A. S. w ign, ween. 
 wefte (way), awe.ft, awc-fte (away), A. S. w/^, 6weg. 
 
 In eythcrr, ma\\ (maid), and perhaps in twz ycss and tJiri"css, 
 a long vowel of the Anglo-Saxon is found shortened (in the 
 last two, however, not uniformly) in the Ormulum. This 
 shortening may perhaps be explained as the consequence of 
 an effort to make the feeble } more fully audible. The vowel 
 may have been passed over -lightly in order that a greater 
 force of utterance might be brought to bear on the weak con 
 sonant following it, so as to give this a distinct enunciation. 
 ,It would be perfectly natural, too, that the speakers of the 
 language should become at length weary of this effort required 
 for the weak consonant ; and that they should then allow the 
 consonant to be replaced by a mere continuance of the pre 
 ceding vowel, which would thus recover its primitive long 
 quantity. 
 
ENGLISH VO WEL Q UANTITY. 2 / 1 
 
 In the word master (O. r,ia\-stre, A. S. uiagcstre, mogstrc, 
 Lat. magistcr) we do not lengthen the vowel : here the Scot 
 tish waist cr shows the truer (that is, the more analogical) pro 
 nunciation. In saith and said, the lengthened vowel sound 
 which once belonged to them is still indicated by the ai with 
 which they are written. 
 
 The 5; of the Ormulum is not always to be traced to an 
 original *. In some instances it seems to have arisen from 
 the diphthong ei in the Old Norse, the language of the so- 
 called Danes who came as invaders and settlers into eastern 
 England : the vocabulary of the Ormulum shows evident 
 marks of a Norse influence. Thus the plural pronouns thc^l 
 (they), tJie\\rc (their), tJie\\m (them) are not to be 1 explained 
 from A. S. tlia, tliara, t/idni, but from O. N. their, tJicirra, 
 theim / rc\\scnn (to raise], not from A. S. rasian, but from 
 O. N. rcisa ; hc\\lcnn (to hail, salute), not from A. S. Jial, 
 but from the corresponding O. N. /^zY/(sanus, salvus), which, 
 like E. hail, was often used in salutations. So a]] (aye, al 
 ways, ever) is perhaps to be explained from O. N. czi, ei, ey, 
 which correspond to A. S. awa j while na\\ (nay} may be a 
 mere compound of nc and a\\. The genitive Kc\\scrcss, usu 
 ally Kascrcss, from Kascre (Ccesar, Emperor), might be ac 
 counted for in the same way ; but for the, 55 in bcy-.sannz (be- 
 zcfnts, .coined in Byzantiuwi), and in the proper name R\\noc 
 (Enoch), we have no explanation to offer. 
 
 Yet again, where a consonant h has been suppressed in ut 
 terance, the short vowel before it has become long in many 
 words. I say " a consonant /i," for the Anglo-Saxon //, 
 where it stands at the end of a syllable or is followed by t or 
 th (dJi), must be regarded as a true consonant. Thus in 
 
 O. bohhtc (bought), from biggenn (to buy], A. S. bolitc, from bycgan. 
 brihht (bright], A. S. bcorht, byrht, bryht. 
 brohhte (brought}, from bringcnn (to bring), A. S. brohte, from 
 
 bringan. 
 
 cnihhtess (servants, soldiers, cf. knight], A. S. cniht. 
 dohhtcrr (daughter}, A. S. dnhtor. 
 drohh (drew), also droh, from draihenn (to draw], A. S. drog, from 
 
 drag an . 
 drulihthe (drougJif], A. S. drugadh. 
 
 i^ (virtuous, cf. doughty], A. S. dyhtig. 
 
ENGLISH VOWEL QUANTITY. 
 
 O, chhtc (eight), ehhtennde (eighth}, A. S. cahta, eahtodha, O. N. attundi. 
 / ^ // t?/ 6 ; .(revenflq, money, cf./tr), A. S.feoh,fe6. 
 fili-hfenn-^fight), A. S.feohian. 
 Jlihht (flight), A. S.^j///. 
 
 bi-kahht (caught) , also bikccchcdd (catched) ; of doubtful origin. 
 /////// (//^/, Icvis), A. S. /^/// (/ ///). 
 /////// (//>///, lux), A. S. leoht (lyht). 
 
 mahhi, mihht (might}, ;;///////< (he might}, A. S. meaht, mi /it, mihtc. 
 nahht, nihht (night), A. S. neaht, niht. 
 
 nohht (nought, not}, A. S. naht, nlijUit, from ne and aiuiht. 
 ohJit (aught}, A. S. ^///, contracted from awiht. 
 plihht (danger, state, ci. plight}, A. S. pliht. 
 rihht (right}, A. S. riht. 
 saJih (saui), from scon, sen (to ^r), A. S. scah, from j<?/5 y but see 
 
 p. 293. 
 sihhthc (sight, appearance), iunsihht (insight, knowledge), A. S. 
 
 gesihdh. 
 
 sohhtc (sought}, from sckcnn (to seek}, A. S. soJitc, from sccan. 
 tahhtc (taught}, from tccchenn (to teach}, A. S. tcehte, from ttccan. 
 thohh (though}, A. S. //// //. 
 thohhtc (he thought}, from thcnnkcnn (to think}, A. S. thohte, from 
 
 thencan. 
 thuhhtc (seemed, cf. mcthoiighi}, from thinnkcnn($.Q seem), A. S. 
 
 thnhte, from thyncan. 
 
 tuchhtc (wciglif), A. S. innht, gciviht, from vegan (to weigh}, 
 wihht (being, person, cf. vigJit}, A. S. luiJit. 
 ivrihJitc (maker, worker, cf. wright), A. S. ivyrhta. 
 ivroJiJitc (wrong Jit}, from wirrkciin {to work], A. S. ivorhte, ivrohtc, 
 
 from ivy r can. 
 
 In some of these words (doJiJitcrr, drohh, HJiJit, lililit, 
 nohht, ohht, sohhte, tahlitc, ihohJi) we find a shortening of 
 the Anglo-Saxon long vowel, similar to that just noticed in 
 e\\thcrr, ma]], etc., and explainable in the manner then pro 
 posed ; though the combination of consonants (Jit), which in 
 most of them follows the vowel, may have had something to 
 do with its change of quantity. 
 
 Lastly, in a number of words which in the Ormulum have 
 the consonant w repeated, showing that the vowel before it 
 was then sounded short, this consonant is lost in English (or, 
 at least, has no consonant powen, and the vowel sound is 
 long. Such arc 
 
 O. chewwenn (to chcii.}, A. S. ccoivan. 
 
 dawwess (hoofs, cf. claw}, A. S. eld, clawu, pi. clan ". 
 
ENGLISH VOWEL QUANTITY. 
 
 O. cnewwe (Knee), pi. cncwwess, cnes, A. S. crt 
 dcewwenn (to bedew], from daw (dew), A. S. 
 fowwerr^fowwre (four), foivw err ti^ (forty), A. ^Tjeo-My jjdowertig. 
 strawwenn (to straw, straw), A. S. streowiati, strcdwian, 
 throwwinnge (three, suffering), A. S. throwuing. 
 trewwcss (trees), also treos, tres, sing, treo, A. S. treow, treo. 
 trowwe (true}, trowwenn (to trow), trowwthe (trutJi), A. S. treowe, 
 treow ian, trebwdh. 
 
 In most of these words we see an Anglo-Saxon long vowel 
 shortened before the weak w, as we have already seen it be 
 fore h and 3. In nozwwharr (nowhere), the shortening which 
 may be compared to that of no in English nothing really 
 takes place before Ji, the true order of the sounds being that 
 represented in the Anglo-Saxon orthography, nahwcer. The 
 words owwtherr, noivwtherr (A. S. dwdher, ndwdher), have 
 the sense -of either, neither ; but these last connect themselves 
 with A. S. (Bgdher. The form uowwt (cattle, Sc. nowt) is to 
 be explained from O. N. nant, while English neat corresponds 
 to A. S. neat. In these words, where the ww precedes a 
 consonant, its sound can hardly have differed very much from 
 that of the vowel u. Indeed the Latin au is represented by 
 aww in clawwstremann (cloister-man, monk) from Lat. clans- 
 trnm, and in the proper name Sannt Awwstin (Saint Austin, 
 Augustine), though the Emperor (Kas err- king) Augustus ap 
 pears as Angnsstnss. 
 
 We have now reviewed all the cases where a consonant, 
 which in the time of the Ormulum was heard after a short 
 vowel, is lost in modern English ; and we have seen that in 
 all but a very few (should, would, saith, said, master, them, 
 not), the preceding vowel is long in our present pronuncia 
 tion : even for these few, there is evidence that the most 
 have been pronounced with long vowel sound, though that in 
 more recent times has become short. 
 
 The next important point we have to consider is the effect 
 of a weak r on the preceding short vowel. By weak r I mean 
 to designate that peculiar sound of the letter which it has as 
 sumed in our present English, wherever it stands before a 
 consonant or at the end of a word as \i\ far , farm, for , form, 
 fur, firm. This is evidently weaker than the sound of r in 
 18 
 
274 ENGLISH VOWEL QUANTITY. 
 
 farrow, forest, borough, merit, spirit, etc. According to Mr. 
 Ellis, Ben Jonson, in his English Grammar of 1640, is the 
 earliest writer who gives any sign of having recognized this 
 distinction between a stronger and a weaker r in the pronun 
 ciation of our language. Perhaps the phenomenon itself, the 
 weakening of r where it is final or followed by a consonant, 
 may not be much older than that time. In the dialectic pro 
 nunciation of the Irish, which has in many points preserved 
 the older English sounds, it has not yet established itself. 
 This weak r is most easily produced after the neutral vowel, 
 so called, which is heard as short in cub, cud, and as long in 
 curb, curd. Except after the sounds of ah and au (as in far, 
 for), this u sound is always heard before a weak r : thus it 
 comes in, as a brief yet perceptible element, after the proper 
 vowel in fire, flour, fear, four, fare. When the preceding 
 vowel was a short i or e, this has been overpowered by and 
 merged in the following u sound. Thus fir and her are not 
 distinguished in pronunciation from/"?/; and Hur. The short 
 u itself becomes long when this consonant follows it : as we 
 see in comparing burgh with burrow, where the first has a 
 long sound before weak r, the second a short one before 
 common r. The point with which we are now concerned is 
 this, that the weak r, whether heard at the end of a word or 
 before another consonant, is always preceded by a long vowel 
 sound ; if the preceding vowel sound was originally short, it 
 has become long. 
 
 It is altogether improbable that in the time of the Ormulum 
 r in this position had begun to assume its present weak sound ; 
 but it had begun to affect the quantity of a preceding vowel. 
 We find quite a number of words in which a short vowel of 
 the Anglo-Saxon, standing before r in this position, had be 
 come long in the Ormulum, as in recent English : 
 
 O. ard (place, region, Sc. airf), A. S. card. 
 ccrn (earn, eagle), A. S. earn. 
 
 bcern (children, Sc. bairns], also barrncss, A. S. beam, 
 bcernenn (to burn), A. S. beornan. 
 
 birde (lineage, cf. birth}, A. S. gebyrd (birth, lineage). 
 bord (board, table), A. S. bord. 
 cherl (young man, cf. churl), A. S. ceorl. 
 
ENGLISH VOWEL QUANTITY. 2 ?$ 
 
 O. corn (corn}, A. S. corn, 
 cor less (earls), A. S. eorl. 
 eortJic, erf he (earth], A. S. cordhe. 
 forth (forth], also forrth, A. S.fordh. 
 kirrkc-gcerd (church-yard), A. S. geard. 
 hird (company, family, cf. herd), A. S. heard. 
 hirdc (herd, shepherd), A. S. hirde. 
 hord (hoard], A. S. hord. 
 
 leornenn, lernenn (to learn}, but lerrnde (learned), A. S. leornian. 
 skarn (scorn], O. Fr. escorne. 
 stirne (stern, fierce), A. S. styrne. 
 sivcrd (sword), A. S. sweord. 
 word (word], A. S. word. 
 
 ^eornenn, ^ernenn (to yearn], but ^errnde (yearned), A. S. gevrnian, 
 iroi\\georn (desirous, O. ^eorne, ^erne). 
 
 It is not unlikely that some of these words, though length 
 ened in the Ormulum, may have retained their primitive short 
 quantity in the prevailing English, until by the weakening of 
 the r at a much later time they became long. It is certain 
 that in most cases where an original short vowel has been 
 lengthened before r, the change is not so old as the Ormulum. 
 Thus in almost every instance where we have the sound of ah 
 or an before a weak r, the word, if found in the Ormulum, 
 shows a short vowel. Here belong 
 
 O. arrt (art), arm (they are], A. S. eart, O. N. eru (sunt). 
 
 arrctoss (north, cf. arctic], Gr. and Lat. arctos (bear, north). 
 
 arrke (ark], A. S. earc, Lat. area. 
 
 arrmess (arms, brachia), A. S. earm. 
 
 barrli^ (barley), A. S. bcre (Sc. bear). 
 
 berrme (barm, leaven), A. S. beorma. 
 
 bcrrne (barn], A. S. bere-arn, berern, bern. 
 feorr,ferr (far), A. S.feor. 
 forr (for, prep, and conj.), A. S.for. 
 forrme (former), A. S. forma. 
 
 harrd (hard], A. S. heard. 
 
 heorrte, herrte (heart}, A. S. heorte. 
 
 herrberr^he (lodging, cf. harbor), A. S. hereberga. 
 
 herrcncnn (to hearken], A. S. hyrcnian. 
 
 herrfessttid (harvest-timo), A. S. hilrfest. 
 
 horrs (horse], A. S. hors. 
 
 karrte (cart, chariot), A. S. criit. 
 
 marrch (month of March], Lat. and A. S. Martitts. 
 
 marrtirrdom (martyrdom], A. S. martyrdom, Lat. martyr. 
 
 merrke (mark], A. S. mearc. 
 
276 ENGLISH VOWEL QUANTITY. 
 
 O. norrth (north}, A. S. nordh. 
 
 orr (or), from oththr, otherr, A. S. odhcr. 
 
 patriarrke (patriarch), Lat. patriarcha. 
 
 scorrcncdd (scorched), O. Fr. escorchcr (to excoriate). 
 
 sharrp (sharp}, A. S. scearp. 
 
 shorrt (shorf), A. S. sccort. 
 
 spcrrd (closed, barred, cf. j/tor), unnsperrenn (to unclose), A. S. 
 
 sparrian. 
 
 starr c (firm, cf. stark}, A. S. stearc. 
 steorrnc, sterrne (star}, A. S. stcorra, O. N. stiarna. 
 thorrncss (thorns], A. S. thorn. 
 
 thwcorrt, thwerrt (with ?/// throughout, cf. thwart}, A. S. thiveorh. 
 warrm (warm), A. S. wear in. 
 werre (worse, Sc. waur), A. S. wear (evil). 
 wcrrpenn (to cast, cf. to warp}, A. S. wccrpan. 
 wJiarrfenn (to turn, cf. wharf), A. S. Jiwcorfan. 
 lerrde (rod, rf. yard), A. S. gcard (virga). 
 
 Somewhat less numerous are the cases in which the short 
 vowel of the Ormulum corresponds to any other vowel sound 
 than those of aJi and au (far <\\\<\ for) in modern English : 
 
 O. barrncss (children, Sc. bairns}, also bccrn, A. S. beam. 
 
 to-bresstenn (to burst, Sc. bntsi), A. S. bcrstan. 
 
 currsoin (to curse}, A. S. cursian. 
 
 darr (dare}, durrste (durst}, A. S. dear, dorste. 
 ferrs (t crsc}, A. S.fers, Lat. versus. 
 firrst (first), A. S.fyrst. 
 
 JirrtJircnn (to assist, cf. to further}, A. S.fyrdhcrian.- 
 forrtJterr (further), A. ^.furdJior. 
 
 girrdcll (girdle), A. S. gyrdcl. 
 
 hirrtcnn (to //?/;-/), Dutch and M. H. Germ, hurtcn (to dash against). 
 
 irrc (ire), A. S. yrre. 
 
 kirrkc (church), A. S. ryr/V^, Gr. KVPKIKTJ. 
 
 kirrtell (kirtle), A. S. ^/r//?/. 
 
 mirrtJircnn (to murder), A. S. myrdhrian. 
 
 myrrha, myrra, myrrc (myrrh), Lat. myrrha. 
 
 serrfenji (to serve), O. Fr. scrvir, Lat. servirc. 
 
 skerrenn (to jfrt;r), O. N. j^v rrrt (to drive away). 
 
 thirrst (thirst}, A. S. t hurst. 
 
 thriddc (third), thritti^ (thirty), A. S. thridda, thrittig or ///r?/;^. 
 
 turrncnn (to turn}, A. S. tyrnan. 
 
 turrtle (turtle-dove), A. S. /r//<-, Lat. turtur. 
 
 ivarr (aware), A. S. 7wr. 
 
 weorrc, wcrrc (work}, wirrkcnn (ttfwork}, A. S. -rcvwr, ivy r can. 
 
 wcrrse (worse), werrst (worst), wirrsenn (worsening), A. S. w 
 wyrst, wyrsian. 
 
 ivurrm (worm}, A. S. wyrm, wcorm.. 
 
ENGLISH VOWEL QUANTITY. 
 
 277 
 
 O. wurrth (worth, adj.), wurrthshipe, wurrshipe (worship}, A. S. 
 
 wcordh, weordhscipe. 
 wurrthcnn (to become, be, cf. woe worth the day), A. S. wcordhan. 
 
 But r is not the only consonant which has had this effect of 
 lengthening the vowel before it. We find it produced also by 
 /, a liquid and a lingual like the r. Before / at the end of a 
 word or followed by another conspnant, a vowel originally 
 short has often become long. Cases of this kind, in which 
 theOrmulum still retains the short vowel, are the following: 
 
 .O. all (all], allswa, allse (also], allmasst (almost}, A. S. cal, ealswa, 
 
 ealmcest. 
 
 allderrmann (chief, ruler, cf. alderman), A. S. ealdorman. 
 alltcrr (altar], Lat. alt are. 
 built cdd (breed, bread from bolted flour), O. Fr. bulter, bluter, M. H. 
 
 Germ, biuteln. 
 
 fallenn (\.o fall, fallen), A. ^. fcallan,feallen. 
 f allse (false], A. S.fa/s, Lzt./atszts. 
 galle (gall), A. S. gea II a. 
 Jiallp (holp), hollpcnn (holpcn), from Jiellpenn, A. S. Jtealp, holpen, 
 
 from hclpati. 
 
 halite (halt, lame), A. S. healt. 
 pall (cloth, ci.pall), A. S.pCdl,pell, \j&. pallium. 
 sallt (salt], A. S. sea It. 
 
 shulldre (shoulders], A. S. sculdre, pi. of sculdor. 
 stall (stall), A. S. steal. 
 wallcss (iualls), grunndwall (ground-wall, foundation), A. S. weal. 
 
 This lengthening of a vowel before / had already com 
 menced in the time of the Ormulum ; and indeed, in most of 
 the instances found in the Ormulum, where a vowel origin 
 ally short is followed by the combination Id, the vowel ap 
 pears as long in the Ormulum itself. Thus in 
 
 O. aid (old), but elldre (elder, older) and allderrmann, A. S. cald, yldre. 
 bald (bold), beoldenn, beldenn (to embolden), A. S. beald, bealdian, 
 
 by Id an. 
 
 child (child], but chittdre (children), A. S. did, cildru. 
 f aides s (sheep -folds), A. S. gcfcald. 
 f eld (field), A. S.fcld. 
 
 gold (gold), gildene (golden), A. S. gold, gylden. 
 haldenn (to hold, /widen), A. S. healdan, healden. 
 kald (cold), A. S. ceald. 
 
 milde (mild), but millce (mildness, mercy), A. S. milde, milds or 
 milts. 
 
278 ENGLISH VOWEL QUANTITY. 
 
 O. saldenn (they sold), from sellenn (to sell), A. S. sealdon, from sellan. 
 shildenn (to shield}, A. S. scildan. 
 
 talde (he /0/dQ, from /^/<?w (to /*//), A. S. /ra/de, from A?//0#. 
 weldenn (to govern, cf. to wield}, A. S. wealdan. 
 wilde (wild), A. S. w//</. 
 leldenn (to yield], A. S. gieldan, gyldan. 
 
 Indeed, the Ormulum sometimes lengthens a short vowel 
 before /, where the modern English has it short. Thus in 
 wel, also welle (well, A. S. well and wclla, fons) ; and in wel t 
 also well (well, A. S. wel, bene). The variation of quantity, 
 which the Ormulum shows in the last of these words, is seen,, 
 continued to the present day, in Scottish weelj compared 
 with English well. Further, before Id a short vowel is 
 lengthened in the following : 
 
 O. cwaldenn (they quelled, killed), from cwellenn, A. S. cwcaldon, from 
 
 cwellan. 
 dwalde (he dwelt}, from dwcllenn, A. S. dwcaldc, from dwcllan (to 
 
 hinder, delay). 
 
 elde (old age, cf. eld}, A. $. yldoald, eld. 
 oferrgildedd ( gilded ove r) , A. S. ofcrgyldcd. 
 seldcnn (seldom], A. S. scldan. 
 
 In scaldess (minstrels, scalds, from O. N. skald-r, poet), 
 and hcold, held (he held, A. S. Jicold], the Ormulum pre 
 serves an original long sound, which has become short in 
 English. But the first of these is variously pronounced, as 
 scalds and as scaulds. 
 
 Before the liquid m followed by the mute b, as before the 
 similar combination ld y a short vowel is sometimes length 
 ened in the Ormulum. Thus in five words, three of which 
 have a long vowel in English, while two preserve the earlier 
 short : 
 
 O. camb (comb}, A. S. camb. 
 
 climbcnn (to climb}, A. S. climban. 
 
 wambe (womb, belly), A. S. wamb, womb. 
 
 dumb (dumb}, A. S. dumb. 
 
 lamb (lamb}, but pi. lammbre, A. S. lamb, pi. lambru. 
 
 In like manner, before the liquid n followed by the mute d, 
 a short vowel has become long in very many words. In bi- 
 hinndenn (behind, A. S. beliindaii], and hinndsrrling (de 
 generate, retrograde in character, A. S. hinder ling], con 
 nected with E. hinder, the Ormulum still retains the short 
 
ENGLISH VOWEL QUANTITY. 279 
 
 sound ; as it does before nt mfunnt (font, A. S. font, cf. E. 
 fount), munnt (mount, A. S. mnnf], sannt (saint, A. S. 
 sanct], where the vowel has become long in English. But 
 the instances are far more numerous in which a vowel before 
 nd is already lengthened in the Ormulum. Thus in 
 
 O. bindenn (to bind), bundenn (bound], A. S. bindan, bunden. 
 blind (blind), blendenn (to blind], A. S. blind, blendan. 
 findcnn (\.Q find) , fundenn (found], A. ^>.findan,funden. 
 grindenn (\.Q grind}, A. S. grindan. 
 grund (ground}, but grunndwall (foundation), A. S. grund, grund- 
 
 weal. 
 
 hund (hound], A. S. hund. 
 kindc (nature, kind, kindred], A. S. gccynd. 
 
 minde (mind, memory), but minndi^nesse (memory], A. S.gemynd. 
 simd (sound, integer), A. S. sund. 
 -ivindenn (to wind], m attwindenn (to escape), wundenn (wound}, 
 
 but winndeclut and windeclut (winding -clout, swaddling-cloth) ; 
 
 A. S. windan, wunden. 
 wunde (wound, vulnus), A. S. wund. 
 
 In the preterit singular, the Ormulum has band, fand, 
 wand, like the Anglo-Saxon, but with short a lengthened ; 
 these forms, however, are not represented in English, where 
 the vowel of bound, fonnd, ivoitnd, comes from the plural 
 forms, bundenn t fnndenn, etc., A. S. bundon, fnndon, etc. 
 The change of vowel quantity before nd (as before Id and nib] 
 is carried further in tlje Ormulum than in the modern Eng 
 lish, being extended to a number of words in which it failed 
 to establish itself, so that the original short vowel is heard in 
 their present pronunciation. Thus in 
 
 O. band (band), A. S.*bend. 
 
 ende (end, vulgar eend], A. S. ende. 
 
 hand (hand], oftener hannd, but with added -e always hande, A. S. 
 hand. 
 
 land (land}, A. S. land. 
 
 sand (sand], A. S. sand. 
 
 scndenn (to send}, but sennde (he sent}, A. S. sendan, sende. 
 
 shcndenn (to shend, disgrace), A. S. sccndan. 
 
 strande (strand, bank), A. S. strand. 
 
 snnderr-run (private communing, cf. sunder, asunder}, A. S. sundor. 
 
 ivand (rod, wand}, O. N. vand-r, vond-r, Goth, vandus. 
 
 wendenn (to wend, turn, go), but wennde (he turned, went], A. S. wen- 
 dan, wende. 
 
2 80 ENGLISH VOWEL QUANTITY. 
 
 The vtmdfreondyf rend (friend, A. S. f refold) does not be 
 long to this series ; it came with long vowel quantity from the 
 Anglo-Saxon into the Ormulum, and passed thus into the 
 older English, as we see from the spelling with ic, which it has 
 in common with the opposite, but strictly analogous, fiend 
 (O. fend, A. S.feond). 
 
 It is a very curious fact that a lengthening of the vowel be 
 fore ng, similar to that before nd, is frequent in the Ormulum, 
 although unknown to modern English, in which the vowel be 
 fore ng is always short.* Examples are- 
 
 O. gang (journey, ct. gangway), but ganngcnn (to go), A. S.gang,gan- 
 
 gan. 
 gengc (company, cLgang), A. $. gsnge. 
 
 king (king), A. S. cyning, cyng. 
 
 lang (long), bilcngc (belonging to), A. S. lang, gclenge : but lanngc 
 (long, din), Jcnngre (longer}, A. S. hinge, Icngra, Icng. 
 
 langcnn (\JQ long after), forrlangedd (desirous), A. S. langian. 
 man*, amang and amanng (among], A. S. amang, gemang. ,. 
 
 ringenn (to ring), A. S. hringan. 
 
 singcnn (to sing), sungcnn (they sung), sang (song), A. S. singan, 
 sungon, sang. 
 
 springcnn (to spring), sprang (sprang), sprit ngenn (sprung), off 
 spring (offspring], A. S. springan, sprang, spntngen, of spring. 
 
 stingcnn (to sling), stnngenn (stung), A. S. stingah, stungcn. 
 
 strang (strong), strengenn (to strengthen), but strcnncthe (strength), 
 A. S. strang, gestranglan, strcngdJnt. 
 
 swingcnn (to scourge, cf. to suing, swinge}, A. S. sivlngan. 
 
 thing (thing), A. S. tiling. 
 
 bithrungcnn (oppressed, cf. throng, i. e. press of people), A. S. 
 bithrungen, from t hringan, gethrang ; but O. threnngdenn 
 (they thronged), A. S. throigdon, from t/irengan. 
 
 thwang (thong), A. S. tJiwang. 
 
 tunge (tongue), A. S. tunge. 
 
 wengess (wings], O. N. vcng-r. 
 
 wrang (wrong), A. S. wrang. 
 
 ^ung (young), rarely ^unng, comp. yinngre (younger), A. S. geong, 
 gyngra. 
 
 * The Norse grammarians recognize in that language a similar lengthening of 
 primitive short vowels (a, o, ?/, z ), when followed by ng (or uk ; also of a, o, u, 
 when followed by //", tg, Ik, Im, Ip, Is) : thus Idng-r (long], springa (to spring], 
 tunga (tongue], vccng-r or veng-r (wing]. It seems, however, to be question 
 able, whether, or how far, this change belonged to the old language. See Heyne, 
 Kurze Gramui. dcr altgerm. Diah ctc, 2cl ed., p. 82. 
 
ENGLISH VOWEL QUANTITY. 281 
 
 In Jicng (hung, A. S. heng ; but henngde \Jiangcd], A. S. 
 hangode], tile vowel was already long in Anglo-Saxon. 
 Whether this extension to ng of the euphonic analogy which 
 obtains for nd ever gained much currency in the language, 
 may well be doubted. It seems certain that it cannot have 
 prevailed at the time (probably in the fifteenth century) when 
 the old long sounds of i and // (as \\\ pique, prude] began to 
 pass into the diphthongal sounds heard in pike and proud 
 for in that case, instead of saying king, tiling, tongue, sung, 
 as we now do, we should probably be saying k uig, tJnng, 
 toung, soung, with the same vowel utterance as in k ind, sound. 
 It may be observed, however, that in the most recent English 
 there is a noticeable tendency to lengthen somewhat the short 
 sound of o before ng, so that long, song are apt to be pro 
 nounced with much the same vowel sound as or, nor, for. 
 
 If now we have found in combinations such as Id, nib, nd, 
 ng, where the first letter is a liquid and the second its cognate 
 sonant mute, a certain tendency to protract the quantity of a 
 preceding short vowel, it must be remarked that the ordinary 
 tendency of a combination of consonants is in the opposite 
 direction not to the lengthening of a preceding short, but 
 to the shortening of a preceding long. The speaker slights 
 the vowel in order to concentrate his energy of utterance on 
 the following consonants, which thus massed together pre 
 sent some difficulty of enunciation. It is as in the Greek 
 pos (poorer), for Tre^crrepo?, i.e. Trevrjr-Tepos., from Trez^?, 
 s. Of this change a long vowel shortened on account 
 of two or more consonants following it numerous examples 
 are found in the language of the Ormulum, when compared 
 with the Anglo-Saxon. Thus 
 
 O. asskcnn (to ask], A. S. ascian. 
 blosstme (blossoni), A. S. b lost ma. 
 chappmcnn (chapmen, merchants), A. S. ccupmcn ; cf. O. chcpinng- 
 
 bothc (market-booth). 
 dcnnlike (cleanly], dennscnn (to cleanse), from dene (clean], A. S. 
 
 dcenl ice, dcensian, from clccne. 
 dredde (he dreaded], forrdrcdd (alarmed), from drccdenn, dredenn 
 
 (to dread], A. S. dred, drcedcn, from dree dan ; in O., as in E., 
 
 the verb has passed into the weak conj. 
 errnde (errand], A. S. cerende* 
 
2 82 ENGLISH VOWEL QUANTITY. 
 
 O.fedde (he/ 7 ), homfcdenn (to feed), A. S.feddc, iromfedan. 
 fifti* (fifty} from^ (five), A. S.fiftig, from f if. 
 fosstrenn (to foster], fossterrfaderr (foster-father}, A. S. fostcrian, 
 
 fosterfader. 
 godd spell (gospel) belongs here, if the A. S. word is godspel (good 
 
 tidings, =f vayyeXiov) ; but this is now generally believed to be 
 
 godspcl (God s word). 
 hall\Jicnn (saints, cf. Hallow-ceii), hall^hcnn (Shallow), A. S.h&lige 
 
 orhiilge, halgian. 
 hiddcnn (they ///>/), ///dtf (hid), homhidenn (to 7//V&)> A. S. hyddon, 
 
 hyded, from hydan. 
 
 kepptc (he XY//), from kepenn (to /vr^), A. S. <:<>//* , from ccpan. 
 lasstenn (to /j/), also lastenn, A. S. hestan. 
 leddc (he &Y/), /tY/^ (Av/), from Icdcnn (to lead), A. S. laddc, faded, 
 
 from Ice dan. 
 
 mosste (might, cf. jurist), A. S. moste. 
 ncsst (nearest, next), also nest (Sc. niest), A. S. nehst. 
 redd (read, part.), A. S. m/^, from m/^. 
 shaddc (he parted), shadd (parted), from shcedenn (to part, cf. to 
 
 shed), A. S. javW, sccadcn, from sceadan : of weak conj. in O. 
 
 and E. 
 siththcnn (sithcnce, since), A. S. j^/// tham, s uihdhan, perhaps 
 
 sidhdhan. 
 slepptc (he slept), from shepenn (to j/i-^), A. S. j/<^, from slcepan : 
 
 of weak conj. in O. and E. 
 sofftc (soft), A. S. stftc. 
 
 spredd (spread, part.), A. S. sprccded, from sprccdan. 
 thratte (he rebuked, cf. threatened), A. S. tfireatede, from threatian. 
 wcpptenn (they wept), from wcpctin (to weep), A. S. weapon, from 
 
 wcpan : of weak conj. in O. and E. 
 
 wimmann (woman), also wifmann, A. S. wifman, wimman, per 
 haps wimman. 
 wissdotn (wisdom), also wisdom, A. S. wisdom.- 
 
 Probably the short vowel in tvratht/ic (wrat/i), as well as in 
 laththe (loathing, enmity), and kit /it he seen in kiththeli$ (fa 
 miliarly, cf. /////), may be accounted for in this way, from the 
 addition of a suffix //^ (A. S. dh, dJni) the same as in 
 strenncthe (strength), trowwthe (trutJi), from strang (strong), 
 trowwe (true) to the adjectives wrath (wroth, A. S. wrddti), 
 lath (loathsome, hateful, A. S. ladli), and cutJi (known, contli, 
 A. S. cudJi) : compare A. S. Icedhdhu (offence), cydhdhu 
 (home, household). It is possible that in these words, as well 
 as in dredde,fedde, hiddcnn, leddc, shaddc, siththcnn, wimniann, 
 the first vowel may have been pronounced long, the follow- 
 
ENGLISH VOWEL QUANTITY. 283 
 
 ing consonant being written double only because it was sound 
 ed twice, first in the stem, and again in the suffix.* It is also 
 possible, or even probable, that in some^ases the change from 
 long to short, now under consideration, may have taken place 
 already in the Anglo-Saxon ;t but this makes no difference, 
 either in the reality of the change or in the cause from which 
 it arose. 
 
 In the list of words just given, the short sound which ap 
 pears in the Ormulum is maintained in modern English. The 
 case is otherwise with those which follow : 
 
 O. allmasst (almost], though the simple word is nearly always mast 
 
 (most], A. S. mcEst. 
 demmd (judged), from demenn (to judge, cf. doom, deem), A. S. 
 
 denied, from deman. 
 
 dcrre (dearer), from deore, dere (dear), A. S. deorra, from deore. 
 derrlinng (darling), A. S. dear ling, 
 dunnwarrd (downward}, from dun (dowti), A. S. ddunwcard, from 
 
 dun. 
 hehhre (higher}, also hehre,iwya\ heh (IiigJi), A. S. hcaJira, from 
 
 hedh. 
 
 * Such a supposition must, however, be regarded as improbable for these words, 
 on aceourit of the short quantity which they have in English ; and especially im 
 probable for the preterits in -dde, on account of the corresponding participles 
 dredd, hidd, ledd, etc. , in which a really double pronunciation of the d is hardly 
 to be thought of. 
 
 What is here recognized as possible that a vowel before a double consonant 
 may have been long, the consonant being written twice because actually twice 
 sounded must be admitted also for httte (he touted), as well as the comparatives 
 derre (dearer} and nerre (nearer], mentioned in the next paragraph ; and, perhaps 
 with still stronger reason, for the words clcennesse (clean-ness), Jiff aid (fivefold}, 
 l&ffull (belief -full, believing). As we have skilllces (skill-less, ignorant), sefenn- 
 nahht (seven-night, week), unnned (iin-need, without constraint), sunderrrun 
 (sunder-roun, private communing), forrraht (perverted, Germ, ver-ruckt), it seems 
 not unlikely that the r, n, and f would have been written thrice in derre, nerre, 
 clcennesse, fiffald, laffull, if their first vowel had been short in sound; but the 
 spelling of fullike (f till-like, fully), stilli\ (stilly}, idelle^c=idellnesse (idleness), 
 drunnkennesse (drunken-ness], unnitt (useless, Germ, un-nutz), forrsivunden- 
 nesse (remissness), orrath as well as orrrath (inops consilii, O. N. or-radJi), warns 
 us not to lay too much stress on this consideration. 
 
 \ The same possibility is not to be overlooked in other cases where the vowel 
 quantity of the Ormulum differs from what must have been the primitive quantity 
 in Anglo-Saxon. This is particularly true as to that lengthening of vowels in 
 open syllables which is soon to be considered : the change could hardly have gone 
 so far in the language of the Ormulum, if it had not made a beginning in Anglo- 
 Saxon times. 
 
284 ENGLISH VOWEL QUANTITY. 
 
 O. hcrrde (he heard} , herrd (heard] , from hcrcnn (to hear], A. S. hyrde, 
 
 hyrcd, from Jiyran. 
 
 laffdi^ (lady}, A. S. hlcefdige ; cf. O. /^/77V/ (/^r,/), A. S. 7^?/W. 
 licence s (likeness), from //V (///r), A. S. gd icncs, from ^r/fc, but O. 
 
 onnlicnesse (likeness, image). 
 luttc (he bowed, lout at), from littcnn (to /<?///), A. S. /<r/, from /w/Vz / 
 
 of weak conj. in O. and E. 
 nerre (nearer), from ncr (nearly), O. N. n<zrri,from nar ; cf. O. ncr 
 
 (nearer ], from wr// (nig/i), A. S. ;/r^>, from r<?//. 
 oththr, orr, from othcrr (all meaning 0r), A. S. ddher dwdher 
 
 (either) : otJitiir, perhaps, by a confusion of A. S. odhdhe (or) 
 
 with ad her. 
 
 TJntrrsda"\i (Thursday], O. N. Thors-dag-r, A. S. Thunres-dag. 
 lucnnde (lie lueetied}, from wcncun (to 7tvr;/, _ think), A. S. we tide, from 
 
 we nan. 
 ivcsstc (waste, desert, adj. and subst.), A. S. ivtste. 
 
 From a continued working out of the same tendency, the 
 English has a short vowel before two or more consonants in 
 some words where the Ormulum shows the original long 
 vowel quantity : 
 
 O. adk (disease, cf. addle], A. S. Adi. 
 breost, brest (breast], A. S. bre<>st. 
 bretJire (brcthrcti), A. S. br<>dhru. 
 buhsumm (pliant, compliant, cf. biixoui), from A. S. bu.gan (to bend, 
 
 bow] . 
 
 to- chef (he cleft, also clave, cloi e], A. S. clet if, from eld-fan (to cleave), 
 dost (dost, usual pron. dast), also dosst, from don (to ^), A. S. dest, 
 
 from eli ni. 
 
 fifle (JiftJi),jiftendc (fifteenth], A. S.flfta, f if t cod ha, O. N.fimtdndi. 
 freond,frend (friend], A. S.frei md. 
 gom (care, heed, cf. gumption), A. S. geam. 
 Jicesc (command, /test], A. S. //<2\y. 
 //^/^, //^/ (he held), from haldenn (to //^/), A. S. //r<;/^, from /^m/- 
 
 dan. 
 
 monethth (month), A. S. monad Ji. 
 naness, in /<?rr ///^ naness (for the nonce], from ccness (once, Sc. 
 
 aines), A. S. #w. 
 
 In the following words also the shortening may be ex 
 plained on the same principle, since the vowel which is 
 written before their final liquid is little, if at all, represented 
 in their actual pronunciation : 
 O. of re (ever), ncefre (never), A. S. erf re, ncefrc. 
 becnenn (to beckon), A. S. becnan, beacnian. 
 
ENGLISH VOWEL QUANTITY. 285 
 
 O. bosemm (bosom, often pron. with long oo sound), A. S. bosni. 
 
 brotherr (brother], pi. brcthre (brethren], A. S. brodhor, brodhru. 
 modcrr (mother], A. S. nwdor. 
 otherr (other), A. S. vdJier. 
 wcspenn (weapon), A. S. ivcepcn. 
 
 The cases which we have been considering show an ac 
 cented long vowel shortened in a close syllable, where it is 
 separated by more than one audible consonant sound from 
 the vowel of the following syllable. We have next to notice 
 a change which is the converse of this, a change which has 
 cut much deeper into the integrity of the old system of quan 
 tities : the lengthening of an accented short vowel in an open 
 syllable (generally a penult) that is, when separated by only 
 one consonant sound from the vowel of the syllable which 
 follows. This change has been carried to a very great extent 
 in the modern German : gcbcn (to give), nicder (down, cf. E. 
 nether), trageii (to draiv), ncluncn (to take), are examples 
 taken up at random out of an immense multitude. Accord 
 ing to Schleicher, this change of quantity belongs to, and is a 
 prominent feature in, the transition from Middle to Modern 
 High German, which was made in the fifteenth century. In 
 England the change must have commenced its progress 
 earlier, as we find it carried very far in the language of the 
 Ormulum, which belongs to the thirteenth century. We give 
 first the instances in which the lengthened vowel seen in the 
 Ormulum became so established in English usage as to re 
 main long in the pronunciation of to-day. Thus 
 
 O. -ah (ale), in bridale (r/</ - feast), A. S. calu,brydealu. In E. bridal 
 
 it has become short again. 
 a\he (awe), but also cy^c (fear), A. S. ege ; cf. O. N. t cgja (to strike 
 
 with fear or awe], 
 bakcnn (to bake), A. S. bacan. 
 
 bede (prayer, cf. bead, bead-roll, foads-ma.n), A. S. gebed. 
 bcrenn (to bear], borenu (born), A. S. bcran, boren. 
 bidell (crier, messenger, cf. beadle), A. S. by del. 
 brascne (brazen), from brass (brass), A. S. bra sen, from briis. 
 brckcmi (to break}, A. S. brccan. 
 
 bridledd (bridled), with 7 from bride/I (? not found in O.), A. S. bridel. 
 bule (bole, tree-stem), in bulaxc (axe, hatchet, cf. pole-axe], O. N. 
 
 boloxi, from bol-r. 
 care (care], A. S. ccaru, earn. 
 
286 ENGLISH VOWEL QUANTITY. 
 
 O. charii (mournful, anxious, cf. chary), A. S. ccarig, from ccaru. 
 
 choscnn (chosen), from chescnn (to choose), A. S. cor en, from ceosan. 
 clofenn (cloven), A. S. clofen, from clei fan (to cleave), 
 cnapess (boy s, cf. knave), A. S. cnape, cnafa (boy). 
 cnedenn (to knead}, A. S. cncdan. 
 dcekenn (Levite, deacon], A. S. diacon, Lat. diaconus. 
 dale (dale), A. S. rt^// original quantity preserved in E. dell. 
 dra^Jienn (to draw, drawn}, A. S. dragan, dragen. 
 kirrkedure (church-rt^fr), A. S. duru (door). 
 
 cfenn (equal, even}, but pi. effnc, vb. cjfnenn, A. S. /?;;, efenian. 
 ele (oil}, A. S. ^/<?, O. Fr. <?//6 , 0///V, Lat. oleum, 
 etenn (to ^tf/ 1 ), but impv. ett (eat), A. S. ctan, ct. 
 fader r (father), A. ^. fader. 
 
 farenn (to %p,fare}, but impv./^rr, A. S.faran, far. 
 biforcnn (before}, also biforr, A. S. bcforan. 
 bifrorcnn (frozen}, A. S.froren, fromfreosan (to freeze), 
 gate (way, ci. gait}, O. N.gata (way), A. $>. gcat (gate}, 
 grccfess (ditches, cf. grave), A. S. grfif. 
 
 -gume (man), in bridgume (bridegroom}, A. S. guma, brydgnma. 
 hatcnn (to ///t ), /^ /f (hate}, A. S. hatian, Jiete. 
 hefenn (to raise, heave}, hofenn (hove, hoven}, A. S. hebban, hofcn. 
 hire (Jier}, A. S. ////r. 
 hiihcnn (to hasten, cf. to ///>), higian. 
 hope (hope}, A. S. //<?/. 
 if ell (evil), A. S.j^-/. 
 /Cw//. // (f>Cv), O. N. yfo&z. 
 
 ladcnn (to draw out, cf. to /^^- water, also ladle), A. S. hladan. 
 late (late), but ///n- (latter}, lattst (last), A. S. /<?/, /^/;7?, latest. 
 la^he (law), A. S. lagu. 
 forrlorenn (lost, cf. forlorn), iromforrlesenn (to lose}, A. $> . forloren , 
 
 from f or led s an . 
 
 makcnn (to make, Sc. ;;/<?), but impv. wf<r, A. S. macian, maca. 
 mele (meal, flour), A. S. melu. 
 mete (meat, food), A. S. mete. 
 efennmete (commensurate), from mctt (measure, cf. mete}, A. S. 
 
 gcmct : O. met dike (meetly}, A. S. gcmctlice. 
 naked d (naked}, A. S. nacod. 
 
 name (name), but ncmmnenn (to name}, A. S. name, ncmnan. 
 bincthenn (beneath}, but niththrenn (to lower, cf. nether), A. S. &- 
 
 ncodhan, nidherian. 
 ni^hcnn (nine}, nitficnntte (ninth}, A. S. ?^, nigodha, O. N. 
 
 niundi (ninth}. 
 
 of err (over), but also ^r, A. S. ofer. 
 
 openn (open}, but oppnemi (to ^-), A. S. open, openian. 
 retftellboc (rule-book}, A. S. regol, Lat. 
 sake (quarrel, cf. sake}, A. S. sacu. 
 
ENGLISH VOWEL QUANTITY. 287 
 
 O. forrsakenn (\.Q forsake, forsaken), A. S.forsacan,forsacen. 
 same (same), A. S. same (pariter), O. N. sam-r (idem). 
 se^henn, scne (seen}, A. S. sew en, scgen, sen, from seon (to see), 
 shame (shame], but shammfasst (shamefaced], A. S. sceamu, sceam- 
 
 fast. 
 
 shapenn (to form, create, cf. to shape], A. S. sceppan. 
 skathenn (to harm, to scathe), A. S. sceadhan. 
 sla^cnn (slain), from slan (to slay], A. S. slagen, from slcdn. 
 smeredd (anointed, cf. smeared}, A. S. smyrcd, from smyrian, cf. 
 
 smeoru (ointment). 
 
 spekenn (to speak}, A. S. spree an and specan. 
 stelenn (to steal}, A. S. stclan. 
 stircnn (to j/Vr, move, Sc. steer}, A. S. styrian. 
 swercnn (to swear], A. S. swerian. 
 anndsivere (answer}, A. S. andswaru. 
 takenn (to take, taken, Sc. fa), but impv. torr, A. S. tacan, tacen, 
 
 tac. 
 
 tale (reckoning, number, cf. tale}, A. S. talu. 
 tholenn (to suffer, Sc. thole], A. S. tholian. 
 wakenn (to wake, watch), A. S. wacan. 
 
 watcrr (water}, but wattrenn (to water}, A. S. water, wa.terian. 
 weorclld, werelld (world], but gen. wcorrldcss, wcrrldess, A. S. 
 
 weoruld, woruld, world. 
 
 wrekenn (to wreak), wrceche (wreak, revenge), A. S. wrecan, wracu* 
 wuke (week), A. S. wttcu. 
 late (gate], zlsogate, A. S. gcat. 
 
 It is remarkable that this change is carried to a much great 
 er extent in the Semi-Saxon of the Ormulum than it is in 
 modern English. It should seem that there must have been 
 a reaction early established, which set limits to the tendency, 
 and maintained the short vowel in many words where it had 
 begun to be lengthened. Instances of this kind where an 
 accented short vowel in an open syllable is lengthened in the 
 Ormulum, but the same vow r el is found short in English 
 (mostly, indeed, in monosyllables with final consonant sound) 
 are the following : 
 
 .O. abitfenn (above) ,=bufenn, from A. S. d, be, and ufan. 
 
 beodenn, bedenn (bidden), A. S. bodcn (commanded), bedcn (entreat 
 ed. [The A. S. verbs biddan (to entreat) and beodan (to com 
 mand) are pretty much confounded in O.] 
 
 bis scope ss, pi. of bisscopp (bishop), A. S. bisccop ; bisscopcss had a 
 secondary accent on the o. 
 
 bite (bit, morsel), A. S. bite. 
 
288 ENGLISH VOWEL QUANTITY. 
 
 O. bodii (body], A. S. bo dig. 
 
 bide (bull], Dutch bid, O. N. boll. 
 
 clcpcdd (called, cf. dept, yclept}, from ckpcnn, K. S. cleopod,irom 
 
 cleopian. 
 
 cudc (cud] , A. S. rwrtT. 
 <:;;z<v/ (to <w/^, also as part.), but impv. r^ww, cnmm, A. S /- 
 
 ;;/;/, cumen, cum. 
 
 cwike, pi. of rw/Yr (living, quick], A. S. ravV. 
 </*Vfc (rt /V/), from ^ (to do], A. S. ^(M > from don. 
 drake (dragon}, A. S. draca, Lat. </;vrf0. 
 
 drifcnn (driven), from drifcnn (to ;/r/7v), A. S. ^r//?;/, from ^H/iz. 
 frctcnn (to fret, trans.), A. $. frctan (to cat up). 
 ^/V, pi. oigladd (glad}, glade nn (to gladden} , A. . gliid, gladian. 
 godess, gen. tf godd (god], but pi. goddess, A. S. t <W. 
 grcsess (grasses], sing. \Kgrcsshoppc (grasshopper}, A. S. gras,grs, 
 
 giirsJwppa. 
 hafenn, but habbenn (to have), Jiafcsst, but 7/^jr/ (//j/), //rt^fr 
 
 (/.v^/), A. S. Jiabban, Ju Jst, li<"<Jde. 
 hcfil (heavy}, A. S. /^//>. 
 heofenu (heaven}, in comp., but as separate word heoJJ ne, hcjfne, 
 
 A. S. heofon. 
 
 hiderr (hither), A. S. hider. 
 hise, pi. oi hiss (his), A. S. 7//J. 
 ////;/ /^ (honey), A. S. hunig. 
 kidc (/vV/),O. N. /vV/. 
 klness, kuie (comm. kin ness, kinnc}, gen. and pi. of k inn (kin, kind}, 
 
 A. S. ry//. 
 
 lifcthth (Uveth), from libbenn (to //TV), A. S. lifcidh, libban. 
 limess (limbs), sing, not in O., A. S. //;;/, pi. Icomu, limn, 
 lit ell (little), pi. 7//M-, A. S. /j^-7. 
 lokcnn (shut in, cf. locked}, A. S. loccn, from Idcan. 
 lotess, pi. of loft (Jot), A. S. 7/7<p/, pi. /-/^///. 
 7;//<? (love], htfenn (to love}, but htffsumm (pleasant), A. S. 7?//>/, htjian, 
 
 lufsum. 
 
 mani^, mani (many}, A. S. manig. 
 
 mi k ell (great, many, micklc, much}, but pi. miccle, A. S. micel. 
 minctcrcss (money-changers, cf. minter, mint), A. S. myneterc, from 
 
 my net (money}, Lat. moneta. 
 
 muneclif (monk-life), A. S. nntnec (monk), Lat. monachus.. 
 naru (narrow), but pi. narrice, A. S. near n. 
 nile (niH\&, i. c. ivill not), but ;///// ( /// //<?/), A. S. r/7r, AT//. 
 <?/";/<? (crvcn}, dat. of of en n (? not in O.), A. S. 0/t tt. 
 peninng (penny}, A. ^>. pending, pening, pcnig. 
 rathe (quickly, cf, rathe, rather), A. S. hradhc, JiradJior, from hrfidh 
 
 (quick). 
 risenn (risen}, from risen n (to r/j^), A. S. risen, from r tsan. 
 
ENGLISH VOWEL QUANTITY. 289 
 
 O. rotcnrt (to rot}, A. S. rotian. 
 
 Saterrda^ (Saturday), A. S. Saterndiig, Lat. Saturni dies. 
 
 seofenn, sefenn (seven}, but also seoffne, seffne ; seofennti^ (seventy} : 
 
 A. S. scofon, hundseofontig. 
 shetenn (to shut up), A. S. scyttan. 
 sikcrr (sure, Sc. sicker), O. Sax. sikor, O. H. Germ, sihhur, Lat. 
 
 securus. 
 
 slue, rare for sinne (sin), A. S. jrj/#. 
 
 skathelces (unharmed, scat hie ss}, A. S. sceadha (harmer). 
 stafess, pi. of staff (letter, cf. jfo^; old pi. staves), A. S. stiif (staff , 
 
 letter). 
 
 j /<?</<? (place, cf. stead), A. S. stede, 
 
 stekenn (to confine, cf. to J//V/6, remain fast), A. S. stician. 
 stoke (stock), dat. of stocc (? not in O.), A. S. stoc. 
 sume, pi. ofsumm (some), A. S. sum. 
 sumerr (summer), A. S. sumor. 
 sune (son), A. S. suuu. 
 Sunenndafa (Sunday), from sunne, rarely sune (STIH), A. S. sunne, 
 
 Sunnandag. 
 
 thidcrr (thither), A. S. thider. 
 fullthrifcnn (complete, cf. thriven}, O. N. thrifinn, from thrifask (to 
 
 thrive}. 
 thripell (triple), from Lat. triplex, Fr. triple, confused with A. S. tliri ; 
 
 cf. prov. Eng. thribble. 
 
 trcdenn (to tread, trodden), A. S. trcdan, trcden. 
 whiderrivarrd (whitherward), A. S. hwider. 
 ividewe, comm. widdwe (widow), A. S. widwe, ivydeive. 
 ivilenn (to ////), but willt (wilt), A. S. willan, wilt. 
 witenn (to know, cf. to wit, O. E. to weet), but impv. -ze////, A. S. 
 
 wit an, wit. 
 writenn (written), from writenn (to write), A. S. writen, from wr itan : 
 
 cf. O. write ss, pi. of writt (writ), A. S. wr//. 
 wude (wood), A. S. wudu. 
 w une dd (wont), from wunenn (to accustom), A. S. wuna, gewuna 
 
 (custom). 
 
 letcnn (to get), ytt (gets), bi^etcnn (gotten}, A. S. bcgitan, begiten. 
 lifcnn (to give, given), also written with^-, but impv. 37^", A. S. gifan, 
 
 gifen, gif. 
 
 That this change the lengthening of an accented short 
 vowel in an open syllable was still in progress at the time 
 of the Ormulum, so that the usage in respect to it was then 
 unsettled and fluctuating, is apparent from indications in the 
 book itself. A number of the words given in the last two 
 lists have here and there a mark of short quantity written 
 over the vowel, as if the writer, having first given it as long 
 19 
 
290 ENGLISH VOWEL QUANTITY. 
 
 with only a single consonant after it, was afterwards inclined 
 to recall his judgment, to set it down as short, and therefore 
 drew a curve line over it, this being an easier way than dou 
 bling the consonant by interlineation. What makes this ex 
 planation more probable is the fact that, while there are more 
 than forty distinct words which in one place or another have 
 this short mark over them, it occurs in almost every case over 
 an accented vowel in an open syllable. That there was a 
 special vacillation on the part of the writer as to the quantity 
 of such vowels seems a natural, if it is not a necessary, in 
 ference from this fact. Thus bcrcnn (to bear] is once at least 
 written with a mark of short quantity over the accented 
 vowel; and the same is true of bcde (bede), dale (dale], hatenn 
 (to hate], Jictc (hate], ladenn (to lade), late (late), niele(meal), 
 mete (meat), name (name), stelen (to steal], takcnn (to take), 
 tale (tale) ; also bite (bit), cndc (end), kine (kin), lifethth 
 (liveth), sine (sin), stede (stead), thrifenn (thriven), wilenn 
 (to will), witenn (to wit), witcnn (written). 
 
 Under the broad euphonic analogies and tendencies which 
 have now been described come all but a comparatively small 
 number of the cases in which the modern English quantity 
 differs from that in the Ormulum. There remain, however, 
 some few changes which are not altogether of an isolated 
 character. A long vowel of the Anglo-Saxon and the Ormu 
 lum has in a good many instances been shortened before a 
 final mute. This is especially the case with the old long o 
 before a finals-sound. The long quantity of that vowel was 
 indicated in early English by doubling the o : thus boc was 
 written book. The sound afterwards changed to that which we 
 hear in spoon, spool ; and still later was shortened to its pres 
 ent pronunciation. Instances of this kind are 
 O. boc* (book], A. S. boc. 
 
 croc (hook or crook, device), O. N. krok-r. 
 
 lokenn (to look), A. S. tician. 
 
 forrsoc (forsook), from forrsakenn (to forsake), A. ^.forsoc, from 
 for sac an. 
 
 toe (took), from takcnn (to take), A. S. toe, from tacan. 
 * In this word, and in several others, where a long vowel has become short be 
 fore a final consonant, the Scotch retains the earlier long quantity: thus, bulk, 
 bruik, bhiid, gude, etc. 
 
ENGLISH VOWEL QUANTITY. 391 
 
 Occasionally other long vowels have become short before a 
 /--sound, as in 
 
 O. brukenn (to use, enjoy, cf. to brook], A. S. brucan. 
 fie (fig), vhfictre {fig-tree), A. ^>.fic, \^?&.ficus. 
 seoc, sec (sick}, A. S. seoc. 
 
 strac (passed, cf. struck, O. E. strook], A. S. strac, from s trie an. 
 wic (dwelling, street, cf. SwanwzV/t, GreenwzV//), A. S. wdc. 
 
 The same change of quantity has taken place not unfre- 
 quently before d, seldomer before / / thus 
 
 O. blod (blood}, A. S. Mod. 
 fiod (flood], A, S.jtid. 
 god (good], A. S. god. 
 
 stod (stood], from stanndenn (to stand], A. S. stod, from standan. 
 wod (mad, Sc. wud], A. S. wod. 
 breed (bread], A. S. bread, 
 deed (dead], A. S. dead. 
 
 drcedenn, dredcnn (to dread], A. S. drtzdan. 
 hcefedd (head}, A. S. hedfod. 
 shcedenn (to part, cf. to shed], A. S. sccadan. 
 shrcedenn (to shred, pare, cf. Sc. screed}, A. S. screadian. 
 but, comm. butt (but, except), A. S. bdtan. 
 fot (foot], A. S./o/. 
 hat (hot], A. S. hat. 
 Icetenn and Ictenn (to /r/, allow, also as part, let], pf. let (he /^/), A. S. 
 
 ICE tan, Ice ten, pf. /^/. 
 swat (sweat], A. S. swat. 
 
 wcete (drink), from wat (not in O., wet, Sc. weet), A. S. w^A 
 ze//, also watt (wot], from witenn (to know, cf. to w//), A. S. wat, 
 from wit an. 
 
 In bcdcthth or biddcthth (biddctJi], forrbcdethtJi (forbiddetJi], 
 the long form comes from A. S. beodan (to command), the 
 short one from A. S. biddan (to entreat) ; in biddcnn (to com 
 mand, to entreat) of the Ormulum, the forms of the two verbs 
 are very much confounded. 
 
 The few cases in which the difference of quantity between 
 the Ormulum and the modern English is not to be explained 
 from principles already set forth, will be found, so far as I 
 have noted them, in the following list : 
 
 O. amcen (amen], Gr. a/.*?/*;. 
 
 an (an, one], rarely aim; nan (none); onnan, anan, also anann 
 
292 
 
 ENGLISH VOWEL QUANTITY. 
 
 O. (anon) ; A. S. an, n m. Eng. alone, atone, only, and Sc. ;/<?, 
 
 nane, preserve the original long quantity. 
 ani\ (any ], A. S. a; nig. 
 bcon, ben (to be, been], A. S. be<~>n (to /v) : the long sound of been is 
 
 still sometimes heard. 
 
 caritcth (charity], Fr. charite, Prov. caritat, Lat. car Has. 
 chcle, also <r/^7f ( t 7//7/, subs.), A. S. tv/,-. 
 (f/^/// (clothing, cf. f/ ////, but pi. clothes), A. S. ctf.dh. 
 cuthc (could}, A. S. cu.dhe. 
 
 dcef(i:af, cf. Sc. ^-vzzv, to deafen), A. S. d:<if. 
 dceth (death], A. S. detdh. 
 
 deofcll, defell (devil, Sc. dcevil), A. S. ^y^/, Lat. diabolus. 
 dotJi (doth, usual pron. datli], from don (to ;/<?), A. S. dcdh, from 
 
 ( flesh), A. ^.fi<T.sc. 
 glutcrrnessc (gluttony), O. Fr. gloutonnic (from glouton, Lat. gliitn(ii), 
 
 from glutire) : in glutcrrnessc English affixes are attached to the 
 
 root (glut) of the Latin and French words. 
 gy;j (art, device, ci.gin), shortened from O. Fr. engin, but perhaps 
 
 confused with a derivative of O. X. ginna (to deceive). 
 inoh (enough, also enow), A. *$>. genng, genoh, 
 bikcecJicdd (catcJied), also bikahht (caugJit) ; of uncertain origin. 
 prof etc, prophet c (prophet), Lat. propli^ta. 
 publicaness (publicans), Lat. public<~<ni. 
 rcedi\ (ready), also rccdclike, A. S. nr^/^, radl ic. 
 ricJie (rieh), A. S. r ice, 
 
 sari\ (sorry), but j t -z;v (sorely), A. S. snrig, from j/7/- (J^;Y). 
 j^////5 (happily), A. S. scclig, gcscclig (happy, Sc. secly), whence E. 
 
 silly. 
 
 shcphirdc (shepherd], from shep (sheep), A. S. sceap, sccaphirde. 
 shunenn (to shun), A. S. scanian. 
 tene (ten), rarely tcnn, tende (tenth), A. S. tyn, ten, tcndha, O. X. 
 
 t Hindi (tenth}. The old long quantity is preserved in the com 
 
 pounds thirteen, thirteenth, etc., and in Sc. teinds (tithes). 
 onnycness (against, again), outran (again, against), A. S. ongean 
 
 (against). 
 
 let (yet), strangely lengthened in O., A. S. git, get, giet, gyt. 
 drunncnenn (to drown , trans.), in form = A. S. druncnian (to get one 
 
 drunk), cf. O. X. drukna (to be drowned). 
 enngeU (angel), A. S. cngd, O.ngd, Lat. angelus. 
 flumni (river, ct.Jliiinc), O. Yr.Jtum, L^.Jlumcn. 
 funnt (font, cf. fount), A. S.font, Lat. fons, font-is. 
 irrcnc (of iron, fcrrcus), from ircnn (iron), A. S. "trcn (fcrrum and 
 
 ferreus) : rr in irrene perhaps an oversight. 
 munnt (mount), A. S. munt, Lat. mons, vwnt-is. 
 sannt (saint), A. S. sanct, Lat. sanctus. 
 
ENGLISH VOWEL QUANTITY. 
 
 293 
 
 O. sleckenn* slckkenn (to slake), A. S. slcac (slack), gesleccan (to slack 
 
 en), O. N. slokkva (to .?/&-) 
 
 thurrh (through), A. S. thnrh, tlniruh (through, thorough}. 
 luaccnenn (to waken, trans, and intrans.), A. S. wilcnan : lengthened 
 in E. under influence of A? wake, O. wakenn, A. S. wacan. 
 (whom), from w// (who), A. S. hwam, from 
 . geoc, gioc. 
 
 In the case of ^v;/, could, deaf, death, enough, ready, 
 again, against, the spelling shows that they came into Eng 
 lish with the long quantity which they had in the Ormulum. 
 The preterits barr (bare, bore, A. S. bar), bat (bit, A. S. bat), 
 brace (brake, broke, A. S. brae], eonim (came, A. S. cwam, 
 coui), cwathth (quoth, A. S. cwiid/i), saJiJi (saw, A. S. seaJi), 
 space (spake, spoke, A. S. sprac), \aff ( gave, A. S. geaf) % have 
 not been placed in the foregoing lists, because the English 
 forms, though used in both numbers, correspond apparently 
 to the plurals of the A. S. (bceron, biiou, brcecon, cwanion or 
 coinon, cwcsdon, sdwon or scegon, spree con, gedfori) and the 
 O. (bcerenn, comenn, sce\Jicnn, spcekenn, \cefenri) : inr/(he ate, 
 A. S. at) with long e like pi. etcnn (they ate, A. S. cetoii), the 
 same extension of the plural quantity to the singular appears 
 even in the Ormulum. The on in the English preterits bound, 
 found, wound A. S. I, 3 sing, band, fand, wand, pi. bund- 
 on, fundon,wundon ; O. sing, band, fand, wand, pi. bnndcnn, 
 fundenn is to be explained in the same way.-f 
 
 We have not yet attended to the suffixes of inflection and 
 derivation ; but for these only a few words will be necessary. 
 The inflectional endings are all short in the Ormulum : there 
 is reason, indeed, to believe that such as were originally long 
 had become short during the Anglo-Saxon period. Thus the 
 Ormulum has 
 
 -ess in the gen. sing. : as flashess kinde {flesh s kindred, A. S. 
 
 flasces gecynd*). 
 
 -ess in the plural : as Icechcss (leeches, A. S. Icecas, earlier Icecas or 
 lacias) . 
 
 * The digraph ck, in the Ormulum, is equivalent to cc or kk, and marks the 
 vowel before it as short. At the end of a word, or before a consonant, cc is alone 
 used ; if a vowel follows in the same word, ck or kk takes its place. 
 
 f For several of these preterits the Scottish dialect has forms such as brak, 
 cam, spak, fand, etc. which correspond to those here given from the Ormulum. 
 
294 ENGLISH VOWEL QUANTITY. 
 
 I 
 
 -err in the compar. forrtherr (further, A. S. furdhor), itam forth 
 (forth, A. S.ford/t) : in the compar. of adjectives the O. has -re 
 (A. S. -ret, -re), as fu Ire (fouler, A. S. falrci). In E., -er has 
 become long by change of r. 
 
 -esst in the superl. : as dcrcsst (dearest, A. S. defrost, dearest). 
 
 -esst in the 2cl person of verbs : as hcrcsst (hearest, A. S. hyresf). 
 
 -ctJitli in the 3d person of verbs : as loketlitJi (looketJi, A. S. locadk}. 
 
 -enn in the past part. : as haldenn (Jiolden, A. S. Jiealdeii). 
 
 -edd\\\ the past part. : as tuundcdd (wounded, A. S. wundod). 
 
 The -de of the weak preterit is usually added directly to 
 the stem, as drcdde (he dreaded). The present participle ter 
 minates in -enndc (A. S. -ende], but is rare in the O., the only 
 instances where it is a proper participle being bcernennde 
 (burning), dwallkennde (misleading), glowcnndc (glowing), 
 and stinnkennde (stinking). The suffix -inng is very frequent, 
 but always forms a verbal substantive ; while -uiig, which in 
 Anglo-Saxon is more used for this purpose, is in O. confined 
 to the word rcoivwsunng (rueing, repentance) and two or three 
 others. Suffixes of this kind, found both in O. and in E., are 
 
 -ell (K. S. -cl) : ^girrdell (girdle). 
 
 -ene (A. S. -en} : as brasene (brazeii). 
 
 -ere (A. S. -ere) : as mineteress (winters, money-changers) ; very 
 rare in O. ; another instance perhaps \\\forrle\errnesse (fornica 
 tion). 
 
 -inng (A. S. -ing, -ung] : as biginninng (beginning), lafcrrdinngess 
 (lord ings). 
 
 -issh (A. S. -isc) : as shcpisshc (sheeplike, sheepish}. 
 
 -i\ (A. S. -ig) \ as modi^ (woody). 
 
 -linng (A. S. -ling) : as derrlinng (darling). 
 
 -nesse (A. S. -ncs) : as godnesse (goodness), iviftncss (witness). 
 
 -stcrc (A. S. -stre) : only in huccstercss (huckster s). 
 
 In final -ene and -ere, the first e, short in A. S., is length 
 ened by the open syllable : the English -er is of course long. 
 Final -i\ is long in consequence of the partial vocalization of 
 the^- (compare the effect of a weak r\ while the correspond 
 ing suffix -y of the later language, discarding the semivowel, 
 has returned to the short quantity of A. S. -ig. The same 
 changes appear in the numerals tivcnnti\, thritti\ t fowwerrti^ 
 etc. (twenty, thirty, forty, etc., A. S. twcntig, tJirittig, fco- 
 wertig, etc.) But in the suffix -//} (our-/j/) the vowel was 
 
ENGLISH VOWEL QUANTITY. 295 
 
 ! 
 
 originaliy long, the words which contain it being compounds 
 of the adj. like (O. lie, A. S.gclic). Suffixes of this kind 
 (really words in composition) are the following, found both in 
 the Ormulum and in English : 
 
 -dom (A. S. -dom) : as horedom (whoredom}. 
 -fald (A. S. -fcald) : as t href aid (threefold]. 
 
 -fasst (A. S. -fUst) : as stedcfasst (steadfast) ; in shammfasst (shame 
 faced} the form has been changed by mistaken popular ety 
 mology. 
 
 -full (A. S. -ful} : as sinnfutt (sinful}. * 
 
 -had (A. S. -had} : as ma fodennhad (maidenhood, maidenhead}. 
 -lac (A. S. -lac) : only in weddlac (wedlock}. 
 -Ices (A. S. -leas) : as child Ices (childless} ; rarely -less, as endeless 
 
 (endless}. 
 
 -lie, -like or -li] (A. S. -lie) : as eorthlic, eorthlikc, corthlii (earthly). 
 -manii (A. S. -man) : as allderrmann (alderman). 
 -shipc (A. S. -scipe) : as wurrthshipc and wurrshipc (worship). 
 -summ (A. S. -sum) : as halsumm (wholesome). 
 -warrd (A. S. -weard} : as affterrwarrd (afterward). 
 
 In -/#/# and -shipe, the Ormulum has lengthened an Anglo- 
 Saxon short vowel, from the influence of Id in the first case, 
 and of an open syllable in the second ; as to the last, the 
 English agrees with the Anglo-Saxon. The short vowel of 
 -warrd has been lengthened in English by the weak r ; while 
 the long vowel has been shortened in -dom, -had, -lac, Ices, 
 two of which end with mute sounds. In Ices, the change had 
 commenced in the thirteenth century. The Ormulum has in 
 many words a suffix -le\\c as seen in godle\\c,=godncsse 
 (goodness) which corresponds to -leik-r (=A. S. -lac), a fre 
 quent suffix in the Old Norse. Peculiar cases are rihhtwis 
 (righteous, A. S. rihtwis), and stall wprrthli$ (stoutly, cf. 
 stalwart, A. S. stdlweordh). 
 
XVI. 
 TENNYSON S PRINCESS.* 
 
 1849- 
 
 MANY readers hav& confessed the disappointment which 
 they felt upon their first acquaintance with <( The Prin 
 cess :" and perhaps nothing but the want of equal frankness 
 has kept back many others from the same confession. Though 
 [in 1849] not extensively read in this country, Mr. Tennyson 
 had come to be rated, according to his fame at home, as first 
 among the English poets of the present generation. A large 
 class of readers, who had taken this opinion upon trust, were 
 looking to see it confirmed in his forthcoming poem, which, 
 as the public were assured in sundry notices, was to be the 
 longest and the most elaborate of his productions. Great in 
 many cases was their disappointment, when, instead of a 
 second Paradise Lost, they found what seemed to them only 
 a grotesque extravaganza about " woman s rights." But 
 there were others, old admirers of the poet, familiar with his 
 earlier pieces ; who had dwelt delighted on the splendid pomp 
 of his " Morte d Arthur," the epic breadth of his " Ulysses," 
 the energy and passion of his " Locksley Hall," that miracle 
 of condensation ; and who were now expecting to find all the 
 merits of their favorite s youthful genius, united and exalted in 
 this effort of his riper years. But lo ! instead of the Gothic 
 cathedral or the Egyptian pyramid on which their hopes were 
 fixed, they see, much to their surprise, only a glittering castle 
 in the air. Chagrined to find the work so different from their 
 preconceptions, they shut their eyes to its indisputable merits. 
 Its grace and gayety, its genial humor, its aptness of expres 
 sion, its brilliancy of coloring go for little with those whose 
 minds were set on greater things. Qualities which mieht have 
 
 " The Princess : a Medley ; by Alfred Tennyson. Boston : William 1). Tick- 
 nor & Co. 
 
TENNYSON S PRINCESS. 297 
 
 pleased them in a Christmas romance or a fairy tale for chil 
 dren, seem out of place in what they had predestined for the 
 master-piece of a great poet. Still they read on, hopeful of a 
 change ; hoping that suddenly, as by some flourish of an en 
 chanter s wand, the fantastic air-castle may settle down into 
 the solidity and solemnity of the Egyptian pyramid. Nor are 
 there wanting here and there tokens of such a metamorphosis : 
 the colors seem to deepen ; the forms to take on fixity and 
 definiteness ; laughing extravagance to give signs of earnest 
 ness and truth. These appearances, however, prove illusory : 
 the edifice, though more imposing than it looked at first, has 
 yet neither substance nor foundation ; it is thin air, and not 
 genuine brick and mortar ; " the baseless fabric of a vision," 
 without strength or unity or grandeur. And so the disap 
 pointed reader shuts his book, doubting his past convictions 
 of its author s genius, and renouncing, for the time at least, 
 his faith in Tennyson. 
 
 Not a few, it is believed, will recognize in this description 
 a tolerably accurate rehearsal of their own experience. Even 
 of those whose first impressions were more favorable, few 
 perhaps would say that the work fully satisfied their expecta 
 tions. On the other hand, fewer still would venture to deny 
 that there is much in it which is excellent and admirable. 
 The severest critic must acknowledge that its faults, however 
 serious, are redeemed by many beauties of detail. A recol 
 lection of its beauties has won back the complainer, recover 
 ing from the first flurry of his disappointment, to a reperusal 
 of the poem. Reading now as one who having formed his 
 judgment is no longer forced to play the critic, he proceeds 
 in a more cheerful mood, with a mind more open to all 
 sources of enjoyment. Things which at first offended him 
 are grown familiar, so that if they do not please they at least 
 cease to be offensive. At the same time, new felicities and 
 beauties, hitherto unnoticed, rise before him ; they gather 
 and grow thick about him as he advances; and when he has a 
 second time attained the goal, he is ready to retract, if he 
 has not quite forgotten them, his former disparaging criti 
 cisms. 
 
 It is impossible to do justice to the work in any abstract 
 
298 TENNYSON S PRINCESS. 
 
 of its story ; and yet we know not how to make the criticisms 
 which we have to offer intelligible without some statement of 
 the plot. We shall follow closely in the author s track, and 
 tell the tale as much as possible in his own words. 
 
 The performance opens with an overture which for its airy 
 gracefulness of movement and dexterous-announcement of the 
 theme could hardly be surpassed. Here the poet shows him 
 self a true man of the nineteenth century ; one not so enam 
 ored of the past as to have lost all consciousness and sym 
 pathy for the present. He can enjoy the fresh breath of the 
 passing day, and drive gayly on with the living currents of con 
 temporaneous thought and action. His "Prologue" con 
 tains a spirited miniature of the age, dashed off in a few 
 characteristic strokes which bring the very form and pressure 
 of the time before us. 
 
 The scene opens in the grounds of Sir Walter Vivian : 
 
 " No little lily-handed Baronet he, 
 A great broad-shoulder d genial Englishman, 
 A lord of fat prize-oxen and of sheep, 
 A raiser of huge melons and of pine, 
 A patron of some thirty charities, 
 A pamphleteer on guano and on grain, 
 A quarter-sessions chairman, abler none ; 
 Fair-hair d and redder than a windy morn." 
 
 This variously accomplished knight, we are informed, 
 
 "all a summer s day 
 
 Gave his broad lawns until the set of sun 
 Up to the people ; thither flock d at noon 
 His tenants, wife and child, and thither half 
 The neighboring borough with their Institute, 
 Of which he was the patron. I was there 
 From college, visiting the son the son 
 A Walter too with others of our set." 
 
 The young Walter takes his friends about the house, show 
 ing its curiosities, a heterogeneous collection, and producing 
 an old chronicle of its ancient owners, where they read the ex 
 ploits of Sir Ralph, a crusading hero of the family, and of 
 
TENNYSON S PRINCESS. 299 
 
 " a lady, one that arm d 
 
 Her own fair head, and sallying thro the gate, 
 Had beat her foes with slaughter from her walls." 
 
 The party at length set forth on a visit to the ruins of an 
 Abbey in the Park. On their way they view the multitude 
 scattered in groups over the meadow, and variously occupied ; 
 some listening, while " the patient leaders of their Institute 
 taught them with facts," or watching illustrative experiments 
 with mimic fountain, fire-balloon, steam-engine, and electric 
 telegraph, while others are engaged in the less scientific 
 amusements of cricket-playing, fiddling, and dancing. Ar 
 rived at the Abbey, our youths fall in with Aunt Elizabeth, 
 an elderly maiden lady of grave deportment, and sister Lilia, 
 "half child, half woman," who had just attired the broken 
 statue of Sir Ralph in scarf and shawl, the gay costume of a 
 modern belle. The sight of this " feudal warrior lady-clad" 
 brings up the old chronicle before mentioned, and with it the 
 heroic lady and her martial daring : and then the question 
 rises, " Where lives such a woman now ? " 
 
 " Quick answer d Lilia, There are thousands now 
 Such women, but convention beats them down : 
 It is but bringing up ; no more than that : 
 You men have done it : how I hate you all ! 
 O were I some great Princess, I would build 
 Far off from men a college of my own, 
 And I would teach them all things : you should see. 
 
 And one said smiling, Pretty were the sight 
 If our old halls could change their sex, and flaunt 
 With prudes for proctors, dowagers for deans, 
 And sweet girl-graduates in their golden hair. 
 
 Yet I fear, 
 
 If there were many Lilias in the brood, 
 However deep you might embower the nest, 
 Some boy would spy it. 
 
 At this upon the sward 
 She tapt her tiny silken-sandaPd foot : 
 That s your light way ; but I would make it death 
 For any male thing but to peep at us, " 
 
3 00 TENNYSON 1 S PRINCESS. 
 
 The conversation passes on to college recreations : Christ 
 mas tales are mentioned, extemporaneous fictions in which 
 the whole company participate, each one taking up the story 
 where his neighbor left it. A new thought presents itself: 
 
 " Why not a summer s as a winter s tale ? 
 A tale for summer, as befits the time ; 
 And something it should be to suit the place, 
 Grave, moral, solemn, like the mouldering walls 
 About us. " 
 
 The hint is followed up ; a tale resolved on, and the " first 
 person" in our Prologue named as leader, with the direction, 
 "be, if you will, yourself your hero." 
 
 " Look then, added he, 
 
 < x Since Lilia would be princess, that you stoop 
 No lower than a prince. 
 
 To which I said, 
 
 Take care then that my tale be followed out 
 By all the lieges in my royal vein : 
 But one that really suited time and place 
 Were such a medley, we should have him back 
 Who told the Winter s Tale to do it for us : 
 A Gothic ruin and a Grecian house, 
 A talk of College and of ladies rights, 
 A feudal knight in silken masquerade, 
 And there with shrieks and strange experiments 
 For which the good Sir Ralph had burnt them all, 
 The nineteenth century gambols on the grass. " 
 
 He then commences, representing himself as a Prince 
 whose father rules with rigid sway over a wide kingdom 
 situated somewhere in high northern latitudes. While yet 
 an infant he had been betrothed to the infant Princess of a 
 neighboring country, equally indefinite in situation, being 
 described only as lying south of the former. But when the 
 time has come to execute the contract, and an embassy is 
 sent with costly gifts to fetch the lady, there appears an un 
 expected obstacle : the Princess cannot be induced to come : 
 "she has a will and maiden fancies; loves to live alone 
 among her women: certain will not wed." The northern 
 king, furious at this violation of a solemn compact, " sware 
 
TENNYSON S PRINCESS. 
 
 301 
 
 that he would send a hundred thousand men and fetch her in 
 a whirlwind." Our hero remonstrates ; urges milder courses ; 
 offers to go himself and ascertain the truth. Failing to gain 
 permission, he takes the matter into his own hands ; steals 
 from court " with Cyril and with Florian, his two friends ; " 
 flying southward crosses the frontier, and so journeys on 
 until he 
 
 "gained the mother-city thick with towers, 
 And in the imperial palace found the king. 
 His name was Gama ; cracked and small his voice : 
 A little dry old man, without a star, 
 Not like a king." 
 
 The Prince speaks of his betrothed, and Gama answers 
 
 11 1 would you had her, Prince, with all my heart, 
 With my full heart ; but there were widows here, 
 Two widows, Lady Psyche, Lady Blanche : 
 They fed her theories, in and out of place, 
 Maintaining that with equal husbandry 
 The woman were an equal to the man." 
 
 Acting under such influences, she had begged from her 
 father a certain summer-palace on the northern frontier of his 
 dominions, and established there "an University for maidens," 
 from which all men, even her own brothers, were rigorously, 
 shut out. Gama is ready to give the Prince letters to his 
 daughter, though he rates his chance with her almost at 
 naked nothing. Our hero takes the letters, and sets out 
 again with his associates : but on coming near the summer- 
 palace, " a thought flash d thro him which he cloth d in act." 
 Having " tweezered otit what slender blossom lived on lip or 
 cheek of manhood," the three friends array themselves in 
 female gear, and ask admission to the College. They are 
 brought before the Princess Ida, whose appearance surpasses 
 the expectation, and confirms the passion of her lover. She 
 notices the tallness of the new-comers, but without suspecting 
 their disguise : and at their own request enrols them pupils of 
 the Lady Psyche; a pretty widow, described as " a quick 
 brunette, well -moulded, falcon -eyed, and on the hither side, 
 or so she looked, of twenty summers." This attractive per- 
 
302 
 
 TENNYSON S PRINCESS. 
 
 son is " lady of three castles in the land," and withal sister to 
 Florian the Prince s friend. Ushered into her lecture-room, 
 Cyril falls in love at once with the lady and her three castles, 
 his regard being pretty equally distributed between them. 
 After her lecture, a report of which is given, she meets her 
 new pupils, and soon recognizes her brother with his two 
 companions. She tells them of the inscription on the gate, 
 "let no man enter in on pain of death," which, approaching 
 after nightfall, they had failed of reading ; but she readily 
 agrees to keep their secret, on condition that they take the 
 earliest opportunity to quit the place. This conversation is 
 accidentally overheard by Melissa, Lady Blanche s lovely 
 daughter, whose charms make an instant impression on the 
 heart of Florian. Melissa promises silence, but her mother, 
 artful and suspicious, divines her secret, and succeeds in 
 worming out a confirmation. Cyril, however, by playing on 
 the tutoress s wounded vanity, and holding out bright hopes 
 to her ambition, bribes her to concealment. Yet our adven 
 turers do not long preserve their incognito. They attend the 
 Princess on an excursion " to take .the dip of certain strata to 
 the North : " after which at evening the party is gathered for 
 rest and refreshment in a superb pavilion. Here, after two or 
 three songs, 
 
 " Did Cyril, with whom the bcll-mouth d flask had wrought, 
 Or master d by the sense of sport, begin 
 To troll a careless, careless tavern-catch 
 Of Moll and Meg, and strange experiences 
 Unmeet for ladies." 
 
 Angry at his freedom, the Prince cries, "forbear, Sir," and 
 smites him on the breast. The ladies on the instant take to 
 flight : but coming to a stream which crossed their way, the 
 Princess blind with rage misses the plank, and is swept down 
 by the current towards a cataract not far below. Apprized of 
 her peril by the shrieks of her attendants, our hero, as in 
 duty bound, plunges in, and with some trouble bears her 
 safely to the shore. The Princess returns to her palace, ascer 
 tains the treachery of her assistants, and summons them into 
 her presence. Lady Psyche, it appears, has fled away with 
 
TENNYSON 1 S PRINCESS. 303 
 
 Cyril ; and Lady Blanche, in spite of a fluent and ingenious 
 vindication of her conduct, is sternly banished from the place. 
 Finally the Prince is brought up for judgment, and though 
 the bloody penalty denounced against all masculine intruders 
 is remitted, he is ignominiously thrust out of doors. 
 
 Wandering forth he finds himself presently in his father s 
 camp, where his appearance in female garb 
 
 " drench d with ooze, and torn with briers, 
 More crumpled than a poppy from the sheath, 
 And all one rag, disprinced from head to heel," 
 
 provokes ungovernable laughter. The old king, it seems, 
 learning the situation of his son, and fearing for his life, had 
 crossed the frontier and was now beleaguering the summer- 
 palace. King Gama too, who having just found out his 
 daughter s sanguinary edict, was coming " all in haste to 
 hinder wrong," has fallen into the enemy s hands and is held 
 a hostage for the safety of the Prince. But soon the rumor 
 comes that Gama s sons, Prince Arac and his two stout broth 
 ers, are approaching with a powerful army. A consultation 
 is held and peace resolved on : only the Prince s claim to his 
 betrothed shall be decided by a grand tournament between 
 fifty champions on either side. To this arrangement Ida 
 gives consent, and appears on the battlements of the palace a 
 spectator of the conflict. Our hero s party is utterly worsted : 
 he himself and his two friends bear down two of the royal 
 brothers on the opposite side, but are all three vanquished by 
 the prowess of the resistless Arac. The struggle over, Ida 
 comes forth with her ladies to give her wounded champions, 
 " brethren of her blood and cause/the tender ministries of 
 female hands and hospitality." As she passes by the spot 
 where the Prince is lying, his aged father mourning over him 
 as dead, she is struck with sudden sorrow and remorse : 
 
 " Her iron will was broken in her mind, 
 Her noble heart was molten in her breast." 
 
 Placing her finger on the Prince s brow, she soon announces 
 that he lives, and offers to tend him with her own wounded 
 brethren. A striking colloquy ensues, which ends in the ar- 
 
TENNYSON S PRINCESS. 
 
 rangement that all the wounded, friend and foe without dis 
 tinction, shall be taken to the palace, and there receive the 
 treatment which their hurts require. Meantime the ordinary 
 college business is suspended, the inmates for the most part 
 returning to their homes. 
 
 The result is easily foreseen. Cyril and Florian soon re 
 cover from their wounds and prosper in their loves. The 
 fate of the Prince is long doubtful : 
 
 " And twilight dawn d ; and morn by morn the lark 
 Shot up and shrill d in flickering gyres, but I 
 Lay silent in the muffled cage of life ; 
 And twilight gloom d ; and broader grown the bowers 
 Drew the great night into themselves, and Heaven 
 Star after star arose and fell, but I 
 Lay sunder d from the moving universe, 
 Nor knew what eye was on me nor the hand 
 That nursed me, more than infants in their sleep." 
 
 All this time Ida is unremitting in her attendance : 
 
 " And still she fear d that I should lose my mind ; 
 And often she believed that I should die : 
 Till out of long frustration of her care 
 And pensive tendance in the all-weary noons, 
 And watches in the dead, the dark, when clocks 
 Throbb d thunder thro the palace floors, or called 
 On flying time from all their silver tongues 
 And out of memories of her kindlier days, 
 And sidelong glances at my father s grief, 
 And at the happy lovers heart in heart 
 And out of hauntings of my spoken love, 
 And lonely listenings to my mutter d dream, 
 And often feeling of the helpless hands, 
 And wordless breedings on the wasted cheek 
 From all a closer interest flourished up 
 Tenderness touch by touch, and last, to these, 
 Love." 
 
 The long contest between life and death is at length decided 
 in favor of life. The Prince returns to consciousness ; dis 
 covers the new-born affection of his mistress ; converses 
 eloquently with her on the true position, office, destiny of 
 woman, mixing old-world wisdom with high anticipations of a 
 
TENNYSON S PRINCESS. 305 
 
 brighter future : when the curtain drops, and a brief " Epi 
 logue," carrying out the action of the " Prologue," brings the 
 work to a conclusion. 
 
 A large part of the critical objections urged against "The 
 Princess" relate to the person of the heroine. Her character 
 is charged with inconsistency : "a virago in the progress of 
 the work, she appears at the close in all the modesty and soft 
 ness of her sex : the reader is offended at the outset by her 
 severity, by the self-will and obstinacy which she shows ; and 
 afterwards, when the metamorphosis occurs, when the lioness 
 becomes the lamb, he feels the incrcdulus odi, he cannot trust 
 the transformation." This criticism is not wholly without 
 foundation : yet we feel that it mistakes the earlier phase of 
 Ida s character. The delineation of the poet shows us a being 
 strong in will, with great energy and great persistency, but 
 not destitute of gentleness or tenderness. A generous feeling 
 forms the very basis of her faults. It is not pride and ambi 
 tion which urge her to engage in her great enterprise for ele 
 vating woman : it is the view of wrong, of the weaker suffer 
 ing from the injustice of the stronger, of female degradation 
 and oppression evils which the poet in his reconciling con 
 clusion acknowledges as real : " these were the rough ways of 
 the world till now" evils which we must all deplore in the 
 past, whatever we may think as to the proper remedy for the 
 future. Personally she is not affected by these evils ; her 
 station places her above their reach : but, full of generous 
 -sympathy, she dedicates her life to the redress of injuries of 
 which others are the victims. Surely there is nothing very 
 vixenish in this. She has indeed her theory as to the way in 
 which her objects are to be effected : a method of her own, a 
 grand catholicon for social maladies. It is her college, on 
 which she has " toiled and wrought and thought ; " with which 
 she has become identified in all her feelings, hopes, and wishes. 
 This scheme of hers she comes to regard with the natural 
 fondness of a projector : and now doubtless pride of opinion, 
 ambitious longings for the triumph of her principles and plans, 
 appear among the motives of her conduct. Still, the noble 
 object which at first attracted her is never absent from her 
 mind : a high enthusiasm flashes through mists of vanity and 
 20 
 
3 o6 TENNYSON S PRINCESS. 
 
 selfishness, which dim but cannot hide its splendor. If she is 
 severe against all that menaces her darling scheme, it is not 
 alone because the scheme is hers, but because it involves to 
 her view a millennium of happiness. The faults of the Princess 
 illustrate, as they were certainly intended to illustrate, the 
 havoc which devotion to a theory may work in the best nature. 
 The reader is made to feel that they are not essential elements 
 of the character, but excrescences upon it, the noxious growth 
 of a mistaken system. The poet exposes them with con 
 scientious love of truth ; but at the same time shows us here 
 and there many traits of natural amiable feeling, which pre 
 pare us for his heroine s last development, and make us ready 
 to believe the Prince, when he says 
 
 " Ere seen I loved, and loved thee seen, and saw 
 Thee woman thro the crust of iron moods 
 That mask d thee from men s reverence up, and forced 
 Sweet love on pranks of saucy boyhood." 
 
 Thus, for instance, speaking of the Prince, she says 
 
 " To nurse a blind ideal like a girl, 
 Methinks he seems no better than a girl : 
 As girls were once, as we ourselves have been : 
 We had our dreams j perhaps he mixt ivitti them ." 
 
 and shortly after, 
 
 " Yet will we say for children, would they grew 
 Like field flowers everywhere ! we like them well 
 Children that men may pluck them from our hearts, 
 Kill us with pity, break us with ourselves 
 O children there is nothing upon earth 
 More miserable than she that has a son 
 And sees him err." 
 
 When the Lady Psyche is forced from college precincts, 
 her infant daughter is retained : 
 
 " For this lost lamb (she pointed to the child) 
 Our mind is changed : we assume it to ourselves." 
 
 After the tournament, when urged by friend and foe to restore 
 the child, she thus addresses it: 
 
TENNYSON S PRINCESS. 307 
 
 " Pretty bud ! 
 
 Lily of the vale ! half opened bell of the- woods ! 
 Sole comfort of my dark hour, when a world 
 Of traitorous friend and broken system made 
 No purple in the distance, mystery, 
 Pledge of a love not to b,e mine, farewell ; 
 These men are hard upon us as of old ; 
 . We two must part : and yet how fain was I 
 To dream thy cause embraced in mine, to think 
 I might be something to thee, when I felt 
 Thy waxen warmth about my milkless breast 
 In the dead prime." 
 
 Even in her more heroic strain, the attentive reader will 
 discern a something not altogether natural, a nerving of the 
 heart to heavy toils and painful sacrifices, an inward struggle 
 to stifle softer feelings because their indulgence seems incon 
 sistent with the claims of solemn duty, She is apparently 
 forcing herself on to a hardness alien from her real character, 
 but regarded as essential to her arduous enterprise. Hence 
 when circumstances come to modify her views, and a new 
 passion springs up in her heart, and the fanaticism, noble but 
 mistaken, of a narrow system passes away, the reader is not 
 surprised to witness the full blossoming of a tenderness, the 
 germs of which he had before seen, chilled but not destroyed 
 by the frosts of theory and prejudice. The irrepressible 
 passion of her lover, and the hearty fondness of her stout 
 brother Arac have made him feel from the first that there 
 must be something truly lovable about her. This feeling of 
 his gains confirmation from her generous sacrifices for the 
 good of others, from the honest enthusiasm which animates 
 her most unwomanly words and actions ; it is strengthened 
 into certainty by the beautiful and natural expressions which 
 from time to time force their way out against all opposition ; 
 and it has its full realization in the final purified and softened 
 development of her character. 
 
 Some are displeased with the pedantry, as they call it, of 
 the Princess, and vote her an intolerable blue. It is no 
 wonder, certainly, that one shut out from the ordinary 
 objects of attention and pursuit among her sex should seek 
 to fill their place by science : especially as a severer mental 
 
308 TENNYSON 1 S PRINCESS. 
 
 discipline is an important feature in her favorite scheme. 
 That it should figure in her conversation follows from the 
 same conditions. What has she left to talk of but her studies 
 and her plans ? Her principal theme, however, is not science, 
 but social regeneration : she is not a pedant, but a reformer. 
 She is eloquent on woman s rights, their violation in the past, 
 their vindication in the future : and if when discoursing on 
 these topics she falls into the style of a lecturer or haranguer, 
 it is only a natural result of her theory, and well illustrates 
 the inherent opposition of that theory to true womanly 
 decorum. Yet here, it must be owned, we find a serious 
 difficulty, belonging to the subject of the poem and insepa 
 rable from it, one which the poet himself must have felt most 
 keenly : how to engage his heroine in this Amazonian move 
 ment without unsexing her, without making her unfemininc, 
 and so breaking" the charm with which she was to be invested. 
 
 t> 
 
 His success in the attempt, if not complete, seems to be as 
 nearly so as the nature of the case admits. It should not be 
 forgotten that the prime mover in the business is the Lady 
 Blanche, an artful, selfish, and ambitious dowager, who re 
 ceives charge of the girl Princess after her mother s death, and 
 by her talents and acquirements has gained an ascendency 
 over her pupil s mind. With her the college scheme 
 originates, designed of course solely for her own aggrandize 
 ment. Working on Ida s lively sympathies and taking 
 advantage of her inexperience, she enlists her in the under 
 taking, which once undertaken her own force of character 
 and sanguine temper, and withal her natural pride, will not 
 allow her to abandon, will not allow her even to weigh in the 
 balance of an impartial judgment. 
 
 The two collaborators of the Princess serve admirably as 
 foils, to show the superiority of their head. Lady Blanche 
 is hollow and heartless, plausible in appearances, false in 
 professions, governed in all by sordid motives. Lady Psyche 
 on the other hand is swayed by personal attachment ; it is 
 her affection for the Princess which has led her to embark in 
 the grand enterprise : she has indeed some understanding of 
 its objects and a certain sympathy for them ; but there is no 
 basis of principle, no disposition to suffer martyrdom for the 
 
TEA NYSON S PRINCESS. 
 
 309 
 
 cause in which she is engaged, though for the person of her 
 mistress she would willingly endure it. The Princess Ida 
 alone has a genuine enthusiasm founded on a firm conviction ; 
 she is sustained through difficulty and danger by an unshaken 
 faith in the goodness of her cause ; she believes strongly and 
 therefore acts strongly ; she rises before us in the dignity of a 
 superior nature, great in its excellences, great even in its 
 errors. 
 
 The poem abounds in trios ; the most distinctly drawn, and 
 altogether the most striking, being the lady-trio just de 
 scribed. r But we have again the three brothers of the 
 Princess, the invincible Arac and the doughty twins 
 " fortisque Gyas fortisque Cloanthus" undistinguished in 
 character, and serving only as opposites in tournament to the 
 Prince and his companions. Here in this suitor-trio, on the 
 other hand, we find more marked diversity. 
 
 " That morning in the presence-room I stood 
 With Cyril and with Florian, my two friends : 
 The first a gentleman of broken means 
 (Mis father s fault), but given to starts and bursts 
 Of revel ; and the last, my other heart, 
 My shadow, my half-self, for still \ve moved 
 Together, kin as horse s ear and eye." 
 
 Cyril is a light sketchy character, but drawn with great 
 spirit. His unfailing gayety, his careless humor, the mask of 
 recklessness which he contrives to throw over a nature honest 
 and honorable at bottom, remind us strongly of the dashing 
 heroes of Beaumont and Fletcher. As for Florian, he is a 
 mere double to the Prince, convenient in the conduct of the 
 plot, but without any striking specialty of mind and char 
 acter. 
 
 The Prince himself has failed of satisfying critics. His 
 character has been censured as at once meagre and incon 
 sistent. On the one hand, it is urged that there is nothing 
 distinctive about him, no individual traits, no personal 
 peculiarities : he is a mere stock hero of romance. But we 
 should be careful lest we exact too much. What more can 
 justly be demanded of a young Prince than that he be frank 
 and generous and faithful, brave and gentle, beautiful and 
 
3 1 o TENNYSON S PRINCESS. 
 
 loving ! Do not these qualities suffice to constitute a noble 
 character ? do they not afford fair augury for a good life and 
 glorious actions ? " True, but then they are too common : 
 we find them in every novel : in a great poem we have the 
 right to claim and to expect something different." But this 
 wide diffusion of the character only proves that it awakens 
 universal interest, and is therefore well suited to the purposes 
 of romantic fiction. Such a hero, it should seem, forms in 
 general the most convenient centre for the various figures 
 and actions of the piece. Nor is it right to quarrel with the 
 character for its ideal perfection, to claim that for the sake of 
 variety it should be made crafty, irascible, or garrulous ; as 
 well might you quarrel with the symmetry of an Apollo 
 Belvedere, and insist on giving it a more marked appearance 
 by the truncation of the nose or the elongation of the 
 chin. 
 
 But the Prince, again, equally with his betrothed, stands 
 accused of inconsistency. " He starts, a boy, a mere love 
 sick boy, who steals disguised and undetected into a female 
 college, who finds no difficulty in passing himself off for a 
 woman : yet anon we see him tilting in the lists, a champion 
 of no mean note : soon, recovering from protracted sickness, 
 he appears mature in mind, a poet-sage, uttering maxims of 
 profoundest wisdom." As regards the question whether it 
 is physically possible for one who could successfully disguise 
 himself in woman s weeds to bear his part with honor in the 
 deadly conflict of the tourney, we shall not presume to de 
 cide. We are content with referring to the martial heroines 
 of Tasso and Ariosto, whose exploits seem to prove that, 
 whether true or not in actual life, in the imaginary world of 
 the romancer such a thing is not impossible. And in the 
 same imaginary world, the act of donning such disguise 
 assumed for no base end may be looked upon as little worse 
 than what our Prince himself has called it, "a prank of saucy 
 boyhood." The effervescence of a frolic humor, it does not of 
 necessity imply a want of manly energy or honorable feeling. 
 It is not inconsistent even with that noble peroration ; in 
 which, however, let it be observed, the quality most promi 
 nent is not hoary-headed wisdom, but truth of natural feeling. 
 
TENNYSON S PRINCESS. 3 1 1 
 
 The Prince s ideal of woman is founded on the example of a 
 mother, "who was mild as any saint, and almost canonized 
 by all she knew, so gracious was her tact and tenderness ; " 
 and whatever such a model could not furnish would be 
 readily supplied by the instincts and the experience of a 
 loving heart. Thus inspired and guided may our hero, 
 without incurring the charge of wisdom beyond his .years, 
 break out into a strain of poetry, which. we cannot deny 
 ourselves the pleasure of quoting : 
 
 " For woman is not undevelopt man, 
 But diverse : could we make her as the man, 
 Sweet love were slain, whose dearest bond is this, 
 Not like to like, but like in difference ; 
 Yet in the long years liker must they grow : 
 The man be more of woman, she of man ; 
 He gain in sweetness and in moral height, 
 Nor lose the wrestling thews that throw the world ; 
 She mental breadth, nor fail in childward care : 
 More as the double-natured poet each : 
 Till at the last she set herself to man, 
 Like perfect music unto noble words." 
 
 " And this proud watchword rest 
 Of equal ; seeing either sex alone 
 Is half itself, and in true marriage lies 
 Nor equal, nor unequal : each fulfils 
 Defect in each, and always thought in thought, 
 Purpose in purpose, will in will, they grow, 
 The single pure and perfect animal, 
 The two-cell d heart, beating with one full stroke 
 Life." 
 
 But the objections take a wider range. Much is said of the 
 improbability \vorse than that, the impossibility and sheer 
 absurdity of the story, which, as no one can suppose it true, 
 or make it real to his mind, is declared unfit to awaken inte 
 rest and call forth sympathy. We cannot deny our skepti 
 cism as to the occurrences narrated in " The Princess : " we do 
 not find them in the history of past or present, nor do we look 
 for them in the developments of the future. And doubtless, 
 were we closely questioned, we should be forced to make a 
 similar admission as to the Orlando Furioso and the Midsum 
 mer Night s Dream, poems which are nevertheless acknowl- 
 
312 
 
 TENNYSON S PRINCESS. 
 
 edged to possess high poetic excellence. The truth is, that 
 the probability required in such a work is not absolute but 
 relative ; it is a probability founded in the assumptions of the 
 author. Is the work adequately motived? does it contain the 
 power necessary for its own propulsion ? do the forces given 
 at the outset, with those afterwards brought in, harmonize in 
 all their operation ? will their united action account suffici 
 ently well for the effects described ? We have a right indeed 
 to claim, as this statement obviously implies, that the forces 
 shall be generally intelligible, such as we can understand and 
 feel the meaning of: they must have a true relation to our 
 common nature ; they must be consistent with the universal 
 principles that govern human conduct. But the particular 
 conditions under which they are to act are left to the fancy 
 or the judgment of the poet. If you would form a critical 
 estimate of " The Princess," assume the external conditions of 
 the work. Do not concern yourself about the situation of the 
 northern and southern kingdoms: take them for granted. 
 Do not be surprised that woman s rights have come to be 
 acknowledged even in princely halls : assume the fact of such 
 a recognition. Do not vex your mind with the manifold ob 
 stacles which hinder the organization of a female college : 
 leave them to the Princess Ida, and suppose them happily sur 
 mounted. But inquire whether, on the given conditions, the 
 enthusiasm of the Princess, her ardor in the prosecution of a 
 great moral work, her attachment to a system of agencies re 
 garded as the necessary means of its accomplishment whether 
 these motives operating on a nature such as hers are not suf 
 ficient to account for her conduct. Let the same method be 
 extended to all the persons and actions of the poem. We 
 are persuaded that the result of such a criticism would be to 
 weaken very much the charges of improbability, so often 
 loosely brought against our author s plot. 
 
 It is said, however, that we cannot thus accept the external 
 conditions of the work, because they are inconsistent in them 
 selves ; presenting an aggregate of incongruities, a cabinet of 
 curiosities filled with ill-assorted specimens from every time 
 and country, a confused mass of forms, principles, ideas such 
 as never have been and never can be united in the actual 
 
TENNYSON 1 S PRINCESS. 3 1 3 
 
 world. This view of the poem is strongly stated by an able 
 critic in the North British Review for May, 1848 : 
 
 " The Princess ; a Medley, upon the first reading has a very curious 
 effect. It is so thoroughly a medley, its heterogeneousness is so com 
 plete, that we wonder how any mind should have been able to escape the 
 apparently inevitable continuity with which feelings and ideas suggest 
 themselves. Tragedy, comedy, love, satire, the old and the new, mo 
 dern conventionalisms and outrageous fancies, all contrarieties come 
 together, and at first appear to clash." 
 
 And again, after a synopsis of the story : 
 
 " There are numerous other instances and characters, all wonderfully 
 elaborated, which we have not noticed because they have no connection 
 whatever with the main plot : but, had we done so, we should still have 
 fallen far short of giving the reader a notion of the utter want of interest, 
 unity, and purpose in this production, considered merely as a narrative 
 poem ; and of its miserable weakness and want of integrity, if regarded, as 
 some regard it, as a satire upon learned women. Now by regarding it as 
 neither the one nor the other, and attributing to it some significance of 
 which the incidents and characters are merely symbolical expressions, 
 we at once do away with an overwhelming amount of difficulty and con 
 tradiction, and are enabled to reconcile its composition with the quality 
 of Mr. Tennyson s genius." 
 
 If this were so, the work might perhaps pass for a tolerable 
 allegory or enigma ; but it would be a wretched poem. It 
 could advance no claim to be regarded as a work of art : for 
 art can never be indifferent to outward form ; it requires not 
 merely a unity of idea, but a certain order, proportion, and 
 harmony in the symbols by which ideas are represented. 
 The fantastic painting in the Ars Poetica would still be a mis 
 erable picture, though you should give it symbolical signifi 
 cance, and show that the intellectual objects denoted by the hu 
 man head, the horse s neck, and the various plumage stand in 
 the most natural relation to each other. On the other hand, let 
 men, birds, and horses be grouped together in a painting 
 which has consistency and beauty, and we recognize the merit 
 of the picture without much reference to its spiritual meaning. 
 A certain degree of coherence is essential even to the allegory : 
 we should have been displeased with Bunyan, if his Pilgrim 
 had appeared as a monkey or a mouse, in order to adumbrate 
 some moral transformation. 
 
3 1 4 TENNYSON S PRINCESS. 
 
 There is no difficulty in admitting that our author in his 
 present work intended to portray the passing time in many of 
 its aspects. This design is clearly indicated in his Prologue, 
 and it accords well with the views expressed in numerous pas 
 sages of his earlier poems. He is by no means one of those 
 poets who live wholly in the ideal world. He is alive to all 
 that is living and stirring around him ; watches with eager in 
 terest the great progress of thought and knowledge and action ; 
 and stamps upon his works the fire-new impress of the current 
 day. He not merely appreciates the present, but delights in 
 it, as the grand result of all past ages, rich in the spoils of 
 time, in its own acquisitions, and still more in " the promise 
 that it closes." He does not reclaim in impotent vexation 
 against a progress which he cannot check : but welcomes in 
 every hour "a. bringer of new things; " acknowledges that 
 " meet is it changes should control our being, lest we rot in 
 ease ; " holds it " better men should perish one by one, than 
 that earth should stand at gaze like Joshua s moon in Ajalon ; " 
 and kindling with enthusiasm cries, " Forward, forward let us 
 range : let the nations spin forever down the ringing grooves 
 of change." That the works of such a poet should present 
 manifold reflections of the present age can occasion no sur 
 prise. Many passages of "The Princess" are direct expres 
 sions of ideas, principles, and feelings peculiar to our own 
 time. In many instances he has made use of representative 
 forms : that is, he has taken some particular phenomenon, and 
 made it stand for a large class of similar phenomena to which 
 it is related. This is especially the case with his " university 
 for maidpns," which may be viewed as representing a multi 
 tude of rampant radicalisms. All this, however, is very differ 
 ent from that continuous allegory, that perpetual succession of 
 metaphors, embracing every detail of the poem, which our 
 Scottish contemporary assumes, though he candidly admits 
 his inability to interpret the greater part of them : 
 
 " Let us frankly confess that an unusually careful study of this poem 
 has not enabled us to discover any such distinct connection between the 
 greater portion of its details, and what we conceive to be the central 
 thought; upon which, if the poem be a truly artistical work, they must 
 every one of them depend for their primary meaning and value. Very 
 
TENNYSON? S PRINCESS. 3 j 5 
 
 many are the thoughts, allusions, traits of character and incidents, the 
 true meaning of which we seem to perceive fully : very many appear to 
 us to possess only some half-perceived capacity of application to the cen 
 tral thought : but very many more have proved too enigmatical for our 
 patience or our powers." 
 
 We hardly know why the critic should have given up these 
 latter points. A writer intent on this sort of interpretation, 
 and gifted with a tolerable share of ingenuity, can have little 
 difficulty in educing meaning out of any text ; he can even find 
 for the same text as many meanings as Swedenborgians at 
 tribute to the words of Scripture. But the whole attempt 
 strikes us as a profitless expenditure of thought, by which no 
 thing like certainty could ever be attained. It is a wander 
 ing through labyrinthine forests with an ignis fatuus for your 
 only guide. The work which cannot be justified except by 
 such a guess-work process may be pronounced, in the lan 
 guage of debating societies, " wholly .unjustifiable." Davits 
 sum, 11011 CEdipus. We shall wait till the author publishes a 
 key to his own riddle-book, before we meddle with its puzzles ; 
 before we concede that it was designed by the author and is to 
 be treated by the critic only as a puzzle. 
 
 It cannot be denied that the poem shows in different parts 
 a difference of "tone and coloring. The author himself ad 
 mits the fact in his " Epilogue : " 
 
 " Here closed our compound story, which at first 
 Perhaps but meant to banter little maids 
 With mock-heroics and with parody : 
 But slipt in some strange way, crost with burlesque, 
 From mock to earnest, even into tones 
 Of tragic, and with less and less of jest." 
 
 This mixing up of grave and gay, serious and mirthful, tra 
 gedy and comedy, has given some offence, and in the judg 
 ment of the ingenious critic quoted above can only be de 
 fended on the ground of allegorical significance. Would he say 
 the same of the Midsummer Night s Dream, or Hamlet Prince 
 of Denmark ? The example of our great dramatist has shown 
 that the close association of the tragic and the comic, which 
 so constantly presents itself in real life, is not inadmissible in 
 
3 1 6 TENNYSON S PRINCESS. 
 
 the creations of art ; that they may aid each other as the 
 lights and shadows of a picture ; that the lighter portions are 
 useful to relieve the eye, while by contrast they throw the 
 dark into still deeper gloom. And whether it be correct ap 
 preciation or misjudging fondness we know not ; but in this 
 poem of Tennyson s, it appears to us that the two elements, 
 so far from neutralizing one another, are blended into harmony, 
 and receive each a heightened beauty from their union. 
 
 After all that may be said about the absurdity and inco 
 herence of the story, it certainly produces the impression of 
 reality in a degree which, when the nature of the incidents is 
 considered, must be thought truly wonderful. So vividly and 
 clearly does the poet delineate the creatures of his fancy that 
 we cannot help viewing them as actual existences. We find 
 ourselves sympathizing with the Prince, and wishing him suc 
 cess in his arduous suit. We feel the rush of breathless ex 
 pectation in the hot inclic of the tourney. We wait anxiously 
 the turn of fate beside the sick-bed of the wounded lover. 
 We give him our heartiest congratulations on his eventual 
 recovery and success. It is only when we set ourselves to 
 criticizing that we are struck with the improbability of that 
 which moved us, and become ashamed of our former feelings. 
 In no former production has the author suc<*ecded in giving 
 so much the air of reality to the objects of his imagination; 
 nor has he shown in any one so much delicacy and distinct 
 ness in the delineation of character. 
 
 The poetry of our day has been almost exclusively lyrical. 
 Moving hotly and hurriedly in the career of politics, or swal 
 lowed up in business, or prosecuting science with a zeal and 
 success never before paralleled, we have found no time for 
 lengthened poems. Only now and then could we snatch a 
 moment for a brief utterance of feelings which belong to 
 human nature, and can never be utterly lost even in the mad 
 dest vortices of life. As for great constructive poems, vast 
 systems of narrative, meditation, and description, built up in 
 the deeps of an ideal world, they have well-nigh disappeared. 
 In America, where the influences that oppose their construc 
 tion are the strongest, we have nothing of the kind ; our poets 
 have to a singular extent been song-writers ; while the occa- 
 
TENNYSON 1 S PRINCESS. 3 1 7 
 
 sional attempts which we have seen in epic and dramatic com 
 position have been generally unsuccessful. Yet this has been 
 almost equally the case in England. What long poem of any 
 note has appeared there since the " Excursion ? " and what is 
 the Excursion itself but a long lyric ? no narrative of action ; 
 no development of character ; no plot or story ; no complica 
 tion of incidents ; no catastrophe or denouement ; no unity, but 
 that of a rosary ; a series of lyrical exercises, for the most part 
 didactic, strung loosely on the very slightest thread of story. 
 Byron wrote narrative poems of some length ; but his genius 
 was essentially lyrical. The plots of his Corsair, Lara, Giaour, 
 etc., will not stand the lightest touch of criticism ; barren and 
 confused, they serve only as openings for fine description or 
 eloquent declamation. Nor does Don Juan in the aimless 
 ramblings of its hero show anything more of constructive 
 power. In Southey s epics, if we may use a name which the 
 poet himself rejected with contempt, we find our best recent 
 specimens of epic art. Yet we know not how the critics who 
 blame "The Princess" for its inconsequences and improba 
 bilities could defend the grotesque wonders of " Thalaba 
 the Destroyer" or the " Curse of Kehama." Is it said that 
 these are to be tolerated as forms of actual belief, importa 
 tions from Pantheons of mythologies currently received among 
 our fellow-men ? But how does that lessen the incredulity 
 with which we look upon such wild arid monstrous fictions ? 
 Or would the college of the Princess Ida seem to us one whit 
 more probable, could we find some Hottentot or Eskimo who 
 held it for undoubted verity ? And again, the inconsistency 
 arising from a mixture of incompatible ideas, which is urged 
 against Tennyson, may be charged with equal justice on the 
 elder poet, who has by no means treated the religions of 
 India and Arabia in the spirit of their devotees, but has al 
 lowed the notions and feelings of the Christian to appear 
 among the symbols of the Mohammedan and the Brahman. 
 It may be thought, however, that in naming his work "a 
 Medley" the poet has given up the whole point ; that he has 
 decided the question, and decided it against himself. But the 
 name indicates no more than the widely diverse character of 
 the materials employed in the building of his work ; it can- 
 
3 1 8 TENNYSON S PRINCESS. 
 
 not be justly construed as a confession that no care was used 
 in the selection, and no art in the arrangement. And if a 
 writer modestly apply a slighting name to his own production, 
 we should not press the circumstance against him. It is not 
 to be supposed that an artist such as Mr. Tennyson has 
 shown himself to be would inflict a mere medley on the pub 
 lic. He would feel it beneath himself to waste the treasures 
 of his fancy and the classic riches of his diction on a heap of 
 unconnected, aimless, and incongruous absurdities. Nor 
 would he consent to forfeit a well-earned reputation by bring 
 ing such a work, were he even capable of writing it, before 
 the bar of intelligent criticism. We may rest assured that 
 there is some point of view from which the poem will appear 
 as other than a medley ; from which we shall be able, not 
 perhaps to justify the composition in all particulars, for the 
 highest genius must still fall short of absolute perfection ; 
 but to comprehend at least how it was possible for a man of 
 genius ts be the author of such a composition. 
 
 Mr. Tennyson has evidently taken extraordinary pains with 
 the construction of his verse. He seems to have felt that a 
 single measure running through a long poem must of neces 
 sity become monotonous and wearisome, unless great care be 
 taken to diversify its rhythm. He may have thought also 
 that a narrative piece, where the poet must rely less upon the 
 thought and more upon the form than in other species of 
 poetical composition, requires peculiarly the aid of metrical 
 resources. Certain it is that, iii affluence of means and in 
 variety of effects, the blank verse of "The Princess" surpasses 
 all its author s previous attempts in the same kind of measure ; 
 nor would it be easy to find its equal in these respects since 
 the time of Milton. To the versification of the Paradise Lost, 
 the greatest exemplar of versification in the English language, 
 Mr. Tennyson, it is clear, has given no little attention ; and 
 from this poem, and from the older English poetry in general, 
 he has adopted many rhythmical and metrical expedients 
 liberties or licenses, as they are sometimes called which the 
 too finical taste of later times, and the undue passion for uni 
 formity, have generally discarded. Among these we mention 
 the so called elision more truly, the blending of a final vowel 
 
TENNYSO f N S PRINCESS. 3 1 9 
 
 with the vowel initial of a following word into a single syllable, 
 or at least what passes for such in the rhythm. Thus we 
 have 
 
 " That made the old warrior from his ivied nook 
 Glow like a sunbeam." 
 
 " The violet varies from the lily as far 
 As oak from elm." 
 
 "O Swallow, Swallow, if\ could follow, and light 
 Upon her lattice." 
 
 So, too, where the second word begins with a weak conso 
 nant easily elided in pronunciation : 
 
 " Fly to her, and pipe and woo her, and make her mine." 
 " You must not slay him : he risked his life for ours." 
 
 The same fusion occurs often in a single word, and not only 
 in such forms as lovelier, sapience, etc., where all our poets 
 have employed it, but in many instances where the last two 
 centuries have renounced its use. Thus, in the following 
 lines, the words seeing, crying, highest go for monosyllables 
 in the rhythm : 
 
 "And Cyril seeing it, push d against the Prince." 
 11 Some crying there was an army in the land." 
 "And highest among the statues, statue-like." 
 
 The combinations in the, of the, etc., are often treated as 
 filling but one rhythmical place : 
 
 11 Better have died, and spilt our bones in the flood." 
 " Poets, whose thoughts enrich the blood of the world." 
 " When the man wants weight, the woman takes it up." 
 
 In many instances a short syllable is neglected that is, does 
 not count as forming by itself a place in the metre. In the 
 following quotation, the words enemy, general, soluble are 
 treated as dissyllables : 
 
 " Now she lightens scorn 
 
 At the enemy of her plan, but then would hate 
 The general foe. More soluble is the knot." 
 
320 TENNYSON S PRINCESS. 
 
 Especially does this occur where a short final syllable is fol 
 lowed by a word beginning with a vowel : 
 
 " A palace in our own land, where you shall reign." 
 " A tent of satin, elaborately wrought." 
 
 We could distinguish other cases, in which a reader unfa 
 miliar with the earlier English rhythms might be offended by 
 supernumerary syllables : but to enter upon long details would 
 perhaps be more tedious than profitable. In none of these 
 instances, if we may judge of Mr. Tennyson s pronunciation 
 from his way of writing, would he omit a syllable in reading : 
 nor does the rhythm of the verse (let metrical doctors, like 
 Mr. Guest, say what they please about it) require of us the 
 use of any such expedient. Yet, as the same word or combi 
 nation may receive in different places a different metrical ar 
 rangement, the reader, unless he proceeds with unusual fore 
 sight, will sometimes be obliged to take a new start and try 
 it over again. This difficulty might be obviated by the use 
 of some appropriate notation ; the elisions and apostrophes of 
 the old books are of course objectionable, if nothing is to be 
 elided ; but it is much to be wished that some method might 
 be substituted for them, to make the proper mode of reading 
 obvious at first sight. 
 
 The poem presents us with many passages in which the 
 rhythm is very studiously adapted to the sense. In some of 
 these perhaps the thing is overdone ; but in most the effect 
 is admirable. We instance the following fine exhibition of 
 " the pomp and circumstance of glorious war:" 
 
 " They made a halt ; 
 
 The horses yell d ; they clashed their arms ; the drums 
 Beat ; merrily blowing shrill d the martial fife ; 
 And in the blast and bray of the long horn 
 And serpent-throated bugle, undulated 
 The banner ; anon to meet us lightly pranced 
 Three Captains out." 
 
 The overwhelming onset of Prince Arac is" described in 
 verses not unfit for the exploits of divine Achilles : 
 
TENNYSON S PRINCESS. 3 2 1 
 
 " But that large-moulded man 
 Made at me thro the press, and staggering back 
 With stroke on stroke the horse and horseman, came 
 As comes a pillar of electric cloud, 
 Flaying off the roofs and sucking up the drains, 
 And shadowing down the champaign till it strikes 
 On a wood, and takes, and breaks, and cracks, and splits, 
 And twists the grain with such a roar that the earth 
 Reels, and the herdsmen cry ; for everything 
 Gave way before him." 
 
 A waterfall is thus represented in words which address 
 themselves at once to mind and sense : 
 
 " And up we came to where the river sloped 
 To plunge in cataract, shattering on black blocks, 
 A breadth of thunder." 
 
 In the following passage, describing the rescue of the Prin 
 cess, the verse is managed with much art : 
 
 " There whirled her white robe like a blossom d branch 
 ^ Rapt to the horrible fall : a glance I gave, 
 
 No more ; but woman-vested as I was, 
 Plunged ; and the flood drew : yet I caught her j then 
 Oaring one arm, and bearing in my left 
 The weight of all the hopes of half the world, 
 . Strove to bujfet to land in vain. 1 1 
 
 The rhythm of the last line (which occurs quite frequently 
 in the Paradise Lost) is admirably adapted to express la 
 borious and unsuccessful effort. Not unlike it is the follow 
 ing : 
 
 " While now her breast, 
 
 Beaten with some great passion at her heart, 
 Palpitated, her hand shook, and we heard 
 In the dead hush the papers that she held 
 Rustle." 
 
 We are tempted to assume the same movement in another 
 place : 
 
 " they to and fro 
 
 Fluctuated as flowers in storm, some red, some pale" 
 21 
 
322 TENNYSON 1 S PRINCESS. 
 
 and to suppose that the poet, intentionally or inadvertently, 
 extended the line beyond the regular standard. It can be 
 read indeed as a verse of five accents, but not without tak 
 ing from the aptness of the rhythm. Breathlessness is thus 
 described : 
 
 11 And on a sudden ran in 
 Among its, all out of or eat Ji, as pursued, 
 A woman-post in flying raiment." 
 
 In this instance, however, the poet seems -to have been more 
 ingenious than successful : the verse breaks down under his 
 menage ; the line, as naturally read, has but four accents, and 
 it is not without difficulty that the reader comes to under 
 stand the metrical intention of the writer. 
 
 In his use of language Mr. Tennyson has shown himself in 
 this work uncommonly adventurous ; but he has in general 
 adventured with a happy boldness. He abounds in striking 
 novelties, in words, meanings, and constructions seldom or 
 never found elsewhere ; and thus, it must be owned, pro 
 duces on the reader s mind a first impression that his style is 
 more or less artificial and affected. On a second reading this 
 impression for the most part disappears. Many phrases 
 which at first seem quaint and odd are soon perceived to 
 have a propriety and a beauty which justify their strangeness. 
 We accept the language as the appropriate vesture of the 
 thought, and feel that no other vesture would become it half 
 so well. One thing certainly is true : the words, be they 
 good, bad, or indifferent, are never the unmeaning substitutes 
 for thought : they are always richly freighted with sense and 
 sentiment. Mr. Tennyson is not one of those copious and 
 fluent writers who can turn off verses by the ream ; he has no 
 patent machine for the manufacture of poetry. His compo 
 sitions are widely different from the voluminous and flimsy 
 job work of literary artisans ; widely different, too, from the 
 productions of many real artists, who are wont to use the best 
 materials, but weave them hastily in thin and ill-compacted 
 fabrics. Our recent poetry even of the better sort has been 
 too generally written currente calaino in many cases it 
 hardly mounts above the reach of a ready-witted improvisa- 
 
TENNYSON 1 S PRINCESS. 323 
 
 tore. You read it, possibly, with interest and pleasure ; but 
 it neither makes a definite impression on the mind, nor gains 
 an abiding home in the memory. You may remember the 
 subject of the work and something of its plot or course of 
 thought ; but words and images and sentences are gone ; 
 quotation is impossible. Seldom, too seldom, do you meet 
 with expressions singularly felicitous, or lines which realize 
 Coleridge s ideal, where every change conceivable would be 
 a change for the worse ; or passages which seem like the 
 spontaneous crystallization of great sentiments and principles ; 
 or utterances which might make one think he heard the Muse 
 herself pronouncing oracles for the instruction and delight of 
 every coming age. It is a singular circumstance that Cole 
 ridge, who was so much of an improviser in his prose, should 
 have been, almost alone among his contemporaries, so spar 
 ing and elaborate in his poetry. In Tennyson we see the 
 same frugality of verse, the same studious care and conscien 
 tious toil in execution. He does not spread himself out upon 
 paper, but brings his whole material within the narrowest 
 compass. Far from dispersing the rays of his genius in di 
 vergent lines, he gathers them in luminous centres, which 
 shine clear and bright like the stars. To one who writes in 
 this way, each word becomes a matter of importance, a sub 
 ject for careful thought and nice selection. He cannot take 
 whatever comes first or lies nearest ; he must ransack the 
 treasury of language, to find that which will most aptly, 
 tersely, forcibly express his meaning. Sometimes it may be 
 an old-world word, which men have latterly forgotten, but 
 may well bring back to memory ; sometimes a word from 
 the living mintage of the poet s own brain, carrying its warrant 
 in its use and justly claiming place among the currency of 
 older dates ; sometimes a colloquialism that deserves to be 
 ennobled ; sometimes an expressive idiom coming from 
 abroad, and naturalized with full rights of native citizenship. 
 A style thus diligently selected and compacted can hardly 
 fail to have a certain quaintness ; and some such quality may 
 have been apparent, even from the first, in the writings of 
 Spenser and Jonson and Milton. The time has been when 
 it was regarded as the very beau ideal of style that it should 
 
324 TENNYSON S PRINCESS. 
 
 resemble fashionable conversation. Fortunately, the feeling 
 now-a-days is very different. We claim for conversation a 
 freedom greater than it once enjoyed, and for writing a free 
 dom greater than we yield to conversation. We allow the 
 writer to consult his own taste as to the colors and the 
 fashions irr which he shall present himself before the public. 
 We do not regard even language itself as a thing inflexible 
 and immutable, which all are to accept and use exactly alike. 
 We acknowledge in the writer, especially in the man of genius, 
 a certain power over language ; a right of origination, not to 
 make himself unintelligible, but to make himself more in 
 telligible ; a right to share actively in that progress which, 
 spite of all conservative resistance, is the inevitable condition 
 of a living language, and the cessation of which proves a 
 language to be really dead, unfit for the living use of living 
 men. 
 
XVII. 
 
 THE NUMBER SEVEN.* 
 1858. 
 
 IT is well known that men of different times and nations 
 have associated with particular numbers the idea of a pe 
 culiar significance and value. It is also well known that, of 
 all numbers, there is no one which has exercised in this way 
 a wider influence, no one which has commanded in a higher 
 degree the esteem and reverence of mankind, than the num 
 ber Seven. The mystic preeminence of this sacred number 
 is as ancient as it is venerable. It belongs to the simple 
 wisdom of a primitive age. It had its native home in the 
 East, near the springs of light and of day. True, we find it 
 also in later times, and upon occidental ground, pervading 
 the mind and literature of modern Europe. But we must 
 remember that an Oriental book, an Asiatic book, the Sacred 
 Scripture of the Hebrew, has leavened may we not add 
 that it has sevened the mind and literature of modern 
 Europe. But before this influence began, before a new re 
 ligion coming into Europe from the East brought with 
 it the Oriental feeling for the Seven, the case was widely 
 different. Among the ancient Greeks and Romans, but 
 little prominence upon the whole is given to this number. 
 Let us look at the first great monuments of western literature, 
 the poems of Homer. Here we find a number of sevens. 
 Seven talents are more than once bestowed as a present. 
 Seven tripods, seven women, and seven towns are among 
 the gifts by which Agamemnon seeks to propitiate the 
 enraged Achilles. Then there are seven ships of Philoctetes, 
 seven brothers of Andromache, seven sons of Polyctor, seven 
 gates of the Boeotian Thebes, seven layers of ox-hide in the 
 
 * Von Hammer-Purgstall, Ueber die Zahl Sieben. (Wiener Jahrbiicher der 
 Literatur. 1848. CXXII. 182-225. CXXIIT. 154. CXXIV. 1105.) 
 
326 THE NUMBER SEVEN. 
 
 impenetrable shield of Ajax, seven herds of cattle belonging 
 to the sun-god Helios, seven roods of ground covered by the 
 fallen war-god Ares. Seven years the murderer ^Egisthus 
 reigns upon the throne of Agamemnon ; seven years Ulysses 
 is kept a prisoner by the fondness of the nymph Calypso ; 
 seven years in his romancing story to Eumaeus he professes 
 to have spent in Egypt. In four instances in the Odyssey, 
 some action is described as continuing for six days and 
 terminating 1 on the seventh in some critical event a curious 
 
 o 
 
 circumstance, in which we might almost be tempted to trace 
 either a dawning or a vanishing of the week. This is the list 
 of Homeric sevens, nearly complete : it may appear some 
 what long ; but there arc quite as many tens in Homer, and 
 of twelves almost twice as many. In the Greek mythology 
 and it is the mythology of a nation that most faithfully 
 reflects its early thinking and feeling the Seven is quite rare, 
 and is nearly confined to the worship of Apollo ; the Twelve 
 again is greatly more important as a mythological number. 
 From Greek philosophy, however, the Seven has received a 
 more respectful attention. The Pythagoreans, who in general 
 laid much stress upon the mystic properties of number, had a 
 special regard for the seven. Thus Philolaus, the contempo 
 rary of Socrates, and the first to set forth, in writing, an 
 extended exposition of Pythagorean doctrine, says concerning 
 God, the author and governor of all things, that " he is with 
 out variation, ever like himself and like no other, even as the 
 number seven." But it will be recollected that Pythagoras, 
 according to the general tradition, had travelled in the East, 
 and was supposed to have drawn from thence, to a greater or 
 less extent, the elements of his system. All have x heard of 
 the Seven Sages, or Wise Men of Greece ; men who, about 
 six centuries before Christ, were highly distinguished among 
 their contemporaries for wisdom and experience ; of whom 
 the most celebrated were the Milesian Thales and the 
 Athenian Solon, It may have been a philosopher who first 
 conceived the idea of selecting out just seven such men to 
 form the group. It is at any rate a philosopher Plato, 
 himself also a traveller in the East who first gives us the 
 seven names, selected out and grouped together. But the 
 
THE NUMBER SEVEN. 327 
 
 Seven in this case does not seem to have taken very strong 
 hold of the Greek mind, or to have possessed inviolable 
 sanctity ; for Dicsearchus substitutes ten sages for the seven, 
 and Hermippus enumerates seventeen. Again, the idea of 
 Seven Wonders of the World, which we find among the 
 Greeks, appears to have originated in the East, with the 
 Egyptian Greeks of Alexandria. One of the seven wonders 
 belongs to Alexandria itself, the light-house in its bay. A 
 second is also Egyptian, the pyramids. Of the remaining 
 five, only one is European, the statue of Zeus at Olympia ; 
 while four are Asiatic the Ephcsian temple of Diana, the 
 Mausoleum of Artemisia, the Colossus at Rhodes, and the 
 hanging gardens of Babylon. The conquests of the Mace 
 donians, and after them the Romans in -the East, the wide 
 dispersion of the Jews, and, following that, the still wider 
 diffusion of Christianity, all had the effect of acquainting the 
 European mind more and more with Oriental ideas. And 
 hence it comes that, in the centuries after Christ, we find a 
 large .number even of heathen writers who render homage to 
 the sacredness and dignity of the Seven. We will not dwell 
 upon them here ; but will rather turn to lands where a vener 
 ation for the Seven appears unborrowed and original. We 
 will look first to the far East, to the banks of the Indus and 
 the Ganges, to the votaries of the Brahman religion. 
 
 According to the conceptions of the Indians, Mount Meru, 
 the sacred mountain of the gods, is surrounded by the seven 
 rocks called Cakravarta, rising in the mystic regions of the 
 atmosphere. Indra, the great lord of the sky, governs by 
 seven vice- kings the seven regions of the heaven. Agni, 
 the- god of fire, is thus addressed in one of the sacred books : 
 " Seven are thy fuels, seven thy tongues, seven thy holy 
 sages, seven thy favorite haunts, in seven ways thy worship 
 pers adore thee, seven are thy sources ; be graciously content 
 with thy clarified butter ! " The sun has his seven rays, 
 which are themselves described as suns, and pour down their 
 sevenfold heat on the torrid land of the Hindu. The earth - 
 itself has its seven Dvipas, seven islands encompassed by as 
 many seas the seas of milk, sugar, honey, salt, salt water, 
 sour water, and butter. Over it blow the winds, Maruts, 
 
328 THE NUMBER SEVEN. 
 
 seven times" seVerr in number. The earth is renewed every 
 seven thousand years, or, as others hold, every seven times 
 seven thousand. There are seven Apsaras or nymphs of 
 Paradise ; seven Saktis, or incarnated attributes of the divin 
 ity ; seven Rishis, or holy sages ; seven Munis, or holy 
 hermits. These last are not unknown to Christendom : at 
 least, the name of holy recluses might be given with propriety 
 to our old friends, the seven sleepers of Ephesus. All have 
 heard the story : how, in the great persecution under Decius, 
 seven Christian youths fled for refuge to a mountain near the 
 city of Ephesus, and there hid themselves in a cave. They 
 slept during the night without disturbance, and woke, as they 
 supposed, on the following morning. Venturing out, after 
 some time, to obtain provisions, they attracted notice by 
 their uncouth garb and appearance. Having purchased what 
 they wanted, they offered in payment some strange-looking 
 antique coins. Suspicion was aroused, and they were 
 brought before a magistrate. They then told their story, 
 from which it appeared that the supposed night s sleep had 
 lasted well-nigh two centuries. The bishop Martin was 
 called in, and even the emperor Theodosius II., brought by 
 express from Constantinople ; in whose presence they re 
 peated their narrative, and then, praising God, with the halo 
 of sanctity visibly encircling them, gave up the ghost. The 
 reader is not required to believe the story ; Baronius, the 
 famous church historian, though he swallows a wagon-load 
 of marvels, is squeamish as to this one. It is, indeed, almost 
 as much a Mohammedan as a Christian tradition ; the Koran 
 is all but the earliest authority we can quote for it. Even at 
 this day the Ottoman navy is under the especial guardianship 
 of the Holy Seven Sleepers ; and in sleepiness, if not in 
 holiness, does credit to its sainted patrons. But again, the 
 Brahman system has its seven paradises, and seven hells ; 
 these re-appear in the religion of the Moslim, who, however, 
 adds another paradise, on the ground that God s mercy ex 
 ceeds his vengeance. Nor are they unknown to Christendom. 
 The celebrated Pico clella Mirandola left among his manu 
 script remains a treatise on the seven heavens and the seven 
 earths, and another on the seven places of hell. In a German 
 
THE NUMBER SEVEN. 
 
 poem of the middle ages ( Wr ^r^ttffiJ^i *f&*m\d the 
 couplet 
 
 Das ist die Hell mit irem Slund, 
 
 Darin wol siben Kammer greulich sind erzund. 
 
 That is hell with its pit of woe, 
 
 Where in fearful flame seven chambers glow. 
 
 It would be easy to extend this enumeration of Indian sevens. 
 But the specimens already given will suffice ; especially as we 
 are not in condition to determine how far they belong to the 
 earlier forms of the Hindu religion, or what proportion they 
 bear to other mythological numbers in the same system. Let 
 us now turn westward to Central Asia, to the countries which 
 formed the heart and strength of the ancient Persian empire. 
 Here in ancient times the prevailing religion was that of Zo 
 roaster, which owns the Zend-Avesta for its Bible, and is pro 
 fessed at the present day only by the scanty remnants of the 
 Parsees. The two great divinities of this religion are Ormuzd 
 (Ahura-mazda), the divinity of light and good, on the one 
 hand ; and, on the other, Ahriman (Angra-mainyus), the di 
 vinity of darkness and evil. Ormuzd is surrounded by his at 
 tendant spirits, the seven Amshaspands, who may be compared 
 with the seven throne-angels, that, according to the book of 
 Tobit, go in and out before the glory of the Holy One. Ahri 
 man in like manner has his court, composed of seven arch-devs 
 or demons, whom, as regards the number, we might compare 
 with the seven that haunted Mary of Magdala, or with the 
 seven more wicked than himself, whom the evil spirit, after 
 his restless wandering through the desert, took back with him 
 to his former habitation, which he found ready, swept, and 
 garnished " and the last end of that man was worse than 
 the first." 
 
 The modern literature of Persia abounds in sevens. Native 
 dictionaries enumerate above a hundred septenaries, groups of 
 objects designated as the seven so-and-so. We will not under 
 take to name them. We could not say, what it would be most 
 interesting to know about them, how far they have sprung out 
 of the spontaneous feeling and invention of the Persian, or how 
 far they are due to the Arabian influence,, itself saturated with 
 
330 THE NUMBER SEVEN. 
 
 sevens, which entered Persia with the religion of Mohammed 
 in the seventh century of our era. 
 
 But let us proceed to ground which has for us the double in 
 terest of more familiar acquaintance and more sacred associa 
 tions. The preeminent importance of the number seven 
 throughout the Bible is seen in the extraordinary frequency 
 of its occurrence. It is found in the Old and New Testaments 
 not less than three hundred and eighty-three times. This count 
 includes the ordinal scvcntJi, as well as the compound seven 
 fold but does not include the higher numbers which contain 
 a seven, such as seventeen, twenty-seven, seventy, seven hun 
 dred, and the like. If, now, we count the sixes in the same 
 way, we find them to be one hundred and eighteen. The 
 eights counted in the same way are fifty-eight. The sixes and 
 the eights taken together amount only to one hundred and 
 seventy-six, or less than one-half of the sevens. 
 
 The preeminence of the Seven is a fact which meets us at the 
 threshold of the Bible, in the first chapter of Genesis, or in what 
 should be the first chapter. It is well known that an unfortu 
 nate blunder in the division has deprived the opening chapter 
 of three verses which justly belong to it. The real break is af 
 ter the third verse of chapter ii. : for the fourth verse, so far 
 from being connected with those before it, is the commence 
 ment of a distinct narrative, composed probably at a different 
 time, and indeed, according to the opinion of many good Bibli 
 cal scholars, composed by a different writer. Now the first ac- 
 .count of the creation, as we find it in the opening chapter 
 with the first three verses of the second, represents to us a 
 sevenfold process, which occupies the first week of world-his 
 tory, and is made up of six successive acts of creation dis 
 tributed through six successive days, and terminated like the 
 Hebrew week by a day of rest. It is indeed conceivable that 
 in this account the sevenfold arrangement may belong rather 
 to the form and drapery of the narrative than to its veritable 
 substance : and in fact we find no hint of it in the fourth, 
 fifth., sixth, and seventh verses of the second chapter, which 
 also contain a separate account of the creation. Even thus it 
 would remain true that a sevenfold arrangement was adopted 
 for the more elaborate narrative which was to embody the 
 
THE NUMBER SEVEN. 33! 
 
 great truth of an original divine creation ; and this circum 
 stance alone would be highly significant ; it would demon 
 strate the preponderant value given by the Hebrew mind to 
 the week and to its number the Seven. Or, again, we may 
 suppose with others that an actual week was spent in this 
 way ; that, after an immeasurable past of geological mutations, 
 of trilobites, ichthyosauri, and batrachians, when a new and 
 nobler resident was to be introduced upon our planet, then 
 its existing arrangements were thrown into temporary confu 
 sion, and a week was passed in fitting and furnishing it anew 
 for the habitation of man. Such a view might perhaps be re 
 garded as still more honorable to the Seven. Or, yet 
 again, with others, we might swell days till they assume 
 the dimension of ages seven ages, each with its own sep 
 arate record, written on the tablet of the earth, in characters 
 which science has been able to decipher, and has found to be 
 identical with the words of the Mosaic description. This 
 view, it is clear, gives the highest exaltation to the Seven. 
 For it makes geology a Seven, and so the mightiest of all 
 sevens. It offers the crowning attestation of science to the 
 secular predominance of this majestic number. It is a view 
 adopted by very eminent scientific men ; with whose judg 
 ment we could not think of matching our own in a question of 
 this nature. Yet we must not seem to glorify our subject at 
 the expense of honest dealing. The confession must be made, 
 that with the best wishes we have not been able to satisfy 
 ourselves in regard to that view ; to be certain that it has a 
 positive basis, that it is more than an ingenious and interest 
 ing speculation. If it were otherwise, if we could overcome 
 our doubts on this head, then would we profess ourselves 
 with a more unquestioning faith even than now, votaries of 
 the world-regulating Seven. 
 
 But let us proceed with the Old-Testament sevens. The 
 Seven bears an important part in ritual observances. In many 
 sacrifices the sprinkling of the blood was to be repeated seven 
 times ; and in many we find seven mentioned as the number 
 of the victims to be offered. But it is still more important in 
 reference to holy times and seasons. Not only is the seventh 
 day of the week honored as a Sabbath with perpetual rcmem- 
 
332 THE NUMBER SEVEN. 
 
 brance, but the seventh week of the year brings its festival, 
 the Penteecst or fiftieth day : for it is separated from the Pass 
 over by forty-nine days, a week of weeks, and is therefore 
 sometimes called the feast of weeks. So the seventh month 
 has its festival, the feast of tabernacles, and its solemn fast too 
 upon the tenth day, the great day of atonement. Again, 
 .there is a week of years terminating in the seventh or Sab 
 batical year, when the land was to cease from labor and to lie 
 untilled. Arid once more, after seven weeks of years, forty- 
 nine years, came the great fiftieth or Year of Jubilee. 
 
 Returning to the book of Genesis, we find Noah command 
 ed to receive .clean beasts and fowls into the ark by sevens. 
 Seven days after this command the rain begins. Wearied 
 with .long imprisonment, Noah sends forth a dove, which 
 returns, having found no rest for the sole of her foot. After 
 seven days he sends her forth again, and she returns w r ith an 
 olive-leaf in her mouth. After seven days more he sends her 
 forth the third time, and she returns to him no more. The 
 patriarch Jacob, in his protracted courtship, after . serving 
 seven years for the wife he did not want, was forced to 
 serve another seven for the wife he wanted. In Pharaoh s 
 dream interpreted by Joseph, there are seven fat kine and 
 seven lean, the seven years of plenty and of famine. Seven 
 years occur repeatedly as the duration of a famine. The 
 one in Elisha s time lasts seven years, and among three al 
 ternative evils offered to David s option, one is a seven years 
 famine. The descendants of Jacob return at last to the land 
 of their fathers, from which God had promised to drive out 
 seven nations greater and mightier than they. In the siege 
 of Jericho, the people for seven successive days march round 
 the city, headed by seven priests blowing on seven rams 
 horns. Only on the seventh day they marched seven times 
 round, and at the seventh time the priests blew, the people 
 shouted, and the wall fell down flat. Samson, when he gave 
 his riddle to the Philistines in Timnath, allowed them the 
 seven days of his wedding feast to make out the solution. 
 When the Philistines were endeavoring through Delilah to 
 discover the secret of his prodigious strength, he first directed 
 that they should bind him with seven green withs, and again 
 
THE NUMBER SEVEN. 333 
 
 that she should weave the seven locks of his head with the 
 web. These methods were tried without success ; but when 
 at last the seven locks were quite shaved off, his strength 
 went from him ; in the well-known language of the hymn, he 
 "shook his vain limbs with sad surprise, made feeble fight 
 and lost his eyes." But it would be tedious to pursue this 
 enumeration through the whole Bible. Let us pass on to 
 notice a somewhat different class of cases. 
 
 In numerous instances the Seven appears to be used, as 
 we say a score or a dozen, for a large indefinite number. 
 The great prominence given to the Seven, and the great 
 respect in which it was held, made it natural .that it should 
 be used in this way. Thus in Daniel, the fiery furnace was 
 to be heated for the three recusant Hebrews " one seven 
 times more than it was wont to be heated." In Proverbs we 
 are told that a just man falleth seven times and riseth up 
 again. The Psalmist says, " Seven times a day do I praise 
 thee because of thy righteous judgments." Apparently this 
 was not intended for an exact numerical statement ; although 
 being interpreted in that way it is relied upon as author 
 ity for the seven canonical hours of devotion ; prima (or 
 prime), matutina (matin), teriia, sexta, nona (noon), vesper a 
 (vesper), and completa. Lamech, who was, like Enoch, 
 seventh from Adam, but in the line of Cain, says to his wives 
 Adah and Zillah, " If Cain shall be avenged sevenfold, truly 
 Lamech seventy and sevenfold." And our Lord, when asked 
 by Peter, "how oft shall my brother sin against me and I 
 forgive him? till seven times?" replies, "I say not unto 
 thee, until seven times, but until seventy times seven." 
 
 But of all the inspired books, the last in the series, the 
 Apocalypse, is the one which displays most frequently and 
 prominently the mystic sacredness of the number seven. 
 At the outset John addresses himself to seven churches in 
 Asia, greeting them from the Lord and from the seven spirits 
 which are before his throne. He describes his vision on the 
 isle of Patmos, when he saw one like unto the Son of Man, 
 in the midst of seven golden candlesticks, and holding in his 
 right hand seven stars. The golden candlesticks are ex 
 plained as being the seven churches, and the seven stars the 
 
334 THE NUMBER SEVEN. 
 
 angels of those churches. In the following visions, a throne 
 is set in heaven, and in the right hand of Him that sat on it 
 is a book sealed with seven seals. Then a lamb with seven 
 horns and seven eyes, which are the seven spirits of God 
 sent forth into all the earth, takes the book and opens the 
 seven seals one after another, the opening in each case being 
 followed by different prodigies. When the seventh seal is 
 opened, seven angels appear with seven trumpets, which 
 they blow one after another, and the blowing is followed in 
 each case by new prodigies. Before the seventh angel sounds, 
 seven thunders utter their voices. Afterwards appears a 
 dragon with seven heads and ten horns and seven crowns 
 upon his heads : and anon a beast rising up out of the sea, 
 also with seven heads and ten horns, but with ten crowns 
 upon his horns. Again another sign in heaven great and 
 marvellous, seven angels having the seven last plagues, who 
 pour out one after another the seven vials of God s wrath 
 upon the earth. Then is seen a woman seated on a scarlet 
 colored beast which has seven heads, these being seven 
 mountains on which the woman sittcth ; and it is added, 
 there are seven kings, of whom five are fallen, and one is, 
 and one is yet to come. In the last vision, that of the 
 heavenly Jerusalem, the prevailing number is not seven but 
 twelve, derived evidently from the twelve tribes of Israel. 
 But with this exception seven is everywhere the prevailing 
 number of the book ; so that, as Von Hammer-Purgstall ob 
 serves, there are two sevens in the greeting, seven churches 
 and seven spirits ; and in the body of the work there are 
 fuund besides two sevens of sevens : viz., first, seven candle 
 sticks, stars, seals, horns, eyes, trumpets, thunders ; and 
 second, seven angels, heads, crowns, plagues, vials, moun 
 tains, kings. 
 
 Such being the rank and dignity of the number Seven 
 throughout the Bible, it is not surprising that the nations of 
 Christendom, with whom the Bible is at once the best known 
 and most revered of all books, should have attached special 
 importance to the number. Illustrations of this fact, drawn 
 from the literature of modern Europe, might be multiplied 
 to almost any extent. But I shall confine myself to a single 
 
THE NUMBER SEVEN. 335 
 
 author, one who may be regarded with more propriety than 
 any other as the representative of modern European litera 
 ture. 
 
 The following are specimens of the Shakspearian sevens. 
 In the Merchant of Venice (II. 9), the Prince of Arragon, 
 who comes as a suitor for the hand of Portia, having unfor 
 tunately for himself made choice of the silver casket, reads 
 this schedule : 
 
 The fire seven times tried this ; 
 Seven times tried that judgment is, 
 That did never choose amiss. 
 
 In Hamlet (IV. 5), Ophelia appears fantastically drest, and 
 cries : 
 
 heat, dry up my brains ; tears seven times salt 
 Burn out the sense and virtue of mine eye. 
 
 Of Coriolanus (II. i) it is said: "He received in the re 
 pulse of Tarquin seven hurts in the body." 
 
 In Measure for Measure (II. i), we find an allusion to the 
 seven mortal sins : 
 
 Sure, it is no sin, 
 Of the deadly seven it is the least. 
 
 In Julius Caesar (III. i), the servant of Octavius says of 
 his master : 
 
 He lies to-night within seven leagues of Rome. 
 So in Midsummer-Night s Dream (I. i), Lysander says : 
 
 Hear me, Hermia ; 
 
 1 have a widow aunt, a dowager, 
 
 Of great revenue, and she hath no child : 
 From Athens is her house remote seven leagues. 
 
 This may remind us of the famous seven-league boots, so 
 long current in popular tradition, which, having now been 
 vamped up with new art by the author of Peter Schlemihl s 
 Wonderful History, may be expected to travel down to the 
 remotest posterity. 
 
336 THE NUMBER SEVEN. 
 
 In As You Like It (III. 2), Rosalind, speaking of the pro 
 gress of time, says, " Marry, he trots hard with a young 
 maid between the contract of her marriage and the day it is 
 solemnized; if the interim be but a sevennight, Time s pace 
 is so hard, that it seems the length of seven years." 
 
 The designation " seven years " is very frequently repeated. 
 Thus in Much Ado About Nothing (III. 3), the Watchman 
 says, " I know that Deformed ; a has been a vile thief this 
 seven year." 
 
 In Coriolanus (II. i), Menenius says, " A letter for me? it 
 gives me an estate of seven years health." And again : 
 
 If I could shake off but one seven years 
 
 From these old arms and legs, by the good gods, 
 
 I d with thee every foot. 
 
 In Pericles (IV. 6), Boult says, " Go to the wars, would ye, 
 where a man may serve seven years for the loss of a leg, and 
 have not money enough in the end to buy him a wooden one." 
 
 In King Lear (III. 4), Edgar sings: 
 
 But mice and rats and such small deer 
 Have been Tom s meat for seven long year. 
 
 In As You Like It (II. 7), is found the celebrated passage : 
 
 All the world s a stage, 
 
 And all the men and women merely players ; 
 They have their exits and their entrances, 
 And one man in hi*s time plays many parts, 
 His acts being seven ages. 
 
 Then follow the seven infant, school-boy, lover, soldier, 
 justice, old age, and second childhood. Everybody knows 
 the passage ; but everybody does not know that this division 
 of human life into seven ages is an idea prevalent long before 
 the time of Shakspeare, as far back even as the Greek physi 
 cian Hippocrates, more than four centuries before our era. 
 
 In the same play, we have the seven degrees of offence in 
 affairs of honor. Touchstone says, " I have had four quar 
 rels, and like to have fought one. Jaques. And how was 
 that ta en up ? Touchstone. Faith, we met and found the 
 
THE NUMBER SEVEN. 337 
 
 quarrel was upon the seventh cause. Jaques. How did you 
 find the quarrel on the seventh cause ? Touchstone. Upon a 
 lie seven times removed as thus, Sir : I did dislike the cut 
 of a certain courtier s beard ; he sent me word, if I said his 
 beard was not cut well, he was in the mind it was. This is 
 called the Retort courteous. If I sent him word again, it was 
 not well cut, he wo if Id send me word, he cut it to please him 
 self. This is called the Quip modest. If again it was not 
 well cut, he disabled my judgment. This is called the Reply 
 churlish. If again it was not well cut, he would answer, I 
 spake not true. This is called the Reproof valiant. If again 
 it was not well cut, he would say, I lie. This is called the 
 Counter-check quarrelsome : and so to the Lie circumstantial, 
 and the Lie direct. Jaques. And how oft did you say his 
 beard was not well cut ? Touchstone. I durst go no further 
 than the Lie circumstantial ; nor he durst not give me the Lie 
 direct, and so we measured swords and parted." 
 
 Further on in the same colloquy, Touchstone says, " I 
 know when seven justices could not settle a quarrel, but when 
 the parties were met themselves, one of them thought but of 
 an If, as, If you said so, then I said so. And they shook 
 hands and swore brothers. Your If is the only peacemaker : 
 much virtue in If." 
 
 A word now in regard to the Mohammedans. The religion 
 of the Koran, like that of the Old and New Testaments, made 
 its first appearance among a Semitic people ; the Israelite has, 
 in race and tongue, no nearer kinsman than the Ishmaelite. 
 The Koran, too, like the Old and New Testaments, has car 
 ried the ascendency of the number Seven over vast regions 
 of the earth. Its preeminence in the Moslem scriptures, 
 though less marked and conspicuous than in ours, is yet not 
 to be mistaken. The sevens of the Koran counted up greatly 
 outnumber the eights and sixes put together. It is possible 
 that in this point Mohammed may have felt the influence of 
 the Bible, with which, though unable to read it, he had cer 
 tainly picked up some acquaintance. But it is more likely 
 that he gave expression to a feeling which belonged already 
 to his Arab fellow-countrymen, in common with their kindred 
 of Palestine. The most venerated monument of their ante- 
 22 
 
338 THE NUMBER SEVEN. 
 
 Mohammedan literature, the collection of poems known as the 
 MoallakatTpozms written in golden letters, and suspended 
 at the Holy Kaaba in Mecca was the production of seven 
 authors. At all events, it is certain that the Mohammedan 
 Arabs have shown an extraordinary predilection for the 
 Seven, not yielding in this respect either to Jews or Chris 
 tians. They have discovered or imagined an immense num 
 ber of septenary groups, in religion, history, art, philosophy, 
 and indeed all branches of human knowledge. We shall not 
 undertake to exhibit even specimens of these. They are col 
 lected to a great extent in a remarkable work, the Sukkerdan, 
 or Sugar Box, as it is called, in which this confectionery of 
 the Arab mind is sorted and stored in innumerable parcels 
 of seven. It is the composition of an African scholar, Ibn 
 Khojle, who wrote in the year of the Hejra 757 ( A - D - I 35fy, 
 and died in the year of the Hcjra 776. We may suspect that 
 in this last date a mistake has been made of one year, and 
 that his life actually passed away in 777. 
 
 Ibn Khojle is not the only writer, nor is he by many centu 
 ries the earliest, who composed an elaborate work on the 
 number seven. Among the voluminous writings of Philo 
 Judaeus Philo, the learned Jew of Alexandria, a contem 
 porary of our Saviour we find a special dissertation, De 
 Scptcnario, on the number seven ; while in another work, on 
 the Mosaic History of the Creation, he dwells at length on the 
 dignity and sacredness of the same number. Nor has the 
 subject been overlooked by Christian writers. The Fathers 
 have frequent allusions to it, though no one of them, so far as 
 we know, has made it the theme of a separate treatise. But 
 in modern times we find the work of Wurffbain which bears 
 the following title : De numero scptcnario variarum lectionum 
 collcctioncm hanc philologicam elaboravit Leonhartt Wurffbain 
 Noribergcnsis Doctor, anno salntis 1630, cetatis suce septies 
 septimo. Const at septeno qnicquid in orbe fuit. Nirinbergce, 
 163$ This philological collection of various readings on the 
 number seven hath Leonhartt Wurffbain, of Nuremberg,* 
 Doctor, elaborated in the year of salvation the i63Oth, of his 
 own age the seven times seventh. Whatsoe er on earth 
 existeth, in a seven it consisteth. Nuremberg, 1633. Within 
 
THE NUMBER SEVEN. 339 
 
 the last few years, WurfTbain has found among his own coun 
 trymen a more distinguished successor. The Wiener Jalir- 
 buchcr for 1848 contains a series of articles on the number 
 seven, the contribution of a veteran Orientalist, whose la 
 mented death occurred about two years ago the Baron Von 
 Hammer-PurgstalL We have placed his name at the head of 
 our own remarks, and take pleasure in acknowledging that we 
 are dependent to a great extent upon his learned labors. If 
 any of our readers find their curiosity athirst for further de 
 tails, we can refer them to his two hundred pages. Only a 
 word of caution. We would advise our friends not to attempt 
 those articles, unless consciously animated by a genuine inte 
 rest in the subject, profoundly impressed by the mystic pre 
 dominance of this venerable number ; otherwise they may 
 find Von Hammer-Purgstall s ocean-flood of septenary erudi 
 tion somewhat too overwhelming for them, and may prefer to 
 take up with the small specimen-phial which we have the 
 honor of exhibiting. 
 
 It remains to say a word in regard to the cause or causes of 
 the honor so early and so widely paid to the number seven. 
 Arithmetical reasons have been assigned for it ; we find them 
 drawn out at length by Philo the Jew. If we take the ten 
 primary numbers that is, the series from I to 10 (leaving out 
 of view the I , which is regarded as the basis of all number, 
 but hardly a number itself) we shall find that some of them 
 are produced by multiplication : as 4 by multiplying 2 and 2 ; 
 6 by 2 and 3 ; 8 by 2 and 4 ; 9 by 3 and 3 ; 10 by 2 and 5. 
 Some again are not themselves produced by multiplication, 
 but by their multiplication produce others of the series : thus 
 2 helps to produce 4 and 6 and 8 and 10 ; 3 helps to produce 
 6 and 9 ; 4 to produce 8, and 5 to produce 10. Thus we 
 have the products of multiplication 4, 6, 8, 9, 10, and the pro 
 ducers 2, 3, 4, 5. Seven alone belongs to neither class; it 
 neither produces nor is produced ; and is thus clearly distin 
 guished from its fellows. Another property : if we start with 
 unity, and go on doubling, we form the geometric series 1,2, 
 4, 8, 16, 32, 64. In this series the seventh term 64 is at the 
 same time an exact square number, and an exact cube number, 
 its square root being 8, and its cube root 4. Now let us start 
 
340 THE NUMBER SEVEN. 
 
 again with unity and go on trebling : we form the series I, 3, 
 9, 27, Si, 243, 729; and here again the seventh term 729 has 
 the same remarkable property of being an exact square and 
 cub e number at the same time ; its square root being 27 and 
 its cube root 9. And so if we form our series by quadrupling 
 or quintupling, or with any other ratio, the seventh term will 
 still have the same property ; which is easily accounted for by 
 our algebra, the seventh term of such a series being the sixth 
 power of the ratio, which is of course the square of the third 
 power and the cube of the second power. It is evident, how 
 ever, that these properties of the number seven will not ex 
 plain the origin of the feeling under consideration : they are 
 very far from being obvious ; they would probably have 
 passed without notice, or at all events without special atten 
 tion, had not the established sacredness of the number set 
 men upon the hunt to find out everything remarkable con 
 nected with it. 
 
 Others rely upon a chronological reason : they derive the 
 veneration for the Seven from the early division of time into 
 periods of seven days that is, from the week. It is certainly 
 probable that the week, if it did not give origin to the feeling, 
 has contributed to give it strength and perpetuity. Even if 
 we regard the week as at first a merely human division of 
 time, suggested by the changes of the moon, though after 
 wards taken up with divine sanctions into the Mosaic economy, 
 still it could hardly fail, when once established, to invest its 
 number with peculiar interest and importance : just as the 
 widely recognized distinction of the number twelve may be 
 ascribed to the twelve months (mooneths), revolutions of the 
 moon, which correspond nearly to a single revolution of the 
 sun. Still more might this effect be looked for, if w r e regard 
 the week as being from the beginning of the world a positive 
 divine institution. But some who take the latter view feel 
 prompted to go further, and to explain why this number 
 should have been selected by the Deity as the number of the 
 week, and thereby as the subject of peculiar dignity and 
 reverence. .Thus Bahr, in his Symbolism of the Mosaic 
 Ritual, observes that the Seven is formed by the union of two 
 symbolic numbers : namely, three, which symbolizes the 
 
THE NUMBER SEVEN. 
 
 341 
 
 divine, since the Godhead is a trinity, and four, which sym 
 bolizes the cosmical, the created universe of space, this being 
 all determined in situation by the four cardinal directions or 
 points of the compass, North, South, East, and West. The 
 Seven, therefore, is in the highest degree symbolic, repre 
 senting the union of the divine and the cosmical, and espe 
 cially representing that reunion of the world with God which 
 is the great aim and crowning consummation of all true re 
 ligion. Kurtz also, another learned and pious theologian of 
 Germany, in the Studicn und Kritiken for 1844, goes into an 
 elaborate vindication of the same view. In like manner, the 
 twelves of the new Jerusalem, which have been already re 
 ferred to, are explained as being the symbolic product of the 
 same symbolic numbers. We have spoken before of arithme 
 tical and chronological reasons : we may describe this as a 
 symbolical reason. Such views will be very differently re 
 ceived by different persons. The example of Bilhr himself 
 shows that minds of a high order can find interest and satis 
 faction in them. At the same time there will always be 
 others, men of positive and critical minds, who will distrust 
 them as wanting an objective basis, or think of them as 
 thin but highly flavored soups, fitted to tickle or to tease 
 the intellectual palate, but affording next to nothing of sub 
 stantial nutriment. 
 
 Next the physiological reasons. It is well known that the 
 importance of the Ten, as the universal numerating number 
 of all languages and peoples- for all men count by tens, and 
 tens of tens, and tens of tens of tens, and so on, not in fig 
 ures only but in words, and not in" words only but in con 
 ceptions : thought and language have always been decimal, 
 though figures have not always been so nobody doubts, we 
 say, that this ascendency of the Ten depends upon a physio 
 logical reason, one which makes it natural and handy for all 
 men to reckon thus ; that our apprehension of number is 
 - more than figurative, it is really a taking hold of it with our 
 ten fingers. The science of number appears to be of all 
 others the least artificial ; yet there is no art, not even the 
 potter s, which shows more clearly the impress of man s 
 hands. Now some would find similar reasons for the pre- 
 
342 THE NUMBER SEVEN. 
 
 eminence of the Seven. Thus we have the seven parts of the 
 human body, the head, chest, and loins, with the four limbs, 
 upper and lower. So, too, we have the seven openings of 
 the head, the three twin pairs of eyes, ears, and nostrils, with 
 the monadic mouth to make the seventh. Further, in many 
 diseases, the seventh, fourteenth, twenty-first days are critical 
 periods, and were so regarded by ancient physicians. To 
 these anatomical and pathological sevens Von Hammer- 
 Purgstall seems inclined to give the foremost place, and to 
 consider them as the most effective agents in creating a 
 reverence for the Seven. To us, we confess, they do not 
 appear very striking. We should ascribe a far earlier and 
 more powerful influence to the astronomical sevens. 
 
 We need not say, that to the men of primitive times the 
 spectacle of the nocturnal heavens was as impressive as it was 
 constant. But in this spectacle there was no object, the 
 moon alone excepted, so striking to the inhabitants of the 
 north temperate zone that is, to all the cultivated nations of 
 antiquity as that group of seven splendid never-setting stars- 
 in which the utilitarian imagination of the Yankee recognizes 
 a dipper. The ancient Greeks saw in it the great northern 
 bear; the Romans, seven plough-oxen, scptcm trioncs ; 
 among Greeks and Romans both, an entire quarter of earth 
 and sky received its name from this constellation. The stars 
 composing it were called by the Persians heft crcng, the 
 seven thrones, seats for the monarchs of the sky. The 
 celestial empire would thus sQem to be, like Anglo-Saxon 
 England, a Heptarchy. But while in this unequaled constel 
 lation the glory of the Seven is most conspicuously blazoned, 
 there are other notable groups which hold forth the same 
 skyey number : as the lesser bear, inferior in brightness to 
 the other, but distinguished as containing that remarkable 
 star, which amid all motions and revolutions of earth and 
 heaven has kept through the ages the same fixed place, the 
 unvarying guide of benighted mortals. Nor must we forget 
 the sweet influences of the Pleiades, " glittering like a swarm 
 of fire-flies tangled in a silver braid;" nor how through 
 scudding drifts the rainy Hyades vexed the dim sea." In a 
 passage of the Iliad, Homer, describing the shield of 
 
THE NUMBER SEVEN. 
 
 Achilles, tells us that the divine artisan represented within 
 its central circle 
 
 the earth, the heaven, the sea ; 
 The sun that rests not, and the moon full-orbed ; 
 There also all the stars, that round about 
 As with a radiant frontlet bind the sky ; 
 The Hyads, and the Pleiads, and the might 
 Of huge Orion, with him Ursa called, 
 Known also by his popular name the Wain. 
 
 It is worthy of notice that, among the constellations here spe 
 cified as most interesting to the Homeric Greeks, three out of 
 the four present the septenary number. 
 
 But w r e have yet to mention the great planetary seven. 
 Most of the heavenly bodies, though revolving daily round the 
 earth, maintain the same position relative to each other. 
 But there are seven which wander without resting through 
 the stationary camp. Among these seven wanderers, or plan 
 ets, are the two greater lights that rule the day and the night, 
 and the two usher stars that herald the morning and the eve 
 ning. Enumerated in the order of their distance from the 
 earth, as determined by the ancient astronomy, they are the 
 Moon, Mercury, Venus, the Sun, Mars, Jupiter, and Saturn. 
 From the earliest times they have been an object of wonder, of 
 curiosity, of study, to star-gazing men. Their movements 
 have been watched with the minutest observation, and every 
 possible device adopted to unravel the mystery of their way 
 ward courses. The feeling toward them is well illustrated by 
 the Persian literature, in which a great variety of honorary 
 titles are applied to them, such as the seven pearls, the seven 
 golden corals, seven eyes of heaven, seven tapers or torches, 
 seven peers, seven sultans, seven great ladies, seven green 
 daughters, seven heart-breaking boys, and so on, and so on. 
 Each of these bodies has his own heaven, or sphere in which 
 he moves about the earth. Hence the idea of seven heavens, 
 which, with the correlative seven hells, we have noticed al 
 ready. For these seven heavens the Persians have an even 
 greater redundancy of titular expressions, such as the seven 
 buildings, the seven temples, the seven roofs, the seyen domes, 
 the seven vaults, the seven blue curtains, the seven watered-col- 
 
344 E NUMBER SEVEN, 
 
 ored sun-screens, the seven castles of gilded enamel, the seven 
 horse-mills or ass-mills (in which the stars go round and round, 
 as the ass in the mill). But we need not multiply illustrations 
 to show how profound, as well as early, were the impressions 
 made by these seven planets upon the minds of men. It is 
 true that science has been making wild work with this ancient 
 and venerable seven. The peerage of England has been more 
 than once menaced with degradation by a large addition of 
 upstart nobles. Something like this has actually happened 
 to the celestial peerage. But let not these parvenu planets 
 Dii minorum gentium whether condemned to outer dark 
 ness like Uranus and Neptune, or huddled together like As- 
 traea, Flora, and fifty more, as if to make up by collective 
 numbers for individual insignificance let them not suppose 
 that they can take rank with the ancient nobility of the skies. 
 There arc prerogatives of the original seven from which they 
 are forever excluded. They can never preside over the re 
 volving week. They can never be the arbiters and expo 
 nents of human destiny. For the seven planets are the great 
 objects, not only of astronomy, but of astrology. They 
 are the lords of life ; by their endless relations to each other 
 and to the fixed stars they determine the endless varieties of 
 human character and fortune. We profess to disbelieve in 
 astrology; our science is against it. Yet still, if anything 
 particularly fortunate befalls us, we bless our stars for it. As 
 Max Piccolomini says in Schiller s play : 
 
 " Still doth the old instinct call back the old names, 
 
 and even at this day 
 Tis Jupiter who brings whate er is great, 
 And Venus who brings everything that s fair." 
 
 At all events, the days of the week still retain their old as 
 trological designations, still own in name the mastery of the 
 planetary seven. We have our Saturn-day, our Sun-day, our 
 Moon-day. And if, instead of having a Mars-day, Mercury- 
 day, Jove day, Venus-day, like the people of France, Italy, 
 and Spain, we speak of Tuis -day, and Woden s-day, and 
 Thor s-day, and Freya s-day, it is a mere translation the 
 
THE NUMBER SEVEN. 345 
 
 translation of Roman names by their supposed representatives 
 in northern mythology. 
 
 And now, who will not admit that the veneration for the 
 Seven is in literal truth a lesson of celestial teaching ? In this 
 respect, as in others, "night unto night showeth knowledge 
 no speech, no language their voice is not heard yet their 
 line is gone out through all the earth, and their words to the 
 end of the world." 
 
DECISIONS OF COLLEGE CLASS DISPUTES. 
 
 XVIII. 
 
 I. ARE THE WRITINGS OF LORD BYRON IMMORAL IN THEIR 
 
 TENDENCY ? 
 
 1847- 
 
 BYRON, it seems to me, was a bad man. Doubtless he 
 had his good impulses who has not ? A mind so vigor 
 ous must have seen the excellence of virtue ; a sensibility so 
 keen must have felt it. But was the will subject to its power ? 
 Did the man live with good and noble aims ? or did he live 
 with vile and selfish aims ; live to gratify pride and passion ? 
 The life and writings of Byron are consistent with each other : 
 they both nlike show a fierce and haughty nature, spurning 
 all restraint, controlled neither by reason nor religion. He 
 never governs himself. lie is swayed like a reed by every 
 gust of passion. It is melancholy to speak thus of a man so 
 highly gifted. But it is the truth. I have read Moore s 
 apology nearly through. It is a book which corrects itself. 
 The writer wishes to convey a favorable impression of Byron. 
 In his descriptions everything is softened, and displayed in 
 the most advantageous light. But unfortunately unfortu 
 nately, I mean, for the subject of the biography ; fortunately 
 for the reader, fortunately for the interests of truth he has 
 not muzzled his charge ; his protege is allowed to speak for 
 himself; and he never speaks without betraying the atra bills 
 that was in him. It is impossible to read that book, as it is 
 impossible to read the writings of Byron, without feeling that 
 the moral nature of the man had become corrupted and dis 
 eased. Would you become convinced of this ? Take another 
 man, and not the highest style of man, either Walter Scott 
 and compare Byron with him. His qualities become more 
 striking when placed side by side with those of a sound and 
 healthy nature. Turn from Walter Scott to Byron, and how 
 
LORD BYRON S WRITINGS. 
 
 347 
 
 great the transition ! It is like passing from summer to 
 winter. The one healthy in feeling and expression, with a 
 sunny genial nature, radiant with charity, kindness, honesty, 
 hearty affection, and loving humor ; a soul at peace with the 
 world, and submitting without a murmur to the laws that 
 govern it. Such is the one. The other is cold and bitter 
 and satirical. According to Moore, his conversation was 
 always spiced with raillery at absent friends ; he had a breast 
 haunted by evil demons, devoured by gloomy passions. He 
 is always kicking against the pricks, dashing himself with im 
 potent spite against the barriers which society, which God 
 himself has placed around him. With all his splendid en 
 dowments, was there ever a man so little to be envied ? 
 
 There may have been writers whose moral qualities did 
 not appear in their writings. Byron was not one of these. 
 He was intensely subjective to use a fashionable term of 
 modern criticism. The man appears in all his works. His 
 personality is everywhere prominent. He is always describ 
 ing himself; or, if describing anything else, stamps himself 
 on that which he describes. His poems of course reflect the 
 immorality of his nature. And this is true of all of them 
 grave or gay, true heroic or mock heroic. Childe Harold is as 
 really immoral as Don Juan. The difference between them 
 is that the one scowls and the other sneers. The one bids a 
 sullen defiance to the laws of moral order ; the other seeks to 
 cover them with a load of blasphemous and indecent ridicule. 
 A poem is not of course moral because it contains no foul 
 words, no expressions but such as could be repeated in a 
 drawing-room. Immorality and indelicacy are different 
 things. Rabelais is indelicate to the last degree, but he is 
 not really immoral. Congreve is far less indelicate, but far 
 more immoral. The book which breathes an immoral spirit 
 is an immoral book. The book which expresses the feelings 
 of an immoral man is an immoral book. The spirit of Byron s 
 poems is destructive. Discontent, skepticism, scorn are 
 written on every page. He is at war with all that is estab 
 lished ; he refuses, at least in spirit, submission to established 
 authority, divine and human. 
 . It is often said that this impatient temper must be forgiven 
 
348 CLASS DECISIONS : 
 
 to a man of his genius. The poet, gifted with a soul larger, 
 deeper, stronger than those of ordinary men, cannot be con 
 fined within the same narrow limits as they. He bursts the 
 barriers which shut in others ; or, if he cannot burst them, 
 chafes angrily and impatiently against them. Was it so with 
 Shakespeare and Milton ? The poet, if he sees farther than 
 common men, more clearly than they, will see the necessity 
 of those restraints which God and man have placed about his 
 path their necessity to himself; for, if he is a genius, he is 
 yet a man, erring and sinful like all his race their necessity 
 to others; and, seeing this, he will beware how he shakes 
 others faith or encourages in them a feeling of discontent and 
 insubordination. And if not, if instead of this he chooses to 
 wage war against the laws of moral order, the grand condi 
 tions of human life, individual and social, then his guilt will 
 be proportionate to those endowments of the Almighty which 
 he has perverted to wicked and unworthy ends. 
 
 To criticize the literary merit of Byron s poems does not 
 come within the scope of this discussion. I am ready to ac 
 knowledge his extraordinary powers though I think that his 
 moral character has had a very injurious effect upon his 
 literary character ; that with the calmness, the clearness, the 
 conscious effort, the lofty purposes of a truly moral nature, 
 he would have soared far higher, and produced works of far 
 greater value even as works of poetic art. 
 
 I do not say that Byron should not be read at all. The 
 man who would investigate the morbid anatomy of the hu 
 man mind will find in Byron s works an admirable study. It 
 may be well on some accounts to know how a haughty and 
 malignant spirit will express itself in good set English. 
 
 My own course of reading was subject to very little guid 
 ance from without, and was at first confined to a small library, 
 to which many of the British classics had not gained admit 
 tance. Under, such circumstances I read much the greater 
 part of Byron s works. I now regret, not that I read him at 
 all, but that much of the time which I gave him was not de 
 voted to other and healthier writers especially to those con 
 temporaries of Byron whom he so much affected to despise, 
 but whom he had so little reason to despise, his equals, his 
 
ANCIENT AND MODERN ELOQUENCE. 349 
 
 superiors even in many intellectual gifts, greatly his superiors 
 in moral excellence Robert Southey and William Words 
 worth. 
 
 2. IS ANCIENT ELOQUENCE SUPERIOR TO MODERN? 
 
 1849- 
 
 TEIERE are two ways of viewing this subject. We might 
 commence with the circumstances and conditions on which 
 eloquence depends, and, by comparing these in ancient and in 
 modern times, judge by a sort of d priori judgment which 
 period should naturally have the higher eloquence. We often 
 see and hear the question discussed in this way. On the 
 other hand, we may consider the actual specimens of ancient 
 and modern eloquence, and by direct comparison determine 
 the superiority of this or that, leaving it for after work to de 
 termine, if possible, the causes and conditions of that superi 
 ority. The form of our question would lead us rather to the 
 latter mode of treatment. But there is a preliminary point to 
 settle : and that is the standard of excellence. What are we 
 to look at is it elegance, is it feeling, is it passion, is it 
 argument, is it logic or rhetoric ; or all combined ? Some 
 have a simple test, and that is persuasiveness ; the best ora 
 tion is the most persuasive, and vice versa, the most persuasive 
 is the best ; for it best fulfils the end of eloquence, which is 
 persuasion. But there is still a difficulty here. The elo 
 quence of Mike Walsh has an effect as persuasive on the col 
 lective blackguardism of New York as the eloquence of Daniel 
 Webster has on the collective dignity and learning of the 
 Senate or the Supreme Court. Shall we therefore decide 
 that the one is no higher than the other ? Now persuasive 
 ness, tendency or adaptation to secure that course of action 
 which the orator desires, is indeed an indispensable element 
 in true eloquence. But there is another element, of great im 
 portance, which may vary much in orations equally persuasive : 
 and that is artistic perfection, aptness to satisfy the aesthetic 
 sense, the critical faculty in man. An oration may be re- 
 
350 CLASS DECISIONS: 
 
 garded as a work of art, and subject more or less to the pre 
 dicates symmetrical, perfect, beautiful, or the reverse. In 
 deed, no cultivated mind can help taking this view. He whcr 
 has taste that is. the critical faculty which judges of form- 
 must be shocked with what is out of taste, must admire that 
 which good taste recommends and sanctions. Now it is pre 
 cisely here that I would take my stand. In everything else 
 in vehemence of passion, in depth of feeling, in cogency of 
 argument, in attractiveness of persuasion I admit that the 
 moderns have shown themselves not inferior to the ancients. 
 But in finish and perfection, in symmetry and beauty of out 
 ward form, I bow to the superiority of the ancient masters. 
 I regard the orations of Demosthenes as not only unsurpassed, 
 but as unrivalled" specimens of oratoric art. And now, having 
 pronounced this opinion, if you ask me why, why this was 
 so, why an old Athenian should have borne away from all 
 the world the prize of art in oratory, I cannot say, I cannot 
 satisfy your curiosity. I do not ascribe it to a genial climate, 
 or a productive soil ; to successful wars with foreign, powers, 
 or noble struggles for national independence ; nor to agricul 
 tural, mechanical, professional, political activity ; nor to de 
 mocratic institutions, to annual magistracies, to vote by ballot 
 or by bean. I do not deny that these things had their in 
 fluence, as furnishing field and opportunity and encourage 
 ment to intellectual development. But that matchless 
 aesthesis of the Greeks, that unrivalled taste, that wonderful 
 sense of beauty and harmony and proportion where that 
 came from is more than I can tell ; I greatly doubt whether it 
 is in any proper sense the creature of external circumstances. 
 I hold, then, that the same causes which made a Homer the 
 first of epic poets, and a Phidias the first of sculptors, and a 
 Pindar the first of lyric bards, and the builder of the Parthe 
 non the first of architects the same causes rendered Demos 
 thenes the first of orators. Only let circumstances offer fair 
 occasion, and the nation which has the keenest and truest 
 sense of the beautiful and the perfect will present the highest 
 master-works of art. 
 
 It fills me with wonder to see how completely some of 
 those old Greeks were under the dominion of their taste, and 
 
ENGLISH OR THO GRAPHY. 3 5 1 
 
 how much they must have sacrificed to it. Every oration of 
 Demosthenes is to my mind a heroic exhibition of self-denial. 
 So many graces of style and felicities of expression, so many 
 details sublime, pathetic, brilliant must have occurred to his 
 rich and copious mind, which ever and anon shows when such 
 things are necessary that there was no lack of power to ori 
 ginate them and yet, in general he rigorously rejects them, 
 and, if he ever employs them, employs them most sparingly. 
 Everywhere he sacrifices the part to the whole, beauty of de 
 tail to general effect : and so he holds on in his severe sim 
 plicity, rarely allowing any embellishment, just as an archi 
 tect decorates only the capitals of his columns or the cornice 
 of his roof, trusting to the grand and stately proportions of 
 the completed edifice. How different all this from the gene 
 ral character of modern oratory, and above all of American 
 oratory, I need not say. The introduction to one of Colonel 
 Benton s speeches is longer than the three Olynthiacs put to 
 gether, and contains, I dare say, more figures of speech and 
 more grandiloquent phrases, and questionless more egotism 
 than all three ; and yet the world of cultivated, educated men 
 would hardly consent to barter one Olynthiac oration for all 
 the windy declamations of the Colonel, from the fight with 
 General Jackson to the present time. 
 
 3. IS A REFORM DESIRABLE IN THE METHOD OF WRITING 
 THE ENGLISH LANGUAGE ? 
 
 1852. 
 
 THE Chinese more than any other people have need of a 
 phonographic revolution. With less than five hundred differ 
 ent syllables in their language, they have an immense num 
 ber of written characters forty thousand, it is said, in all, 
 though not more than ten thousand are in general use. 
 These characters are most of them very complicated, and re 
 quiring, as they do, to be written with perfect accuracy in 
 order to avoid confusion between such as are of similar form, 
 
352 
 
 CLASS DECISIONS : 
 
 they make it a most arduous work to master them, and it is 
 literally true that multitudes in China are all their lives long 
 learning to read and write. The case is not so bad with us ; 
 there is a general relation between the alphabetic signs we 
 use and the vocal sounds for which we use them. But it is a 
 relation so variable, so irregular, so capricious, so subject to 
 manifold and endless exceptions, as to impose grievous diffi 
 culties on a learner. It does not matter so much with us who 
 are brought up under it ; line upon line, lesson upon lesson, w r e 
 get hold of the system, if that name can be given to anything 
 so unsystematic ; through much weariness, through many 
 blunders, through some floggings, it may be, the successful 
 boy at length learns to spell fails now and then on a hard 
 word, such as phthisic; has to consult his Dictionary now and 
 then, to find how many^ s there are in waggon, or whether 
 sibyl is spelt sibyl or sybil but on the whole is entitled, as 
 things human go, with some degree of unavoidable imperfec 
 tion, to the dignity of Master of Spelling. That is the suc 
 cessful boy ; but all boys are not successful, nor all young 
 men either as college compositions not unfrequently attest. 
 I know a man of talents and of literary cultivation who cannot 
 learn to spell the words which come from the Latin cede, 
 some of which have cccd, and others cede such as accede and 
 exceed, precede and proceed, secede and succeed, etc. ; he never 
 can tell which is which. And I have heard several well-edu 
 cated persons say that there are particular words or classes of 
 words which always puzzle them. But, as I was saying, it is 
 of less consequence to us who are born and bred under this 
 sort of spelling ; the years which we spend upon it are of no 
 great value to us ; we should not accomplish very much be 
 side except that one might learn two or three decently- 
 spelled languages while he is learning to spell his own ; and, 
 if he did not, how much pleasanter to be playing long ball or 
 ranging through the woods ! We get a certain discipline, no 
 doubt, from our drudgery ; only would it not be better to get 
 our discipline in mastering some useful thing, when there are 
 so many useful things to be mastered ? But with us it is less 
 matter ; it is for foreigners that my sympathies are most 
 strongly enlisted, for men who in adult years set themselves 
 
ENGLISH ORTHOGRAPHY. 353 
 
 to learning English. In itself our language is a very easy 
 one. Its sounds are not difficult, except the two t/i s, which 
 are easy for a modern Greek, but hard to all the other Euro 
 pean nations ; but in general they are easy, though some 
 times rendered difficult by harsh accumulations as in Jiosts, 
 asked. Our system of declension and conjugation is an easy 
 one, as the inflexions are reduced very nearly to a minimum. 
 Nor has the language very much of troublesome idiom, like* 
 French and Latin ; there is a twist about will and sliall in the 
 future, which proved fatal, it is said, to a drowning French 
 man, whose agonizing cry, "I will be drowned ; nobody shall 
 help me," was rather calculated to repel than invite assistance 
 but this distinction is ill-observed by Scotch and Irish, and 
 by Americans, too, out of New England, and seems to be 
 gradually fading out from the language. I am convinced 
 (without particulars) that, apart from its orthography, the 
 English is easier of acquisition than any other idiom of Europe. 
 But the orthography is a very serious obstacle ; every for 
 eigner who has learned English in adult years will speak feel 
 ingly of his trouble and occasional despair at its numberless 
 and capricious exceptions and irregularities. Perhaps it 
 would not be going too far to say that a uniform orthography, 
 with notation of the accent, would lighten by half the labor of 
 learning English. And many foreigners are studying Eng 
 lish now-a-days ; the astonishing development of the English 
 race is carrying their language far and wide in every conti 
 nent. English is now the language of North America, with 
 its rapidly increasing millions ; it will soon be the language of 
 millions in the ancient peninsula of India ; it will soon, if it 
 . does not now, carry a man farther over the world, bring him 
 into connection with greater multitudes and wider regions, 
 than any other tongue. This commanding and ever-growing 
 importance of the language is making it more and more the 
 object of study among foreigners, who might not have been 
 attracted by the treasures of a literature which I do not 
 hesitate to pronounce the noblest of modern Europe. But 
 when you put all things together, its noble literature, its 
 political importance, its simplicity of structure, you will 
 see that, if its acquisition were facilitated by a uniform 
 23 
 
354 CLASS DECISIONS: 
 
 orthography, it might almost come to be ere long a universal 
 language. 
 
 It cannot be denied that the English language is shocking 
 ly spelled. The original difficulty lay in the mixture of dif 
 ferent languages, Saxon and Norman French, out of which 
 came English ; the confusion of different systems, varying 
 and conflicting analogies, is everywhere to be seen in our 
 .^orthography. 
 
 Besides, the French, which makes one element of 
 English, does itself enjoy, next to the English and perhaps 
 the Gaelic, the honor of being the worst spelt language of 
 Europe Franklin used to say that what we call false spelling 
 of the vulgar was really true spelling. I do not know that I 
 should say that, for vulgar spelling is sometimes most ingeni 
 ously absurd. But I certainly feel a good deal of hesitation 
 about saying in regard to any man that he spells badly : I 
 say that he does not spell like most of us ; he spells singu 
 larly, peculiarly ; but I do not see, on the whole, that he 
 spells worse than the spelling-books and newspapers. 
 
 It is very unfortunate that Johnson s Dictionary should 
 have come to be such a standard of spelling. For the con 
 sequence has been that the processes which were going on 
 before gradual progressive processes, to root out anomalies 
 and bring in greater regularity, processes which went on 
 naturally and almost without notice were at once arrested, 
 and the system which had before been somewhat flexible be 
 came at once a cast-iron affair. That such processes were 
 going on, to the great advantage of the system, will be plain 
 enough to any one who takes a book printed in the 
 Elizabethan age say one of the earlier translations of the 
 Bible and compares it with the books printed in the first 
 part of the last century. That such processes were arrested 
 by the appearance of Johnson s Dictionary is evident from 
 the outcry which is raised against any spelling that departs 
 from the prescription of that autocratic lexicographer. No 
 philologist needs to be reminded that Johnson had little fit 
 ness for the work of legislation in orthography. This would 
 be evident enough from his famous dictum that it is absurd 
 to regulate your spelling by your pronunciation, for pronun- 
 
ENGLISH ORTHOGRAPHY. 355 
 
 elation changes all the time, and your standard is therefore 
 variable and fluctuating. He did not see that this is one of 
 the strongest reasons for regulating spelling by pronunciation ; 
 for if pronunciation changes all the time while spelling re 
 mains fixed, the two will diverge more and more widely from 
 each other, until they cease to have any relation, and we shall 
 write in hieroglyphics. Most evidently the proper aim and 
 object, the ideal of alphabetic writing, is to furnish an exact 
 reflexion of the spoken language, a faithful representation of 
 what we hear in daily utterance. In its most advanced per 
 fection, every elementary sound will be represented by a 
 special character, and each character will be used in every 
 case to represent the same sound. There is no great objec 
 tion, however, to a combination of characters used to repre 
 sent a sound different from either as, for instance, ch in 
 church, provided always that it is used with perfect consistency. 
 The reform recently attempted has taken high ground, avoid 
 ing such combinations of characters, and representing the 
 same sound always by the same alphabetic sign. Perhaps 
 this is the best course, though I have always felt that the in 
 troduction of new letters, which this plan requires, might op 
 erate pretty strongly to prevent its adoption. 
 
 The objections commonly urged against a new system ot 
 phonography have in my view very little weight. It is often 
 said that, if this plan were adopted, all books printed hitherto 
 would be useless. It is certain that people would not read 
 them quite so readily as now ; but only for this reason, that 
 they would spend less time in acquiring the power. It would 
 be really as easy as ever for people to learn the old system : 
 or rather, far easier ; for then it would be necessary only that 
 they should learn to read, to recognize the words when they 
 see them ; and not to learn what is far harder, to spell, to 
 reproduce the words when you do not see them. Another 
 objection, which has considerable influence, is that a new 
 system would obscure the etymology of words, which is now 
 shown in many cases by the spelling. But as regards this, 
 the etymology of words is of little practical value except to 
 scholars, who could always get it out of books of lexicography ; 
 it is not worth while for their benefit to impose a heavy bur- 
 
356 CLASS DECISIONS : 
 
 den upon the world at large. But our common spelling is 
 often an untrustworthy guide to etymology. Take the word 
 .sovereign : the people who first spelt it so supposed no doubt 
 that it had something to do with reign / but it most certainly 
 has not. It comes from Latin super, through Italian sovrano, 
 etc. But I will go further, and say that the wants of the 
 philologist require a different system. What is important for 
 him is that he should know the condition of a language at any 
 given period of the past, that he may be able to trace it 
 through its successive changes to its latest form. Now in 
 doing this he must depend mainly on the spelling, the writ 
 ing ; if this be maintained invariable from age to age ami-d all 
 mutations of spoken words, the philologist is deprived of his 
 most serviceable guide. I would give a good deal to get a 
 Fonctic Nuz of Chaucer s time, that I might know how far 
 some important phenomena of the modern language as for 
 instance the change of a, to e, of e to ?, and of \ to ai had es 
 tablished themselves five centuries ago. 
 
 There are many points of which I should be glad to speak 
 more at length, but I must stop. You will see from what I 
 have said that I recognize fully the evils of our present 
 orthography (as men sarcastically term it), and that I sym 
 pathize in the objects of a phonographic revolution. But in 
 regard to the feasibility of such a revolution I am far from 
 being sanguine ; a political revolution, I suspect, would be 
 a much easier undertaking. Yet I have no desire to damp 
 the ardor of those who are more sanguine than myself; on 
 the contrary, I wish them all success in their work, being 
 sure at least of this that, whatever imperfections may belong 
 to their systems, they cannot be so bad as ordinary good 
 spelling. 
 
 4. WAS CIVIL LIBERTY IN EUROPE PROMOTED BY THE 
 CAREER OF NAPOLEON ? 
 
 1852. 
 
 THE selfishness of Bonaparte is now as universally admitted 
 as his transcendent abilities. There was an English officer 
 
NAPOLEON AND EUROPEAN LIBERTY. 357 
 
 who wrote a book, some years ago, to prove that Bonaparte 
 was a bungler in war, that he violated every rule of general 
 ship, and won all his great battles by mistake, A preposter 
 ous attempt as though the voice of a pettifogging critic could 
 drown the paean of praise that rises to Napoleon from twenty 
 glorious battlefields ; as though Jena and Austerlitz, Boro 
 dino and Eylau were not stronger arguments than any pos 
 sible accumulation of professional technics ! Equally pre 
 posterous would it be to dispute the selfishness of Bonaparte, 
 to pretend that he cared for anybody but himself, and, most 
 of all, to maintain that he cared a straw for civil liberty. His 
 government was the government of the strong arm. It was 
 of espionage, of repression, of censorship, of anything but 
 civil liberty. If Bonaparte did anything to help the cause of 
 freedom, he must have done it undesignedly, by accident or 
 by mistake. This is possible, certainly. A man may work 
 toward ends which he has no idea of. The old Greeks, when 
 they planted their colonies on all the coasts of the Mediter 
 ranean, and under Alexander and his successors established 
 their civilization and their language in wide regions of Europe 
 and Asia, had no thought of Christianity, and the progress 
 by which it would one day supplant the gods of Hellas ; yet 
 they were contributing all the time to this great progress of 
 Christianity. In the same unconscious way may Bonaparte 
 have contributed to the advance of civil liberty. 
 
 Look then to his own country, to France, the special 
 theatre of an activity which extended over the world. What 
 did he do for civil liberty in France ? Did he limit or destroy 
 the absolute authority of the throne ? He established a 
 throne as absolute as the most despotic court of the Bour 
 bons. Did he erect the fabric of a free representative govern 
 ment ? He would as soon have shattered the throne on which 
 he sat, and blown up the palace in which he held his court. 
 Did he break the bonds of an antiquated and oppressive feu 
 dalism ? He found them broken by the Revolution, and it 
 would have required a hand stronger even than his to fasten 
 them again. He promulgated a body of law, the famous 
 Code Napoleon. But a code of law is intended to secure the 
 people in person and in property, not to secure them in the 
 
358 CLASS DECISIONS: 
 
 rights and privileges of citizens. It is designed to protect 
 social rights, not civil. The Roman Law attained its highest 
 perfection in the time of its emperors, when civil liberty was 
 at an end. Its latest and most finished form appears in the 
 Justinian Code ; yet nobody speaks of Justinian as having 
 promoted civil liberty, in Europe or elsewhere. It may be 
 said that a good code, giving as it does personal security, will 
 lead men to aspire further after political privilege. But the 
 experience of the world is rather against this. Men who are 
 personally well off, secure, easy, comfortable in their circum 
 stances, are apt to be content with this, and to submit to the 
 existing government, however absolute it may be. It was 
 physical suffering, hardship, and starvation which made the 
 Revolution in France, not the aspiration after citizenship. 
 
 But let us look beyond France. The Germans in general 
 speak very bitterly of Napoleon. But Godfrey Hermann, the 
 famous scholar, used to say that it was a good thing to have 
 a thorough shaking up now and then. This is the great benefit 
 which Bonaparte conferred on Europe : he gave it a thorough 
 shaking up. His actions were too great to be hushed up 
 among princes and diplomatists, to be concealed from the 
 common people ; they penetrated every class of society, and 
 roused the popular mind from its slumbers. They set men 
 thinking : and it was not strange that, among other things, 
 they should think on civil liberty ; that they should raise a 
 thousand questions as to their actual and their proper condi 
 tion which had never occurred to them before. Men saw 
 common men, I mean they saw that the government placed 
 over them was not an immutable fixity, not an inevitable 
 necessity ; when Napoleon appeared, it was neither fixed nor 
 necessary it was obliged to change and yield according to 
 his direction. Might it not feel the influence of some other 
 force as (say) the declared will of the people ? The same 
 lessons, to be sure, had been taught by the American Revolu 
 tion. But America was far away beyond the ocean, and there 
 were multitudes in Europe who had scarcely heard the name. 
 The same lessons had been taught by the French Revolution 
 a very great event, the greatest since the Reformation, and one 
 that could not fail to make a deep impression on the general 
 
NAPOLEON AND EUROPEAN LIBERTY. 359 
 
 mind of Europe ; and it is not easy always to distinguish be 
 tween the agitation which Napoleon effected, and that which 
 would have been produced without him by the Revolution ot 
 which he was himself in some sort the creature. But this we 
 may say, that whatever lessons were taught by the French 
 Revolution, it was Napoleon who by his wars and conquests 
 carried them to every door in Europe. 
 
 In Germany the influence, the indirect influence, of Napo 
 leon went further than this. The legitimate sovereigns of 
 Germany, defending themselves against the Conqueror, found 
 it convenient to make their appeal to a sentiment of German 
 patriotism. Napoleon had said, there are no Germans ; there 
 are Hessians, Saxons, Bavarians, but there are no Germans. 
 He found in his own experience that there were Germans. 
 They rose against him in their anger and their might, and 
 thrust him out of their land, their own German fatherland. 
 This Germanic unity is a strange sort of thing. The people 
 are one in their language, their physical peculiarities, their 
 general type of character, their traditions, and their feelings. 
 The sentiment of their oneness is very strong among them ; 
 it is a positive enthusiasm ; and they are capable of doing 
 great things under its influence, as witness their heroic resist 
 ance to Napoleon in 1813. But when it comes to regular, 
 systematic political action, the oneness vanishes, the old di 
 visions reappear ; and, as Napoleon said, we have only Hes 
 sians, etc. , no longer any Germans. These things were very 
 plain in the movements of the year 1848 ; there was a great 
 deal of talk about a united Germany ; and there was war with 
 Denmark, in order to wrest from her Holstein, which is Ger 
 man, and attach it to this united Germany ; but though 
 Prussia led the attack, Denmark, little as she is, opposed a 
 stout resistance, and has at last got her own again. But this 
 German unity, though it could produce an aggressive war, 
 could not keep the Frankfort parliament together, or soothe 
 the jealousies of thirty petty states. And the case is much 
 the same with Italy. 
 
 But, imperfect and defective as it is, this feeling is very 
 strong in Germany, and must some day work out great results 
 in the history of that country. It received a very great start 
 
360 CLASS DECISIONS : 
 
 in the Napoleon times ; and whatever it may accomplish in 
 the future will be due in some measure to Napoleon. Small 
 thanks to him for it. 
 
 But Germany went further still. The common activity of 
 the people, their efforts, contributions, sacrifices for their 
 country, awoke in them the consciousness of their own manly 
 worth, a consciousness of political capacities and energies 
 which they had never felt before ; and they began to feel the 
 honorable yearning after citizenship. They had not the po 
 litical experience and education necessary for the enlarged and 
 thoroughly successful exercise of civic rights ; but they were 
 not sensible of their deficiencies ; and, if they had been, 
 might have felt that to supply them, the best way was to be 
 gin at once under some system of free institutions. And be 
 sides, their demands were not extravagant. They did not 
 propose to overthrow their monarchical governments ; they 
 desired only constitutions, with provision for popular repre 
 sentation. This the German princes were ready enough to 
 promise in the hour of their extremity ; and ready enough to 
 forget their promise in the hour of their restored security. 
 But they could not make their people forget it : again and 
 again have they been reminded of their promise, until the 
 King of Prussia, reluctant as he was, felt himself at last 
 obliged to yield and establish a system of popular represent 
 ation. It was greatly qualified, indeed, and restricted ; still, 
 it was something ; it conceded at least the principle ; and it 
 might serve as an entering wedge to open the way by slow 
 and sure degrees to greater things. The spirit which extorted 
 this concession is not dead ; it is not confined to .Prussia. It 
 must make itself felt more and more in the politics of Ger 
 many ; and as it gained its first strength in the Napoleon times, 
 so whatever it may do hereafter will be due in some measure 
 to Napoleon. And again, small thanks to him for it. 
 
 While I hold, then, that Napoleon did something to pro 
 mote the advance of civil liberty Jn Europe, my great sur 
 prise would be, that one could hold his place and do so little 
 for it. The truth is that progress in civil liberty is the des 
 tiny of Europe ; it belongs to the inevitable, irresistible tend 
 ency of things ; no human power can arrest it ; and even its 
 
REPUBLICANISM IN EUROPE. 361 
 
 greatest enemies cannot in spite of themselves help doing 
 something for its furtherance. 
 
 5. IS EUROPE TENDING TO REPUBLICANISM? 
 1852. 
 
 I REGARD this question as one not of mere curiosity, but of 
 real practical importance ; important not only to the nations 
 immediately interested, but to us also, who are fellow-actors 
 with them in the great historic drama. No man of intelli 
 gence and education ought to hold himself apart no man of 
 enlarged mental activity and moral earnestness can hold him 
 self aloof from the great world- movements of his time. He 
 must take his attitude in regard to them ; he must have his 
 sympathies ; he must be prepared to approve the right and to 
 condemn the wrong, to uphold the beneficent and to oppose 
 the harmful : and in order to this, he must have his reasons, 
 intelligible to himself and statable to others, his theory of 
 things, his conception of the world as it is, its condition, pur 
 port, and tendencies. Public opinion is only the general re 
 sultant of such individual conceptions ; it is the grand sum of 
 such private opinions : and if the great aggregate is an 
 immense force in the world, for good or for evil, it is the duty 
 of every man to look carefully to that particular element of it 
 which he himself contributes. Nobody, certainly, can doubt 
 that our government has and will have an important influence 
 in shaping the world as it is to be : what this influence shall 
 be, is a thing to be determined ultimately by the views and 
 feelings of the individual citizens. And therefore it is that I call 
 this a question of real practical importance to every one of us. 
 
 The writings of Thomas Carlyle contain a theory of 
 government such as it is, undeveloped, unsystematic, yet 
 still a theory of government which I wish to speak of, be 
 cause Carlyle exerts a powerful influence on the public mind, 
 and especially on the mind of young men ; an influence which 
 I have felt powerfully myself, and see frequent traces of in 
 others ; an influence in many respects most valuable, but all 
 the more needing correction where it is wrong. Carlyle holds 
 
362 CLASS DECISIONS: 
 
 that the great object in government is that the people should 
 be well governed, good laws passed with good sanctions, 
 good roads and bridges made, good schools established, good 
 connections formed with foreign countries, good measures 
 taken for war and for peace in short, that things should be 
 w r ell done for the people, that they should be thoroughly and 
 effectually governed. Now this requires that the best man 
 should be made governor ; the king, the true king, according 
 to Carlyle s derivation, is the canning or able man, the 
 * cunning or knowing man ; he it is whom the Saxons called 
 cyniug, or king ; he is dc jure the king, the sovereign of his 
 people. Gifted with faculty to discern what is best for the 
 people, and with commanding energy to execute what he dis 
 cerns to be best, it is his right to rule ; it is the duty of others 
 to obey ; and the more implicit and submissive the obedience, 
 the better. This is the political creed of a man who deifies 
 power, and above all power of will ; who does not scruple to 
 say that even in this world might is right in the long run a 
 principle which has its relation to truth, but is as dangerous 
 as its converse, that right is might in the long run, is always 
 noble and elevating. That Carlyle should deride republican 
 ism, and especially democracy, is a natural consequence of 
 this way of thinking. But in his scheme there are two fatal 
 deficiencies ; he provides no answer for these two momentous 
 questions: i. How we are to ascertain this canning cunning 
 man who is to be our king if without previous trial, by what 
 criteria ; if with trial, by what system that shall afford fair 
 field for all competitors ; 2. How we are to shift in the mean 
 time ; how we are to get along, if he does not exist, until 
 he does ; or, if he does exist unknown to us, how we are to 
 get along until we know him. 
 
 But the theory of Carlyle, like all absolutist theories, has 
 another, deeper and more fatal deficiency. It assumes that 
 the good of a government lies only in its ends its acts, 
 laws, measures and so on, and that when these are given 
 it matters little by what means they come about ; only, 
 the simpler and more effectual the means, the better. You 
 have only to consider the grist ; if that is good, the way 
 of grinding is a matter of but little consequence. But gov- 
 
REPUBLICANISM IN E UR OPE. 363 
 
 ernment in its ideal perfection is valuable not only for its 
 ends, the acts of government, but for its means also, the 
 system of agencies and operations by which these ends are 
 reached, these acts performed. One great design of govern 
 ment should.be to serve as a grand educational establish 
 ment for the people, to call out, exercise, and train their 
 mental activities. Of course, the highest earthly end of man s 
 existence is the full harmonious development of all his facul 
 ties. For this end he should use every means that earth and 
 man and human society can furnish him. Such means he finds 
 in his own private life, its interests and cares and duties ; they 
 are making continual demands upon his energies, and giving 
 them the strength and discipline that come of constant exer 
 cise. But the community in which he lives furnishes a wider 
 field ; the man who can take in its interests and provide for 
 them, feel its cares, discharge its obligations, has a higher 
 and more valuable cultivation ; he becomes a stronger, wiser, 
 better man. Now go as far as you can go in this way let a 
 man feel that his country with its millions rests in part on him, 
 that he has a substantive influence on its affairs, that he must 
 throw that influence on the right side, that he must gain the 
 information and insight necessary for this purpose and see 
 how far you have widened his sphere of thought and action, 
 and what a powerful stimulus you have given him to mental 
 exertion. This, now, is an object never to be lost sight of, an 
 object of the highest moment, which is practically disregarded 
 in every absolute government. An absolute government 
 may provide as well as a republic for the material wants and 
 interests of the people ; but the absolute government holds 
 them in tutelage, and calls them and keeps them children ; 
 the republic holds them in partnership, treats them as men, 
 and makes them men. If I am a democrat, it is on this 
 ground ; not because a democracy is the safest, strongest, 
 quickest government, but because it is the manliest it is a 
 gymnasium for making men, strong-minded, vigorous, ac 
 tive men. This gymnastic training has its dangers, its pecu 
 liar difficulties and embarrassments ; but the advantages are 
 more than worth the risks. It must be a gradual training, 
 which does not at any time impose upon a people more than 
 
364 CLASS DECISIONS: 
 
 they can do, which goes on slowly from point to point, in 
 creasing the extent and difficulty of its exercises as the people 
 advance in aptitude and strength. 
 
 As popular self-government exercises the mental activities 
 of a people, so these activities, if general and strong, require 
 self-government as an arena for their development and train 
 ing. They may indeed seek and find for themselves some 
 other vent as, among the Spaniards of the i6th century, in 
 wandering, adventure, and discovery ; or as, among the 
 Germans of the present century, in an unprecedented cultiva 
 tion of literature and science. Yet even here among the Ger 
 mans we find this mental activity bursting the bounds within 
 which despotic government would confine it, overflowing into 
 politics, and working slowly and heavily but powerfully to 
 ward the republic of the future. You may find thus great 
 mental activity under a despotism ; and it may seem peace 
 able enough for a time ; but there is always clanger ; it is a 
 subterranean fire, which may become too fierce and strong 
 for the vent that is left open, and so open for itself a vent by 
 some great shock of earthquake that shall shatter the political 
 edifice to its foundations. It was the mental activity that il 
 luminated France with wit and science under Louis XV. 
 which shook down the throne of his successor Louis XVI. , 
 and buried the monarch in its ruins. It was the mental activity 
 of the Elizabethan age that broke out in the age of Charles 
 into open resistance, and at length into successful rebellion. If 
 we had the early history of Greece, we should probably find 
 that the mental activity which produced the Homeric Epos in 
 an age of monarchy was soon after occupied in overthrowing 
 monarchy and setting up republican governments throughout 
 the country. You may think of the Augustan age .at Rome 
 as a case that makes against me. But we have learned to re 
 gard the Augustan age as a period of outward show but real 
 decline ; its culture was not the result of a genuine popular 
 progress, an original movement of the national mind, but of 
 an imitative fashionable dilcttantei?m. 
 
 If we ask where it is that monarchy has been permanent 
 from age to age, from the beginning until now, a thing undis 
 puted and indisputable the answer must be, In the East, 
 
REPUBLICANISM IN EUROPE. 365 
 
 where human life has a stereotype character, where thought 
 and feeling and action stiffen in their moulds and grow more 
 and more solid and unyielding. That the permanence of 
 monarchy in the East is connected with this want of free 
 spontaneous motion in the Oriental mind seems to me quite 
 certain, though I would not say that this is a full explanation 
 of the fact. There is a mystery about it which has not yet 
 been penetrated that monarchy should be so universal and 
 indefeasible in the East, while in the West it has been so 
 fluxing and unstable. But in the East monarchy is monarchy, 
 worthy of the name ; it is the absolute supremacy of a single 
 will. That will may be subject to a hundred influences, of 
 court or camp, or priests or harem ; but such influences are 
 irregular and accidental ; they do not belong to the system, 
 they impose no constitutional restriction. The enthronement 
 of a single will as the all-controlling centre of political life and 
 action that is the one political idea of the Eastern world. 
 This absolutism is in fact essential to monarchy as a perma 
 nent system. If a constitutional restriction is interposed, it 
 proves that the king is a king no longer ; that a power has 
 arisen stronger than the king, able to limit and restrain him 
 the power that brought in the restriction. And in the natu 
 ral course of things this power goes on from more to more 
 till it becomes predominant in the state. In most instances, 
 this first encroaching and restricting power has been an aris 
 tocracy. The Homeric king was supplanted by his council of 
 chiefs. The old Roman king was supplanted by the Senate 
 of the Patricians, who reigned more despotically than the kings 
 themselves. It was the nobles of England who extorted 
 their great charter of privileges from the feeble John. 
 
 The nobles of France were marching on to the same ascend 
 ency when the monarchs learned the art of pitting them against 
 the communes, so as to exhaust them both, and establish that 
 absolutism which rose to its highest glory under Louis XIV. 
 If absolutism is to be overthrown in Russia, it will not be by 
 the people, who are devotedly loyal to the Czar, but by the 
 nobles, who have long been and still are disaffected and turbu 
 lent, ever ready to engage in conspiracy and rebellion. 
 
 The gradual decline of royal power from steady progres- 
 
366 CLASS DECISIONS: 
 
 sivc encroachment is best seen in the political history of Eng 
 land. We cannot here trace its progress ; but we must in 
 quire into the result : What is the throne in England ? It is 
 a political nullity, it is a mere titular dignity, it is a name, the 
 shadow of a name stat nominis umbra it possesses no real 
 authority, no substantial strength. It is like an old oak from 
 which the heart has rotted out, which stands a mere shell, 
 with the same form, but not the same strength which it had 
 of old. It stands while the air is calm, but it will go down 
 before the first great storm ; and so will the English mon 
 archy. The only sure support of a monarchy is loyalty 
 among its subjects ; this is sufficient, and this is essential. 
 But loyalty must itself have a foundation. What foundation 
 has English loyalty ? Loyalty may be founded on the person 
 al qualities of the sovereign : on his superiority, real or sup 
 posed, in strength or courage or talents or character. Doubt 
 less the English sovereigns at present arc worthy and respect 
 able people ; but such people are common in England ; 
 something more than this is necessary. And even this may 
 not continue ; it is little more than twenty years since George 
 IV., a worthless debauchee, vacated the throne. Loyalty 
 may be founded on the possession of power by the sovereign ; 
 and this is most commonly its main foundation. The despots 
 of the East have at least this claim on the devotion of their 
 subjects ; it is reason enough for being loyal to a man if he 
 can cut your head off when he pleases. But the king of Eng 
 land, or the queen of England, is powerless, and is coming 
 to be more and more recognized as such. You may say that 
 the king is still head of the State, and that this is a sufficient 
 basis for loyal feeling ; certainly, if he were really so, and not 
 a mere ornamented figure-head on the ship of state. The 
 real head is the real sovereign ; and that is not the king, it is 
 the House of Commons, or the constituency that elect the 
 House. What then are the props of English loyalty ? They 
 are tradition and habit. The queen, if not powerful herself, 
 is descended from powerful Tudors and Plantagenets, and the 
 feeling that belonged to them has come down to her, a tradi 
 tion and a habit. Evidently this is an insecure foundation ; 
 less so in England than anywhere else, owing to the peculiar 
 
REPUBLICANISM IN EUROPE. 367 
 
 tenacity of the people, their attachment to established forms, 
 their submission to prescription and precedent. Still, it is in 
 secure, as anything must be which is founded on the past and 
 has no root in the present. From the very constitution of 
 our nature, the present must in the end prevail over the past ; 
 it is always with us, and cannot be overlooked or forgotten. 
 This English loyalty, resting as it does on tradition, could 
 never be transferred to any new dynasty. If any collision 
 should overthrow the present reigning house, no new one 
 could be established in its stead ; the idea of setting up Wel 
 lington as king or Cobden as king in England would be 
 simply ridiculous. Nor is such a collision strongly improba 
 ble. Suppose that an ambitious prince should arise, who 
 should feel the degradation of his situation, and attempt to 
 change his nominal kingship into a real one. The result would 
 be a revolution, in which the princely aspirant must fall, as 
 fatally perhaps as Charles I., and with no hope of restoration 
 for his family. It would be folly to undertake to say when 
 monarchy shall cease in England ; but the result in my view 
 is inevitable, and probably not very far distant. English 
 royalty has come to a minimum, and cannot be reduced with 
 out being destroyed. The situation of the country, the dis 
 tress of the laboring classes, the decline of manufacturing and 
 commercial industry, must sooner or later bring about a 
 crisis. The abolition of royalty, however, will involve no 
 essential change of government, for in substance the govern 
 ment is already republican. 
 
 France is less fortunate. When monarchy was swept away 
 there w r as nothing to replace it ; no republican institutions had 
 existed before ; and it was not possible and never will be 
 possible to build them up in a day, so to build them up as 
 to make them strong and permanent. This, however, is cer 
 tain : that loyalty as a pervasive national sentiment is dead ; 
 and therefore no substantial monarchy can be set up in 
 France neither Bourbons, Orleaners, nor Bonapartes can be 
 for any long time sovereigns of France. 
 
 Italy is notoriously without loyalty. Every throne in the 
 peninsula rests on a basis of force, and is essentially a tyranny. 
 Spain has more of loyalty ; Germany perhaps still more. 
 
368 CLASS DECISIONS: 
 
 Yet any one who should compare this century with the last, 
 this generation with the last, would see an immense and most 
 significant change in Germany a change which is sufficiently 
 evinced in the general demand of the German states for con 
 stitutional governments. 
 
 The case is different with Russia : her people, half ori 
 ental in their character, have a perfectly oriental devotion to 
 the person of their emperor. The throne in Russia is secure, 
 to all human appearance, for a long time to come ; and it is 
 the only secure throne in Europe. I used to think differently 
 in regard to this whole subject. I wrote a disputation when 
 in college to show that the progress of Europe was on the 
 whole anti-republican. But the evidence is too strong for 
 me ; I have been forced to acknowledge and recant my errors ; 
 I am coming to feel more and more that the present is an 
 age of transition ; the old order breaking up, the new one 
 not yet established. You may compare it with the two cen 
 turies of Grecian history between established monarchy and 
 established republicanism. Centuries of strife and tumult and 
 revolution and usurpation, the period of tyrants. It would 
 not be strange if this remarkable feature of ancient Greek 
 history should reappear in Modern Europe. France, indeed, 
 has already seen two tyrants, in the Greek sense of the word ; 
 men who raised themselves by military force to an unconsti 
 tutional sovereignty. In this sense, Napoleon Bonaparte and 
 Louis Napoleon are of the same class with Periander of 
 Corinth and Pisistratus of Athens. Such governments, 
 founded on force and not on law, cannot be permanent ; they 
 can last at most only one or two generations, and will at 
 length give place to a well-established republicanism* The 
 same progress may be expected in Italy, and the same also 
 in Germany. The only chance here is that the existing dy 
 nasties may make timely concessions, and so keep themselves 
 on the throne until men are prepared for complete republican 
 ism. If so, these countries will have the experience of Eng 
 land and not of France they will make the transition from 
 absolutism to republicanism through limited monarchies, and 
 not through tyrannies. Many English writers have spoken 
 of limited monarchy as if it were a great permanent form of 
 
DA Y-DREAMING. 369 
 
 government, like absolutism, or like republicanism ; and find 
 it unpardonable in the ancients that they did not adopt it. 
 Whereas it is rather, like tyranny, a transition form, assumed 
 in the course of a nation s progress, but wanting the condi 
 tions of permanence, and subverted at length by the natural 
 operation of the causes that produced it. We have passed 
 through this stage ; and it was a fortunate circumstance for us, 
 as it saved us from a succession of military tyrannies such as 
 we see in our sister republics of South America, which had 
 known nothing but absolutism up to the time when they 
 began to struggle for their independence. Limited monarchy 
 with us must have grown very feeble ; for the crisis of the 
 Revolution, which in other respects made no great change in 
 our institutions, was sufficient to destroy monarchy among us 
 at once and forever. I believe that any great crisis in Eng 
 land would show that monarchy has but little more tenacity 
 there. 
 
 The general tendency of things in Europe appears to me 
 certain and undeniable. How fast that tendency is advanc 
 ing, how soon it will reach its consummation, I for one will 
 not venture to predict. Providence, says Guizot, is strangely 
 and fearfully slow in its great enterprises. 
 
 Neither is it possible to say what is ultimate. Republican 
 ism I believe will prevail overall western Europe ; but whether 
 finally or forever is a different question. There have been 
 causes gravitating in the opposite direction. Rome was a re 
 public ; her conquests in Italy, Greece, Africa, her great 
 conquests were republics, And yet Rome became a mon 
 archy, the greatest that the world has ever seen, and her ex 
 ample has done more to sustain monarchy in modern Europe 
 than all other causes put together. 
 
 6. SHOULD DAY-DREAMING BE INDULGED IN ? 
 
 1853. 
 
 THE human mind moves in two different worlds the world 
 of reality, and the world of illusion. The objects of the 
 24 
 
370 CLASS DECISIONS : 
 
 former are matters of fact, things obvious to the senses or 
 cognizable by logical understanding, susceptible of proof, 
 certified by testimony or other evidence more or less con 
 vincing. The objects of the latter are imaginations and 
 fancies, sentiments and presentiments, aspirations and en 
 thusiasms. They are not apprehended by logical under 
 standing, but recognized by intuitive perception ; certified, 
 if certified at all, not by processes of reasoning, not by argu 
 ments which the understanding can analyze and appreciate ; 
 whatever warrant of truth they have lies in their native 
 power of attraction, in the response which they call out from 
 unperverted feeling. They belong thus to what Tennyson 
 calls " The truths that never can be proved." 
 
 The objects of the first are substantial, positive, definite in 
 outline, capable of being described by precise and intelligible 
 statements. The objects of the last are vague and shadowy, 
 admitting no exact definition, bright often as rainbow colors, 
 but like these without distinct boundaries, melting into each 
 other by insensible gradations. Passing beyond the sphere 
 of sense, transcending the experience of life, often laying 
 hold of the infinite, they have no finite measures by which 
 their dimensions may be ascertained, and their forms laid 
 down in plot. 
 
 I have called these two the worlds of reality and illusion 
 rather in conformity with ordinary views, than as describing 
 their inherent nature. I do not mean to represent the latter 
 as wholly illusory. It may be that our ideas, even those 
 which we regard as least substantial, have a root in verity, 
 and are destined to be realized in some form, in other spheres 
 of existence and under other conditions of being ; and on the 
 other hand, it may be doubted whether the reality, what we 
 call, such, what presents itself as such to our senses and intel 
 ligence, does after all conform so perfectly and exactly as 
 we are apt to suppose to the absolute, interior, underly 
 ing reality of things as it appears to the view of God. Do 
 not understand me as assuming the attitude of a Pyrrho- 
 nist, a universal skeptic, calling in question the reality of 
 everything which is generally received as reality. Our 
 conceptions of the external world and of the properties which 
 
DAY-DREAMING. 371 
 
 belong to it are necessary conceptions not only that, they 
 are necessary beliefs : even the Pyrrhonist cannot help be 
 lieving and acting on the belief that they are true, that they 
 correspond to outward, objective reality. It is the height of 
 folly to reject them, not to act upon them, to think of seeing 
 round and beyond them. And yet we cannot help admitting 
 there may be something beyond them, a deeper reality to 
 which our faculties cannot penetrate. 
 
 Not to discuss this point, let us go back to our two worlds 
 as we have endeavored to describe them. They belong, I 
 say, to every rriind. Perhaps no man can be found (unless 
 wholly insane) who confines himself exclusively to either : 
 none so absorbed in matter of fact that gleams of imagination 
 and enthusiasm never penetrate the crust that surrounds him ; 
 and none so devoted to imagination and enthusiasm as not 
 sometimes to touch the terra firma of substantial realities. 
 But if this exclusive attachment to one or the other sphere 
 is never fully carried out, we occasionally find a near approx 
 imation to it ; and it is important for us to note the effects 
 thus produced on mind and character. Rev. Mr. Hudson 
 Shakespeare Hudson, as he has sometimes been called, from 
 his lectures on Shakespeare in an allusion to Dr. Tyng s 
 preaching, described it rather uncharitably, I fear, and un 
 truly, or with an unwarrantable exaggeration of truth as 
 "lean, hard, dry, bloodless, and bilious." These words de 
 scribe pretty well what is likely to become the character of 
 the man who confines himself to the world of reality. The 
 spirit of life evaporates, and the residuum is stale and flat: 
 the finer elements of character give place to the coarser. The 
 graceful, delicate, generous and honorable in feeling and ac 
 tion are connected with ideas and sentiments that lie with 
 out the sphere of plain substantial matter of fact. Their 
 springs are in the ideal world ; if cut off from these, they dry 
 up and disappear. On the other hand, the man who confines 
 himself to the world of illusion, the world of day-dreams, 
 loses robustness and power of action. Dwelling in the crea 
 tions of his own mind, without practical objects to call out his 
 powers in vigorous and healthy exercise, he is something like 
 Hudibras s sword, ".whose trenchant blade, Toledo trusty, for 
 
3/2 CLASS DECISIONS. 
 
 want of using k a$ gtown rusty, and ate into itself, for lack of 
 somebody "to- hew and hack." He becomes lachrymose, 
 sentimental, querulous, and declamatory. His enthusiasm 
 runs to waste for want of objects. Even imagination suffers 
 by being deprived of food which it finds or ought to find in 
 the objects of the real world. 
 
 We see then that both these worlds are essential to the true 
 development of mind and character. The man does best for 
 himself who keeps strongest hold on both ; who combines 
 sharp perception and keen enjoyment of practical realities 
 with the glowing visions and high-soaring aspirations of the 
 day-dreamer ; who walks in the clear light of the common 
 everyday sun that rises and sets for all men, while lighted 
 inwardly by " the light that never set on sea or shore, the 
 imagination, and the poet s dream." But it is necessary to 
 bear in mind constantly the essential distinctness of the two 
 worlds, never allowing the imagination and enthusiasm to in 
 vest their objects with the character of reality, always remem 
 bering that air-castles have no foundation on terra firma, and 
 are not built of solid brick and mortar. This is a besetting 
 evil of day-dreamers, and most fully and elaborately illustrated 
 in the immortal prose poem of Cervantes Don Quixote, a day- 
 dreamer who confounds the two worlds of illusion and reality. 
 
 It is necessary in another way to recognize their proper 
 distinctness, by not mixing the language of one with the 
 business of the other. The fault or weakness to which I refer 
 is perhaps most often exhibited in Germany. It was a say 
 ing of Richter s, in the early part of this century (in the times 
 of Napoleon), that to the British belonged the empire of the 
 sea, to the French of the land, to the Germans of the air. 
 Their kingdom is that of the day-dreamer. As a conse 
 quence one often finds among them, both in literature and in 
 life, an overflow of enthusiasm. They are apt to mix up the 
 business of common life with expressions of enthusiastic sen 
 timent, raptures and transport, which the English mind can 
 not help regarding as somewhat ridiculous. The mixture of 
 transports and tea-drinking, roast beef and raptures, appears 
 to it incongruous and absurd. The English perhaps go too 
 far in the opposite direction, excluding the sentimental ele- 
 
... .. 
 
 IMMORTALITY. 373 
 
 ment from everyday life, and the ordirl^^jbj^course of so 
 ciety. But I must confess, my own sympSSteBr^rrlf with 
 them. This day-dreamer life is or ought to be an inner life ; 
 its light should not be beacon or bonfire for public gaze, 
 but hearth-fire burning in the penetralia of mind, to light and 
 warm the mental home. Its objects of faith and hope, its 
 Lares and Penates, if you choose to call them so, are too 
 sacred to be brought into rude and familiar connection with 
 the coarseness and trivialities of everyday life. A man 
 should not forget, indeed, in the midst of practical life and 
 business, that he has this inner world of thought and feeling. 
 Its influence thus present with him will preserve him from the 
 unfeeling coarseness of the merely practical man. But there is a 
 great difference between this and an open exhibition or free 
 public display of the inner life ; a great difference between 
 showing in speech or action that one has feeling, and talking 
 long or loud about one s feelings. It is the latter, not the 
 former, that violates the instinct, and, as I believe, the true 
 and proper instinct, of the Anglo-Saxon mind. 
 
 We see then that day-dreaming is not only not detrimental 
 to intellectual development, but, under right conditions and 
 control, indispensable to the best mental development. 
 
 So that I may conclude with the concluding words of 
 Thekla in Schiller s wonderful lyric, which I must translate at 
 great sacrifice of its ethereal beauty : 
 
 " Every thought of beautiful, trustful seeming 
 Stands fulfilled in Heaven s eternal day ; 
 Shrink not then from erring and from dreaming 
 Lofty sense lies oft in childish play." 
 
 7. CAN IMMORTALITY BE SHOWN FROM THE LIGHT 
 OF NATURE ? 
 
 1855. 
 
 THE obvious fact about death is, that a certain living or 
 ganism, the living human body, in connection with which 
 certain processes or functions manifest themselves such as 
 respiration, nutrition, muscular motion, nervous sensation, 
 
374 CLASS DECISIONS : 
 
 perception, volition this organism becomes disorganized, 
 and all these processes or functions cease to be manifested. 
 Now the claim of immortality may be based on one or the 
 other of two assumptions : either, I. the same organism will 
 be reproduced hereafter, and the same functions or part of 
 them again manifested in connection with it, and accompanied 
 by consciousness of continued identity ; or, 2. the same func 
 tions may be exercised and accompanied by consciousness 
 of identity, though not connected with the same organism as 
 before ; may, in fact, go on without interruption, without 
 being even suspended by death, though no longer manifested 
 to us. Take an illustration. A mill is an organism not in 
 deed a living organism ; something far less complex and mys 
 terious than that, but still an organism and its function is 
 grinding corn. The mill may be blown down by a hurri 
 cane, and the function of corn-grinding ceases to be mani 
 fested. But perhaps the scattered timbers will be put to 
 gether as before, and the same mill will perform the same 
 corn-grinding function. Or perhaps the same identical corn- 
 grinding which belonged to the demolished mill may be 
 going on elsewhere in connection with another mill, unknown 
 to us, or in connection with some wholly different and unim- 
 agined system of apparatus. The assumptions in reference 
 to the mill and the man, it may be conceded, are possible. 
 But the prima facie probability is against them. We cannot 
 entertain them even as probable, until supported by some 
 positive evidence. I know there are some who deny this, 
 who maintain that mortality furnishes no presumption even 
 against immortality, that the certain dissolution of the organ 
 ism with which all our human functions are connected, and 
 apparent cessation of all the functions connected with it 
 that this does not constitute even & prima facie probability of 
 their real or permanent cessation. I cannot take the same 
 view of the question. To me, one of the most striking proofs 
 of the inherent elasticity of the human mind, its irrepressible 
 buoyancy, is, that death does not drive it to despair ; that, 
 with such fearful indications of approaching annihilation, it 
 dares to hope for immortality. What then are the evidences 
 of immortality which can countervail the opposite presump- 
 
IMMORTALITY. 375 
 
 tion furnished by the phenomena of physical death ? I think 
 it must be conceded, as to most of them, that singly they are 
 not very strong, though their collective force may be, and in 
 my judgment is, far greater. Let us pass them in review, or 
 at least those most commonly urged, that we may see how 
 far from decisive they are when separately taken, how plausi 
 bly they can be met and answered. 
 
 And, in doing this, we shall assume that the light of nature 
 is sufficient to prove, notwithstanding all apparent difficul 
 ties, both the wisdom of God and the goodness of God. If 
 the light of nature does not prove them, no light can ; for the 
 strongest proof of Christianity, most decisive of its truth as a 
 divine system of revealed religion, is its conformity to our 
 natural reasonable ideas of God as all-good and all-wise. 
 
 Thus it is said, the universal belief in immortality is evi 
 dence of its truth. We answer : the universal belief in im 
 mortality proves only that the cause from which it springs is 
 universal. But this cause, it is said, can only be the obvious 
 truth of the thing believed in. We answer : that cannot be 
 obvious truth which lies so far remote from human knowledge 
 and experience, and concerning which so many wise men 
 have had their doubts. Then it is said, we must believe it to 
 have come as an original divine communication to the parents 
 of our race. We answer : there is no proof of such communi 
 cation, and no necessity for assuming it. The universal belief 
 in immortality arises from the universal desire for it. Our 
 faith is often the child of our wishes. When we long to find 
 something true, we shut our eyes to all opposing evidence, we 
 dwell constantly upon the arguments which favor it, till at 
 length our judgment is satisfied, our belief is established. 
 
 But, it is said, God would not allow us to have such desires, 
 if they were not to be realized. We answer: God allows us 
 to have many desires which are never realized. Such desires, 
 though not directly satisfied, are still useful, by stimulating to 
 exertion, by developing faculties, or in some other way. And 
 even so in this case. The impulse which makes us desire 
 immortality is not only useful but necessary. For love of 
 immortality is at bottom love of life. We cling to life while 
 we can, and when death plucks it from us, we cling to the 
 
3/6 CLASS DECISIONS. 
 
 hope of recovering it, Alcestis-like, from the grasp of death. 
 Now the love of life, which appears also as the love of immor 
 tality, is essential to the well-being of man and of society, 
 essential even to their being at all. If love of life were ex 
 tinguished, the world would come to an end. Again it is said, 
 the mind of man does not attain its perfect development on 
 earth. We answer : it is not the order of nature that every 
 thing should have perfect development. Thus of plants, how 
 many perish in the earlier stages of growth, how many are 
 dwarfed and stunted by unfavorable circumstances, by barren 
 soil, ungenial climate, by ravages of insects, and a thousand 
 other causes : how few probably attain the highest perfection 
 of which they are in their own nature capable ! So it is with 
 the physical constitution of man and beast ; so it is with the 
 mental faculties of brutes. Not every dog or elephant is al 
 lowed to cultivate to the utmost the often wonderful sagacity 
 and intelligence with which he is endowed by the Creator. 
 Would you have an immortality for each and all of these, that 
 they may come to their maximum development ? But again 
 it is said, there are in nature analogies for immortality, the 
 germination of seed after its rotting in the ground, the trans 
 formation of insects, the phenomena of sleep and fainting. 
 We answer : there are no analogies in nature for life after 
 death; i. c. for continuance of the function, after the organism 
 on which it depended has become disorganized. Grind the 
 seed, and it will never germinate ; crush the chrysalis, and 
 no butterfly will burst from it. Dr. Beaumont s man digested 
 his food, though he had lost a bit of his stomach ; but could 
 digestion have continued if the whole stomach had perished ? 
 A man may feel and think with only a part of his brain ; but 
 can thought and feeling go on without any brain at all ? 
 
 This brings us to what is generally thought the strongest ar 
 gument for a post-mortem existence : viz. the immateriality of 
 the soul. Thought and feeling, it is said, are not attributes of 
 bodily organism, but attributes of something very different, 
 something which uses bodily organism for its various purposes ; 
 especially for communicating with other somethings like itself. 
 Of course, when its instrument gives out, it can no longer 
 make such communication to others, but it is not impaired in 
 
IMMORTALITY. 
 
 377 
 
 its own inherent powers and activities. Now I am far from 
 denying the immateriality of the soul, yet I think this harder 
 to prove than its immortality. That which is used as proof 
 of immortality, the premise, is more difficult to be established 
 than the conclusion. The question, observe, is not whether 
 thought and feeling are material ; they are attributes, and of 
 course immaterial; but the question is, whether the substance 
 to which the attributes belong is material : i. e. whether it is 
 the same substance which possesses the attributes of gravity, 
 impenetrability, and the like. Now it is said, these two sets 
 of attributes are so different thought and feeling on the one 
 hand and gravity and impenetrability on the other that we 
 cannot suppose them to inhere in the same substance. This 
 is the great argument, in fact nearly the whole argument, for 
 immateriality. But it is greatly weakened by the fact that in 
 regard to substance we know nothing whatever, can know 
 nothing whatever, except that certain attributes belong to it. 
 What its internal nature and constitution may be we know 
 not ; what its capacity for taking on other widely different 
 attributes may be, we know not. The organic life of a plant 
 is something very different from gravity and impenetrability. 
 That subtle and mysterious activity which converts shapeless 
 water and lime and potash, the materials of the dung-hill, into 
 beautiful structures of leaf and flower and fruit, is surely very 
 different from the ordinary properties of matter. Shall we in 
 fer, then, that the life of a plant is not an attribute of the mat 
 ter composing it, that the plant has an immaterial soul ? Or, 
 if we shrink from this, where shall we begin to find the imma 
 terial soul ? Is it in the sponge, which occupies the debatable 
 border between the animal and vegetable kingdoms ; in the 
 polyp, which is stomach only, a membranous bag you may 
 turn him inside out, and he will live as well as before ; in the 
 tape-worm, which you may cut into twenty pieces, and each 
 will live for itself, twenty souls carved out of one ; in the bi 
 valve, which lies at the bottom of the sea, allowing water to 
 stream through his pulpy body, making his living by digest 
 ing the animalcules it contains ? All these have properties 
 very different from the ordinary attributes of matter ; proper 
 ties which pass by insensible gradations, as you ascend the 
 
378 CLASS DECISIONS: 
 
 scale of being, to the complex mental phenomena, thought, 
 feeling, and intelligence, of the dog or elephant. Shall we 
 give them all immaterial souls, and wonder whether human 
 arithmetic can enumerate the spiritual units contained in that 
 stratum of mosquitoes, twenty feet thick, which has rested for 
 ages on the banks of the Orinoco ? Now I say these things, 
 not as denying the immateriality of the soul, but to show how 
 little it can be depended on as a positive argument for immor 
 tality. On the other hand, the materiality of the soul, if ad 
 mitted, is far from a decisive argument against immortality. 
 For if the material inheres in a body as substance, this body 
 may eventually be reproduced with all the conditions essential 
 to conscious identity. Or, very possibly, the substance in 
 which the material soul inheres may be a mere attendant form 
 of matter, capable of existing independently of the body, 
 though not capable of manifesting itself to us except in con 
 nection with an organic body. 
 
 I have thus far omitted intentionally what seem to me the 
 two weightiest arguments for immortality in fact, sufficient 
 in my view to countervail the apparent presumption from the 
 phenomena of death : first, the apparent disciplinary char 
 acter of human life, which certainly appears like a course of 
 training and preparation for something beyond it, and be 
 comes much more intelligible if regarded in this light ; and 
 second, the difficulty of reconciling the moral order of the 
 world with our ideas of the divine justice, natural and reason 
 able ideas of the divine justice, if human existence is con 
 fined to the present life. This difficulty is shown very strik 
 ingly in the two books of the Bible, Job and Ecclesiastes. It 
 is an extraordinary fact that the Old Testament Hebrews, 
 though not wholly without the idea of existence after death, 
 had yet no distinct idea of future reward and punishment. 
 An extraordinary fact, I say, considering that the Egyptians 
 and Phoenicians, heathen nations around them, had this idea 
 fully developed, and ex citing a powerful influence, especially 
 on the Egyptians. But the Hebrew, in his habitual conceptions 
 as seen in the Old Testament, looked for manifestations of 
 divine justice in this life without reference to another. It is 
 impossible, however, that the reflective should remain per- 
 
VEGETABLE DIET. 379 
 
 manently insensible to the difficulties of this view. That they 
 did not remain so appears from those two books, which in 
 fact turn upon these very difficulties. Both probably belong- 
 to the later books of the Old Testament : Job, written per 
 haps not very long before the captivity in Babylon ; Eccle- 
 siastes, probably some time after the return to Palestine. In 
 both, the great burden is this : the differences of human life 
 and fortunes do not correspond to differences of moral char 
 acter. How then can we vindicate the justice of God, and the 
 claims of duty ? In Job the conclusion is : God is strong 
 and wise ; man is weak and ignorant ; presume not to ques 
 tion where you cannot understand. In Ecclesiastes the con 
 clusion is : virtue on the whole is the best and happiest for 
 this life ; practise it with moderation of desires and cheerful 
 enjoyment that and only that is true wisdom. That neither 
 of these should have fallen upon the solution furnished by 
 the doctrine of immortality is certainly extraordinary, and 
 may perhaps go to show that the natural evidence of this is 
 not so clear as is frequently supposed. That is my own be 
 lief that the light of nature, when all directed to this ques 
 tion, does furnish a presumption in favor of immortality ; 
 but not so strong a presumption as to exclude great and 
 reasonable doubt upon the subject. 
 
 8. IS AN EXCLUSIVELY VEGETABLE DIET ADVANTAGEOUS ? 
 
 1856. 
 
 A VEGETARIAN (dietarian) book was written by a Dr. 
 Cheyne, one hundred years ago. He was a high liver, and 
 tended to corpulency, till his weight rose to four hundred 
 pounds, so that he could not enter a coach door. He then 
 thought it time to retrench, starved himself to two hundred 
 pounds, was much encouraged by success, and wrote a book 
 on the advantages of starvation. I have never seen it, but can 
 
380 CLASS DECISIONS: 
 
 easily believe it useful for gentlemen of four hundred pounds 
 weight, who cannot enter a coach door. 
 
 The dietetic movement in this country, so far as I know, 
 commenced with a work published in 1827 by President, then 
 Professor, Hitchcock, of Amherst College, entitled " Dys 
 pepsia Forestalled and Resisted." Dyspepsia was then epi 
 demic, spreading gradually over the country and becoming a 
 fashionable disease. It appeared first in Boston, and travelled 
 eastward. The doctors in Springfield and Hartford were 
 much amused when they first heard of the new disease in 
 Boston ; they thought it a new Boston notion. By-and-by it 
 came to Springfield ; the Hartford doctors were still incre 
 dulous. Then it reached Hartford, and the Middletown doc 
 tors thought it a city fashion. But they got it themselves in 
 time, as country people get all city fashions. Dyspepsia is 
 now on the decrease, there being much less of it than twenty 
 years ago. 
 
 Dr. Hitchcock was a dyspeptic in the palmy days of dys 
 pepsia, when the disease was a name of dread. His work 
 made a great sensation. I can recollect hearing it talked of 
 in early youth. I remember well the astonishment and horror 
 with which we heard that every man, woman, and child in 
 the United States ate on an average at least twice as much 
 as he, she, or it ought to eat. Another statement I remem 
 berwas, that somebody had sustained life a year on a hen s 
 egg per diem. Anything more than this, it was implied, 
 tended to evil. They may have drawn the argument from 
 the Roman description of dinner, as proceeding ab ovo usque 
 ad mala. 
 
 The name more commonly associated with the vegetarian 
 movement in the popular mind is that of Sylvester Graham. 
 His abilities, not inconsiderable, were perhaps more than 
 equalled by his own estimate of them. If any man ever lived 
 and died in assured confidence of posthumous fame, it was 
 he. Though neglected, he said, by contemporaries, posterity 
 would do him justice. There was really some good reason 
 for hope. His name will be in many mouths, along with 
 his bread and crackers. These articles are favorites with not 
 a few who neutralize their virtues by the deleterious decoc- 
 
VEGETABLE DIET. 381 
 
 tions of coffee and chocolate, and the poisonous poultry or 
 pastry, swallowed at the same time. He had some ground 
 of complaint against his contemporaries and neighbors, who 
 did not always treat him with proper respect. One of them, 
 passing his house one morning, saw the doctor picking up a 
 basket of chips : " Ah," said he, " baking-day at your house ? " 
 Dr. Graham was much offended, but without remedy. 
 
 Prominent among dietists is Dr. Alcott, the author of sev 
 eral pleasantly written books, and conductor of the monthly 
 Library of Health. In the latter, I remember an article on 
 mince-pie, showing that, if its component parts were to ar 
 range themselves in the stomach each by itself according to 
 its specific gravity, there would be no less than eighteen 
 layers or strata in the stomach. As it was not shown that 
 they do assume any such stratification, or that, if they did, 
 harm would come of it, the argument was a little inconclusive. 
 
 The most learned man among vegetarians was Dr. Mussey, of 
 Dartmouth, afterward of Cincinnati. It is said that he made him 
 self dyspeptic by high living, and then cured himself by ab 
 stemious regimen abstemious in quality, I mean ; for the quan 
 tity of potatoes he consumed was astonishing to ordinary eat 
 ers. His dietetic lectures I had the pleasure of hearing. They 
 were very capital lectures : quiet, clear, and finished, with 
 plenty of anecdote and illustration, and with a certain 
 dry, quaint, subdued, but caustic humor, peculiar to himself. 
 His wit and wisdom were too much for a youth of sixteen, 
 and I must confess myself to have been carried away for a 
 time. I gave up tea and coffee, beef and butter, mince-pie, 
 even, for a month, perhaps ; possibly for two ; quite up to the 
 average length of his conversions and then went back to the 
 old way, adapted my principles to my appetites, and have so 
 continued to this day. I need not be ashamed of being con 
 verted to vegetarianism by Dr. Mussey, for it is said that the 
 Doctor himself was later converted from vegetarianism by an 
 Orang Outang. Two of those mock humanities were* put on 
 shipboard for this country, and fed of course on vegetable 
 food, as being that which they have in their native woods. 
 One fell sick, and, in spite of all therapeutics, died. The 
 other also sickened, and seemed to be going the same way 
 
382 CLASS DECISIONS: 
 
 as his companion. One day, however, his chain being 
 loosened, he made a spring for a table not far off, on which 
 was a roasted chicken. He devoured the chicken with the 
 greatest eagerness, and, instead of languishing, seemed to 
 feel the better for his meal. The hint was taken, the same 
 diet was continued, and the fellow reached this country at 
 length in capital health and spirits. Dr. Mussey laid great 
 stress on the dietetic habits of Orang Outangs, Chimpanzees, 
 and the other nearest congeners of man ; and it is said that, 
 hearing these facts, he gave up his opposition to flesh meats 
 whether permanently, I am not informed. 
 
 Two faults of reasoning are to be observed in Dr. Mus- 
 sey s lectures. He assumes, first, that if a thing is injurious in 
 a large quantity, it is injurious also in a small quantity, 
 though in a less degree ; if a pound will do injury, an ounce 
 will do a sixteenth of the same injury. Dr. Mussey, so as 
 serting in the American Temperance Convention, was op 
 posed by Bishop Potter of Pennsylvania, with a parallel case. 
 An atmosphere of carbonic acid gas would kill you ; and 
 hence the minute fractional part which is always and necessarily 
 present in the atmosphere must be proportionally deleterious. 
 Second, to determine the wholesomeness of an article of food, 
 he looks at its effect on a weak stomach. Again and again 
 in his lectures, he says of a particular article of food that a 
 weak stomach cannot digest it, and therefore it should not be 
 used. He might as well say that, as weak persons cannot 
 take long walks, long walks are fatiguing and exhausting to 
 the physical energy ; and therefore long walks should not be 
 taken. He might as well say that the blacksmith should not 
 use his sledge-hammer, because a little boy would break 
 down in the attempt to wielcl it. That the weak should rule 
 the strong, the healthy stomach submit to the unhealthy, is 
 against the law of nature. Let dyspeptics legislate for them 
 selves, but let them not seek to establish a dyspeptic 
 despotism over the eupeptic world. 
 
 The principal argument of the vegetarian is that man s phy 
 sical organization is analogous to that of herbivorous, not car 
 nivorous animals. There is no doubt that our teeth, our sto 
 mach, and our alimentary canal are widely different from those 
 
VEGETABLE DIET. 383 
 
 of the dog- or wolf; and it is a fair inference that dog s meat 
 is not proper for us. But what is a dog s meat, adapted to 
 his teeth and stomach, adapted to his active habits and in 
 stinct? It is raw flesh. The dog wants flesh, and cannot 
 cook it ; he will take it raw ; and for this he is fitted by his 
 physical organization. The vegetarians resolutely ignore the 
 effects of cookery, not noticing that a similar argument could 
 be used against their favorite diet. For we might say that, as 
 raw potatoes are unwholesome, the use of potatoes as food 
 should be discarded. 
 
 Examples are brought to show that the highest degree of 
 physical vigor can be maintained without animal food. The 
 strongest case is the story told by traveller Buckingham, of 
 powerful fellows from Nepaul, the lower slopes of the Hima 
 laya, who had come down to Calcutta to exhibit feats of 
 strength. They were matched against British sailors, and 
 were always superior. The Indian athletes are restrained by 
 their religion from eating flesh meat, of which the British sailor 
 makes free use. There is, perhaps, some room for doubt, but 
 two things may be said of the comparison : first, the inclina 
 tion and necessity for meat is far less within the tropics than 
 without ; and second, warm climates are far more enervating 
 to men of northern origin than to the native of India. 
 
 The Scripture argument which the vegetarians bring for 
 ward is, that in the grant to Adam only vegetable food is 
 named, no animal. The omission is probably not accidental, 
 for the idea of maintaining life by the slaughter of animals is 
 repugnant to our conceptions of the Paradisiac state ; but the 
 whole condition, physical, spiritual, and moral, of the Paradi 
 siac world is so different from that of the actual, that no argu 
 ment can be drawn from it. In the grant to Noah, animals 
 are allowed as food ; but, as the vegetarians say, only because 
 men would have it like divorce in the Mosaic law, allowed 
 on account of the hardness of men s hearts. Only, cookery is 
 required ; since this is supposed to be meant by "flesh with 
 the life thereof, which is the blood, them shalt not eat." Still, 
 though cookery diminished the evil, human life, under the de 
 leterious diet, fell off from Methuselah s nine hundred and 
 sixty-nine years to the threescore and ten of the Psalmist. 
 
384 CLASS DECISIOA^S. 
 
 The hope is that, by going back to Methuselah s Graham 
 bread and potatoes, we shall gradually, in the course of gen 
 erations, get back to his longevity the new millennium of life 
 on earth. 
 
 We have never been much troubled with vegetarian fanati 
 cism here, but in some places it has been mischievous. It has 
 often happened that a young man has got an idea that the less 
 he eats, the better for him, and so has confined himself to 
 brown bread, and a minimum allowance even of that. If very 
 strong, he bears up under it, until he overlives his delusion, 
 or is brought by appetite, example, or custom, into the ordi 
 nary way. But, if not strong, he sinks under it, and especi 
 ally if he has any tendency to constitutional disease ; which 
 there is nothing more likely to develop. I have in mind the 
 case of an excellent young man, Avho in my belief hastened, 
 if he did not cause, his own death in this way. 
 
 Students need animal food not less than laboring men ; 
 rather, more ; they do not need more of it, but they have more 
 need of it. A man who can digest a peck of potatoes at a 
 time can live upon potatoes alone, for he can get enough nutri 
 ment to sustain life and vigor. But the less hardy stomach 
 requires more concentrated diet, so as to have nutriment 
 enough without being overloaded. 
 
 If I were to legislate on diet and regimen, my rules would 
 be few and simple. In regard to exercise, take a good deal 
 of it, unless from early days accustomed to a very sedentary 
 life ; but not too soon after eating a hearty meal. 
 
 In regard to eating eat what you want ; eat nice things, 
 eat as much as you want of them, unless you find by clear 
 experience that something disagrees with you ; but eat only 
 at meal times, and take no meals, edible or potable, late at 
 night. And, last not least, trust to your stomach ; do not 
 be continually watching over it and criticising it, but believe 
 it will get along without your help ; and, above all, think as 
 little as possible about your diet. 
 
XIX. 
 
 ON THE HEBREW CHRONOLOGY FROM MOSES 
 TO SOLOMON. 
 
 1857. 
 
 \ ^ 7E find in the first book of Kings, sixth chapter and 
 V V first verse, a definite chronological statement, assigning 
 the interval in years between the departure of the Israelites 
 from Egypt and the commencement of work in the erection of 
 Solomon s temple. " And it came to pass in the four hundred 
 and eightieth year after the children of Israel were come out 
 of the land of Egypt, in the fourth year of Solomon s reign 
 over Israel, in the month Zif, which is the second month, 
 that he began to build the house of the Lord." In this pas 
 sage, however, the Septuagint version, the oldest and beyond 
 comparison the most important of the versions, presents us 
 with a different number four hundred and forty years, in 
 stead of four hundred and eighty. This is a remarkable 
 variation, and we shall have occasion to return to it in the 
 sequel. At present, it is enough to observe that the number 
 of our Hebrew text is unquestionably ancient, since it ap 
 pears in other ancient versions ; it appears, indeed, in all the 
 other ancient versions which were made directly from the 
 Hebrew ; we may therefore fairly conclude that in the early 
 centuries of our era, when those versions were made, the 
 number four hundred and eighty was the general and current 
 reading of the Hebrew manuscripts. Whatever discrepancy 
 may have existed among MSS. in the earlier period of the 
 LXX. translators, the number four hundred and eighty had 
 apparently become the approved and established reading. 
 
 It is a well-known fact, however, that this number, four 
 hundred and eighty years, for the interval between the Exo 
 dus and the building of the temple, is at first view inconsist 
 ent (and the smaller number of the LXX. would still more be 
 25 
 
3 86 HEBRE W CHR ONOL OGY. 
 
 inconsistent) with the particular designations of time scatter 
 ed through the earlier historical books of the Old Testament. 
 Let us look at these numbers, in the order of the Biblical 
 history. First, for the prolonged desert wanderings under 
 the lead of Moses, which followed the Exodus from Egypt, 
 the Pentateuch assigns a period of forty years. Next, the 
 book of Joshua (xxiv. 31) informs us that " Israel served the 
 Lord all the days of Joshua and all the days of the elders that 
 overlived Joshua," but does not assign any definite term of 
 years for the days of Joshua himself, or for those of the sur 
 viving elders. After those unassigned periods, the chronolo 
 gy of the book of Judges opens with the first apostasy of the 
 Hebrew people or rather, with the punishment which follow 
 ed that apostasy. In this book we find the following series 
 of numbers, which, according to the obvious appearance and 
 natural impression of the work, if not according to its real 
 intent, denote successive periods of time : 
 
 YK3. 
 
 Servitude under Chushan Rishathaim, king of Mesopotamia. ... 8 
 
 Deliverance by Othniel, son of Kenaz, and subsequent rest. ... 40 
 
 Servitude under Eglon, king of Moab 18 
 
 Deliverance by Ehud, son of Gera, and subsequent rest 80 
 
 Ascendency of Shamgar, son of Anath, time not stated 
 
 Oppression of Jabin, king of Canaan, who reigned in Hazor. ... 20 
 
 Deliverance by Deborah and Barak, and subsequent rest 40 
 
 Oppression by the Midianites 7 
 
 Deliverance by Gideon, and subsequent rest 40 
 
 Reign of Abimelech, son of Gideon 3 
 
 Tola, son of Puah, a man of Issachar, judges Israel 23 
 
 Jair, a Gileadite, judges Israel . 22 
 
 Oppression by the Ammonites 1 8 
 
 Deliverance by Jephthah the Gileadite, who judges Israel 6 
 
 Ibzan of Bethlehem judges Israel 7 
 
 Elon, a Zebulonite, judges Israel 10 
 
 Abdon, son of Hillel, a Pirathonite, judges Israel 8 
 
 Oppression by the Philistines 40 
 
 Samson, son of Manoah, judges Israel 20 
 
 In regard to this last number, it is not certain that the writer 
 intended to represent the twenty years of Samson as succeed 
 ing the forty of Philistine oppression : for his language is, 
 Samson "judged Israel in the days of the Philistines twenty 
 
HEBREW CHRONOLOGY. 387 
 
 years." It is possible, therefore, that he conceived of Sam 
 son s twenty years as included in the forty of Philistine op 
 pression. However this may be, it is the only instance in 
 which the language of the book lends any plausible color to 
 the assumption that the times above stated are not to be re 
 garded as successive each to the preceding. Let us suppose 
 for the present that the twenty here was thought of as follow 
 ing the forty ; and let us add together the whole series of 
 numbers presented by the book, from the invasion of Chu- 
 shan-rishathaim to the death of Samson. We obtain an ag 
 gregate of four hundred and ten years, with an uncertain 
 blank for the time of Shamgar, the successor of Ehud. If, 
 now, to this four hundred and ten we add forty years 
 for Eli, who according to I. Samuel iv. 18 judged Israel 
 forty years, we have a total of four hundred and fifty years 
 for the period of the Judges, as reckoned from the first 
 chronological indication in the book of that name on to the 
 opening of Samuel s judgeship. Now in Acts xiii. 20, we 
 find Paul in the synagogue of Antioch, while describing in 
 brief outline the early history of his people, saying that God 
 " gave unto them judges about the space of four hundred and 
 fifty years until Samuel the prophet." This is an important 
 passage, not only as showing that the numbers which now 
 appear in the book of Judges stood there in the first century 
 of our era, but also because it proves that they were then re 
 garded as designating consecutive periods of time. For, ac 
 cording to the only natural and unforced interpretation of the 
 apostle s words, they assign the duration of a period during 
 which God was giving judges to his people, a period begin 
 ning at the time when he began to give them judges, and 
 ending with the accession of the prophet Samuel. And the 
 duration assigned for this period, viz. four hundred and fifty 
 years, is precisely that which results from the numbers in 
 Judges and Samuel, on the supposition, which is strongly 
 favored if not absolutely required by the language of those 
 books, that the numbers stand not for contemporaneous but 
 for successive times. 
 
 The death of Eli was followed by the judgeship of Samuel 
 and the kingship of Saul. For these, no term of years is as- 
 
388 HEBREW CHRONOLOGY. 
 
 signed in the Old Testament narrative. Paul, however, in 
 the address just quoted, says (Acts xiii. 21) that when the 
 Israelites desired a king, " God gave unto them Saul the son 
 of Cis, a man of the tribe of Benjamin, by the space of forty 
 years." It is possible that this term of forty years may have 
 been intended to cover the whole interval from the close of 
 the four hundred and fifty years mentioned just before to the 
 accession of David, which is mentioned directly after. If so, 
 it would include the time of Samuel s ascendency, as well as 
 that of Saul s dominion. To David, the Old Testament it 
 self ascribes a reign of forty years. If we add this forty and 
 the forty years of wandering in the desert to the sum of four 
 hundred and fifty before obtained for the Judges, we have an 
 aggregate of at least five hundred and thirty years between 
 the Exodus and Solomon, or fifty years more than the four 
 hundred and eighty named in I. Kings vi. i. We might, 
 indeed, avail ourselves of the doubt in regard to Samson, and, 
 by reckoning his twenty years of judgeship under the forty 
 years of Philistine oppression, might reduce our aggregate to 
 five hundred and ten. But even this is thirty years beyond 
 the mark of I. Kings ; and it must be still further increased, 
 when we come to consider the two certain and important gaps 
 in our series, one for Joshua and the elders after him, the 
 other for Samuel and Saul. 
 
 The statement in the book of Acts gives forty years appa 
 rently to Saul alone, though meant possibly both for Samuel 
 and Saul. At all events, we could not allow less than fifty 
 years to Joshua and the elders, Saul, and Samuel together. 
 We should thus swell our aggregate to an amount exceeding 
 by eighty or one hundred years the interval assigned by the 
 writer in I. Kings. A discrepancy at once so obvious and so 
 large could not fail to attract attention. Much effort and in 
 genuity have been expended in harmonizing the discordant 
 statements. The methods which have been suggested for 
 this purpose, though considerably numerous, rest for the 
 most part upon the same leading assumptions. These are : i. 
 that the number four hundred and eighty years given in I. 
 Kings must be accepted as the true and exact statement of the 
 interval in question ; and 2. that the period apparently in- 
 
HEBRE W CHR ONOL OGY. 389 
 
 eluded in the book of Judges must be shortened by making 
 the times in that book, not all successive, but more or less 
 contemporaneous. I do not propose to criticise these 
 schemes, or any one of them, in detail. But two or three re 
 marks may be made which are applicable to all : I. In itselt 
 it is not by any means improbable, but rather, considering the 
 disordered state of Palestine after the Hebrew conquest, the 
 imperfect subjugation of the old inhabitants, the isolation of 
 the different tribes, the want of national centres, of common 
 government and of concerted action considering these facts, 
 I say, we must acknowledge it as not improbable that the 
 different parts of the country were subject to different for 
 tunes, that they were oppressed at the same time by different 
 assailants, were delivered by the heroic conduct of different 
 chiefs, and recognized the authority of different judges. In 
 itself, it is not improbable that the oppressions, deliverances, 
 authoritative judgeships, described in the book of Judges, were 
 partial and not general ; confined each to particular sections 
 of the people, and going on contemporaneously in different 
 portions of the land. To such an extent might we concede 
 this, without any violation of historic probability, as to reduce 
 the period of the Judges even more than is required to har 
 monize its chronology with the statement in I. Kings. 2. 
 Nevertheless, this view of the events recorded in the book of 
 Judges, as if they were partial and contemporaneous, seems 
 scarcely consistent with the conception of the writer. From 
 beginning to end, from the oppression of the Mesopotamian 
 king to the heroic death of the Danite champion, he treats 
 the people as a whole. As a people they apostatize from the 
 Lord, as a people they are punished by oppressive invasion, 
 as a people they return to the Lord with repentance and 
 confession, as a people they receive deliverance from the 
 foreign yoke and submit to the authority of the deliverer. 
 The more we read the book, and the more attentively we 
 study its plan as announced in the second chapter and car 
 ried out in the sequel, the more strong becomes our impres 
 sion that the writer in his own view is recording a series of 
 national judgments, which fall successively for good and for 
 evil on the whole Hebrew nation, the children of Israel. In 
 
390 HEBREW CHRONOLOGY. 
 
 passing from one to another, he almost invariably uses lan 
 guage which implies succession. The only exception is at 
 the beginning of the sixth chapter, where he says : " and 
 the children of Israel did evil in the sight of the Lord." 
 Though, even here, the connection would seem to show con 
 clusively that he views this evil-doing as subsequent to the 
 rest which, according to the previous verse, the land (that is, 
 the whole land there is nothing to suggest the idea of apart) 
 had enjoyed for forty years. But everywhere else he says 
 " and the" children of Israel did evil again in the sight of the 
 Lord ; " thus marking their conduct as a succeeding and later 
 instance of apostasy. So he says: " And after Abimelech 
 there arose Tola;" "And after him arose Jair ; " "And 
 after him (Jephthah) Ibzan judged Israel;" "And after him 
 Elon ; " " And after him Abdon." We cannot be surprised, 
 therefore, when we find St. Paul assigning as the period of 
 the Judges a number of years equal to the aggregate of all the 
 numbers contained in this book (increased, of course, by 
 Eli s forty years), as if the events and conditions to which 
 they belong were all successive. 3. The schemes referred to 
 rest upon no positive foundation, and have no absolute prob 
 ability. At most, we can only say of them, it is possible 
 that things may have taken place in this way. The com 
 pleteness with which they solve the problem, and cut down 
 the chronology of the Judges to a dimension consistent with the 
 statement in Kings, is far from being decisiveasto their truth. 
 For the problem, by the nature of its conditions, is essentially 
 indeterminate indeterminate w 7 e might say in algebraic lan 
 guage, because there are more unknown quantities than fixed 
 relations between them ; it admits, as the facts show, a con 
 siderable variety of solutions, and might equally well be solved 
 in several different ways, if the standard number were as 
 given in the Septuagint, four hundred and forty years, instead 
 of four hundred and eighty. 
 
 But 4. I remark, in regard to these attempts at reconcilia 
 tion, that they overlook a very important feature in the 
 chronological statements which they propose to harmonize, 
 and must therefore of necessity fail to give satisfaction. By- 
 taking this into account, we may hope, not indeed to arrive 
 
HEBREW CHRONOLOGY. 391 
 
 at the solution of the problem, but at least to throw light up 
 on its nature and conditions. 
 
 If we recur to the series of numbers which we before 
 gave, as belonging to the period from Exodus to Solomon, 
 and examine them attentively, we cannot fail to be struck 
 with the frequent repetition of the number forty. It stands 
 for the wandering in the desert, the rest under Othniel, the 
 rest under Deborah and Barak, the rest under Gideon, the op 
 pression of the Philistines, the judgeship of Eli, and the king 
 ship of David that is, seven times in a series of twenty one 
 numbers, or once in every three numbers. Now this circum 
 stance alone is enough to show that there is something pecu 
 liar in the chronology we are dealing with. Let us compare 
 it with some other chronological series e.g. the reigns of 
 Solomon s descendants, the kings of Judah. Among these 
 we may omit two or three reigns of a year or less numbers 
 so small could hardly find place in the more sketchy history 
 of the earlier times. There remain seventeen designations 
 of time, of which we find forty years for Jehoash and forty- 
 one for Asa : only two in the seventeen which are anywhere 
 in the vicinity of forty. Or, let us look at the sovereigns of 
 England from the Norman Conquest thirty-four in number 
 extending through eight centuries, and including an uncom 
 monly large proportion of long reigns : here we find no reign 
 of forty years ; one (Henry VI.) of thirty-nine, one (Henry 
 VIII.) of thirty-eight ; and only six out of the thirty-four 
 which come within a half-dozen of the forty, on one side or 
 the other. But the case as regards the use of forty in this 
 early Hebrew chronology is not yet fully stated. The only 
 large number beside forty which occurs in the series, the 
 only other number which rises above five and twenty, is that 
 given for the rest which followed the deliverance by Ehud ; 
 and this is eighty years that is, twice forty. Now this rela 
 tion may be accidental : I mean, there is no absolute impos 
 sibility in the supposition that the periods of time referred to 
 should have had this precise ratio to each other. But it must 
 be owned, I think, that the chances are very greatly against 
 it. That a series of historic times given with historical ex 
 actness, or nearly so, should contain seven forties in twenty- 
 
392 HEBREW CHRONOLOGY. 
 
 one numbers, is exceedingly improbable ; but that the only 
 remaining number which is more than twenty-five should be 
 twice forty, adds much to the improbability. We may fairly 
 say that the appearance of twice forty in this place is, under 
 all the circumstances, a fact as surprising and as much requir 
 ing explanation as if forty itself had been the number ; and 
 consequently, in estimating the character of this chronology, 
 we may fairly consider this as another repetition of the 
 number forty. But yet again, we find twice forty years 
 assigned (Ex. vii. 7) f r the n ^" e f Moses prior to the 
 Exodus, and this a New Testament passage (Acts vii. 23, 30) 
 divides into two periods of forty years each. It is an 
 education of forty years at the Egyptian court, followed by a 
 wandering life of forty years among the Midianites of the 
 desert, that prepares him to act during another forty years, at 
 once as the law-giver of his people and the leader of their 
 desert wanderings. And lastly, as David, the great national 
 hero and conqueror of the Hebrews, has a reign of forty 
 years, so we find forty years assigned to the peaceful but 
 splendid reign of his successor, the builder of the temple 
 and the last monarch of the united people. And we have al 
 ready noticed in the book of Acts a statement, resting per 
 haps on a tradition as ancient as the Old Testament, which 
 gives the same period to the first in the great triad of national 
 Hebrew sovereigns, the warlike but unfortunate founder of 
 the Hebrew monarchy. We find thus in the chronology 
 from Moses to Solomon inclusive no less than eleven repeti 
 tions of the number forty ; or twelve, if the eighty after 
 Ehud s deliverance be reckoned, as it fairly may, in this cate 
 gory. 
 
 Now the view here presented of the facts seems to suggest 
 and not only that, but almost to require the conclusion 
 that we have in this narrative traces of an artificial chrono 
 logy, proceeding by periods of forty. I should call that a 
 natural chronology which, in stating known times, should give 
 them with exactness, or nearly so ; and, in regard to others, 
 should estimate them as nearly as it knew how, if there were 
 means for estimating them, and if not, should leave them un 
 determined, either expressly or tacitly confessing ignorance. 
 
HEBREW CHRONOLOGY. 393 
 
 On the other hand, I call that an artificial chronology in 
 which the times, instead of being stated thus or left unstated, 
 are determined in whole or in part by assumptions or allow 
 ances of an arbitrary or systematic character. Thus, if forty 
 years should be assumed as a general expression for any 
 period of considerable length as for instance the reign of a 
 king who remains some time upon the throne, or the duration 
 of a peace which has some degree of permanence : a chronol 
 ogy proceeding on this assumption, and assigning forty years 
 to particular facts or states of such a character, would be to 
 that extent an artificial chronology. Or, if forty years should 
 be assumed as the length of time for a generation, a chro 
 nology proceeding upon this assumption, and assigning this 
 period to the public activity of particular individuals who 
 should take their place in a series of generations, would again 
 be so far forth an artificial chronology. These modes I sug 
 gest merely by way of illustration. Whether the chronology 
 we are considering involves either of these assumptions, or 
 whether its peculiar character is to be explained in some dif 
 ferent way, I do not at present undertake to decide, nor is 
 it necessary to do so. That it has traces of something dif 
 ferent from a precise historical chronology results from the 
 extraordinary predominance of the number forty, and results 
 equally, whether we can find a plausible explanation for it or 
 not. The best explanation we can suggest may be encum 
 bered by serious difficulties : but those difficulties will not 
 affect the evidence which seems to force upon us the conclu 
 sion already stated that this series of numbers in the early 
 Hebrew history from Moses to Solomon, nearly half of them 
 forties, bears traces of what maybe called in the sense just 
 explained an artificial chronology, proceeding in some way by 
 forties. It is not easy to see how this result can be avoided 
 except by supposing a kind of miracle, by assuming a special 
 divine interference, which gave to an extraordinary proportion 
 of early Hebrew times the exact or approximate length of 
 forty years. It is hardly necessary to add that I use the term 
 artificial here in no invidious sense, and without in the least 
 meaning to imply any unjustifiable deviation from historic 
 accuracy ; much less any purpose of concealing their own 
 
394 HEBREW CHRONOLOGY. 
 
 ignorance, or of misleading others, on the part of the ancient 
 Hebrew chroniclers. It may have been observed that, in 
 speaking of this series of numbers, I do not say it is a chronol 
 ogy proceeding by forties, but that it contains traces of such a 
 chronology. For, mixed up with the forties in the book of 
 Judges, we find other numbers of a definite and apparently 
 historical character thus, for the oppressions by the Meso- 
 potamians, Moabites, Canaanites, Midianites, and Ammo 
 nites, we have the numbers eight, eighteen, twenty, seven, 
 and eighteen years. For Abimelech we have three years, for 
 Tola twenty-three, for Jair twenty-two, for Jephthah six, for 
 Ibzan seven, for Elon ten, for Abdon eight, for Samson 
 twenty. And even in regard to the forty, I would not be 
 understood as saying that it is in every instance unhistorical, 
 or that it is so in any part cular instance ; but only that the 
 concurrence of so many forties in a chronological series of 
 this extent can hardly be looked upon as historical. 
 
 Let us proceed now from what seems nearly certain to 
 that which is more doubtful. We started at the outset with a 
 particular chronological number in I. Kings, and, in comparing 
 this with a series of numbers scattered through the earlier 
 history, we found a discrepancy between them. We then ex 
 amined the discrepant series, and discovered in it the traces 
 of a peculiar chronology, proceeding by forties. It becomes 
 then a fair question whether the discrepant number in I. Kings 
 may not itself have the same character. Now, singularly 
 enough, the number in question (four hundred and eighty 
 years) is the precise aggregate of twelve forties. Of course 
 this fact alone would be of little weight, if in the remaining 
 chronology the number forty had no special prominence. In 
 that case, the fact that four hundred and eighty consists of 
 twelve forties would be no more important than the fact that 
 it contains sixteen thirties or thirty sixteens. The significance 
 of the fact lies in the coincidence that four hundred and 
 eighty should be an exact multiple of the particular number 
 which figures so largely in the detailed chronology. Such a 
 coincidence is d priori improbable. The chance that two 
 events remote in time should be found to be separated from 
 each other by a number of years which is divisible by forty 
 
HEBREW CHRONOLOGY. 395 
 
 is only one chance in forty, if the interval is given with exact 
 ness. And the present case is one in which we should ex 
 pect that the writer, if the precise number of years had been 
 present to his mind, would have stated it precisely, instead 
 of contenting himself with a loose approximation. For the 
 statement is not a casual one, suggested in passing ; it stands 
 prominently forward at the opening of the section, and ap 
 pears to be regarded by the writer as fixing the relation in 
 time between the two grand eras of Hebrew history. More 
 over, the expression is peculiarly distinct and circumstantial ; 
 not " Solomon began building the temple four hundred and 
 eighty years after the Exodus ; " but " it came to pass in the 
 four hundred and eightieth year after the children of Israel were 
 come out of the land of Egypt, in the fourth year of Solo 
 mon s reign over Israel, in the month Zif, which is the second 
 month," etc. But, waiving this point, and supposing that the 
 writer might have contented himself here with a loose state 
 ment of the interval, with giving it in round numbers by the 
 nearest ten, still the d priori chance that the number even as 
 thus stated should be an exact multiple of forty is but one in 
 four. 
 
 We cannot help thinking, therefore, that the relation of four 
 hundredand eighty to forty is calculated to suggest the conjec 
 ture (and that with some degree of force) that the statement 
 in I. Kings may be founded on a chronological method, such as 
 we have before been criticising. And this conjecture receives 
 some color of plausibility from the various reading of the 
 LXX. (four hundred and forty for four hundred and eighty) 
 to which we have already adverted : at least, that variation 
 makes it apparent that a similar idea \vas entertained as early 
 as the time of the Septuagint translators. For, in the first 
 place, this must be regarded as an intentional, not an acci 
 dental variation. That is, whichever of the two was the 
 original number, the other was the result of a conscious and 
 calculated alteration. As it will make no difference in the ar 
 gument, we may assume, what appears to be most probable 
 in itself, that four hundred and eighty was the original read 
 ing, and four hundred and forty either a change made by the 
 LXX. translator or one which he found already made by some 
 
396 HEBREW CHRONOLOGY. 
 
 one else, and represented in the MS. from which he render 
 ed. It is true that Winer and Thenius explain the variation 
 as arising by accident from a confusion between the Hebrew 
 letters eighty and forty. But there are these two things to 
 be said against This : i. That the letters in question, though 
 a good deal like each other in the present Hebrew square 
 character, have much less resemblance in the ancient forms 
 of the Phoenician inscriptions and the Maccabean coins, and 
 therefore, so far as we can judge, were not particularly liable 
 to be confounded in MSS. of older elate than the LXX. ver 
 sions. 2. Aside from such a special resemblance, we must 
 hold it in a high degree improbable that a purely accidental 
 variation should preserve the extraordinary characteristic of 
 the original number that of being exactly divisible by forty. 
 The chances against it can hardly be less than six to one. 
 We think it highly probable, therefore, that we have here an 
 intentional variation such as are well known to exist on a 
 larger scale in the patriarchal chronology of Genesis, ch.v. and 
 xi. ; where there is no doubt that the difference is the result of 
 calculation on one side or the other, and perhaps on both. 
 But if any one designedly reduced the number from four hun 
 dred and eighty to four hundred and forty, he must have re 
 garded the four hundred and eighty as made up by the addi 
 tion of forties, and have satisfied himself in some way that 
 too many forties by one had been allowed in the calculation : 
 in other words, he must have regarded it as belonging to an 
 artificial chronology proceeding by forties. Now it is cer 
 tainly quite supposable that the author of this variation was 
 mistaken in his view of the original number ; but, at any 
 rate, it is a remarkable circumstance, to which we can hardly 
 refuse some degree of weight, that the possible relation (as 
 we have presented it) between the four hundred and eighty 
 of I. Kings and the forty of the detailed chronology , was ob 
 served not more than three centuries, and perhaps much less, 
 after the completion of the books of Kings. 
 
 But there is an important circumstance which lends a fur 
 ther degree of plausibility to the supposition that the number 
 in I. Kings vi. I is to be regarded as an aggregate of so 
 many forties. We have reason to believe that the most 
 
HEBREW CHRONOLOGY. 
 
 397 
 
 complete and exact genealogical registers preserved among 
 the Hebrews gave eleven or twelve generations for the inter 
 val from the Exodus to the building of the temple. If among 
 the family lists for that period which we find in the book of 
 Chronicles \ve had to choose the one which could with most 
 probability be relied upon for exactness and completeness, we 
 should have no hesitation in taking that which gives us the 
 descendants of Aaron, the family which inherited the preemi 
 nent dignity and consequence of the high-priesthood. Now 
 here we find (I. Chron. vi. 50-53) twelve generations enu 
 merated, beginning with Aaron and closing with Ahimaaz the 
 son of Zadok. Ahimaaz, it will be remembered, was a con 
 temporary of David, though certainly a younger contempo 
 rary, since his father Zadok was still living at the commence 
 ment of Solomon s reign. If therefore I present it merely 
 as a supposition if on the basis of this genealogy one should 
 undertake to assign the number of generations between the 
 Exodus and the building of the temple, he would give twelve 
 or eleven, according as he included or excluded Ahimaaz, 
 the last in the series. But in the same connection we find 
 two other genealogies, commencing with Levi and ending re 
 spectively with Asaph the Gershomite and Ethan the Mera- 
 rite, whom " David set over the service of song in the house 
 of the Lord, after that the ark had rest." In the former of 
 these we find from first to last fifteen generations. Now in 
 the high-priestly series just mentioned, the fourth from Levi is 
 Aaron. If, then, in the lineage of Asaph we suppose the one 
 named fourth from Levi to have been a contemporary of 
 Aaron, we find here twelve generations subsequent to the 
 Exodus. The lineage of Ethan gives us, on the same suppo 
 sition, eleven generations. It is true that the lineage of 
 Heman the Kohathite, a third chief of the temple singers, 
 which we find in the same chapter of I. Chronicles, appears to 
 give us twenty-two generations from Levi, or nineteen from 
 the Exodus to the time of David. It is also true that the re 
 maining genealogy which spans the chasm of the Judges, I 
 mean that of David s ancestors, assigns only eleven genera 
 tions from Juda the brother of Levi to David himself, which 
 on the same supposition as before would give eight genera- 
 
398 HEBREW CHRONOLOGY. 
 
 tions from the Exodus. Whatever may be thought of these 
 series, either absolutely or in their relation to each other, the 
 interesting fact remains, that a majority of the whole num 
 ber, and among them that one which would seem most likely 
 (judging d priori) to be exactly and completely given, make 
 either eleven or twelve generations between the Exodus and 
 the erection of the temple. Now it is a well-known fact that 
 among the Hebrews the number forty was used as a score is 
 with us, for a large indefinite number. It would not be sur 
 prising, therefore, if a Hebrew author, in making a chrono 
 logical estimate, should allow forty years for a generation, 
 though this is certainly beyond the real mark ; the true aver 
 age for a generation being probably not much more than 
 thirty years. In this manner we might find a plausible ex 
 planation for the four hundred and eighty of our Hebrew text, 
 and a plausible explanation at the same time for the four hun 
 dred and forty, the early and remarkable variation of the 
 LXX. The difference of those numbers would be accounted 
 for by the recognition either of twelve or of eleven genera 
 tions, both of them naturally suggested by the genealogical 
 series, for the interval in question. 
 
 But if the number four hundred and eighty may have been 
 founded on a reckoning of twelve generations at forty years 
 each, we can hardly doubt that in the view of him who first 
 assigned it, whether the compiler of I. Kings or some one 
 before him, it would be connected with the forties of which 
 so many are .found in the Pentateuch, Judges, and Samuel. 
 The forty years wandering under Moses, the forty years of 
 Othniel, the twice forty of Ehud, or perhaps Ehud and Sham- 
 gar together, the forty of Deborah and Barak, the forty of 
 Gideon, the forty of Philistine oppression, including perhaps 
 the deliverance of Samson, the forty of Eli, and the forty of 
 David, would give nine forties. It can hardly be doubted 
 that forty more would be allowed for the generation of Joshua 
 and the elders who survived him ; .and again, forty more for 
 the generation of Samuel and Saul. This would make eleven 
 forties or four hundred and forty years, the number of the 
 LXX. As for the remaining forty of the Hebrew text, it 
 might be made by allowing two generations for Joshua and 
 
HEBREW CHRONOLOGY. 
 
 399 
 
 the elders, or by allowing two for Samuel and Saul. Both 
 of these suppositions, however, appear improbable. The 
 elders who overlived Joshua would most naturally be viewed 
 as belonging to the same generation ; and Sarnuel would ap 
 pear to have continued his prophetic activity until late in the 
 reign of Saul. Another supposition is suggested by the 
 contents of the book of Judges. The principal figures of 
 this book are Othniel, Ehud, Deborah and Barak, Gideon, 
 Jephthah, Samson. These alone deliver Israel from an actual 
 subjugation by foreign oppressors. Their actions receive a 
 prominence in the narrative, and are rehearsed for the most 
 part with a minuteness, altogether in contrast with the brief 
 and passing notices of the remaining judges. Each one of 
 these is introduced with a statement that the children of 
 Israel did evil in the sight of the Lord ; that the Lord in con 
 sequence delivered them into the hands of a particular enemy, 
 to whom they were subject for a specified time ; that the 
 people then, in their distress, cried unto the Lord (this particu 
 lar is omitted only in the case of Samson), and that an indi 
 vidual hereupon appeared to act with divine sanction as the 
 deliverer and judge of his people. None of these particulars 
 is found in the case of Shamgar, Tola, Jair, Ibzan, Elon, 
 Abdon. Now it is a curious fact that the number forty 
 occurs only in connection with the former class of judges, and 
 it is found in connection with all of them excepting Jephthah. 
 This coincidence in a large number of particulars may warrant 
 the suspicion that there was once complete coincidence where 
 we now find a single exception. It seems not unlikely that 
 the minute account of the six heroic deliverers may have 
 come from a source in which only they appeared, with a 
 chronology by forties for all of them, Jephthah not excepted ; 
 while the brief notices of the remaining six judges, and per 
 haps all the precise numbers of the book, are derived from 
 other sources. The number four hundred and eighty then 
 may have come, not from our present book of Judges, but 
 from the source referred to, with its chronology of forties 
 only. I am aware how little certainty can attach to conjec 
 tures of this kind, even when suggested by the obvious fea 
 tures of the works to which they relate. But the object is 
 
400 HEBREW CHRONOLOGY. 
 
 not to show how the four hundred and eighty actually was 
 connected with the forties of the earlier narrative for this 
 we could hardly hope to render certain but only to show 
 how it may have been connected. 
 
 Another argument which has its bearing upon the subject 
 is founded upon a connection with Egyptian history. It is 
 well known that Manetho, in a long extract quoted by Jose- 
 phus c. Apion, i. 26, describes a great removal of leprous 
 and unclean persons from lower Egypt under Osarsiph, an 
 apostate priest of Heliopolis, who afterwards (he says) as 
 sumed the name of Moses. This transaction is referred by 
 Manetho to the reign of Menophis or Amenophis (as the name 
 stands in the text of Joscphus), who is identified by what ap 
 pear to be convincing proofs with Menephta, a king of the 
 XlXth Egyptian dynasty. His reign, according to Lepsius, 
 extended from 1328 to 1309 (ace. to Bunsen, 1325-1307) ; 
 and, if so, preceded the building of Solomon s temple by little 
 more than three centuries a period much less than the four 
 hundred and eighty or four hundred and forty years which 
 we have been criticising, though sufficient to allow about 
 eleven generations of the ordinary or average length, and pos 
 sibly not incompatible with twelve generations in a particular 
 case, where the average may have chanced to be rather 
 smaller than usual. But into this argument I do not pro 
 pose to enter. To determine the degree of weight which 
 should be given to it would require an extended discussion, 
 demanding at once more Egyptological science than I can 
 lay claim to, and more time than my hearers could well 
 afford me. I shall content myself with having stated the 
 considerations which appear to be suggested by a critical 
 view of the Hebrew sources; and will only add a word in 
 conclusion, to guard against a possible misconception of 
 my meaning. 
 
 When pointing out in the detailed Chronology from Moses 
 to Solomon the extraordinary repetition of the number forty, 
 and inferring thence the peculiar character of this chronology, 
 I represented the conclusion as one which it seemed to me 
 not easy to avoid. As to the other point that the same 
 peculiar character may perhaps belong to the four hundred 
 
HEBREW CHRONOLOGY. 
 
 401 
 
 and eighty of I. Kings I am far from intending to speak 
 with anything like equal confidence. I have aimed to repre 
 sent this as a conjecture, to which a number of circumstances 
 lend more or less of plausibility, but by no means as a proved 
 or ascertained result. It is no doubt much more difficult to 
 establish any such conclusion as to a single number, which 
 overleaps centuries at a bound, than for an extended series, 
 where the disproportionate frequency of one designation fur 
 nishes evidence hardly to be resisted. But the intrinsic diffi 
 culty of obtaining evidence in regard to the historical exact 
 ness of the four hundred and eighty does not justify us in as 
 serting that it is inexact or unhistorical without sufficient evi 
 dence to prove it so. And I confess that the evidence in this 
 case apart from that which may be furnished by the Egyp 
 tian history, for that I do not feel myself prepared to criticise 
 the evidence to be derived from biblical sources alone, does 
 not appear to me by any means decisive against the histori 
 cal character of the number. At the same time, I am free to 
 acknowledge that I cannot look upon it as I should if the cir 
 cumstances here brought together had no existence ; and that 
 my faith in the commonly received chronology of the early 
 ages, which relies on this number as a bridge to overpass the 
 chasm that separates historical from patriarchal times, is far 
 from being clear and confident. 
 
 One word more, in acknowledgment of obligations. The 
 idea that the period of four hundred and eighty years in I. 
 Kings was determined by a series of generations reckoned at 
 .forty years each is said to have been suggested by Gehringer, 
 in a program Ueber die biblische Acre, published at Tubingen 
 in 1842. This I have not seen : its author appears to have 
 maintained, nevertheless, the historical exactness of the num 
 ber, and to have rearranged the chronology of the Judges in 
 order to make it harmonize with this statement. The same 
 idea is developed with much clearness and caution by Ber- 
 theau of Gottingen, in the introduction to his Commentary 
 on the Judges, published in 1845, as a part of the Excgctisches 
 HandbucJi for the Old Testament. It is controverted by 
 Thenius, in his Commentary on Kings published in 1849 as a 
 part of the same manual: It is taken up and carried out still 
 26 
 
402 HEBREW CHRONOLOGY. 
 
 further, with abundance of learning, by Lepsius, in his great 
 work, Die Chronologie der sEgypter. These are the sources 
 from which I have derived in great part the materials for this 
 article, 
 
XX. 
 
 ON THE LANGUAGE OF PALESTINE AT THE 
 TIME OF CHRIST. 
 
 1868. 
 
 THE subject of this article is not the language used by 
 the writers of the New Testament, but the language of 
 its speakers, the actual language of the discourses and conver 
 sations which stand reported in the Greek of the New Testa 
 ment. 
 
 On the question, WJiat was the prevailing language of 
 Palestine in the time of onr Saviour? there has been great 
 difference of opinion and much earnest controversy. Some 
 have maintained that the mass of the people spoke Aramaic 
 only ; others that they spoke Greek only ; and yet others that 
 they were acquainted with both languages, and could use this 
 or that at pleasure. To understand the merits of the case, the 
 simplest way will be to take up each of the two languages in 
 question, and trace the indications of its use among the Pales 
 tine Jews of the first century. 
 
 We begin then with THE ARAMAIC (the Jewish- Aramaic 
 or Chaldee, in distinction from the Christian-Aramaic or 
 Syriac dialect). It is not unlikely that the long intercourse, 
 friendly and hostile, between the Kingdom of Israel and its 
 Aramaean neighbors on the north, especially the Syrians of 
 Damascus, may have produced some effect on the language 
 of the northern Israelites. But the effect must have been 
 much greater when the Kingdom of Israel was overthrown by 
 the Assyrians, the higher classes carried into other lands, and 
 their places filled by importations from tribes of Aramaean 
 speech. In the siege of Jerusalem by the Assyrians, a few 
 years later, it appears from the proposal of the Jewish chiefs 
 to Rabshakeh (II. K! xviii. 26) that the Aramaean language 
 was understood by the leading men of the city, though unin- 
 
404 LANGUAGE OF PALESTINE. 
 
 telligible to the people at large. The course of events during 
 the next century must have added to the influence of the 
 Aramaic in southern Palestine, until at length the conquest 
 by Nebuchadnezzar and the Babylonian captivity gave it a 
 decided preponderance. Surrounded for two generations by 
 speakers of Aramaic, the Juda^an exiles could not fail to ac 
 quire that language. It may be presumed that many, perhaps 
 most of them, still kept up the use of Hebrew in their inter 
 course with one another ; but some, doubtless, forgot it alto 
 gether. After the return to their own land the Aramaic was 
 still required for communication with many brethren out of 
 Palestine or in it, and with the officers or agents of the Persian 
 government, which seems to have made this the official lan 
 guage for the provinces between the Tigris and the Mediter 
 ranean (comp. Ezra iv. 7, 8). The progress of the change 
 which made the Hebrew a dead language, and put the Ara 
 maic in its place as a living one, cannot be distinctly traced 
 for want of literary monuments. But the result is certain ; it 
 was complete at the Christian era, and may have been so two 
 or three centuries earlier. It is true that the New Testament 
 in several passages speaks of the Hebrew as if still in use ; 
 but in some of these (John v. 2 ; xix. 13, 17) it is evident 
 from the form of a word described as Hebrew (BriOecrSd, 
 ra/3/3a0a, To\yo9a) that the Aramaic is meant, the current 
 language of the Hebrew people. In many other cases, where 
 words of the popular idiom are given in the New Testament, 
 but without being called Hebrew, they can only be explained 
 from the Aramaic : thus Matt. v. 22, paicd; vi. 24 (Luke xvi. 
 9, 13), /AajjLwvds ; xvi. 17, ftap Iwva ; Mark v. 41, Takuda /eotyu ; 
 vii. 34, ty^afld; xiv. 36, A/3/3d ; John i. 43, K^a? ; Acts i. 
 19, AKe\$a/nd ; I. Cor. xvi. 22, papdv dOd to which add the 
 words paftfti, pajSftovvi, fieao-ia^^ Trda^a, and proper names 
 beginning with Bar ( son ). By Josephus, too, the name 
 Hebrew is often used to denote the popular Aramaic : thus 
 ISftjyLta, red (Ant. ii. I, i), %avawi9, priests (iii. 7, i), 
 AaapOd, Pentecost (iii. 10, 6), e/uW, priest s girdle (iii. 
 7, 2), all of which he designates as Hebrew, are evidently 
 Aramaic. 
 
 That this Jewish-Aramaic was not confined to a fraction of 
 
LANGUAGE OF PALESTINE. 405 
 
 the people, but was in general and familiar use among the 
 Jews of Palestine in the first century, is proved by a variety 
 of evidence, outside of the New Testament as well as in it. 
 Josephus speaks of it repeatedly (B. J. pr. I, v. 6, 5, v. 9, 
 2) as TI Trdrpios YXwcrcra, the tongue of the fathers and father 
 land, or, as we should say, the mother tongue, the native, ver 
 nacular idiom. As such he contrasts it with the Greek, which 
 he describes (Ant. pr. 2) as a\\oa,7rr)V rjplv KOL gevrjs 
 8ia\efCTOv crvvrfQeiav, a mode (of expression) alien to us and 
 belonging to a foreign language. From Josephus we learn 
 (B. J. v. 6, 3) that, in the siege of Jerusalem, when the 
 watchman on the towers saw a heavy stone launched from the 
 Roman catapults, he cried in the native tongue " the missile 
 is coming ; " he \vould, of course, give warning in the language 
 best understood by the citizens at large. Josephus himself, 
 when sent by Titus to communicate with the Jews and per 
 suade them to surrender, addressed the multitude in Hebrew 
 (B. J. v. 9, 2), which he would not have done if the language 
 had not been generally intelligible and acceptable. For fur 
 ther proof we might appeal to the Targums or Chaldee para 
 phrases of parts of the Old Testament, of which the oldest, 
 that of the Pentateuch by Onkelos, was probably written not far 
 from the time of Christ ; but it is possible that these Targums 
 may have been composed, not for the Jews of Palestine, but 
 for those of Babylonia and the adjacent countries ; as Josephus 
 states (B. J. pr. i) that the first edition of his own history 
 was composed in the native tongue (rfj Trarplw) for the Bar 
 barians of the interior (rot? avco ffapfidpots). Of more 
 weight as proof of a vernacular Aramaic in Palestine is the 
 early existence of a Hebrew gospel (/. e. an Aramaic, or, as 
 Jerome calls it, Syro-Chaldaic gospel, " Chaldaico Syroque 
 scrmone conscriptum "), commonly ascribed to the Apostle 
 Matthew. Papias, bishop of Hierapolis, who flourished in the 
 first half of the second century, speaks of such a book, and 
 holds it for the composition of the Apostle. He may have 
 been mistaken as to the authorship ; but as to the existence 
 of an Aramaic gospel at a very early period, there is no 
 sufficient ground to discredit his testimony. It appears then 
 that there was a body of people in Palestine during the first 
 
406 LANGUAGE OF PALESTINE. 
 
 century to whom it seemed desirable to have the gospel in 
 Aramaic, perhaps not solely as being more intelligible, but as 
 recommended also by patriotic or sectarian feeling. 
 
 Turning to the New Testament, we find it stated (Acts 1.19) 
 that when the catastrophe of Judas became known to the inhab 
 itants of Jerusalem, the place where it occurred was called A/ceX- 
 Sapd, field of blood/ a name clearly Aramaic ; and that it was 
 called thus rfj i$ia Sia\efcr(D avra)i>, in their own dialect. 
 This does not imply that the Aramaic belonged to the inhab 
 itants of Jerusalem exclusively, so as to be spoken by no 
 other population ; nor that it belonged to them as their only 
 language, so that no other tongue was spoken in the city ; 
 but that it belonged to them more properly than any other 
 tongue which might be spoken there, which could only be 
 true of the native vernacular, rj Trdrpios y\wo-cra. A strong 
 light is thrown on this whole subject by the account of Paul s 
 address to the people of the city (Acts xxi. 27 ff.). The 
 Apostle, having been rescued by the chief captain from a mob 
 who sought to kill him, \vas about to be taken to the castle ; 
 but was allowed at his own request to address the multitude. 
 " And when there was made a great silence, he spake unto 
 them in the Hebrew tongue." "And when they heard that 
 he spake in the Hebrew tongue to them, they kept the more 
 silence." (Acts xxi. 40; xxii. 2.) It is plain that he took 
 them by surprise. If they did not know him for a native of 
 the Greek city Tarsus, they had heard him charged with 
 bringing Greeks into the temple ; and they expected him to 
 use the Greek. When they found him speaking Aramaic, 
 they showed by their greater attentiveness that they were not 
 only surprised but gratified ; not that a Greek address would 
 have been unintelligible, and perhaps not on account of any 
 prejudice against the language, but because the speaker, by 
 adopting an idiom that was peculiarly their own, evinced his 
 respect for their nationality, his sympathy with their feelings, 
 and, as it were, made himself one of their number. 
 
 Of our Lord himself it is expressly stated that on three oc 
 casions he made use of the Aramaic : when with the words 
 rakiOa Kovjjii he raised the daughter of Jairus (Mark v. 41) ; 
 when with e^aOd he opened the ears of the deaf man (Mark 
 
LANGUAGE OF PALESTINE. 407 
 
 vii. 34) ; and when upon the cross, paraphrasing the first 
 words of Psalm xxii., he cried eXou, eX&K, \a^a 
 (Mark xv. 34 ; in Matt, xxvii. 46, 7J\,i, rj\,i, \rjfjua 
 It is hardly supposable that among all his utterances recorded 
 in the Gospels these three were the only ones for which he 
 used the native idiom of the country. Yet it is not easy to 
 say why out of a larger series these alone should be given in 
 the original form. In the last case it seems probable that the 
 Aramaic words actually uttered by our Lord were given by 
 the writer, to explain how it was that some of the bystanders 
 conceived him to be calling on Elias. As to the other two, it 
 is noteworthy that they appear in only one of the Evangelists. 
 The miracle wrought with the word efrfiaOd is found in Mark 
 alone : the miracle wrought with ra\(,6a Kovfja is found in Luke 
 also, but the words ascribed to our Lord (viii. 54) are Greek, 
 77 Trot?, eyetpov showing how unsafe it is in other cases to 
 conclude that he spoke Greek because he is not said to have 
 spoken Aramaic. It is not an unlikely supposition that in 
 these two instances the narrative of Mark reflects the impres 
 sions of an individual, whose mind was peculiarly struck by 
 the stupendous effect instantly following, and seemingly pro 
 duced by, the utterance of one or two words, so that the very 
 sound of the words became indelibly fixed in his memory. 
 That the same subjective impression was not made in other 
 cases of the same kind, or that being made it did not find its 
 way with uniformity into the narrative, are both easily con 
 ceivable. There is, however, yet another instance in which 
 our Lord is expressly stated to have spoken Hebrew (Aram 
 aic) : in his appearance to Paul when journeying to Damascus. 
 Of this event there are three narratives (Acts ix., xxii. , xxvi.) ; 
 and here again it is worth noticing that among the parallel 
 accounts only one (xxvi. 14) alludes to the fact that the lan 
 guage used was Hebrew. An able writer, who holds that 
 Christ seldom spoke Hebrew, suggests that he used it on this 
 occasion to keep his words from being understood by Paul s 
 companions. But if these companions failed to hear or to 
 understand the voice (Acts ix. 7 ; xxii. 9), it is not safe in an 
 event of this nature to infer their ignorance of the language. 
 And it is quite supposable that the use of Hebrew here be- 
 
408 LANGUAGE OF PALESTINE. 
 
 longed to the verisimilitude of the manifestation, Jesus ap 
 pearing to this new apostle not only with the form in which 
 he was known to the Twelve, but with the language in which 
 he was accustomed to converse with them. 
 
 The influence of THE GREEK in Palestine began with the 
 conquest by Alexander. The country fell under the power 
 of Macedonian rulers, the Ptolemies of Egypt, and afterwards 
 the Seleucidae of Syria, with whom Greek was the language 
 of court and government. It was used for the official cor 
 respondence of the state ; for laws and proclamations ; for 
 petitions addressed to the sovereign, and charters, rights, or 
 patents granted by him. The administration of justice was 
 conducted in it, at least so far as the higher tribunals were 
 concerned. At the same time, commercial intercourse be 
 tween the countries under Macedonian rule came into the 
 hands of men who either spoke Greek as their native tongue 
 or adopted it as the means of easiest and widest communica 
 tion. Partly for purposes of trade and partly as supports for 
 Macedonian domination, colonial cities w r ere planted in these 
 regions, and settled by people who, if not all of Hellenic 
 birth, had the Greek language and civilization, and bore the 
 name of Greeks. Such influences were common to the 
 countries about the eastern Mediterranean ; and their effect 
 in all was to establish the Greek as the general language of 
 public life, of law, of trade, of literature, and of communica 
 tion between men of different lands and races. It did not in 
 general supplant the native idioms, as the Latin afterwards 
 supplanted those of Gaul and Spain : it subsisted along with 
 them, contracting but not sw r allowing up the sphere of their 
 use. Its position and influence may be compared with those 
 possessed, though in a much inferior degree, by the French 
 language in modern Europe. The sway of the Greek ex 
 tended to lands never conquered by Alexander. To a lan 
 guage so capable, so highly cultivated, so widely diffused, so 
 rich in literature and science, the Romans could not remain 
 indifferent, especially when the regions where it prevailed 
 became part of their empire. Long before the Christian era 
 a knowledge of Greek was an indispensable element in the 
 training of an educated Roman. In the reign of the Em- 
 
LANGUAGE OF PALESTINE. 
 
 409 
 
 peror Tiberius, under whom our Lord suffered, we are told 
 (Val. Max. ii. 2, 3) that speeches in the Roman Senate 
 were often made in Greek. The emperor himself, acting as 
 judge, frequently heard pleadings and made examinations in 
 it. (Dion. Cass. Ivii. 15.) Of the Emperor Claudius, a few 
 years later, it is said (Sueton. Claud. 42) that he gave audi 
 ence to Greek ambassadors speaking in their own tongue, and 
 made replies in the same language. 
 
 The people of Palestine were subjected to Hellenizing in 
 fluences of a special character. Their Seleucid rulers, not 
 content with the natural operation of circumstances, made 
 strenuous efforts to impose upon them the Greek culture and 
 religion. The great national reaction under the Maccabees, 
 provoked by these efforts, was of no long duration. The Ro 
 mans became masters of the country ; and must have given 
 new force to the Greek influences to which they had them 
 selves yielded. It cannot be doubted that the Roman ad 
 ministration of state and justice in Palestine was conducted in 
 the Greek, not the Latin language. The first Herod who 
 reigned for many years under Roman supremacy was mani 
 festly partial to the Greeks. Caesarea, which he founded, and 
 made, after Jerusalem, the greatest city in the land, was 
 chiefly occupied by Greek inhabitants. Of many other cities 
 in or near the Holy Land, we learn, mostly from incidental 
 notices, that the population was wholly or partly Greek. 
 Thus Gaza, Ascalon, Joppa, Ptolemais, Dora, as well as Cae- 
 sarea, on the western sea- coast ; Tiberias and Sebaste in the 
 interior ; and on the east and northeast, Hippos, Gadara, 
 Scythopolis (or Bethshan), Pella, Gerasa, Philadelphia, and 
 perhaps the remaining cities of the Decapolis. It is obvious 
 that the Jews must have been powerfully affected by so many 
 Greek communities established near them and connected 
 with them by manifold political relations, and especially the 
 Jews of Galilee, surrounded as they were and pressed upon 
 by such communities. 
 
 While many Greeks were becoming settled in Palestine, 
 Jews in yet larger number were leaving it to establish them 
 selves in all the important places of the Grecian world. With 
 out losing their nationality and religion, they gave up their 
 
410 LANGUAGE OF PALESTINE. 
 
 Aramaic mother-tongue for the general language of the people 
 round them. Had the Jews of Egypt retained the native 
 idiom, the first translation of the Scriptures would probably 
 have been made in Aramaic, and not in Greek. Even Philo 
 of Alexandria, an older contemporary of our Lord, gives no 
 evidence in his voluminous and learned writings of an ac 
 quaintance with either Hebrew or Aramaic. But these Jews 
 of the dispersion frequently returned to their fatherland ; they 
 gathered in crowds to the great national festival ; and in per 
 sonal communication with their Palestinian kindred, did much 
 to extend the use of their adopted language. In many cases 
 they continued to reside in Palestine. Thus we hear (Acts 
 vi. 9) of one or more synagogues of Libertines (Jewish freed- 
 men from Italy), Cyrenians, Alexandrians, Cilicians, and 
 peoples from western Asia Minor. That many would con 
 tent themselves with their familiar Greek, as being sufficient 
 for the ordinary purposes of communication, without taking 
 the trouble to learn Aramaic, is a fact which can hardly be 
 doubted. It is generally believed that the Hellenists men 
 tioned in Acts ix. 29, and (as converts to Christianity) in Acts 
 vi. I, were persons of this sort separated from those around 
 them not by speaking Greek (for most others could do so), but 
 by speaking only Greek. The satisfaction which Paul gave 
 by his use of Aramaic (Acts xxii. 2) makes it easy to under 
 stand how such persons, who being settled in Palestine dis 
 dained to acquire the native idiom, might be looked upon 
 with coldness or disfavor as a class by themselves, especially 
 if they showed, as may often have been the case, a weakened 
 attachment to other features of the national life. 
 
 The Greek version of the LXX. did much to make the 
 Greek known and familiar to the Jews of Palestine. The 
 original Hebrew was an object of scholastic study ; a learned 
 acquaintance with it was highly valued in popular estimation 
 (Jos. Ant. xx. n, 2) ; and the number of scribes, lawyers, 
 etc. who possessed such knowledge was probably not incon 
 siderable ; but to the mass of the people the Hebrew Scrip 
 tures were a sealed book. Nor was there, so far as we know, 
 prior to the Christian era, any Aramaic version. To the com 
 mon man the man of common education if he had any 
 
LANG UA GE OF PALESTINE 411 
 
 V \ 
 
 knowledge of Greek, the most natural and eSfcg^Jyay to gaui 
 
 a knowledge of the Scriptures was by reading ~tlie Greek 
 translation. That such use w r as made of it by great numbers 
 of the people cannot well be doubted. Of the quotations 
 from the Old Testament made by the writers of the New, the 
 greater part are in the words of the LXX. Comparatively 
 few give any clear evidence that the writer had in mind the 
 Hebrew original. This familiarity with the Greek version 
 makes it probable that it was used not only for private read 
 ing, but *in the public services of the synagogue. In many 
 places there may have been no one sufficiently acquainted 
 with the ancient Hebrew to read and translate it for the con 
 gregation ; but in every community, we may presume, there 
 were persons who could both read the Greek and add what 
 ever paraphrase or explanation may have been needed in 
 Aramaic. It is apparent in the case of Josephus, that even 
 men of learning who had studied the Hebrew were familiar 
 with the version of the LXX. ; in his Antiquities Josephus makes 
 more use of the latter than of the former. To the influence 
 of the LXX. must be added that of a considerable Jewish- 
 Greek literature, composed mainly in the last two centuries 
 before Christ, the so-called Apocrypha of the Old Testament. 
 It is true that one of these books, the Wisdom of Jesus the 
 son of Sirach, is declared in its preface to be the translation 
 of a work composed in Hebrew (i.e., not improbably, in 
 Aramaic) by the grandfather of the translator. There is 
 much reason for believing also that the First Book of Macca 
 bees was written in Hebrew ; and the same may perhaps be 
 true of some other apocryphal books. The fact, however, 
 that no one of them is extant in that language seems to show 
 that in general use (except in countries east of the Syrian 
 desert) the Hebrew (or Aramaic) original was early super 
 seded by the Greek version. A case nearly parallel is seen in 
 Josephus s History of the Jewish War. It was composed (ac 
 cording to the statement of the preface) in the native tongue 
 for the barbarians of the interior, i. e. beyond the Syrian de 
 sert, the limit of the Roman power. But for those under the 
 Roman government he translated it into Greek (rot? Kara TTJV 
 rfj *E\\d8t, 7X060-0-77 neTa(3a\a)v). And 
 
412 LANGUAGE OF PALESTINE. 
 
 this translation has so thoroughly superseded the original 
 work that, but for the statement of its author, we should not 
 have known, or perhaps even suspected, its existence. 
 
 That Greek was generally understood by the people of 
 Jerusalem is evident from the circumstances of Paul s address 
 in Acts xxii. The multitude who listened with hushed atten 
 tion while he spoke to them in Aramaic were already atten 
 tive when expecting to hear him in Greek. It does not follow 
 that all understood him in the former language, or that all 
 would have understood him in the latter. To gain attention 
 it would be enough that a large majority could understand 
 the language of the speaker ; those who could not might 
 still get some notion of the speech, its drift and substance, by 
 occasional renderings of their fellows. 
 
 The Greek New Testament is itself the strongest proof of 
 the extent to which its language had become naturalized 
 among the Jews of Palestine. Most of its writers, though not 
 belonging to the lowest class, to the very po6r, or the quite 
 uneducated, were men in humble life, in whom one could 
 hardly expect to find any learning or accomplishment beyond 
 what was common to the great body of their countrymen. 
 We are not speaking of Saul or Luke or the unknown writer 
 of the Epistle to the Hebrews ; but of Peter, Jude, James, 
 John, and Matthew, if (as is most probable) we have his 
 gospel in its original language. Yet we find them not only 
 writing in Greek, but writing in a way which proves that they 
 were familiar with it and at home in it. They do not write it 
 with elegance or with strict grammatical correctness ; but they 
 show a facility, a confidence, an abundance of apt and forci 
 ble expression , which men seldom attain in a language not 
 acquired in early life. Some have found in the Hebrew idioms 
 which color their style an indication that they thought in 
 Hebrew (or Aramaic), and had to translate their thoughts 
 when they expressed them in Greek. But similar idioms oc 
 cur in the compositions of Paul, who as the native of a Greek 
 city must have been all his life familiar with the Greek lan 
 guage. When Greek began to be spoken by Hebrews, learn 
 ing it in adult years, they had to go through a process of 
 mental translation ; and the natural result was the formation 
 
LANGUAGE OF PALESTINE. 4! 3 
 
 of a Hellenistic dialect, largely intermixed with Semitic idioms, 
 which they handed down to their descendants. The latter, 
 as they did not cease to speak an Aramaic idiom, were little 
 likely to correct the Aramaic peculiarities in the Greek re 
 ceived from their fathers. Josephus speaks with emphasis of 
 the difficulty which even a well-educated Jew found in writing 
 Greek with idiomatic accuracy. The Greek style of a Jew, 
 especially when writing on religious subjects, was naturally 
 affected by his familiarity with the LXX. , which copied from 
 the original many Hebrew forms of expression, and kept them 
 alive in the memory and use of the people. 
 
 In view of these proofs, the conclusion seems unavoidable 
 that, as a general fact, the Palestine Jews of the first century 
 were acquainted with both languages, Greek and Aramaic. 
 It is probable, indeed, as already stated, that some were not 
 acquainted with the Aramaic ; and it is by no means im 
 probable, though the proof is less distinct, that some were 
 not acquainted with the Greek. Of both these classes the! 
 absolute number may have been considerable. But apparently/ 
 they were the exceptions, the majority of the people having a 
 knowledge more or less extended of both languages. Other 
 instances of bilingual communities, of populations able for the 
 most part to express themselves in two different tongues, are 
 by no means wanting. One of the most striking at the pres 
 ent day is to be found in a people of Aramaean origin with a 
 firmly held Aramaic vernacular, the Nestorian Syrians or 
 Chaldee Christians. "In Persia most of the Nestorians are 
 able to speak fluently the rude Tatar (Turkish) dialect used 
 by the Mohammedans of this province, and those of the 
 mountains are equally familiar with the language of the 
 Koords. Still they have a strong preference for their own 
 tongue, and make it the constant and only medium of inter 
 course with each other." (Stoddard s Preface to Modern 
 Syriac Grammar, in Journal of Amer. Oriental Soc., vol. v.) 
 
 It is a common opinion that by the pentecostal gift of 
 tongues (Acts ii.) the Apostles were miraculously endowed 
 with a knowledge of many languages and the power of using 
 them at pleasure. But this gift would seem from the tenor 
 of the accounts to have been a kind of inspiration under 
 
414 LANGUAGE OF PALESTINE. 
 
 which the speaker gave utterance to a succession of sounds, 
 without himself willing, or perhaps even understanding, the 
 sounds which he uttered. It does not appear from the sub 
 sequent history that the Apostles in their teaching made use 
 of any other languages than Greek and Aramaic. It is not 
 necessary to suppose that Paul spoke Latin at Rome, or Mal 
 tese in Melita (Acts xxviii.), or Lycaonian at Lystra (Acts 
 xiv.). In the transactions at Lystra it is pretty clearly im 
 plied that Paul and Barnabas did not understand the speech 
 of Lycaonia, and therefore failed to perceive and oppose the 
 idolatrous intentions of the people until they had broken out 
 into open act. 
 
 In choosing between the two languages which they un 
 doubtedly possessed, the Apostles were of course guided by 
 the circumstances. Outside of the Holy Land they would 
 generally, if not always, make use of the Greek. In Syria, in 
 deed, a considerable part of the people the same for which 
 the Peshito version was made in the next century would 
 probably have understood an address in the Aramaic of Pal 
 estine ; but in Antioch, the capital where the disciples were 
 first called Christians, Greek must have been the prevalent 
 language. Even in Palestine, Paul s addresses to the Roman 
 governors Felix and Festus would naturally be made in 
 Greek. This is not so clear of the address to Agrippa, who 
 had enjoyed a Jewish education. In the meeting of apostles 
 and elders at Jerusalem (Acts xv.), occasioned by events in 
 Antioch and attended by delegates from that city, the pro 
 ceedings were probably in Greek, as also the circular letter 
 which announced its result to " the brethren which are of the 
 Gentiles in Antioch and Syria and Cilicia." When Peter on 
 the day of Pentecost addressed the multitude of Jews gath 
 ered from many different countries, he would naturally use 
 the language which was most widely understood. It is true 
 that the " Parthians and Medes and Elamites and Arabians," 
 if no others, would have been most accessible to an Aramaic 
 address ; so we judge from the fact that Josephus, writing 
 for readers in these very lands, composed his history in the 
 native tongue. Still, when we consider the " dwellers in 
 Cappadocia, in Pontus and Asia, Phrygia and Pamphylia, in 
 
LANGUAGE OF PALESTINE. 
 
 415 
 
 Egypt and in the parts of Libya about Cyrene, and strangers 
 of Rome," it is probable that more would have understood 
 Greek than Aramaic; so that if there was only one address in 
 one language (which perhaps the terms of the narrative do 
 not require us to suppose), it was probably made in Greek. 
 
 The difficulty of determining the language used for each 
 particular discourse is even greater in the Gospels than in the 
 Acts. It seems reasonable to suppose that conversations be 
 tween kindred and friends, and the familiar utterances of 
 Christ to his disciples, were in Aramaic ; the native idiom of 
 the country,. if not wholly given up, would naturally be em 
 ployed for occasions like these. Yet as long as speakers and 
 hearers had another language at command, there always re 
 mains, in the absence of express statements, a possibility that 
 this, and not Aramaic, may have been used for any given 
 conversation. And if, on the other hand, it seems reasonable 
 to suppose that our Lord in his more public discourses spoke 
 Greek, there is a similar difficulty about being sure in particular 
 cases that he did not use the other language which was familiar 
 to him and to the mass of his hearers. A recent writer as 
 sumes that every discourse which, as reported to us, contains 
 quotations from the Old Testament in the words of the LXX., 
 must have been pronounced in Greek ; and this criterion, 
 were it trustworthy, would decide many cases. But if an 
 Aramaic speech containing Scripture quotations were to be 
 reported in Greek by a writer familiar with the LXX., who 
 seldom (if ever) read the Scriptures in any other form, is it 
 not probable that he would give the quotations for the most 
 part according to the LXX.? Sometimes, it is likely, he 
 would depart from it, because he did not correctly remember 
 its phraseology ; and sometimes because he remembered that 
 the Aramaic speaker gave the passage a sense varying from 
 that given by the LXX. As the writers of the Gospels were 
 probably in this condition of persons familiar with the LXX. 
 who seldom (if ever) read the Scriptures in any other form 
 it is unsafe from the way in which they give the Scripture 
 quotations to infer anything as to the language used by the 
 speakers who quoted them. There are instances, however, 
 in which the circumstances of the case afford some indications 
 
4i 6 LANGUAGE OF PALESTINE. 
 
 on this point. Thus in communicating with the people of 
 Gadara, which Josephus calls a Greek city, our Lord would 
 use the Greek language. Among the crowds who followed 
 him before the Sermon on the Mount and who seem to have 
 stood about the mountain while he was speaking, were some 
 from Decapolis (Matt. iv. 25). As already stated, the ten 
 cities of that region were (most, if not all, of them) Greek. 
 As our Lord had thus in the surrounding multitude of his 
 auditors some who probably were unacquainted with Ara 
 maic, there is plausible ground for believing that on this im 
 portant occasion he made use of the Greek language. In the 
 closing scenes of his life, when he was brought before the 
 Roman governor for judgment and execution, it is nearly cer 
 tain that Greek was used by Pilate himself and by the various 
 speakers about his tribunal. 
 
 It is stated in the Mishnah (Sotah, c. 9, n. 14) that, when 
 the war of Titus broke out, an order was issued in which 
 fathers were forbidden to have their sons instructed in Greek. 
 Whether this is true or not, it would be only natural that the 
 excited patriotism of such a time should cause the Jews to 
 set a higher value on their national tongue. Perhaps those 
 who spoke Greek and Aramaic were now inclined as far as 
 possible to discard the use of Greek ; the Targums, which 
 seem to have made their first appearance or to have assumed 
 a permanent shape about this time, would be a help in 
 doing so. At all events, there is reason for believing that 
 after this period there was a considerable population in Pales 
 tine who did not understand Greek. The general opinion of 
 the Fathers (from Clement of Alexandria down), that the 
 Epistle to the Hebrews was composed in Aramaic, had 
 probably no other foundation than the belief that it would 
 otherwise have been unintelligible to the Jews of Palestine 
 for whom it was designed. This belief is of little weight as 
 regards the original language of the epistle ; but as regards 
 the prevailing language of Palestine in later times it may not 
 be without value. Eusebius of Caesarea, a native and life 
 long resident of Palestine, declares (Dem. Evang. lib. iii.) 
 that the Apostles before the death of their Master understood 
 no language but that of the Syrians ; this he would hardly 
 
LANGUAGE OF PALESTINE. 
 
 417 
 
 have clone if Greek had been generally spoken by the Gal 
 ileans of his own day. 
 
 The discussion as to the language of Palestine in our 
 Saviour s time has been quite generally connected with the 
 question whether Matthew wrote his Gospel in Hebrew or 
 in Greek. Most defenders of the Hebrew original (as Du 
 Pin, Mill, Michaelis, Marsh, Weber, Kuinocl, etc.) have 
 maintained that this was the only language then understood 
 by the body of the people. And many champions of the 
 Greek original (as Cappell, Basnage, Masch, Lardncr, 
 Walaeus, etc.) have made a like claim for the Greek. For 
 a full list of the older writers, see Kuinocl in Fabricius, Bihl. 
 Grceca. ed. Harles. iv. 760. We add the names of some 
 writers who have treated the subject more at large. Isaac 
 Vossius (Dc Oraculis Sibyllinis, Oxon. 1680), though a 
 staunch believer in the Hcbre\v original, held that Greek was 
 almost universal in the towns of Palestine, and that the Sy- 
 riac still spoken in the country and in villages had become so 
 corrupted as to be a kind of mongrel Greek. He found an 
 opponent in Simon (Hist. Crit. du Tc.rtc du N. T., Rotterd. 
 1689), who allowed that Greek was the common language 
 (languc vulgairc") of the country, but contended that the Jews, 
 beside the Greek, had preserved the Chaklee which they 
 brought with them from Babylon, and which the}* called the 
 national language. Dioclati of Naples (Dc Christo Greece 
 loqucntc, 1/67 ; reprinted, London, 1843) went farther than 
 Vossius, asserting that Greek in the days of our Lord had 
 entirely supplanted the old Palestinian dialect. Replies to 
 this work were put forth by Ernesti (/// Ncucstc Tliccl. Bibl, , 
 1771) and De Rossi (Delia Lingua propria di Christo, Parma, 
 1772). De Rossi s work was adopted by Pfannkuche as the 
 basis of his essay on the Aramaean language in Palestine (in 
 Eichhorn s Allgem. Bibl., 1797), translated by E. Robinson 
 (in Am. Bibl. Rcpos., 1831) with an introduction on the 
 literature of the subject. Another translation (by T. G. 
 Repp) is given in Clark s Biblical Cabinet, vol. ii. Against 
 Pfannkuche, who is one-sided in his advocacy of the Aramaic, 
 Hug (Einl. in d. N. T., 4th ed., 1847; 3<3 ed - translated 
 by Fosdick, Andover, 1836) maintained the concurrent use of 
 27 
 
41 8 LANGUAGE OF PALESTINE. 
 
 Greek. His position which is nearly the same with that of 
 Simon is held substantially by most later writers, as Credner 
 (Einl. in d. N. T., Halle, 1836) and Bleek (Einl. in d. N. T., 
 Berlin, 1862). A somewhat more advanced position is taken 
 by -Dr. Alex. Roberts (Discussions on the Gospels , 2d ed., 
 London, 1863), who, while admitting that both languages 
 were in general use, contends that our Lord spoke for the most 
 part in Greek, and only now and then in Hebrew (Aramaic). 
 
INDEX. 
 
 tf, history of English pronunciation 
 of, 244, 246-248. 
 
 ablative case, Greek genitive as, 
 44-55 ; Latin ablative in rf, 47. 
 
 accent, Greek, nature and theory 
 of, 110-127; the various accents 
 and their signs, iio-iii ; differ 
 ence of pitch as an element in 
 Greek accent, 111-114, 127; 
 whether accompanied with stress 
 of voice, 113-117; question of 
 middle tone in, 117-119, 127; 
 laws of position of, 121-122; 
 .theory of cadence explaining them, 
 120-124. 
 
 accent, Latin, laws of, and theory 
 
 usage, 96-97 ; ratios of arsis to 
 thesis, 98. 
 
 aspirate mutes, Indo-European, 
 nature and history of, 168-183; 
 Sanskrit aspirates, 168-172 ; 
 Bactrian, 172; Greek, 172-175; 
 Latin, 172-173, 176; Slavonic and 
 Lithuanian, 176-177; Celtic, 177; 
 Germanic, 177-178; aspirates in 
 original Indo-European language, 
 178-179; their compound charac 
 ter, 179; question whether surd 
 or sonant, 179-183 ; Kuhn s views 
 discussed, 180-183. 
 
 Babylonian captivity, its effect on 
 
 of cadence explaining them, 124- language of Palestine, 404. 
 T ^/c Bactrian aspirate mutes, 172. 
 
 126. 
 
 acute accent, 1 10. 
 ^Eolic dialect, accentuation of, 123- 
 
 124. 
 Ahrens on e*o-roy, 64 ; on loss of 
 
 sigma, 78. 
 ai, early English pronunciation of, 
 
 246. 
 Anglo-Saxon genitive continued in 
 
 English possessive, 222-239. 
 Anglo-Saxon pronunciation, its re 
 lation to later English, 240-262 ; 
 
 relation of its vowel quantity to 
 
 English, 263-295. 
 Anglo-Saxon transliteration of Greek 
 
 passages, 128-140. 
 Aramaic or Chaldee, familiar lan 
 guage of Palestine at the time of 
 
 Christ, 403-408 ; its alternation in 
 
 use with Greek, 413-416. 
 Arauack passive formation, 210. 
 Aristides Quintilianus on Greek 
 
 rhythmic, "85-86. 
 Aristoxenus on Greek rhythmic, 85, 
 
 86, 90-91, 93-95 ; rhythmic feet as 
 
 defined by him, 99-103. 
 Armenian future, 189. 
 arsis and thesis, original meaning 
 
 f> 95 "9^ ; exchanged in modern 
 
 Bekker s cligammated edition of 
 Homer criticised, 56-80. 
 
 Bentley s views of Greek metric, 
 97, 104. 
 
 Biblical sevens, 330-334. 
 
 Bunsen s views of the Ionian Migra 
 tion, 26-27. 
 
 Bursian s views of do., 27. 
 
 Byron, character and tendency of 
 his writings, 346-349. 
 
 Byzantine Greek pronunciation of 
 the loth century, 128-140. 
 
 Csesar, Julius, his labors on Greek 
 rhythmic and metric, 88-89, Io ^- 
 
 Carlyle, defects in his theory of 
 government, 361-362. 
 
 Celtic, aspirate mutes of, 177, 182 ; 
 passive formation of, 202. 
 
 Chaucer, rhymes of, 255-257 ; or 
 thography proposed for, 257-259. 
 
 Chinese, passive expression in, 212. 
 
 chronology of the Hebrews, 385- 
 402. 
 
 circumflex accent, iio-iii. 
 
 Classen s views of the Ionian Migra 
 tion, 21-23. 
 
 Crosby s view of the Greek genitive, 
 45-46. 
 
42O 
 
 INDEX. 
 
 Curtius, E., his theory of Ionian Mi 
 gration explained and discussed, 
 1-36. 
 
 Curtius, G., on the j-sound in 
 Greek, 60, 72-75 ; on the Greek 
 relative stem, 76-77 ; on Greek 
 accent, 119. 
 
 day-dreaming, indulgence in, 369- 
 
 373- 
 Demosthcncs s perfection as orator, 
 
 350-35 1 - 
 
 digamma, its value, 58 ; its appear 
 ance in Homer, 58-80 ; its treat 
 ment in connection with the aug 
 ment, 68-69 ; with the reduplica 
 tion, 70-72. 
 
 digammated edition of Homer, Bek- 
 ker s, criticised, 56-80. 
 
 Dionysius of Halicarnassus on 
 Greek accent, 112-114. 
 
 Dondorff s views of the Ionian Mi 
 gration, 31 . 
 
 Duncker s views of do., 20-21. 
 
 e, history of English pronunciation 
 of, 244-245 ; unaccented final c, 
 257. 
 
 Egyptian records, their bearing on 
 the Ionian theory, 11-13, 26-27. 
 
 f/, early English pronunciation of, 
 246. 
 
 Ellis, A. ]., his account of Anglo- 
 Saxon transliteration of Greek, 
 128-140; his view of Greek pro 
 nunciation in loth century, 138- 
 140 ; his Early English Pronuncia 
 tion reviewed, 240-262. 
 
 eloquence, question of superiority 
 between ancient and modern, 349- 
 
 35 1 - 
 
 English, Ellis s work on early pro 
 nunciation of, reviewed, 240-262 ; 
 history of pronunciation of vowels, 
 244-251; of consonants, 251-255 ; 
 pronunciation of Chaucer and 
 Gower, 255-259 ; his insufficient 
 treatment of vowel quantity, 260- 
 262. 
 
 English orthography, its value, 351- 
 356. 
 
 English vowel quantity in I3th and 
 1 9th centuries compared, 263- 
 295 ; nature of distinction of long 
 and short vowel, 263-265 ; persis 
 tence of English quantities, 265 ; 
 
 evidence of the Ormulum as to 
 this, 265-295 ; general accordance 
 between Ormulum and modern 
 English, 267; cases of discord 
 ance, 268-295 ; vowel lengthened 
 where a consonant lost, 268-273 ; 
 before a weak r, 273-277 ; before 
 /, 277-278 ; before ;;//;, nd, ng, 
 278-281 ; vowel shortened before 
 other consonant combinations, 
 281-284; lengthened in an open 
 syllable, 285-290; other isolated 
 changes, 290-293 ; changes in 
 endings and suffixes, 293-295. 
 Europe, its tendency toward repub 
 licanism, 361-369. 
 
 foot, meaning of the term in metric, 
 8 1 ; composed of arsis and thesis, 
 95-99 ; rhythmic feet according to 
 Aristoxenus, 99-102 ; compound 
 feet, 102-106; irrational feet, 106- 
 108. 
 
 forty, predominance of, in the He 
 brew chronology, 390-402. 
 
 futures, formation of, in Indo-Euro 
 pean languages, 184-198; the fu 
 ture in Semitic, 184-185 ; in Ger 
 manic, 186; in Celtic, 187-188, 
 192 ; relation of future to sub 
 junctive or potential, 188-189; 
 Armenian future, 189 ; Greek, 
 Sanskrit, and Lithuanian future 
 in s, 189-191 ; Latin futures, 189, 
 191-193; Slavonic, 192-194 ; Ro 
 manic, 194; later Germanic, 195- 
 196 ; question of original Indo- 
 European formations, 196-198. 
 
 , its suppression has lengthened the 
 preceding vowel in English, 269- 
 270; softened into the consonant 
 j -sound, 270-271. 
 
 genitive, Greek, as an ablative 
 case, 44-55 ; Kiihner s treatment 
 of the case, 44-46 ; criticisms on 
 it, 46-55 ; relation of Greek case- 
 system to Indo-European, 47-48 ; 
 Greek genitive represents more 
 than one original case, 49-55. 
 
 German, dialectic possessive ex 
 pressions, 230-231 ; passive for 
 mation, 200-201. 
 
 Germanic languages, want of dis 
 tinction between the present and 
 
INDEX. 
 
 future, 186; later auxiliary fu 
 tures., 196. 
 
 Germany, Napoleon s influence on, 
 35S-359 ; tendency to unity in, 
 359-360. 
 
 gk, pronunciation of in early Eng 
 lish, 254. 
 Goldstiicker, his .view of English 
 
 possessive, 226. 
 
 Gothic aspirate mutes, 177 ; pas 
 sive formations, 208-211. 
 government, its ideal office and 
 
 perfection, 362-363. 
 Gowcr, rhymes of, 255-257. 
 grave accent, 1 1 1. 
 Greek accent, nature and theory of, 
 
 110-127 an d see accent. 
 Greek aspirates, their original pro 
 nunciation, 172-175. 
 Greek future, 189-191. 
 Greek genitive as an ablative, 44- 
 
 56. 
 
 Greek language, relation to Italican, 
 141 ; Greek as used in Palestine 
 at the time of Christ, 408-413. 
 Greek, modern, Ross s view of the 
 antiquity of its pronunciation, 161- 
 162; Ellis s do., 138-140; its loss 
 of ancient vowel quantity, 263. 
 Greek passive, 206-209. 
 Greek pronunciation in loth cen 
 tury, 128-140. 
 Greek rhythm and metre, 81-109 
 
 and see metre. 
 
 Greeks and Italicans, Ross on the 
 
 relation of their languages, 141- 
 
 163 ; their true relations, 163-167. 
 
 Greeks, their superiority in arts, 350. 
 
 Greek words and letters : 
 
 au, its sound in loth century, 135- 
 
 136. 
 
 f, origin of its name, 135. 
 
 ZKIICSTOS, 64, 78-80. 
 
 eKure/)or, eKarepdf, 79-80. 
 
 e Xcop. etc., 65-66. 
 
 eoiKa, etc., 72-73. 
 
 fog. 78. 
 
 epvofjiai and e /n o), 66-68. 
 
 CTOS, 65. 
 
 eu, its sound in loth century, 135- 
 
 136. 
 17, its sound in loth century, 132- 
 
 133- 
 fiVcro, 69-70. 
 
 s, derivation of, 38-43. 
 and tiyjLit, 73-74. 
 
 421 
 
 aoi/fj, IO. 
 ot, its sound in loth century, 134- 
 
 f)Ovg, 38. 
 
 v, its sound in loth century, 132 ; 
 
 its name, 135. 
 v\rj, 78. 
 
 Gutschmid s views of the Ionian 
 Migration, 27-31. 
 
 h, its pronunciation in early Eng 
 
 lish, 254; its suppression some 
 
 times lengthens the preceding 
 
 vowel, 271, 272. 
 
 Hammer Purgstall on the number 
 
 seven, 339. 
 Hebrew, a dead language in the 
 
 time of Christ, 404. 
 Hebrew chronology, from Moses to 
 Solomon, account of, 385-402 ; 
 discrepant statement of the He 
 brew text and the Scptuagint, 
 385 ; discordance of both with the 
 items in Judges, etc., 386-388 ; 
 relation to Paul s statement in 
 Acts, 387 ; method of the rec 
 onciliation of the discrepancies, 
 388-389 ; prominence of the num 
 ber forty in the series, 390-392 ; 
 apparent artificiality, in part, of 
 the chronology, 393-402 ; rela 
 tion to number of generations, 
 397-399 ; connection with Egyp 
 tian history, 400 ; sources used, 
 401-402. 
 
 his, question of its use in formation 
 of the English possessive case, 
 221-239. 
 
 Homer, Bekker s digammated edi 
 tion of, 56-80. 
 
 Homeric sevens, 325-326. 
 
 Hungarian passive formations, 210. 
 
 /, history of its pronunciation in 
 
 English, 244-246. 
 Ibn Khojle on the number seven, 
 
 33^- 
 immortality, whether demonstrable 
 
 from the light of nature, 373-379. 
 Indian sevens, 327-329. 
 Indo-European aspirate mutes, na 
 
 ture and history of, 168-183 an d 
 
 see aspirates. 
 Indo-European futures, 184-198- 
 
 and see futures. 
 Indo-European languages, their re 
 
 lation to one another, 145-146. 
 
422 
 
 INDEX. 
 
 intransitive verbs, expressing pas 
 sive, 210-213. 
 
 Ionian, origin and application of 
 the name, 10-11, 22, 23-24. 
 
 Ionian Migration, E. Curtius s the 
 ory respecting, 1-19; discussions 
 of it by other scholars, 20-36. 
 
 lonians, their position in early 
 Greece, 16-18. 
 
 Irish future, 192. 
 
 is being, English continuous pas 
 sive, 212. 
 
 Italian passive reflexive, 203. 
 
 Italicans, relation of their language 
 to Greek, 141. 
 
 Italicans and Greeks, Ross on the 
 relation of their languages, 141- 
 163 ; their true relations, 163-167. 
 
 Josephus, his testimony as to the 
 language of Palestine, 404-405, 
 411. 
 
 k, its pronunciation in early Eng 
 lish, 253. 
 
 Knight s digammated Iliad, 56. 
 
 Koran, its sevens, 337. 
 
 Kuhn, his views upon primitive as 
 pirates, 180-182. 
 
 Kiihner s treatment of the Greek 
 genitive, 44-46. 
 
 /, vowel lengthened where it is lost in 
 
 English, 269 ; other cases of pro- I 
 
 traction before, 277-278, 281. 
 labialized consonants, Ellis s views ; 
 
 of, 253. 
 
 Latin ablative in </, 47. 
 Latin accent, laws and theory of, ! 
 
 124-126. 
 
 Latin aspirate mutes, 172-173, 176. i 
 Latin future, expressed by potential 
 
 or optative, 188-189, I 9 I ~ I 93 5 i 
 
 future in bo, 191-192; future! 
 
 perfect, 191. 
 
 Latin passive, 201-202, 205-206. 
 Latin subjunctive, uses of, 215-220. ! 
 Layamon s Brut, English genitive or 
 
 possessive in, 233-235. 
 Lepsius s views on the Ionian Mi- i 
 
 gration, 13, 26 ; on aspirate mutes, j 
 
 171. 
 liberty, civil, Napoleon s influence j 
 
 on , 35 6 -3 61 - 
 Lithuanian future, 190; passive,! 
 
 204. 
 
 Madvig, his treatment of Latin sub 
 junctive, 219-220 ; of genitive 
 case, 52. 
 
 Manning s essay on English posses 
 sive case, reviewed, 221-239. 
 
 mb, English short vowel lengthened 
 before, 278-281. 
 
 metre, ancient Greek, 81-109; re 
 lation of rhythmic, metric, and or 
 chestic, 81-82 ; Greek writers on 
 metric, 82-84 ; relation of metric 
 and music, 83, 89-92 ; labors of 
 Rossbach and Westphal on Greek 
 metric, 87-109; arsis and thesis 
 and their ratios, 95-99 ; metrical 
 feet according to Aristoxenus, 
 99-102; compound and extended 
 feet, 102-106 ; irrational feet, 
 106-108 ; protracted syllables and 
 pauses, 82, 84, 103, 108, 109. 
 
 middle voice, endings of, 207. 
 
 Misteli on Greek accent, 119. 
 
 Napoleon s influence on civil liberty 
 
 in Europe, 356-361. 
 nd, short English vowel lengthened 
 
 before, 278-281. 
 Nestorians a bilingual community, 
 
 4i3- 
 New Testament, vernacular of its 
 
 writers, 403-418. 
 
 ng, short English vowel lengthened 
 "before, 280-281. 
 
 o, history of English pronunciation 
 of, 244-245. 
 
 Old Slavonic passive, 203. 
 
 orchestic, its relation to rhythm and 
 metre, 81. 
 
 Ormulum, English genitive or pos 
 sessive in, 235-236 ; its orthogra 
 phy, 265-266; its vowel quantities 
 compared with those of modern 
 English, 268-295. 
 
 orthography, English, its value, 
 
 351-356. 
 
 ou, early English pronunciation of, 
 245-246. 
 
 palseotype, Ellis s system of, 241- 
 242. 
 
 Palestine, language of, at the time 
 of Christ, 403-418 ; use of Ara 
 maic in, 403-408 ; of Greek, 408- 
 413 ; special Hellenizing in 
 fluences in Palestine, 409-410 ; 
 influence of the Septuagint, 410 ; 
 
INDEX. 
 
 423 
 
 the two languages used together, 
 and each according to circum 
 stances, 413-416; authorities on 
 the subject, 417-418. 
 passive formations, 199-214 ; Eng 
 lish, 199-200 ; German, 200-201 ; 
 Latin, 201-202, 205-206; Celtic, 
 
 202 ; Italian, 203 ; Old Slavonic, 
 
 203 ; Wallachian, 203 ; Lithuani 
 an, 204 ; Scandinavian, 204-205 ; 
 Greek, 206-209 ; Gothic, 208- 
 211 ; Sanskrit, 208-209; Hun 
 garian, etc., 210 ; Arauack, 210 ; 
 Semitic, 211 ; Tibetan and Chi 
 nese, 212-213 5 Indian and Afri 
 can, 212-213 ; Philippine, 214. 
 
 pentameter, rhythmic character of, 
 
 84. 
 
 Persian sevens, 329. 
 Philippine passive formation, 214. 
 Philo Judseus on the number seven, 
 
 338-339- 
 
 pitch, difference of, main element 
 in Greek accent, 110-113. 
 
 position not felt in English as mak 
 ing length, 264. 
 
 possessive case, English, origin of, 
 221-239. 
 
 pracJi as a Greek root, 37-43. 
 
 Princess, Tennyson s, review of, 
 296-324. 
 
 Priscian on the Latin_/, 172-173. 
 
 quantity of English vowels see 
 English vowel quantity. 
 
 r, pronunciation of in early Eng 
 lish, 251-252; effect of weak r 
 on preceding vowel, 273-277. 
 
 reflexive origin of passive forma 
 tions, 202-210; its rationale, 269- 
 
 210. 
 
 relative, Greek, original form of, 
 
 74-78. 
 republicanism, tendency of Europe 
 
 towards, 361-369. 
 rhymes of Chaucer and Gower, 
 
 255-257. 
 rhythm, ancient Greek, 81-109; 
 
 Greek writers on rhythmic, 84- 
 
 87 and see metre. 
 Romanic futures, 194. 
 Ross on the relation of Italicans and 
 
 Greeks, and of the Latin and 
 
 Greek languages, reviewed, 141- 
 
 167. 
 
 Rossbach and Westphal, their la 
 bors on Greek rhythmic and met 
 ric, 87. 
 
 s of the English possessive, its 
 origin from Anglo-Saxon genitive, 
 2^1-234. 
 
 Salesbury, W., on English pronun 
 ciation, 242-243, 246-247, 250. 
 
 Sanskrit, its use in Indo-European 
 etymology, 144-145 ; its relation 
 to other Indo-European lan 
 guages, 145-146 ; its accents, 
 119, 120; aspirate mutes, 168- 
 172 ; future, 190; passive, 208-209. 
 
 Savelsberg on the Greek relative, 
 74-76. 
 
 Scandinavian passive, 204-205. 
 Schumann s views of Ionian Migra 
 tion, 23-24. 
 Semitic future, 184-185 ; passive, 
 
 211. 
 
 Septuagint, influence of, on lan 
 guage of Palestine, 410-411. 
 
 seven, its peculiar sanctity, 325- 
 345 ; Homeric sevens, 325-326 ; 
 other Greek, 326-327 ; Indian, 
 327-329 ; Persian, 329 ; Biblical, 
 330-334; Shakspearian, 335-337 ; 
 Arabic, 337-338 ; authors who 
 have written on the seven, 328- 
 329 ; reasons of its prominence, 
 339-345 ; arithmetical, 339-340 ; 
 chronological, 340-341 ; physio 
 logical, 341-342; astronomical, 
 
 342-345. 
 
 Seven Sleepers, story of the, 328. 
 
 .y //-sound in early English, 255. 
 
 Shakspearian sevens, 335-337. 
 
 sigma, its disappearance in Greek, 
 78-79. 
 
 Slavonic futures, 192-194. 
 
 Spanish passive reflexive, 203. 
 
 spelling of English, question of re 
 form of, 351-356; difficulty of 
 learning, 352-354 ; Johnson s dic 
 tionary as standard of, 354-355 ; 
 objections to a new system con 
 sidered, 355-356. 
 
 stress of voice as element in Greek 
 accent, 113-117. 
 
 subjunctive, relation of future to, 
 188-189; uses of in Latin, 215- 
 220. 
 
 Suscmihl on Greek rhythmic, 89. 
 
424 
 
 INDEX, 
 
 Targums, evidence of, as to lan 
 guage of Palestine, 405, 416. 
 
 Tartaric passives, 210. 
 
 Tennyson s Princess, review of, 
 296-324. 
 
 thesis see arsis. 
 
 /#, English infinitive sign, 54. 
 
 tone, high, low, middle, in Greek 
 and Latin accent, 110-127; tone 
 in Chinese, 127. 
 
 ?/, history of English pronunciation 
 of, 244, 248-250. 
 
 vegetable diet, whether advantage 
 ous, 379-3 8 4- 
 
 tUj its suppression lengthens the 
 preceding vowel, 272-273. 
 
 Wallachian passive, 203. 
 
 Wallis, J., on English pronuncia 
 tion, 243, 247-249. 
 
 Weil on Greek metric, 87, 104. 
 
 Welsh future expression, 187-188; 
 Welsh glosses in Greek MSS., 129. 
 
 Westphal s labors on Greek rhyth 
 mic and metric, 85-109. 
 
 7C //, early English pronunciation 
 of, 252. 
 
 worth, become, 200. 
 
 wr, early English pronunciation of, 
 
 Wurffbain on the number seven, 
 33 s -339- 
 
 ly-sound, consonant, its origin in 
 English from ^, 270 ; from other 
 letters, 271. 
 
 
 
U.C. BERKELEY LIBRARIES