BF 
 
 leas 
 
 TRICHOTOMY IN 
 
 ROMAN LAW 
 
 BY HENRY GOTJDY 
 
 IRLF 
 
 551 
 
 in 
 
 LTD 
 
 sD 
 CO 
 
 u 
 
 OXFORD 
 
 AT THE CLARENDON PRESS 
 
 1910 
 

TRICHOTOMY IN 
 ROMAN LAW 
 
 BY 
 
 HENRY GOUDY D.C.L. 
 
 REGIUS PROFESSOR OF CIVIL LAW IN THE UNIVERSITY OF OXFORD 
 
 OXFORD 
 
 AT THE CLARENDON PRESS 
 1910 
 
HENRY FROWDE, M.A. 
 
 PUBLISHER TO THE UNIVERSITY OF OXFORD 
 
 LONDON, EDINBURGH, NEW YORK 
 
 TORONTO AND MELBOURNE 
 
TO 
 
 HIS COLLEAGUES OF 
 
 "THE SOCIETY OF PUBLIC TEACHERS OF LAW 
 IN ENGLAND AND WALES" 
 
 THIS LITTLE BOOK IS DEDICATED 
 
 BY THE AUTHOR 
 
 IN GRATEFUL RECOGNITION OF THE 
 
 HONOUR CONFERRED ON HIM BY HIS ELECTION AS 
 
 FIRST PRESIDENT OF THE SOCIETY 
 
 1908-10 
 
 M105768 
 
PREFACE 
 
 SOME years ago I contributed an article to a 
 Festschrift in honour of Professor Carlo Fadda of 
 Naples, in which I examined the classifications 
 of obligations in Justinian's Institutes, and I 
 endeavoured to show that the peculiar divisions 
 into groups of four that we there meet with may 
 be traced to a desire for artificial symmetry on 
 the part of the Editors, and were neither based 
 on any considerations of logic nor the result of 
 mere coincidence. This article was published in 
 the Studi Fadda in 1906 under the title ' Artifi- 
 ciality in Eoman Juristic Classifications ', and its 
 conclusions, though not generally accepted, have 
 been received with favour by some eminent 
 Romanists who have been good enough to write 
 to me about it. In that article I stated that I 
 was of opinion that many of the Roman lawyers 
 were influenced by the symbolism of numbers, as 
 regards both the external arrangement of their 
 works and the classifications of the subject-matter 
 contained in them, though I did not attribute the 
 
4 PEEFACE 
 
 particular fourfold classifications in the Institutes 
 to symbolism but rather to a desire for artificial 
 symmetry. I also said (Studi Fadda V, p. 212) 
 that I should at a future date, if possible, com- 
 ment on the symbolic influence of the number 3. 
 The present monograph is the fulfilment of that 
 undertaking. 
 
 For the views expressed in the following pages 
 I am solely responsible, but I desire to express 
 here my thanks to my friends Professor Cook 
 Wilson, Mr. Warde Fowler, Fellow of Lincoln 
 College, and Emeritus Professor Strong, of Liver- 
 pool University, for some valuable suggestions to 
 which I have given effect. To Professor Cook 
 Wilson in particular, who read over the whole 
 of the proof prints of the work and made a 
 number of acute criticisms, I am much indebted. 
 
 ALL SOULS COLLEGE, 
 OXFORD, 1910. 
 
TRICHOTOMY IN ROMAN LAW 
 
 IN modern times, at any rate among European 
 nations, one does not attach any particular signifi- 
 cance to numbers as such, outside the realm of 
 mathematics. One does not at all regard them 
 as influencing actions or events either in public 
 or private life. Least of all have modern writers, 
 if we except perhaps Hegel, been consciously 
 influenced by them in the composition of their 
 writings. 1 What literary or scientific author 
 nowadays, in dividing his treatise into parts, 
 books, &c., or dividing his subject-matter into 
 heads and categories or genera and species, would 
 attach any special importance to what the number 
 of these might be ? But in former times it was 
 different: to the ancients certain numbers had 
 a mystical value or fixed traditional force, and 
 ancient writers frequently made the distributions 
 of their subject-matter, as well as the external 
 
 1 But see Hofmann, Zeitschr. fur Rechtsgeschichte, xi, 
 p. 344, n. 20 . Hegel's triads however are not properly classi- 
 fications of genera into co-ordinate species, but rather a con- 
 sequence of his method of philosophy. 
 
6 TEICHOTOMY 
 
 arfarigenient ,of their works, depend upon their 
 ajgctiEdahoa -with one or other of these numbers. 
 The influence of this symbolism, or whatever it 
 be called, can be demonstrated to have existed, 
 in greater or less degree, among all ancient 
 nations with whose culture we are acquainted, 
 especially the Hindus, Hebrews, Egyptians, 
 Greeks, and Komans. 1 The numbers 3, 4, 7, and 
 12 seem to have been above all sacred, but 9 
 and 10 were also universally symbolic. By the 
 vulgar they were generally regarded as lucky 
 numbers, while philosophers sought a rational 
 basis for them. It is in the last degree im- 
 probable, for instance, that the Commandments 
 given to Moses would ever have been published 
 as, say, eight or eleven, or the decemviral code of 
 the Eomans in, say, thirteen tables. 2 How far 
 this mysticism of numbers continued to affect 
 mediaeval writers or left its impress on mediaeval 
 institutions it is not my purpose to inquire. 
 That it still continued to exercise a certain, 
 
 1 See Usener, 'Dreiheit,' Rhein. Mus. f. Philol (N. P.), Iviii, 
 pp. 32 sq. The Celtic races also showed great partiality for 
 them. See K. Meyer, as in p. 7, n. 1 infra and p. 9, n. 3 . 
 
 2 Ausonius describes the law of the XII Tables as 
 
 lus triplex, tabulae quod ter sanxere quaternae, 
 Sacrum, privatum, et populi commune quod us- 
 quam est. 
 
 Edyll. xi, line 61. 
 
IN KOMAN LAW 7 
 
 though gradually decaying, influence is un- 
 doubted. On this point it will suffice to quote 
 a passage from our own Lord Coke, a dry but 
 not wholly unimaginative writer. Speaking of 
 the English jury, he says, ' It seemeth to me 
 that the Law in this case delighteth herself in 
 the number of twelve. For there must be not 
 only 12 jurors for the trial of matters of fact, 
 but 12 judges of ancient time for trial of matters 
 of law in the Exchequer Chamber. . . . And 
 that number of twelve is much respected in Holy 
 Writ as 12 Apostles, 12 stones, 12 Tribes, etc.' 1 
 
 The Pythagorean school of philosophy was 
 apparently the earliest to propound a system of 
 symbolic numbers. Possibly they may have 
 borrowed it from India, for, at all times in the 
 philosophy and religion of the Hindus, such 
 symbolism has been prominent. The Pytha- 
 gorean views on the subject influenced Plato, 
 as we see from the Timaeus, 2 and had an un- 
 
 1 Institutes, Part I, lib. 2, cap. xii 234. The Ancient 
 Laws of Ireland, among others, afford unmistakable evi- 
 dence of symbolic numbers, a whole book being composed 
 in heptads. See vol. v, pp. 117 sq. of Ancient Laws of 
 Ireland, Dublin, 1901, and K. Meyer, The Triads of Ireland, 
 in Todd Lecture Series of Koyal Irish Academy (1906), 
 p. xii. 
 
 8 See trans, by Jowett, pp. 463, 468, 474, &c. ; Proclus, 
 
8 TKICHOTOMY 
 
 doubted influence on Greek mythology. Perhaps 
 strongest of all was the influence exercised upon 
 the Neo-Platonists. The conquests of Alexander 
 brought the Greeks into direct touch with the 
 religion and philosophy of the East and affected 
 powerfully, it is thought, the speculations of the 
 later Greek Schools. 
 
 It is not my purpose, and this is not the place, 
 to enter upon a discussion of theories regarding 
 the origin and meaning of the symbolic numbers 
 above mentioned. My concern is with the 
 number 3. That a special mystical value was 
 attached by ancient writers to this number is of 
 course familiar to scholars. 1 To quote what I said 
 in the Festschrift for Carlo Fadda, referred to above 
 in my Preface : l To many of the leading nations 
 of antiquity it represented divine power. As evi- 
 
 Commentary on Tim. iii. 181 ; cf. Plato, Legg. x. 888 E. 
 Aristotle seems on the whole free from the same charge. 
 But see De Caelo, i. 1 init. 268 a, where he discusses the 
 completeness of the number 3 with reference to the line, 
 plane, and solid in magnitude, and Ehet. i. 2, 3. 
 
 1 For an explanation of the origin of the mystic value 
 of the number 3 see Diels, Archiv f. G-esch. d. Philos., x, 
 pp. 232-3 ; Usener, op. cit., pp. 348-62. They trace it back 
 to the earliest periods in the history of races, when 3 was 
 the highest number that could be reckoned and signified 
 many. Tylor, Primitive Culture, i, p. 220. Tribus, meaning 
 a community, may perhaps be taken as an example of this. 
 
IN KOMAN LAW 9 
 
 dence of this it is enough to refer to the three Gods 
 of Hindu mythology Vishnu, Siva, and Brahma, 1 
 and the Trinity of Christian doctrine. By some 
 ancient philosophers, again, it was regarded as 
 the most symbolic of numbers, because it repre- 
 sented the beginning, middle, and end of all things, 
 and also the dimensions of space.' 2 To this 
 may be added that among the Welsh and Irish 
 the influence of the number 3 was unbounded. 
 The name triad was given in these countries to 
 a species of literary composition in which the 
 subject-matter was arranged in groups of three. 
 K. Meyer observes, * I am not aware that this 
 kind of composition has ever attained the same 
 popularity elsewhere as in Wales and Ireland, 
 where the manufacture of triads seems at times 
 almost to have become a sport.' 3 He also gives 
 one or two curious illustrations of modern triads, 
 said to have been in current use among the 
 
 1 And the three goddesses Doorga, Lakshmi, and Saras- 
 vati. Egyptian triads (e. g. Amon, Mut, and Khonsu) were 
 also very common. See Maspero, Etudes de Mythologie 
 jZgyptiennes, vol. ii, pp. 386 sq. 
 
 2 Studi Fadda, v, p. 210. See Usener, ' Dreiheit,' Rhein. 
 Mus. f. Philologie (N. F.), Iviii. 1, pp. 31 sq. ; Lobeck, Aglao- 
 phanus, pp. 384-9. For the further conception of trinity in 
 unity see Usener, 1. c. ; cf. Martial, J^p. v. 24 l Hermes 
 omnia solus, et ter unus.' 
 
 5 Todd Lecture Series, p. xv, cited supra. 
 1211 B 
 
10 TKICHOTOMY 
 
 common people within living memory. The two 
 following are amusing and may be quoted : 
 
 ' Three disagreeable things at home : a scolding 
 wife, a squalling child, and a smoky chimney.' 1 
 
 1 The three finest sights in the world : a field 
 of ripe wheat, a ship in full sail, and the wife 
 of a MacDonnell with child/ 
 
 But at any rate it is certain that anlong the 
 Eomans from the earliest times the number 3 
 was by far the most significant. 3 As an odd 
 
 1 With this compare the old German couplet 
 
 ' Drei ding im haus sind ungelegen : 
 der rauch, ein bos weib und der regen.' 
 
 2 Ibid. p. ix. f ln England and Germany, also, as stated 
 by me in the Studi Fadda, relics of the old superstition about 
 the number 3 still linger. Thus in England one says * the 
 third time is the lucky one ', and in Germany there are 
 proverbs l aller guten Ding sollen drei sein ' and * dreimal 
 ist Bubenrecht '. For other instances of the symbolism of 
 the number in Germany see Grimm, Diet. v. ' Drei '. 
 
 3 The symbolism of the numbers 4 and 7 may just be 
 noticed. The Pythagoreans regarded the 4 (the rerpaKTv?) 
 as the perfect number, the root of all nature and ground of 
 all things. See Themist. Phys., lib. iii, and other references 
 in Koby, Introd. to Digest, p. xxix. Its importance in Holy 
 Writ is also well known and needs no illustration. Not 
 less symbolic was the number 7 (probably due to its being, 
 as Philo says, a combination of 3 + 4, or, as some hold, 
 3 + 1+3, and not, as more commonly supposed, based on 
 astrological reasons, just as the symbolism of the 9 and 12 
 may be due to their being 3x3 and 3x4 respectively). 
 Thus in Holy Writ the whole book of Revelation is domi- 
 
 
IN KOMAN LAW 11 
 
 number it enjoyed of course the good omen 
 which the Eomans attached to such numbers 
 1 numero deus impare gaudet V But the evidence 
 of its influence meets us on every hand. The 
 Roman mythology presents it to us unmistak- 
 ably. 2 Thus we have the Capitoline triad of 
 
 nated by the number 7 7 churches, spirits, golden candle- 
 sticks, &c. while elsewhere we are given 7 days of the 
 week and endless other examples. (See e.g. Ezekiel xlv. 
 23-5.) Among the Greeks the 7 was sacred to Apollo and 
 Dionysus. Among the Komans we have also clear evidence 
 of its influence, as in the 7 kings, hills, seals of testaments, 
 &c. ; see Aulus Gellius, N. A. iii. 10, 17 ; Apuleius, Meta- 
 morph. xi (238). Justinian himself, in his introduction to 
 the Digest (Tanta, 1), tells us that he made the division of 
 that work into 7 parts 'non perperam neque sine ratione 
 sed in numerorum naturam et artem respicientes '. Consult 
 especially Smith, Diet, of Bible, voce seven ; Hofmann, Z. fur 
 Eechtsgeschichte, pp. 343 sq. ; Hadley, Essays Philological 
 and Critical, pp. 325 sq. See also on the symbolism of the 
 numbers 7 and 9 in Greek mythology, Eoscher, ' Sieben und 
 Neunzahl' in Abhand. d. Konig. Sachs. G. d. W. (1904). 
 
 1 See Virg. Eclog. viii. 73-77. Numa's supposed enactment 
 regarding the Calendar is given by Macrobius thus l Numa 
 in honorem imparis, numeri adjecit diem '. See also Cen- 
 sorinus, d. d. Nat. xx. 4. Cf. Plato, Laws, iv. 717. 
 
 2 Similarly among the Greeks. According to Usener's 
 reckoning, the Theogony of Hesiod contains fifteen triads 
 of Gods and lesser divinities ; see Bhein. Mus. f. Philologic 
 (N. F.), Iviii. 1, p. 4. Tpn-oyeWa was a name for Athene 
 and Tpwr/xeyto-Tos for Hermes. For the influence of the 
 number 3 in the cult of mysteries, see Hildebrand's Prole- 
 gomena to his edition of Apuleius, Metamorph., p. xxxvii. 
 
12 TRICHOTOMY 
 
 Jupiter, Juno, and Minerva, the three Furiae, the 
 three Parcae, the three Gratiae, and others l most 
 of them originally borrowed from the Greeks, 
 but some of them native. But every department 
 of Roman life, secular as well as sacred, private 
 as well as public, was profoundly affected by it. 
 In the singular poem of Ausonius, ' GripJius 
 ternarii numeri/ 2 he has given, following in 
 the steps of earlier writers from Pythagoras 
 downwards, an immense number of instances of 
 all kinds from which the nature and force of 
 the number 3 among the Romans can be readily 
 appreciated. 
 
 In short, the number 3 seems to have repre- 
 sented to the Romans periodicity, and hence by 
 a natural transition completeness : it was to 
 #iem the perfect number Tres numerus super 
 omnia.* The traditional origin of the Romans 
 
 1 The Capitoline triad, as such, seems to have been 
 Graeco-Etruscan. Another notable triad was Ceres, Liber, 
 Libera ; corresponding to Demeter, Dionysus, Persephone. 
 On the three Graces, see Seneca, de Beneficiis, i. 3. 3. He 
 discusses why the number should be three as held by the 
 Stoics (instead of two, as some held), and explains that they 
 corresponded to the tria beneficiorum genera, viz. 'pro- 
 nierentium, reddentium, simul accipientium reddentium- 
 que.' Cf. note by Bouillet in Lemaire's ed. 
 
 2 Ausonius, Edyllia xi. 
 
 3 Ausonius, Edyll. xi, 1. 88 ; cf. ibid., 1. 52. 
 
IN KOMAN LAW 13 
 
 from 3 races (Kamnes, Titles, and Luceres) and 
 the consequent political division of the early 
 people into tres tribus is a noticeable instance. 1 
 Evidence of the early symbolism of the num- 
 ber is also clearly shown in the XII Tables. 
 One of the most remarkable provisions of the 
 decemviral code is expressed thus : ' Si pater 
 filium ter venumduuit, films a patre liber esto.' 2 
 This means that if a paterfamilias sells his son 
 three times his potestas over the son shall end. 
 So fundamental was the power of a father over 
 his son conceived to be involving the notion 
 of imperium as well as dominium that the sale 
 of the son to another person only suspended 
 and did not destroy it. If such son were after- 
 wards manumitted by the purchaser he fell back 
 into patria potestas. Were he sold a second time 
 and again manumitted, the result was the same. 
 But if sold a third time the potestas was regarded 
 as extinguished and, if now manumitted, the 
 son became sui juris? Another rule of the XII 
 
 1 It also appears in the tripudium of the Carmen Arvale. 
 On this see Warde Fowler in Classical Review, xvi, p. 211. 
 
 2 Tab. iv. 2 (Schoell). 
 
 3 Whether this rule was first laid down by the Decemvirs 
 or was due to earlier custom it is impossible to say. Gaius 
 (iv 79) speaks of it as if introduced by the Decemvirs 
 for purposes of emancipation and adoption, but these applica- 
 
14 TKICHOTOMY 
 
 Tables was the so-called usurpatio trinoctialis. 1 
 1 Lege XII Tab. cautum erat,' says Gaius, * si 
 qua nollet eo modo [usu] in manum mariti 
 convenire, ut quotannis trinoctio abesset atqfue 
 ita usum cuiusque anni interrumperet.' 2 That 
 is to say, a wife who remained away from her 
 husband's abode for three consecutive nights in 
 each year prevented manus being acquired over 
 her by prescription. 3 Another law of the XII 
 Tables, illustrative of my point, was that relating 
 to a judgement debtor arrested by manus injectio 
 of his creditor. He had to be kept for 60 days 
 in prison by the creditor, and produced during 
 that period on three successive nundinae (ternis 
 nundinis) before the praetor in the market-place. 
 At the close of the third nundinde, failing settle- 
 ment of the debt or intervention of a vindex, he 
 was addicted to the creditor who might do what 
 
 tions of it seem rather due to interpretatio, and it is more 
 probable that it was originally intended to relieve sons from 
 the hardship of being repeatedly sold for a price, or pledged 
 by their patresfamilias. For another view see Girard, Manuel 
 d. d. r., p. 183. 
 
 1 Tab. vi. 4 (Schoell). 
 
 2 Gai. i 111 ; cf. Gellius, Hi. 2. 13. 
 
 8 As the law speaks of trinoctium (and not triduum) it may 
 be presumed that the wife might spend the day at the 
 domestic hearth and attend to her duties, provided that she 
 absented herself during the night. 
 
IN KOMAN LAW 15 
 
 he Jiked with him. 1 Another illustrative law of 
 the Tables was in these words : l Cui testimonium 
 defuerit, is tertiis diebus ob portum obvagulatum 
 ito.' 2 Of this law various explanations have 
 been given, of which one is that it refers to a 
 plaintiff, in an action calling aloud upon his 
 debtor outside the latter's house, to perform his 
 obligation. If this were done for three consecu- 
 tive days the recalcitrant or impecunious debtor 
 (against whom ordinary in jus vocatio could not 
 be made) could then be arrested by manus in- 
 jectio even within the house itself. 3 Besides the 
 instances just mentioned there are others in the 
 XII Tables in which the number 3 figures. 4 In 
 
 1 Tab. iii. 5. Also 'tertiis nundinis partis secanto* in 
 iii. 6 ; cf. Gellius, xx. 1. 49. 
 
 2 Tab. ii. 3 ; Festus, s. v. vagulatio (ed. Mueller), p. 375. 
 
 3 See Voigt, XII Tafeln, i, p. 535. If this be right, the 
 Dharna of Hindu law offers an interesting analogy. Puchta, 
 however (Inst. 160 a), and others think the obvagulatum 
 relates to a recalcitrant witness who, after the third day, 
 might be seized by manus injectio. 
 
 4 e. g. the dies comperendinus, the three kinds of delict, 
 and the three forms of personal injury. It has been held 
 by some writers that, prior to Justinian, three adjudications 
 in separate liberates causae were required to fix the status of 
 slavery on an alleged freeman. Such a rule may have been 
 in the XII Tables, but the evidence on the subject is obscure 
 and does not warrant the conclusion that there were then 
 actually three actions ; see Cod. vii. 17. 1 pr. ; Buckland, 
 Law of Slavery, p. 668. 
 
16 TEICHOTOMY 
 
 later times and among the classical jurists the 
 notion of completeness or perfection involved in 
 it is signified by numerous phrases, rules, and 
 maxims. Such are, for example, the jus trium 
 liberorum, tria onera tutelarum dant excusationem, 
 the tria verba (do, dico, addico), tres faciunt colle- 
 gium, the ex tribus maribus of the Senatusconsult 
 Afinianum (or Sabinianum ?). 
 
 That Eoman writers literary, philosophic, and 
 juristic were alike profoundly influenced by the 
 ternarius numerus, as indeed they were by other 
 symbolic numbers, in the external arrangement 
 and composition of their works, cannot be doubted. 
 It has been frequently pointed out. 1 It continued 
 down to Justinian, and even to mediaeval times, 
 as already indicated. Thus Justinian, in his 
 constitution ' Tanta ', says : ' Omni igitur Eomani 
 juris dispositione composita, et in tribus volu- 
 minibus, id est Institutionum, et Digestorum seu 
 Pandectarum, necnon Constitutionum perfecta, et 
 in tribus annis consummata.' 2 That a similar 
 
 1 See Koby, Introd. to Digest, pp. xxix-xxx; Hofmann, 
 Zeitschrift fur R. G. xi, pp. 340 sq., and xii, pp. 180 sq. 
 Hofmann calls it ' Zahlenspielerei '. He gives reasons for 
 thinking that Justinian's partiality for mystic numbers was 
 influenced by the Neo-Pythagorean philosophy. 
 
 2 Tanta, 12 ; cf. 23, and Omnem reipublicae, 7. As 
 to what he says about the number seven see supra, p. 10, n. 3 . 
 
IN EOMAN LAW 17 
 
 influence affected the treatment of their subject- 
 matter, particularly their divisions and classifi- 
 cations, is to my mind equally clear, and it is 
 this that I will attempt to prove, from juristic 
 texts, in the following pages. This influence 
 frequently led the jurists, or at least some of 
 them, to sacrifice logic and proportion in dealing 
 with legal topics, and thereby to produce unsatis- 
 factory and sometimes even misleading doctrines. 
 As I have elsewhere said, ' Almost necessarily 
 classifications influence the treatment of subject- 
 matter, and, if they are bad, the result is that 
 subjects are placed in juxtaposition that are not 
 truly co-ordinate, worthless distinctions are pro- 
 duced, and otherwise the legal system is inju- 
 riously affected.' 1 
 
 This influence of the ternarius numerus is best 
 illustrated from the Institutional treatises of the 
 jurists, as it is in them we naturally look for 
 classifications and categories, but it appears in all 
 their writings. Most of the jurists are, more or 
 less, chargeable with it, but with Ulpian, infected 
 as he was, apparently, by Stoic philosophy, it is 
 
 On the possible symbolism of the mediaeval division of the 
 Digest into three parts (vetus, infortiatum, novum), see Puchta, 
 Inst. i, pp. 644 sq. 
 
 1 Studi Fadda, v, p. 209. 
 
 C 
 
18 TKICHOTOMY 
 
 specially remarkable. It is perhaps not too much 
 to say that wherever Ulpian gives a classification 
 of an institution or doctrine into genera or species 
 we may expect to find it tripartite, if the subject- 
 matter admits of it or can be forced into it. 1 
 
 Keeping in view, then, the symbolism of per- 
 fection attached by the Eomans to the number 3, 
 illustrations may now be given of how it influenced 
 the jurists in their classifications. By the term 
 t classification ' I mean a division of the matters 
 or topics, which constitute any doctrine or insti- 
 tution, into classes that should be mutually exclu- 
 sive or, in other words, a division into genera and 
 species ; it does not apply to mere varieties or forms 
 included in a species, which are of course generally 
 indeterminate and may be infinite. 2 These latter, 
 being incapable of classification, are generally 
 introduced by such words as velut (veluti), sicut 
 (sicuti), ut, utputa, quorum in numero, or the like, 
 though no doubt some of these words were also 
 
 1 The same charge may he to a large extent made against 
 Paul, who, if we may judge from his writings, was a Stoic of 
 the Stoics. But, strictly speaking, the charges, as we shall 
 see, are rather to be made against the earlier writers from 
 whom Paul and Ulpian drew. 
 
 2 Distinctions however, in jurisprudence, between species 
 and varieties are often as arbitrary and difficult to be made 
 as in the field of natural science. See, as to the latter, Life 
 and Letters of Charles Darwin, ii. 88. 
 
IN EOMAN LAW 19 
 
 occasionally used where an exhaustive enumera- 
 tion was intended. 1 
 
 The partiality of the Koman jurists for the 
 tripartite classification 2 has not been unobserved 
 by modern writers, 3 but, so far as I know, no one 
 has sought for a reason for it or attempted to give 
 any explanation of it. 
 
 I shall now arrange in juxtaposition some lead- 
 ing tripartite classifications, confining myself to 
 the primary divisions of law and the topics dealt 
 with in the first books of the Institutes of Gaius 
 and Justinian and the cognate texts in the Rules 
 
 1 See Gai. i 120, and iii 141 ; Cohn, Beitrage, Heft 2, 
 p. 45. On partition and division of genera see Cicero, Top. 
 30-4. The jurists unfortunately use the terms genera 
 and species without precision, e. g. species of obligations and 
 genera of contracts in Gai. iii 88. Cf. Inst. iii. 13 2. 
 We get genera instead of species furtorum in Inst. iv. 1 3. 
 Cf. Gai. iii 182-3. 
 
 2 Tripertitus (or tripartitus), formed from the numeral and 
 pars (partior), is quite classical (Cicero, de Off. iii. 2. 9 ; 
 Top. 90). As is well known, the earliest treatise on Roman 
 law, of which there is any distinct record, was the so-called 
 Tripertita of Sextus Aelius, a work said by Pomponius 
 (Dig. i. 1, 2 38) to have been composed of three parts, 
 which he describes as harmonizing with his own threefold 
 division of the sources of the law (Ibid. 6). Besides triper- 
 titus there are numerous other Latin words meaning three- 
 fold, similarly formed by composition with the numeral tres, 
 as triplex, ternarius, trifarius, triplasius, triplusj triplaris. 
 
 3 See Savigny, Systenij i, p. 395 ; Mommsen, Bom. S.-R. 
 iii. 1, p. 8n. 
 
20 TRICHOTOMY 
 
 of Ulpian and in the Digest ; afterwards I shall 
 comment upon each of them separately. It will 
 be observed that the great majority of these classi- 
 fications are given by Ulpian ; several of them are 
 given by him alone. 
 
 1. The precepts of law are honeste vivere, alte- 
 rum non laedere, suum cuique tribuere. 1 
 
 2. Public law consists in sacra, in sacerdotibus, 
 in magistratibus. 2 
 
 3. Private law is composed of jus naturale, jus 
 gentium, and jus civile. 3 
 
 4. All law was either made by consent, or esta- 
 blished by necessity, or settled by custom. 4 
 
 5. Praetorian law was introduced by the prae- 
 tors for the purpose either of assisting, or supply- 
 ing deficiencies in, or correcting the rigour of the 
 jus civile, in the public interest. 5 
 
 6. Constitutions of the emperors, as a source of 
 private law, are decreto vel edicto vel epistula. 6 
 
 7. The subject-matter of private law relates 
 either to persons, or to things, or to actions. 7 
 
 1 Ulpian in Libro primo Eegularum, Dig. i. 1, 10 1 ; Inst. 
 i. 1 3. 2 Ulp. in Lib. i Institutionum, Dig. i. 1, 1 2. 
 
 3 Ulp. in Lib. i Instit., Dig. i. 1, 1 2 ; Inst. i. 1 4. 
 
 4 Modestine in Lib. i Eeg., Dig. i. 3, 40. 
 
 5 Papinian in Lib. ii Definitionum, Dig. i. 1, 7. 
 
 6 Gai. i 5 ; Ulp. in Lib. i Inst., Dig. i. 4, 1 1 ; Inst. 
 i. 2 6. 7 Gai. i 8 ; Inst. i. 3 pr. 
 
IN EOMAN LAW 21 
 
 8. All law is concerned with either the acquisi- 
 tion, or conservation, or restriction of rights. 1 
 
 9. Persons are grouped into (I) liberi and servi, 
 (2) cives, latini, and peregrini, and (3) sui juris and 
 alieni juris. 2 
 
 10. Persons are divided into lib&ri, servi, and 
 libertini. 3 
 
 11. Freedmen are divided into cives, latini, and 
 dediticiorum numeral 
 
 12. The conditions upon which slaves might 
 acquire civitas are three, and the modes of manu- 
 mission under jus civile are also three, viz. vindicta, 
 censUj and testamento. 5 
 
 13. Persons alieni juris are either in potestate, in 
 manu, or in mancipio. 6 
 
 14. Marriages under jus civile are contracted 
 farreo, coemptione and usu.* 1 
 
 15. Emancipation and adoption of filiifamilias 
 
 1 Ulp. in Lib. ii Inst., Dig. i. 3, 41. 
 8 Gai. i 9, 12, 48. 
 
 3 Inst. i. 5 pr. ; Ulpian in Lib. i Institutionum, Dig. 
 i. 1, 4. 
 
 4 Gai. i 12 ; Ulp. Beg. i 5 ; cf. Inst. i. 5 3. 
 8 Gai. i 17 ; Ulp. Beg. i 6. 
 
 6 Gai. i 49 ; Ulp. Beg. xix 18. 
 
 7 Gai. i 110. Ulpian doubtless gave this also in Eeg. ix, 
 but the MS. is here defective. It was also contained in his 
 Institutes as cited by Boethius in Topica, 3, 4 ; see Girard, 
 Textes, p. 467. 
 
22 TKICHOTOMY 
 
 is effected by three mancipations and three manu- 
 missions. 1 
 
 16. Of persons sui juris it is said: 'quaedam 
 vel in tutela sunt vel in curatione, quaedam neutro 
 jure tenentur/ 2 
 
 17. Tutors are divided into ' legitimi aut senatus 
 consultis constituti aut moribus introducti '. 3 
 
 18. Status consists of libertas, civitas, familial 
 
 19. Capitis deminutio is divided into maxima, 
 media, and minima. 5 
 
 The above nineteen divisions form the large 
 majority of the leading classifications of the 
 subject-matter dealt with in the first books of 
 the Institutes of Gaius and Justinian or cognate 
 thereto. 6 I have so put them together for the 
 sake of a clear conspectus. 
 
 1 Gai. i 132 ; Ulp. Eeg. x 1. Neither Gaius nor 
 Justinian give the modes of creating patriapotest as formally ; 
 but as described by them in their Institutes there were 
 just three modes, viz. birth in lawful marriage, adoption, and 
 legitimation the last of these having of course a widely 
 different import with Gaius from what it had with Justinian. 
 
 2 Gai. i 142 ; Inst. i. 13 pr. It seems clear that a tri- 
 partite division was here intended. 
 
 3 Ulp. Eeg. xi 2 ; cf. Ulp. in Dig. xxvi. 1, 6 2. Neither 
 Gaius nor Justinian give formally any division, but see 
 infra, p. 47, n. 2 . * Paul in Dig. iv. 5, 11. 
 
 5 Gai. i 159 ; Ulp. Eeg. xi 10 ; Inst. i. 16 pr. ; Paul 
 in Dig. iv. 5, 11. 
 
 6 There are no doubt a considerable number of classifica- 
 
IN KOMAN LAW 23 
 
 It would, of course, be rash to predicate that all 
 of them are due to the mystic influence of the 
 number 3, but I cannot resist the conclusion that 
 the great majority of them are due to that cause. 
 They are mostly, if not altogether, traditional. 
 A critical examination of them, which I now pro- 
 ceed to make, will show that with few exceptions 
 they are neither logical nor suitable to the treat- 
 ment of the topics that they severally represent. 
 
 tions to be found in the Institutes both of Gaiusand of Justi- 
 nian that are not tripartite. The influence of the ternarius 
 numems in fact in these works, apart from traditional 
 classifications, is not so marked as it is in the writings of 
 Ulpian, and possibly neither Gaius nor Justinian's editors 
 were consciously affected by it. Gaius's fundamental divi- 
 sion of law into jus gentium and jus civile is noticed below 
 (p. 27). Another primary division of law into jus publicum 
 and jus privatum (perhaps taken originally from Aristotle, 
 Hhet. i. 13. 2 Xcyw 8e v6p.ov TOV fJiV tSiov TOV of. KOLVOV ) is 
 also noted below (p. 25). Another dual division into jus 
 scriptum and jus non scriptum was, as Ulpian indicates (Dig. 
 i. 1, fr. 6 and Inst. i. 2 3), taken from the Greeks and 
 perhaps from Aristotle (Ehet. i. 13. 2). See Schrader, Com- 
 mentary on the Institutes, hoc loco. Gaius further makes 
 logical twofold divisions of men into liberi et servi and 
 of freemen into ingenui et libertini, while Ulpian, as above 
 noted, has a trichotomy (see infra, p. 39, n. 2 ). Of adoption 
 it is said by Gaius, i 98, ' duobus modis fit ', while Ulpian, 
 Reg. viii. 2, describes it as created ' aut per populum aut per 
 praetorem velpraesidem provinciae ', where, though the con- 
 trast of aut and vel shows perhaps that he does not intend 
 a tripartite division, he seems almost as if constrained by his 
 usual habit. 
 
24 TKICHOTOMY 
 
 1. The praecepta juris are said to be three 
 honeste vivere, alterum non laedere, suum cuique tri- 
 lucre. That this text was originally borrowed 
 from some writer on philosophy can hardly be 
 doubted. It was familiar to the Greeks. 1 It is 
 obvious that none of the three precepts is a rule 
 of positive law. To live honestly, for instance, is 
 a good moral maxim, but the law does not insist 
 on it. The police constable is not called upon till 
 the dishonesty or immorality discloses itself in the 
 infraction of some rule prescribed by the law of 
 the land. ' Non omne quod licet honestum est,' 
 says Paulus. 2 The other two precepts are open 
 to similar criticism. The second would seem to 
 be logically included in the third. Or the honeste 
 vivere, if it is to be taken in a wide sense as in- 
 cluding both negative and positive duties, would 
 alone have sufficiently included the other two. 3 
 
 1 See references in Schrader, Inst. hoc loco. The precepts 
 are rightly, I think, connected with the three cardinal virtues 
 of the Stoics by Thomasius, Not. ad Inst. In. I. cited by 
 Schrader. 
 
 2 Dig. 1. 17, fr. 144 pr. Cf. Dig. xxiii. 2, fr. 42 pr. ; and 1. 
 17, fr. 197 where it is said of marriage ' semper in conjunc- 
 tionibus non solum quid liceat considerandum est, sed et 
 quid honestum sit '. 
 
 3 See the criticism by Savigny, System, i, p. 409, who 
 regards the precepta not as rules of law but as the basis on 
 which rules of law rest. Puchta, Inst. 102 n. c , thinks 
 
IN EOMAN LAW 25 
 
 It would have been more logical still for Justinian 
 to have said that the universal precept of law is to 
 do justice justitia having been already defined 
 as ' constans et perpetua voluntas jus suum cuique 
 tribuendi'. 1 On the other hand, if each of the 
 three precepts was intended to be taken in a re- 
 stricted sense, they seem incomplete, and might 
 have been added to appropriately. Thus Ulpian 
 might have borrowed from Gaius and added ' male 
 nostro jure non uti '. 2 
 
 2. Public law is tripartite, consisting in sacris, 
 in sacerdotibus, in magistratibus. 3 This strange classi- 
 fication is also taken from Ulpian, and is a sub- 
 division of his primary dual division of law into 
 jus publicum and jus privatum. It is interesting 
 to notice that a tripartite division of law into jus 
 sacrum, publicum and privatum was current among 
 writers before Ulpian and is given by Quintilian, 4 
 
 the precept honeste vivere may be a ' Eeminiscenz aus den 
 Stoikern (Cic. de Fin. ii. 11) ', and adds 'Es wiirde dann 
 eine Bestatigung davon sein, dass dergleichen Entlehnungen 
 so gut wie gar keinen Einfluss auf den eigentlichen recht- 
 lichen Gehalt der Thatigkeit der Juristen hatten, denn wir 
 finden nicht dass aus jenen praecepta juris das Kecht wirk- 
 lich abgeleitet worden ware*. On this view see infra, 
 p. 71, n. 1 . 
 
 1 Inst. i. 1 1. . 2 Gai. i 53. 3 Supra, p. 20, n. 2 . 
 
 4 Orat. ii. 4. 33. It is also given by Ausonius, Edyll. xi, 
 
 1. 62 ; supra, p. 6, n. 2 . 
 
 1211 D 
 
26 TEICHOTOMY 
 
 and also appears (though in a somewhat different 
 sense) in the Code. 1 For Ulpian to have adopted 
 this would, of course, have prevented his tricho- 
 tomy of jus publicum. Under magistratibus we must 
 apparently place the vast subjects of the legis- 
 lature, the magistracy, imperial administration 
 and taxation, in short the whole of constitutional 
 and administrative law. But is there any real 
 distinction between sacris and sacerdotibus so as to 
 justify the jurist in placing each of them separately 
 on the same plane with magistratibus ? Either term 
 alone would quite as well have represented all 
 public matters of the jus divinum, as magistratibus 
 represented those relating to the jus humanum. 
 But for any practical consequence the classification 
 was worthless and seems to have been ignored. 
 A twofold division into jus divinum and humanum , 2 
 corresponding to matters of Church and State in 
 modern law, the jus humanum being subdivided 
 into publicum and 'privatum, would have been the 
 natural and logical one. The division into jus 
 divinum and jus humanum was, in fact, quite well 
 known to Ulpian and was adopted by him in 
 dealing with interdicts. 3 
 
 3. Private law is tripartite, viz. law (1) common 
 
 1 Cod. i. 2, 23. 2 See Cic. de Legibus ii and iii. 
 
 3 Dig. xliii. 1. 1 pr. 
 
IN ROMAN LAW 27 
 
 to men with the lower animals (jus naturale), 
 (2) common to all peoples (jus gentium), and (3) 
 binding upon Roman citizens alone (jus civile}, 1 
 This division of private law is from a different 
 point of view to that of public law but is scarcely 
 less objectionable. It can hardly have originated 
 with Ulpian (in his Institutions), but much more 
 probably is due to some unknown Stoic philo- 
 sopher from whom he drew ; and, if we except 
 perhaps two unimportant writers, Hermogenianus 
 and Tryphoninus, 2 it was not adopted by any 
 other jurist. The twofold division of Gaius into 
 jus gentium and jus civile* was of course per- 
 fectly well known to Ulpian, and, so far as the 
 Corpus Juris shows, this was the only one which 
 the jurists distinctly recognized. 4 We can say at 
 
 1 Supra, p. 20. 
 
 2 Savigny (System, Beilage i, p. 415) thinks that these two 
 jurists adopted the trichotomy, but the passages in the Digest 
 bearing on the point are not conclusive, viz. Dig. xii. 6, fr. 64 ; 
 xvi. 3, fr. 31 ; and i. 1, fr. 5. It has been suggested that the 
 first two of these, and a passage from Paul, Dig. xviii. i. 34 1, 
 have suffered interpolation ; see Mitteis, Privatrecht, p. 63 n. 
 This seems improbable. 
 
 3 Gai. i. 1. Cicero, de Off. iu. 5, 23, had said 'jus naturae 
 id est gentium.' 
 
 4 E.g. Marcian in Dig. i. 8, fr. 2 and 4 ; Florentine in Dig. i. 
 8, fr. 3 ; Paul in Dig. i. 1, fr. 11, and xxiii. 2, fr. 14 2, and 
 1. 17, fr. 84 1 ; Modestine in Dig. xxxviii. 10, fr. 4 2. The 
 term j. g. is used by them sometimes in a wide, sometimes in 
 
28 TKICHOTOMY 
 
 any rate that no jurist recognized a distinction 
 between jus naturale and jus gentium so far as 
 practical effects were concerned. Jus naturale is 
 defined by Ulpian as ' quod natura omnia animalia 
 docuit ', such as ' conjunctio maris atque feminae, 
 liberorum procreatio, educatio V This definition 
 elevates the instinctive actions of birds, beasts, 
 and fishes to the domain of law, and astonishment 
 has been often expressed by modern writers that 
 a scientific lawyer like Ulpian should have adopted 
 it. At the same time attempts have been some- 
 times made to defend or explain it on philosophi- 
 cal grounds (e. g. on the Stoic doctrine of living 
 according to nature), but none of these can be 
 admitted by jurisprudence. 2 Law is not con- 
 cerned with instincts that are common to man 
 and the lower animals. 3 The only theoretic con- 
 sequence which Ulpian himself deduced from it 
 was that slavery was contrary to the law of nature 
 but sanctioned by jus gentium 4 , as regards prac- 
 
 a narrow sense. Cf. Mitteis, I. c. and authorities there 
 cited. 
 
 1 Dig. i. 1, 1 3. 
 
 2 See Savigny, System i, p. 416 ; Vinnius, Com. ad Inst. i. 
 2 pr. Ulpian himself says (D. ix. 1, 1 3) ' nee enim potest 
 animal iniuria fecisse quod sensu caret '. 
 
 3 Cic. de Fin. iii. 20. 67. 
 
 4 This conception was undoubtedly borrowed from the 
 Stoics. Even as regards slavery it is hard to suppose that 
 
IN KOMAN LAW 29 
 
 tical consequences he really did not adhere to 
 or make any use of it, but on the contrary posi- 
 tively identified jus naturale and jus gentium in 
 various passages in his writings other than the 
 one above cited. 1 
 
 Justinian in his Institutes, while so far adopt- 
 ing the tripartite division of Ulpian, makes at the 
 same time a variation from it which Savigny 
 rightly describes as senseless (gedankenlos). 2 He 
 first gives Ulpian's definition of jus naturale, and 
 then proceeds to define jus gentium in the words 
 of Gaius as ' quod naturalis ratio inter omnes 
 homines constituit '. 3 He then further somewhat 
 modifies his definition of jus gentium by applying 
 it to the settled usages of mankind and the require- 
 ments of social life ( l usu exigente et humanis ne- 
 cessitatibus '), 4 as shown by the consensual and 
 
 Ulpian can have had in view his own definition. Mediaeval 
 lawyers following Ulpian fell into similar difficulties on the 
 subject. See Odofredus in Digestum Vetus i. 1 6, citing 
 a gloss of Irnerius (lucerna juris). This gloss is given by 
 Savigny, GescMchte des rom. Eechts im Mittelalter iv, p. 387. 
 See also Carlyle, Mediaeval Political Theory in the West, vol. ii, 
 pp. 28-33. 
 
 1 E. g. Dig. ix. 2, fr. 50 ; x. 4, fr. 3 15 ; xvi. 2, fr. 6 ; xliii. 
 16, fr. 1 9 ; xliv. 7, fr. 14 ; 1. 16, fr. 10. 
 
 2 Inst. i. 1 4 ; i. 2 pr. See Savigny, System i, p. 419. 
 
 3 Inst. i. 2 1. 
 
 4 Inst. i. 2 2. 
 
30 TKICHOTOMY 
 
 other contracts, 1 and then finally, as every tyro 
 knows, altogether ignores in the rest of the work 
 (save as regards the origin of slavery) the tripartite 
 division, and even in one passage expressly iden- 
 tifies jus naturale with jus gentium. 2 There was 
 indeed a sense in which, in the time of the classical 
 jurists, jus naturale was distinguishable from jus 
 gentium and might have justified a tripartite divi- 
 sion. The law based on 'naturalis ratio inter 
 omnes homines' (called jus gentium in Gaius i. 
 1, and Inst. i. 2 1) truly represented the jus 
 naturae or naturale of the philosophers and philo- 
 sophizing lawyers of the early empire. 3 It was 
 essentially theoretical and based on speculation, 
 and in this sense was distinguishable from jus 
 gentium historically considered, which formed from 
 an early period part of the jus Romanum.* That 
 dominium might be acquired by occupatio and 
 traditio and that civil contracts might be created 
 by mere consent were illustrations of jus gentium 
 in this latter sense, and to it the name might well 
 have been confined ; while upon the jus naturale^ in 
 
 1 Following in this Ulpian and Hermogenian Dig. 1. fr. 1 
 5 and frs. 4 et 5. 
 
 2 See Inst. i. 2 11, and ii, 1 11. 
 
 3 Of. Cicero, de Off. iii. 5. 23. 
 
 4 This is vaguely contemplated in Inst i. 2 2, but the 
 definition there is too wide for the historic jus gentium. 
 
IN EOMAN LAW 31 
 
 its wide speculative conception, depended inter alia 
 such dogmas (the product mainly of Stoic philo- 
 sophy operating upon the jurists), as the per- 
 sonality of slaves, naturalis cognatio and naturalis 
 oUigatio. 1 Unhappily, however, the jurists, as 
 already indicated, failed to make any precise 
 distinction between the two. However this be, 
 it noways justifies Ulpian's tripartite division, 
 which must ever remain what Savigny calls 
 a Curiositat, and seems only explicable on the 
 hypothesis for which I am contending. 
 
 4. The statement of Modestine regarding the 
 sources of law that ' omne jus aut consensus fecit 
 aut necessitas constituit aut firmavit consuetudo ' 2 
 is very different in its point of view from that of 
 Papinian in Dig. i. 1, fr. 7 (who divides the sources 
 of the written law into five parts), and from the 
 well-known divisions given by Gaius and Justinian 
 in their Institutes. Modestine's classification is 
 obviously traditional, and Cujacius thinks it was 
 taken from the Greek of Menander. Modestine 
 was not the kind of man to originate it. It is 
 
 1 See on this Voigt, Jus naturale, etc., i, pp. 286 sq. ; 
 Krttger, G. d. Quellen, 17. Early writers on International 
 Law, e. g. Grotius and Puffendorf, lay stress on this distinc- 
 tion. For Greek law see Schulin, GescMchte d. rom. Rechts, 
 pp. 85-6. 
 
 2 Supra, p. 20. 
 
32 TKICHOTOMY 
 
 both scientifically and practically worthless, though 
 Cujacius makes an ingenious, if futile, attempt to 
 explain it. 1 No use was ever made of it. 
 
 5. 'Jus praetorium est quod praetores intro- 
 duxerant adjuvandi, vel supplendi vel corrigendi 
 juris civilis gratia.' 2 This, in some respects 
 admirable, definition of Papinian has been much 
 commented upon by modern writers on Eoman 
 law, and treated as if it represented (which it 
 hardly does) an exhaustive summary of the prae- 
 tors' reforms. 3 Yet its three parts what Cujas 
 calls the 'tres virtutes juris praetorii' awaken 
 suspicion, and I cannot resist the supposition 
 that it was borrowed from some quasi-philosophic 
 writer. With it may be compared the definition 
 in Justinian's Institutes, where the expressions 
 confirmandi, emendandi, and impugnandi, juris civilis 
 gratia are used. 4 
 
 1 Cujac. Obs. xiv, c. 16. He thinks that ' omne jus ' must 
 be limited to leges, senatus consulta, and constietudo, to which 
 threefold division of topics the rubric of the Title (Dig. i. 3) 
 is confined. His comments are very strained and uncon- 
 vincing. He also endeavours to explain the text alter- 
 natively on the assumption that it applies to the whole body 
 of law. If he had thought of a Stoic trichotomy he would 
 have saved himself the trouble. See also Otto, TJiesaur. iii, 
 p. 695. 
 
 2 Supra, p. 20. 
 
 3 See e. g. Sohm, Inst, pp. 540 and 547. 
 
 4 Inst iii. 9, pr. 1. 
 
IN EOMAN LAW 33 
 
 6. The imperial constitutions as a source of law 
 are divided into three kinds, viz. decreta, edicta, 
 and epistulae. 1 This purely legal classification does 
 not seem open to serious objection. It may have 
 originated with Gaius and at any rate cannot be 
 much earlier than his time. Nevertheless it is 
 incomplete, as it omits mandata. 2 These mandata 
 though sometimes contrasted with constitutiones 
 (e.g. Dig. xlvii. 11. 6) cannot be sensibly dis- 
 tinguished from, but were a species of, them. 
 Eeasons have been suggested for this omission by 
 modern writers, but none of them satisfactory. 
 Thus it has been said by some that, as they are of 
 the nature of edicta, they may have been intended 
 to be included in that category; by others that 
 they were intended to be included under epistulae, 
 and by others that they were deliberately excluded 
 because they related mainly to matters not of 
 private but of public law. 3 But these views cannot 
 be maintained, it is thought, in face of the texts. 4 
 
 1 Supra, p. 20 ; Krttger, Quelkn, p. 99. 
 
 2 How far leges datae (e. g. lex Salpensana) should also be 
 included may be a question. See Schulin, Gesch. d. rom. 
 Eechts, p. 102. 
 
 3 See Schrader, Inst. i. 2 6 ; Puchta, Inst. as in next note. 
 
 4 See Dig. iv. 1, 1 pr. ; xxix. 1, fr. 1, pr., fr. 2 ; xxxiv. 9. 
 2 1 ; Pernice rev. Wlassak, Z. d. Sav. Stift. vi, pp. 297-8. 
 On the possibility of an interpolation in Dig. i. 4. 1 1, see 
 Puchta, Inst. i ex, note g. 
 
 1211 E 
 
34 TEICHOTOMY 
 
 7. The subject-matter of private law relates 
 to persons, things, and actions. 1 Personae, res, 
 and actiones represent respectively the subjects, 
 objects, and sanctions of rights. I will not enter 
 upon an exhaustive criticism of this famous 
 classification, upon which so much has been 
 already written, and which has so largely domi- 
 nated the method of mediaeval and modern 
 writers upon law. It has its conveniences but 
 it is quite illogical. Apart from the illogicality 
 of co-ordinating the law of actions with that 
 relating to persons and things, to which it stands 
 properly as adjective, the classification forced 
 Gaius to treat property as a res corporalis, and 
 succession and obligations as res incorporates* 
 rights to things being thereby confounded with 
 things themselves. 3 
 
 1 Supra, p. 20. 
 
 2 As is well known, it has been doubted whether obliga- 
 tions were intended in the classification in Justinian's 
 Institutes to fall under the head of res or actiones. Theo- 
 philus in his Paraphrase treated them as falling under 
 actions. 
 
 3 See Austin, Jurisprudence, ii, pp. 764-6. Austin remarks 
 on the Gaian method 'In pursuance of that well-founded 
 distinction between Law of Persons and Law of Things, 
 which Gaius . . . has adopted, the main division (according 
 to Ends and Subjects) should not have been threefold (still 
 less fourfold) but twofold ; the matter of the Law of Actions 
 (i. e. civil procedure), as unmodified by status, . . . being 
 
IN EOMAN LAW 35 
 
 Here also it, is a question whether Gaius 
 borrowed his tripartite division direct from 
 some earlier writer, and the probabilities are 
 all in favour of his having done so. It had 
 already, at any rate, been foreshadowed in non- 
 legal works as, for instance, in Varro's Anti- 
 quitateSj where, in connexion with the doctrine 
 of res divinae, he treats first of persons in rela- 
 tion to things, then of loca sacra vel religiosa, and 
 finally of the Calendar and days of festival. 1 
 All attempts to explain the classification on 
 grounds of abstract jurisprudence seem to me 
 quite vain, 2 as, for example, the old theory of 
 Duarenus that Gaius did not intend the three 
 parts to be mutually exclusive, but used them 
 rather to indicate that the whole law may be 
 
 included under the Law of Things, and the several modifica- 
 tions, under the several status out of which they arise/ 
 
 1 See Girard, Textes, 3rd ed., p. 203 ; Karlowa, E. E. G. i, 
 p. 725. For the view that it was more probably due to 
 Gaius than traditional, see Savigny, Syst., i, pp. 395-7, where 
 he refers to Hugo, Civil Mag. 5, p. 403, and 6, p. 286. Hugo's 
 view that it was taken from some early non-juristic writer 
 seems to me most probable. Karlowa, 1. c., remarks ' diese 
 Eintheilung ist aller Wahrscheinlichkeit nach ein altes 
 Gemeingut der romischen Eechtsgelehrten, auf sie vererbt 
 von der alten pontifizischen Jurisprudenz ', and refers to 
 Booking, Pandekten, i 2 , Anhang vi. 
 
 2 See Austin, Jurisprudence, ii, pp. 764-66. 
 
36 TEICHOTOMY 
 
 studied in relation to any one of them. 1 The 
 author of the division (whether Gaius or another) 
 wanted a tripertita, and the doctrines had to be 
 fitted into it somehow. One may, however, well 
 attribute the division to the traditional influence 
 of the Stoa 2 a view which seems to me to be 
 supported by the fact that Ulpian in his Institu- 
 tions makes, or rather borrows, another scholastic 
 trichotomy of the omne jus, which, as we will see 
 under next head, is without any scientific value 
 whatever. 
 
 8. In his Institutions, as cited in the Digest, 
 Ulpian says : i Totum autem jus consistit aut in 
 adquirendo aut in conservando aut in minuendo : 
 aut enim hoc agitur quemadmodum quid cujusque 
 fiat, aut quemadmodum quis rem vel jus suum 
 
 1 The passage of Duarenus is quoted by Booking in an 
 appendix to his edition of Gaius and by Emerton, The Three- 
 fold Division of Roman Law, p. 19 (London, 1888). Einerton 
 founds on the use by Gaius of vel instead of aut in the 
 division, as confirming the view of Duarenus, but his argu- 
 ments are to my mind quite unconvincing. Cf. Savigny, 
 Syst. , ut sup. 
 
 2 Compare the division of Philosophy by the Stoics into 
 Logic, Natural Science, and Ethics. See Zeller, Stoics, Epi- 
 cureans and Sceptics, chap. iv. (Reichel's Trans.), pp. 62-7; 
 August, de Civ. Dei, xi, c. 25. But for my thesis it is 
 immaterial whether it was taken from a Stoic writer on 
 jurisprudence or was an inheritance from old pontifical 
 traditions. 
 
IN KOMAN LAW 37 
 
 conservet, aut quomodo alienet aut amittat/ ! 
 This is a rather strange limitation of the totum jus, 
 but we may perhaps assume that Ulpian is here 
 referring to the jits privatum alone. Can any 
 one doubt that it represents one of Ulpian's 
 borrowings from some philosophic or pseudo- 
 philosophic treatise ? Apparently he attached no 
 significance or importance to it whatever. Com- 
 mentators have as a rule, so far as I have ob- 
 served, left the passage severely alone, but Savigny 
 makes some observations about it. After stating 
 the proposition that every legal relation has its 
 own special rules according to which in respect of 
 any specific person it comes into existence and 
 again is extinguished, and that these rules are of 
 such importance that in many legal relations they 
 form the sole ground of inquiry, he cites the 
 above passage in a note and remarks : ' Hier wird 
 noch ein neues Moment in die Mitte jener bei- 
 den [i.e. Entstehung and Untergang] gestellt, das 
 conservare. Nimmt man dieses im eigentlichen 
 Sinn fur die Bewirkung der Fortdauer des Kechts 
 selbst, so fallt es mit dem dritten zusammen, 
 indem es dann als die Negation des dritten (oder 
 umgekehrt) aufgefasst werden kann ; dann sagt 
 aber auch das totum ius consistit viel zu viel. 1st 
 1 Supra, p. 21, n. 1 . 
 
38 TEICHOTOMY 
 
 dagegen das conservare als Erhaltung der Aus- 
 iibung, oder als Kechtsverfolgung, gedacht, so um- 
 fassen allerdings jene drei Momente den grossten 
 Theil aller Bechtsregeln iiberhaupt : dann aber 
 sind die drei Momente nicht so gleichartig dass 
 diese Zusammenstellung derselben gerechtfertigt 
 werden konnte. 5 1 All this may be admitted, but 
 the ineptitude of the text and its worthlessness 
 to Ulpian himself remain to my mind perfectly 
 clear. No use is made of it. 2 
 
 9. The grouping of persons into three classes, 
 viz. (1) liberi and servi, (2) cives, latini, and pere- 
 grinij and (3) sui juris and alieni juris, is not formally 
 given by Gaius, but it appears to have been before 
 his mind quite distinctly nevertheless. Whether 
 it was borrowed by him is unknown. In terms, 
 however, the second class is confined by him to 
 liberti, whom he divides into cives, latini, and pere- 
 grini dediticiorum numero. 3 The reason why a 
 similar division of the second class was not made 
 specifically by him for the ingenui is not easy to 
 say. It may have been that he regarded the dis- 
 cussion of civitas and peregrinitas as falling under 
 
 1 System, iii, p. 2, note (b). 
 
 2 Except perhaps the triad of interdicts ? See infra, p. 53. 
 8 Gai. i 12 ; see supra, p. 21, n. 2 ; infra, p. 41 ; Mommsen, 
 
 Staatsrecht, iii. 1, p. 8 2 . 
 
IN EOMAN LAW 39 
 
 public rather than private law. Whatever the 
 reason, he only notices freeborn citizens, latins and 
 peregrins incidentally, and in so far as it enables 
 him to discuss by analogy the three classes of 
 libertini. It is an obvious defect of the above 
 classification that slaves fall under both (1) and (3). 1 
 10. Persons are again treated by Ulpian as 
 divisible jure gentium into three classes, ' liberi 
 et his contrarium servi et tertium genus liberti.' 2 
 This is apparently an illogical division, as liberti 
 are not co-ordinate with the other two but form 
 a subordinate class of liberi. It is moreover in- 
 consistent with Ulpian's own definition of manu- 
 mission as ' datio libertatis ' (Dig. i. 1, 4). Gaius 
 avoids the illogicality and rightly divides persons 
 into servi and liberi, subdividing the latter into 
 ingenui and libertini. 5 Some commentators have 
 suggested that by liberi Ulpian meant ingenm,* 
 but this would also make a bad classification 
 and is merely an attempt to save his logic. 
 
 1 For a logical classification of persons see Savigny, System, 
 ii, pp. 58-9. 
 
 2 Swpra, p. 21, n. s . His language seems to indicate a tradi- 
 tional division. 
 
 3 Gai. i 9. Justinian's editors give first the Gaian 
 classification (Inst. i. 3 pr. 5), and then carelessly that of 
 Ulpian (Inst. i. 5 pr.). 
 
 * See Schrader, Inst. i. 5 pr. 
 
40 TKICHOTOMY 
 
 Moreover, a classical jurist would not have used 
 ingenuitas in this way in contrast with servitus. 
 It is hard to suppose that Ulpian, to whom 
 the dual classification of Gaius must have been 
 familiar, should have been unconscious that 
 he was making a cross division. Whether he 
 borrowed it or not, I see no explanation but the 
 desire for a tripartite division. 
 
 11. Gaius (i J 17) says, regarding acquisition 
 of civitas by a manumitted slave, ' in cujus per- 
 sona tria haec concurrunt, ut major sit annorum 
 triginta, et ex jure Quiritium domini, et justa ac 
 legitima manumissione liberetur, id est vindicta 
 aut censu aut testamento, is civis Eomanus erit ; 
 sin vero aliquid eorum deerit, Latinus erit.' 
 Here we are given, first, three conditions for, 
 and second, three modes of, manumission. As 
 regards the conditions, it was only the second 
 and third that were required by the old jus civile 
 to make a Eoman citizen ; the first was due to 
 the lex Aelia Sentia. Apparently the three con- 
 ditions are intended to be exhaustive, but, if so, 
 they are wanting in accuracy. The age of the 
 master, for instance (he must be over twenty 
 years), had by the Aelia-Sentian law as much to 
 be considered as that of the slave. 1 But this 
 1 Gai. i 18, 38 ; cf. Ulp. Reg. i 12, 13. In the 
 
IN KOMAN LAW 41 
 
 may perhaps be a mere oversight of Gaius or 
 due to his mode of exposition. 
 
 Besides the three forms of manumission men- 
 tioned (which Ausonius apparently gives as an 
 illustration of the ternarius numerus, under the 
 name * triplex libertas ' *), there were other modes 
 of a slave being legally manumitted and obtain- 
 ing civitas, e. g. grant by the State under various 
 enactments. 2 In any case the trichotomy is an 
 old one and purely Koman. 
 
 12. The classification above mentioned of 
 freedmen into cives, latini, and dediticiorum nvr 
 mero* is obviously the symmetrical counterpart 
 of the division of freemen into cives, latini, and 
 peregrini, and is, as we have seen, one of the 
 three groups under whicli Gaius discusses persons 
 
 West Gothic Gaius the regular modes of manumission are 
 different but are kept tripartite, viz. ' testamento, aut in 
 ecclesia, aut ante consulem.' Gai. Epit. i. 1. 
 
 1 EdyU. xi, 1. 65, but see infra, p. 42, n. 1 . 
 
 2 See also Inst. i. 11 12, adoption by a master of one of 
 his own slaves. The irregular modes of manumission by 
 which Junian latinity was obtained were also classified as 
 threefold in the West Gothic Gaius, where it is said ' Latini 
 sunt qui aut per epistolam aut inter amicos aut convivii adhi- 
 bitione manumittuntur '. As to the magistrates before whom 
 manumission could take place cf. Ulp. Reg. i 7 apparently 
 consul, praetor, proconsul with Inst. i. 5 2, and Schrader's 
 note hoc loco. 
 
 3 Supra, p. 21. See also Gai. Epit. i. 1. 
 
 1211 
 
42 TEICHOTOMY 
 
 in the first book of his Commentaries. In the 
 Institutes of Justinian this classification of freed- 
 men is called tripertitus status. 1 It cannot of 
 course be earlier than the Empire, as it was 
 based on the lex Aelia Sentia and lex Junia. 
 
 As we learn from Gaius (iii 74-6) these 
 dediticii were, in the matter of succession to their 
 states, treated sometimes as citizens and some- 
 times as latins, this depending on the nature of 
 their manumission. 2 The only criticism one can 
 make upon the classification is that it is not 
 exhaustive. There were other freedmen who 
 were peregrini, though not in the number of dedi- 
 ticii, viz. those manumitted by peregrin owners 
 in accordance with the laws of the latter's own 
 civitos.* 
 
 13. Persons alieni juris were said to be either in 
 potestate, or in manu, or in mancipio.* In the early 
 period of Eome the only word to express the 
 powers of a paterfamilias over the familia was 
 
 1 Inst. i. 5 3. It is not certain whether the term 
 1 triplex libertas' in Ausonius, Edyll. xi, 1. 65, applies to 
 this or to the three modes of manumission. Probably the 
 latter. 
 
 2 Cf. Gai. i 17. 
 
 3 Pliny, Ep. x. 4 ; Frag. Dosith. 12. See Girard, Manuel 
 d. d. r., 4th ed., p. 126. 
 
 4 Supra, p. 21, n. 6 . 
 
IN EOMAN LAW 43 
 
 probably manus ; potestas was a word of later 
 origin. But in time manus, though still occasion- 
 ally used in a general sense to express power, 1 
 became technically restricted to power over a 
 wife. Mancipium was also a term of early law 
 used in a variety of senses, but in time came to 
 mean (in the expressions 'in mancipio' or 'in 
 causa mancipii ') power of a paterfamilias over free 
 persons mancipated to him, not being liberi or 
 loco liberorum. Potestas was, like mantis in early 
 times, a general term for power of the head of 
 the house, but was more specially applied to 
 power over slaves and children. This tripartite 
 division, however, as given by Gaius and Ulpian, 
 is quite artificial and without solid basis. Con- 
 sidering the fundamental difference in the time 
 of the classical jurists between slaves and chil- 
 dren, one would naturally expect a fourfold 
 division of persons alieni juris, viz. those in domi- 
 nica potestate, those in patria potestate, those in 
 manUj and those in mancipio and in fact Gaius 
 does discuss them as if so classified. 2 It has been 
 
 1 e. g. Inst. i. 5 pr. ; Dig. i. 1, 4 and i. 2, 2 1 ; Cod. viii. 
 48. 6. 
 
 2 Gai. i 49 sq. The classification is doubtless older 
 than Gaius. In Justinian's Institutes, as manus and manci- 
 pium were in desuetude, there is a twofold division into 
 
44 TEICHOTOMY 
 
 justly observed by Poste, * The powers of the 
 head of the family were nominally three, potestas, 
 manuSj mancipium -, potestaSj however, was either 
 potestas dominica ... or potestas patria . . ., which 
 at the period when Koman law is known were 
 different in kind ; so that the rights of a pater- 
 familias were really fourfold.' l The differences 
 between patria and dominica potestas were at all 
 times much greater than those between manus 
 or mancipium and patria potestas. 
 
 14. On Gaius's (and presumably Ulpian's) divi- 
 sion of marriages cum manu into confarreatio, 
 coemptio, and usus, 2 there is only this remark to 
 be made that Gaius gives it as traditional, and, 
 though possibly correct at one time, it was out of 
 date at the time he wrote. It is a purely Eoman 
 trichotomy. Hand marriages were still in the 
 second century of our era contracted by coemp- 
 tion, though rare, and even for certain purposes 
 by confarreation, 3 but usus as creative of manus, 
 as Gaius tells us, had been abolished partly by 
 
 persons in dominica, and persons in patria potestate. Potestas 
 was also applied to the power of guardians over wards. 
 XII Tab. v. 7 ; Inst. i. 13 1 . ; Dig. xxvi. 1, 1 pr. ; Aul. 
 Gell. v. 19. 10. See also Dig. 1. 16, 215. 
 
 1 Poste, Gaius (3rd ed.), p. 40. 
 
 2 Supra, p. 21, note 7 . 
 
 3 Gai. i 112, 113, 136. 
 
IN EOMAN LAW 45 
 
 statute and partly by desuetude before his time. 1 
 Gaius apparently says not a word about the con- 
 stitution of marriage by simple consent, which 
 was the ordinary mode in his time. 2 
 
 15. Extinction of patria potestas over sons was 
 effected by three sales and three corresponding 
 manumissions. I have already sufficiently dis- 
 cussed the symbolism of this. 3 It is probable, 
 and it seems to be the view generally accepted, 
 that originally this rule applied to all descendants, 
 and that it was by casuistical interpretation of 
 the word * filium ' in the law of the XII Tables, 
 that a single mancipation and manumission was 
 subsequently held to be sufficient in the case of 
 daughters and grandchildren. 4 
 
 16. Persons sui juris are distributed by Gaius 
 and Justinian in their Institutes into those in 
 tutory, those in curatory, and those who are in 
 neither the one nor the other. 5 The Eomans 
 had no technical term equivalent to our English 
 
 1 Gai. i 111. 
 
 2 Had this traditional tripartite division anything to do 
 with Justinian's treatment of marriage in the Digest ? See 
 Const. Tanta, 5 ; Hofmann, Ztsch. f. R. G. xi, p. 350, who 
 says, 'Die Zahl drei bedeutet unter anderem auch die Ehe.' 
 
 3 Supra, p. 13 and p. 22, n. 1 . 
 
 4 See Jhering, Geist des rom. Eechts, ii, p. 104. 
 
 5 Supra, p. 22, n. a . 
 
46 TEICHOTOMY 
 
 ' infant ' signifying a person under majority, or to 
 our English 'ward' signifying any person under 
 guardianship, nor had they any general term 
 equivalent to guardianship. Pupilarity and mi- 
 nority were with them fundamentally distinct and 
 had different effects, and it was proper for them 
 to treat tutory and curatory separately. But we 
 have here a strange tripartite division, of which 
 the third class is wholly uncalled for, save in order 
 to be symmetrical with the division of persons 
 alieni juris. The way in which the classification 
 is made shows that it was intended to be three- 
 fold: 'Videamus igitur quae in tutela, quae in 
 curatione sint ; ita enim intellegemus ceteras per- 
 sonas quae neutro jure tenentur.' x It would have 
 been as reasonable for Gaius to have treated of 
 marriage by dividing persons into those married 
 cum manu, those married sine manUj and those who 
 are neither. But a threefold classification, of 
 which one of the heads does not, ex necessitate, 
 admit of discussion in relation to the other two, 
 is worthless. 
 
 17. In the Institutes of Justinian and also in 
 the Digest tutors are treated (without any formal 
 
 1 Gai. i 142 ; cf. Gai. i 50, 51. These divisions 
 rest on the principle of Aristotelian logic ' contrariorum 
 cognito uno, cognoscitur et alterum.' 
 
IN KOMAN LAW 47. 
 
 classification being given) as divisible into testa- 
 mentarii, legitimi, and a magistratibus dati, 1 and 
 this is not objectionable. It was familiar to 
 Ulpian and doubtless used by him. But in his 
 Eules he gives formally a separate tripartite 
 division, viz. legitimi, senatusconsultis constitute and 
 moribus introducti, and this seems highly arti- 
 ficial. 2 It seems to be conceived from the point 
 of view of the sources from which the office of 
 tutor is derived. But the first class (the legitimi) 
 is altogether out of proportion to the other two, 
 embracing as it does each of the three kinds in the 
 above-mentioned Justinianian treatment, i.e. tutory- 
 at-law (agnatic and patronal) as well as testamen- 
 tary and ordinary magisterial tutory. 3 The second 
 
 1 Inst. i, tits. 13-20 ; Dig. xxvi, tits. 2-5. 
 
 2 Ulpian, Eeg. xi 2-24. Ulpian seems at times to lose 
 sight of it. See 5, 6, where he distinguishes apparently 
 tutor fiduciarius from tutor legitimus. As Gaius tells us 
 (i 172), the former was thought to be unable to cede injure 
 his tutory while the latter could do so. See Ulp. Eeg. 17. 
 Like his tripartite division of the totum jus (supra, p. 36), 
 Ulpian evidently attaches no importance to this division. 
 
 3 While Gaius, i 168, subdivides tutors-at-law into three, 
 viz. agnati, patroni, and liberorum capitum manumissores, 
 Justinian in the Institutes subdivides them into agnati, 
 patroni, parentes, and fiduciarii. Ulpian, on the other hand, 
 in a passage in the Digest (xxvi. 4, 1 pr.), ascribes to the XII 
 Tables three species, viz. agnati, consanguine^ and patroni a 
 
48 TRICHOTOMY 
 
 and third classes represent merely such magisterial 
 tutors as were appointed in special cases either 
 by virtue of a senatusconsult or by custom. The 
 senatorial tutors were appointed by magistrates 
 only in a few exceptional cases, as, e.g. where 
 a legitimus patronal tutor of a woman was dumb 
 or insane and auctoritas was required to constitute 
 her dowry, 1 while the tutors by customary law 
 were likewise appointed by magistrates excep- 
 tionally, as, for instance, where a pupil or adult 
 woman was desirous of raising an action against 
 his or her proper tutor. 2 In any proper classifica- 
 tion it is obvious that these special cases would 
 have been treated simply as exceptions in discuss- 
 ing the magisterial tutors, as in fact is done by 
 Gaius. 3 Could Ulpian have had before his mind 
 the Ciceronian 'si partiri velis tutelas, inscienter 
 facias si ullam praetermittas ' ? 4 Gaius, speaking 
 of the species tutelarum, as he calls them, viz. 
 
 plainly objectionable division. See Gothofredus, edition of 
 Corpus Juris (1663), notes on this .passage. 
 
 1 Ulp. xi 21 ; cf. Gai. i 177, 180-3. By the lex 
 Julia de marit. ord. the same course was followed where the 
 patronal tutor was a p.upil. Gai. i 178-9 ; Ulp. xi 20. 
 
 2 Ulp. xi 24 ; cf. Gai. i 184. Such tutor was technically 
 called praetorius. 
 
 3 Gai. i 176-84 ; cf. Inst. i. 21 3. 
 
 4 Cic. Top. 33. 
 
IN EOMAN LAW 49 
 
 the testamentary, agnatic and others, says this 
 (i 188) : ' Si vero quaeramus in quot genera hae 
 species diducantur, longa erit disputatio ; nam de 
 ea re valde veteres dubitaverunt, nosque diligen- 
 tius-hunc tractatum executi sumus et in edicti 
 interpretatione et in his libris quos ex Q. Mucio 
 fecimus. Hoc tantisper sufficit admonuisse quod 
 quidam quinque genera esse dixerunt ut Q. Mucius; 
 alii tria ut Ser. Sulpicius ; alii duo, ut Labeo ; alii 
 tot genera esse crediderunt, quot etiam species 
 essent.' This passage seems to show rather an 
 odd lack of capacity for systematic arrangement 
 among the earlier jurists. 
 
 18. Paulus says, i Tria sunt quae habemus, 
 libertatem, civitatem, familiam.' 1 But obviously, 
 as Savigny points out, there were many other 
 things in private law quae kabemus, such as 
 property, obligations, marriage, &c. What Paul 
 probably however meant was that status, in the 
 sense of caput or civic personality, is composed 
 of three elements, viz. liberty, citizenship, and 
 family. These made up a man's caput under jus 
 civile. The classical jurists seem to have im- 
 properly invented the triad, holding that no one 
 enjoyed full public and private rights who did 
 
 1 Supra, p. 22. Dig. iv. 5, 11. See Savigny, System, ii. 
 Beilage vi. 9 (pp. 508-10). 
 1211 
 
50 TRICHOTOMY 
 
 not possess all three. But civitas included fa- 
 milia. On the other hand, it is clear that in early 
 times the distinction between libertas and civitas 
 was unknown; if a person was not a citizen he was 
 de jure a slave, he had no caput. 1 
 
 19. Capitis deminutio is either the greatest, the 
 middle, or the least. 2 There are various passages 
 in the texts proving that originally there were 
 only two kinds of cap. dem., viz. the greater and 
 the less (magna or major and minor), and that the 
 division into three classes was due to the jurists 
 of the last century of the Republic. 3 The magna 
 cap. dem. meant loss of citizenship and with it all 
 public and private civil rights ; the minor cap. dem. 
 meant loss of caput within the sphere of the private 
 law alone, i.e. loss of existing family rights. This 
 was sufficient to meet all cases. But the jurists, 
 
 1 The word liber signifying both free and a child points 
 to this. See as to liber and libertus, Mommsen, Eom. Staatsr. 
 iii. pp. 8, 62, 421. Commenting on Capitis deminutio, 
 Savigny (1. c. p. 511) says: 'die ganze Stelle des Paulus 
 erscheint demnach nur als ein mislungener Versuch die 
 dreifache Cap. dem. auf eine rationelle Weise zu begrtinden. 
 
 2 Supra, p. 22. Ausonius gives this as an illustration of 
 ternarius numerus Edyll. xi, line 65. 
 
 3 Thus Ulp. in Dig. xxxviii, 16. 1 4, speaks of only 
 magna and minor, but he is not there making any division. 
 For the varying terminology in the texts see Savigny, 
 System, ii 68 d. Cf. Kuntze, Inst. ii. p. 368. 
 
IN EOMAN LAW 51 
 
 founding on the distinction, which gradually grew 
 up, between that loss of citizenship which carried 
 with it loss of liberty, and that which did not, but 
 reduced merely to the status of peregrin, divided 
 the magna cap. dem. into maxima and media or 
 minor, and gave the name minima to cap. dem. 
 within the private law. Thus they got a triad. 
 In both the maxima and media there was loss 
 of citizenship ; but the former, it was said, reduced 
 to slavery, the latter did not. 2 We have thus an 
 attractive but somewhat irrational symmetry 
 maxima includes the other two, media includes 
 minima to which the three elements of status 
 above considered, viz. libertas, civitas, familia, were 
 made to correspond. 3 This view of the origin of 
 the threefold division is, indeed, only conjectural, 
 
 1 Could a non civis be said to undergo cap. dem. maxima? 
 There is no sufficient evidence of it, though it has been 
 maintained by some writers. See Cohn, Beitrage, Heft ii. 
 pp. 46 sq. 
 
 2 Mommsen observes, E. S.-R. iii. 1, p. 8 The Koman 
 jurists, in order to support their threefold division into (1) 
 liberi et servi, (2) ewes, L., p., and (3) sui et alieni juris, have 
 obscured the meaning of cap. dem. by setting up three 
 degrees of it, * obwohl sie selbst diese mehrfach auf zwei 
 zuruckfuhren und zuruckfiihren mtissen, weil dem spateren 
 Privatrecht Freiheit und Burgerrecht zusammenfallt. In 
 der That sind es nicht zwei Grade, sondern zwei verschiedene, 
 obwohl correlate Kechtsbegriffe.' 
 
52 TKICHOTOMY 
 
 but it has received incidentally a good deal of 
 support from the inquiries of recent writers. 1 
 
 If the foregoing critical remarks be sound, it 
 must be admitted that the great bulk of the 
 classifications in the department of the law dealt 
 with are traditional tripertita resting on no logical 
 basis and only explicable on the theory for which 
 I am contending. In the other departments of 
 the law discussed in the three remaining books 
 of the Institutes of Gaius and Justinian the tri- 
 partite division, though it frequently occurs, is' 
 apparently not so prominent as in the first 
 books, while in the department of obligations in 
 Justinian's Institutes it wholly gives place, as I 
 have shown in the Studi Fadda, to artificial 
 quadripartite classifications. However, I will 
 now notice a few threefold divisions in the 
 departments of the law relating to things and 
 actions that have given rise to difficulties and 
 seem to me to support my thesis. 
 
 (a) There are two well-known tripartite classifi- 
 cations of interdicts. By the first, which is called 
 summa divisio, they are divided into exhilitory, resti- 
 tutory, and prohibitory. 2 On this I will content 
 
 1 See Mommsen, Staatsrecht, 1. c. ; H. Krttger, Cap. Dem. 
 vol. i. 
 
 2 Gai. iv, 142; Inst. iv. 15 1 ; Paul in Dig. xliii. 1. 
 
IN KOMAN LAW 53 
 
 myself with quoting the following significant 
 passage from Ulpian : ' Interdictorum autem tres 
 species sunt; exhibitoria, prohibitoria, restitutoria; 
 sunt tamen quaedam interdicta et mixta, quae et 
 prohibitoria sunt et exhibitoria/ l Here Ulpian 
 mentions four classes and yet says 'tres species 
 sunt'. 2 
 
 The second classification is that of the so-called 
 possessory interdicts, they being divided into adi- 
 piscendae, retinendae and recuperandae possessions 
 causa ; 3 Ausonius, in his Griphus above mentioned, 
 gives this as an illustration of the ternarius numerus. 
 He says : 
 
 1 Interdictorum trinum genus ; unde repulsus 
 Vi fuero, aut utrobi fuerit, quorumve bonorum.' 4 
 
 As is well known, this classification of posses- 
 sory interdicts has largely governed the doctrine 
 on the subject. Yet it has often been objected 
 to as illogical, and various attempts have been 
 
 1 1. Gaius specifies, in all, three divisions (tria genera) of 
 interdicts (iv. 142, 143, 156), of which, however, the third, 
 simplicia and duplicia, is bipartite. There were in fact 
 numerous other categories (e. g. Gai. iv. 170). 
 
 1 Dig. xliii. 1, fr. 1 1. See also Dig. xliii. 4, 3 2 and 
 Karlowa, Eom. E.-G. iii. p. 1009. 
 
 2 Of. Dig. ii. 14. 5 cited infra, p. 55. See also Cic. Brutus, 
 43 and 54, where Cicero does much the same thing. 
 
 3 Gai. iv. 143 ; Inst. iv. 15 2 ; Paul in Dig. xliii. 1. 2 3. 
 
 4 Edyll. xi, lines 63-4. 
 
54: TKICHOTOMY 
 
 made, from the time of Donellus onwards, to find 
 a ratio for it. The interdicts adipiscendae posses- 
 sionis, such as the quorum bonorum and the 
 Salvianum, being destined for the obtaining of 
 possession of things by one who makes no claim 
 to having had previous possession of them, are, 
 it is thought, distinct in their character from the 
 retinendae and recuperandae possessionis interdicts, 
 which were intended for the protection of pre- 
 viously existing possession. They are treated 
 apart from the latter in the Digest. Savigny, in 
 the earlier editions of his famous treatise, did 
 not hesitate to exclude the adipiscendae possessionis 
 interdicts altogether from the class of possessory, 
 and charged the Koman jurists with inaccuracy in 
 making the classification, but the idea that the 
 co-ordination of the three was due to the influence 
 of the ternarius numerus does not seem to have 
 impressed him. 1 There existed also, as we learn 
 from Paulus in the Digest xliii. 1. 2 3, and the 
 Vienna fragments of Ulpian's Institutes, interdicts 
 'tarn adipiscendae quam reciperandae possessionis' 
 
 1 Sav. Poss. 37. He slightly modifies his view in later 
 editions. To defend the classification on the ground of its 
 having in each case & jus possessionis i.e. the external object 
 of the suit as its basis would, Savigny holds, be quite 
 unscientific. For a contrary view see Cornil, Possession 
 (Paris, 1905), pp. 307 sq. 
 
IN EOMAN LAW 55 
 
 (such as the quern fundum) ; but these, instead of 
 being given by Paul and Ulpian as a fourth class 
 with the three above named, are classified by 
 them merely as in a sense interdicta duplicia. 1 As 
 Paulus says : ' Sunt interdicta, ut diximus, duplicia 
 tarn reciperandae quam apiscendae possessionis.' 2 
 
 (&) The exception of vitious possession in the 
 possessory interdicts is also a tripertita, viz. vi aut 
 clam aut precario.* This appeared in all interdicts 
 retinendae possessionis, and is of Early Eoman origin. 
 Excluding, as it does, dolus, and for "other reasons 
 it can only be explained on my theory. Against 
 the vitious possessor there seem also to have been 
 three independent recuperatory interdicts, viz. de 
 vi, de dandestina possessione, and de precario, but the 
 nature of the second of these is unknown. 4 
 
 (c) In Digest ii. 14. 5 (de Pactis) we get the 
 following extract from Ulpian's Commentary on 
 the Edict: ' Conventionum autem tres sunt species. 
 Aut enim ex publica causa fiunt aut ex privata : 
 privata aut legitima aut juris gentium.' Here there 
 is apparent disregard of logic ; species is improperly 
 
 1 See Girard, Textes, p. 467. Cf. supra, p. 53, n. 1 
 
 2 Dig. xliii. 1. 2 3. 
 
 3 Gai. iv 150-1 ; c-f. Terence, Eunuch, ii. 3, 28. See 
 Cuq, Nouv. Eev. Hist. 1894, pp. 24 sq., who remarks: 
 ' On tenterait vainement de la justifier.' 
 
 4 Dig. x. 3. 7 5. See Puchta, Inst. ii 225. 
 
56 TKICHOTOMY 
 
 co-ordinated with genus. This is similar to what 
 occurs, as we have already seen, in Ulpian's classi- 
 fication of persons into liberi, servi, and liberti^ 
 and any emendation of the passage, as has been 
 suggested, would be improper. 2 What one would 
 have expected here (and it would have been in 
 accordance substantially with the way in which 
 Ulpian goes on to treat the subject) was a state- 
 ment that there are two kinds of conventions, 
 viz. ex publica causa and ex privata causa, and that, 
 of the latter, there are two species, viz. legitimae 
 and juris gentium. But Ulpian cannot refrain 
 from his tripartite division. 3 
 
 (d) The division of heirs into necessarii, sui et 
 necessarii, and extranei 4 is open to objection. Gaius 
 
 1 Supra, p. 39. Of. the classification of interdicts, supra, 
 p. 53. 
 
 2 See Cujacius, Olservat. lib. xv, c. 33 ; Otto, Thesaur. i, 
 p. 570 and iii, pp. 270, 445. For an independent classifica- 
 tion in accordance with Ulpian's presumed meaning, though 
 not an explanation of the trichotomy, see Voigt, Horn. R.-Gr. ii, 
 pp. 852-3. Cf. Gliick, Erldut. d. PandeJc. iv, p. 278, who 
 describes the action of the Digest editors in connecting the 
 legitima conventio of fr. 5 with that of fr. 6 as ' sehr unschick- 
 lich'. 
 
 3 Cujac. I.e., where it is remarked 'Tres [species] sunt 
 sane, si universas ita numeres, hoc est et genera et species 
 simul, alias duae. Sunt enim generatim pacta, % S^/xoo-ia, 
 17 t&a, aut publica, aut privata.' 
 
 4 Gai. ii 152 ; Inst. ii. 19 pr. ; Ulp. Eeg. xxii 24-5. 
 
 
IN ROMAN LAW 57 
 
 says : ' Heredes autem aut necessarii dicuntur aut 
 sui et necessarii aut extranei.' A grandson in 
 potestate, whose father was also in that position, 
 was, if instituted (the father being disinherited), 
 heres necessarius but not suns', and yet the definition 
 of the heredes necessarii does not include him. 1 
 Persons in causa mancipii, again, instituted heirs 
 cum libertate, though classed as necessarii, had, like 
 sui, the abstinendi potestas which slaves had not. 2 
 The true generic division is into necessarii and 
 voluntarii or into domestici and extranei. Schrader 
 remarks (Inst. hoc loco) : t Abit haec trimembris 
 divisio, ut aliae pleraeque, in bimembrem necessa- 
 riorum voluntariorumque.' 
 
 (e) A division of obligations made by Paulus 
 from a processual point of view is into ' ad dandum 
 aliquid vel faciendum vel praestandum'. He says 
 4 obligationum substantia non in eo consistit, ut 
 aliquod corpus nostrum aut servitutem nostram 
 faciat, sed ut alium nobis obstringat ad dandum 
 aliquid vel faciendum vel praestandum.' 3 This 
 obviously has reference to the formulae for en- 
 forcing them. The meaning of praestare is not 
 
 1 Gai. ii 153 ; Inst. ii. 19 1, 2. Only by a fiction can 
 he be regarded as suus et necessarius. 
 
 2 Gai. ii 160. 
 
 3 Paul in Dig. xliv. 7. 3 pr. ; Gai. iv 2. 
 
 1211 
 
58 TEICHOTOMY 
 
 certain, but it is certain that it only appeared in 
 a formula (if at all) in a few special cases and 
 ought not to have been co-ordinated with the 
 other two. From this point of view Paul might 
 almost as well have said l damnum decidendum ' 
 as 'praestandum'. 1 Praestare was, in any case, 
 comprehended by the generic facere. 2 Com- 
 menting on this Pauline division, Savigny says : 
 ' In diesem Zusammenhang ware das faciendum 
 oder das praestandum (neben dem dandum) allein 
 vollig hinreichend gewesen ; die an sich iiber- 
 fliissige Zusammenstellung der drei moglichen 
 Gegenstande enthalt eine augenscheinliche Anspie- 
 lung auf die drei gleichnamigen Arten der Intentio 
 in den Klagformeln. 5 3 
 
 (/) The frequent trichotomy of actio, petitio, 
 persecutio in the texts creates difficulties which 
 modern writers have in vain attempted to resolve. 
 It is in fact meaningless, and Ulpian and other 
 jurists who adopt it only fall into inexactitude in 
 
 1 Praestare was really a general term embracing dare and 
 facere. See e.g. Gaius, iii 155 and iv 61. For different 
 views as to praestare see Savigny, Syst. vol. v. App. xiv, 
 pp. 599 sq. Of. Lenel, Edict Perpet. 2nd ed. p. 287, n. 12 , and 
 p. 318. 
 
 2 Dig. 1..16, fr. 175, fr. 189, fr. 218. 
 
 3 System, 1. c. p. 602. See also Puchta, Inst. ii 165, note aa . 
 for various explanations by different writers. 
 
IN KOMAN LAW 59 
 
 attempting to explain it. What is the distinction 
 between actio and petitio ? Ulpian x and Papinian 2 
 say that specifically actio is used where the action 
 is based on a personal, while petitio is used where 
 based on a real right. Thus Ulpian says : 
 ' " Actionis " verbum et speciale est et generale. 
 Nam omnis actio dicitur, sive in personam, sive 
 in rem sit petitio; sed plerumque "actiones" 
 personales solemus dicere. " Petitionis " autem 
 verbo in rem actiones significari videntur. " Per- 
 secutionis " verbo extraordinarias persecutiones 
 puto contineri, ut puta fideicommissorum et si 
 quae aliae sunt, quae non habent juris prdinarii 
 exsecutionem.' It is indisputable, however, that 
 petitio and petere are constantly used as general 
 expressions to indicate any kind of action, per- 
 sonal (e. g. in the expression i si certum petetur ') 
 as well as real. In Paul's Sentences we have 
 petitio actually put in contrast with vindication 
 Persecution on the other hand, which Ulpian sup- 
 poses to have applied to the actiones extraordinariae, 
 was used quite indefinitely, 4 and Ulpian is here 
 not supported but rather contradicted by Papinian. 5 
 
 1 Dig. 1. 16, fr. 178 2. In a general sense he makes actio 
 and petitio synonymous. 
 
 2 Dig. xliv. 7 fr. 28. 3 iv. 1. 18. 
 4 e. g. Dig. xxxii. 41 9, 11. 
 
 Cf. references in notes 2 and 4 supra. 
 
60 TKICHOTOMY 
 
 Mitteis, who is the most recent writer to dis- 
 cuss this trichotomy, gives all the principal 
 authorities, and I may refer to what he says. 1 
 After showing the impossibility of making any 
 satisfactory distinction between the three terms, 
 and the consequent inaccuracy of Ulpian's view, 
 he concludes thus : ' Bestimmte Bedeutungen 
 lassen sich hier iiberall nicht erkennen,und so kann 
 man es denn schwer vermeiden, mit Bruns in 
 " actio petitio persecutio " eine sinnlose Haufung 
 der Ausdrucke zu finden' (Bruns, Ztsch. f. E. 
 Gesch. xii. 118). What Bruns and Mitteis have 
 called a ' sinnlose Haufung der Ausdrucke ' it is 
 preferable to call an example of persistent ten- 
 dency to trichotomy. 
 
 (g) Heirs by the jus civile are classified by Ul- 
 pian as sui, consanguine^ and agnati : 2 t Intesta- 
 torum ingenuorum hereditates pertinent primum 
 ad suos heredes . . . ; si sui heredes non sunt, ad 
 consanguineos, id est fratres et sorores ex eodem 
 patre ; si nee hi sunt, ad reliquos agnatos proxi- 
 mos.' Here the introduction of consanguinei is 
 unnecessary, 3 and is not supported apparently 
 
 1 Privatrecht, 1908, p. 39, n. 1 and pp. 89-91. The tricho- 
 tomy, as he points out, was traditional, for it appears in older 
 laws as the lex Salpen. 26, lex Malacit. 58, 62, lex Urson, 129. 
 
 2 Ulp. Eeg. xxvi. 1 ; cf. Collat. xvi. 4 1, 2. 
 
 8 He makes a similar distinction as regards tutors-at-law 
 
IN EOMAN LAW 61 
 
 by any other jurist. The consanguinei are em- 
 braced by the term agnati in Gaius * and all other 
 writers ; in Ulpian's time there were only two 
 classes of heredeSj the sui and agnati. Neither is 
 Ulpian's division in accordance with that of the 
 XII Tables, which Gaius correctly gives as sui, 
 agnati, and gentiles. 2 Is it not in order to avoid 
 a fourfold division that Ulpian refers to the suc- 
 cession of the gentiles as merely matter of history ? 3 
 (h) Ulpian makes a tripartite division of the 
 law into civil, praetorian, and imperial. 4 He is 
 the only jurist who does so, but he is followed by 
 Justinian in the Institutes. 5 This division is in- 
 consistent with the general theory of the classical 
 jurists that the emperors' Constitutions were but 
 
 in Dig. xxvi. 4, 1 pr. (where he also apparently intends a 
 threefold division of tutores legitimi). See Gltick, Pandekt. 
 xxix, p. 325 ; supra, p. 47, note 3 . 
 
 1 Gai. iii 1, 9, 17. 
 
 2 Says Paul (Collat. iii. 3) ' Sane consanguinei, quos lex non 
 adprehenderat, interpretatione prudentium primum inter 
 agnatos locum acceperunt.' 
 
 8 Collat. xvi. 4 2. This reference to the gentiles was also 
 doubtless in the Liber Regularum, but was omitted by the 
 copyist of the MS. 
 
 4 Dig. 1. 16 10 ' Creditores accipiendos esse constat eos 
 quibus debetur ex quacunque actione vel persecutione, vel 
 jure civili ... vel honorario, vel extraordinario/ 
 
 5 Inst. ii. 10 3, where it is said of the imperial testament 
 1 hoc jus tripertitum esse', meaning its sources were threefold. 
 
62 TRICHOTOMY 
 
 the continuation of the jus praetorium in the wider 
 sense. The emperors, prior to Diocletian, were 
 theoretically regarded as but higher magistrates, 
 acting like the consuls and praetors of the Re- 
 public. It has been suggested that the passage of 
 Ulpian may possibly be a Tribonianism, 1 but 
 there is, I submit, no ground for this. On the 
 contrary, the very fact that we have a tripartite 
 classification makes it in my judgement all the 
 more probable that there is no interpolation. 
 
 (i) In Ulpian's Book of Rules it is said, * Actio- 
 num autem quaedam ex contractu, quaedam ex 
 facto, quaedam in factum sunt.' This most illo- 
 gical division has much exercised the commenta- 
 tors, many of whom are disposed to trace in it 
 the interpolating hand of Tribonian. But it is 
 not worse than several of the Ulpianic trichoto- 
 mies above considered. As explained in the 
 Digest text, it places actions in factum conceptae 
 as a species along with actions based on contract 
 and delict. 2 But actions in jus are the proper 
 
 1 Mitteis, Privatrecht, p. 39 n \ Mitteis also observes, ' Es 
 1st moglich class diese nachklassische Eintheilung an die 
 Trichotomie actio, petitio und persecutio in unklarer Weise 
 angeknupft hat.' 
 
 2 1%. xliv. 7, 25 1. In fr. 42 1 h. t. Ulpian makes a 
 similar tripartite division which is almost equally open to 
 objection ' qui aliquam actionem vel civilem habent . . . vel 
 honorariam actionem, vel in factum/ 
 
IN BOMAN LAW 63 
 
 antithesis, as Gaius makes clear, of actions in 
 factum conceptae. The text has been discussed at 
 length by Erman, 1 with characteristic ability, and 
 he arrives at the conclusion that the classification 
 at least is due to Ulpian. He says, with reference 
 to the arguments for interpolation by the Digest 
 compilers, l Wir konnen also die, iibrigens durch 
 nichts beglaubigte, Moglichkeit einer so tief- 
 greifenden Umarbeitung von Ulpian's Eegulae 
 ftir unsere Stelle als durchaus unwahrscheinlich 
 bezeichnen und Ulpian's Autorschaft als sicher.' 2 
 He subsequently to some extent departs from 
 this as regards the text as a whole, but, if 
 I understand him aright, he adheres to his view 
 that the trichotomy is Ulpian's. 3 In any case 
 I hold, in accordance with my thesis, that it is 
 undoubtedly his. 
 
 There are numerous other tripartite divisions 
 in the texts, to which one might refer as resting 
 on no rational basis and creative of embarrass- 
 ment by their inappropriateness or illogicality. 
 
 1 Ztsch. d. Sav.-Stift. xix, pp. 299; but cf. xxiii, pp. 447-9. 
 Erman describes Ulpian's use of the expression ex facto for 
 the usual ex delicto or ex maleficio as a ' Wortspiel '. 
 
 2 Ztsch. d. Sav.-Stift. xix, p. 301. 
 
 3 Ibid, xxiii, pp. 447-9. There is the greatest probability 
 that the last sentence of fr. 25 1 has been altered by the 
 compilers. 
 
64: TKICHOTOMY 
 
 But enough illustrations have probably been 
 given. 1 
 
 Are then these constantly recurring categories 
 of three, which as we have seen are but rarely 
 logical or natural, explicable on no general hypo- 
 
 1 The following may be considered : Tripartite division 
 of (1) praetorian stipulations by Ulp. in ig. xlvi. 5. 1 pr. 
 See explanation by Puchta (Inst. ii. 268 i.) with note by 
 Kudorff. In Van Leeuwen's edition of the Corpus Juris, h. I., 
 it is remarked 'Species recte dixit. Puritas enim, dies, con- 
 ditio, formae sunt stipulationum, non species/ (2) * tribus 
 modis insula in flumine fit ' in Dig. xli. 1. 30 2 ; (3) atrox 
 injuria is ex facto, ex loco, ex persona in Gai. iii 225, cf. Paul, 
 Sent. v. 4 10 and Dig. xlvii. 10. 7, 8 : locus vulneris is added 
 to them in Dig. fr. 8 h. t., Inst. iv. 4 9. (4) 'Edi autem 
 est vel dictare, vel tradere libellum vel codicem proferre' 
 Ulp. in Dig. ii. 13. 6 7. (5) Urere, frangere, and rumpere 
 in the Lex Aquilia which necessitated an extension to 
 corrumpere ; (6) i Inpensarum species sunt tres ' in Ulp. 
 Eeg. vi 14 ; cf. Paul in Dig. 1. 16. 79 ; (7) of obligations it 
 is said (Dig. 1. 16, fr. 19) that quaedam agantur, quaedam 
 gerantur, quaedam contrahantur. This last is attributed by 
 Ulpian to Labeo's Commentary on the Edict and as explained 
 in the Digest is quite illogical and indefinite. See Otto, 
 Thesaurus, i. 672, ii. 403 ; Voigt, Rom. R.-G. ii. 853. (8) 
 Pulsare, verberare, vi domum introire in the lex Cornelia de 
 injuriis. Dig. xxxxvii. 10, 5 pr. See Ehetor. ad Herenn. iv. 
 25, 35, and Girard in Melanges Gerardin, p. 258. (9) Three 
 kinds of receptum, viz. arbitri, nautarum cauponum stalula- 
 riorum, and argentariorum. See Lenel, Ed. Perpet, 48-50. 
 (10) 'Munus tribus modis dieitur' Dig. 1. 16, 18; 'Tres 
 fere causae sunt ex quibus in possessionem mitti solent * 
 Dig. xlii. 4, 1. There are also aliae divisiones trifariae 
 innumerabiles. 
 
IN ROMAN LAW 65 
 
 thesis ? Commentators on the Roman law have 
 for the most part contented themselves with 
 remarking on this or that classification that it is 
 strange or inexplicable or a playing with words, 
 or heaping-up of terms, 1 or else sometimes they 
 have given the most far-fetched and unsatisfactory 
 explanations, as we have seen in relation to the 
 trichotomies of jus naturale, gentium, and civile, and 
 jus personarum, rerum, and actionum. But are they 
 not all attributable to a tendency of the jurists 
 (especially of Ulpian), more or less conscious, to 
 adopt traditional symbolic tripartite divisions 
 (whether handed down from remote times or 
 invented in the later philosophic schools), or, 
 influenced by tradition, to invent them ? I have 
 a conviction that they are due to each of these 
 causes. There is nothing improbable in this. 
 I have shown that the tendency to choose triper- 
 tita was inherent in Roman thinking about law 
 from very early times. I have also indicated here 
 and there that this tendency was increased by the 
 influence of the philosophic schools. The tendency 
 is, I think, to be largely found among non-juristic 
 writings of the late Republic and early Empire. 
 In the writings of the philosophers, both prior to 
 and during the period of classical jurisprudence, 
 1 See supra, p. 60. 
 
 1211 
 
 I 
 
66 TRICHOTOMY 
 
 especially of the Stoics and the Neo-Platonists, 
 but even to some extent of the Peripatetic School, 
 abundant evidence of proneness to trifariae dim- 
 siones exists. It undoubtedly appears in the wri- 
 tings of Cicero, who, though he cannot strictly be 
 described as an adherent of any of these Schools, 
 was, in his later years at least, much influenced 
 by Stoicism. 1 Thus in the Topics he considers 
 the division of law in general (jus) into ' legem, 
 morem, aequitatem *, 2 and he subsequently, with 
 apparent inconsistency, says of aequitas, t institutio 
 aequitatis tripertita est, una pars legitima est, 
 altera con'veniens, tertia moris vetustate firmata. 
 Atque etiam rursus aequitas tripertita dicitur esse : 
 una ad superos Deos, altera ad manes, tertia ad 
 homines pertinere. Prima pietas, secunda san- 
 ctitas, tertia justitia aut aequitas nomiriatur.' 3 
 
 Permce, in commenting on the first sentence of 
 this passage, of which the meaning is, to my 
 mind (as it probably was to that of Cicero), 
 
 1 Cicero is usually described as an Academic. 
 
 2 Top. 81. Of. de Invent, ii. 22 65 sq. and 53 161 sq., 
 where the sources of law are natura, consuetude, lex. 
 
 3 Top. 90. Cf. ibid. 28, where aequitas has a different 
 meaning. The Topica is a work based on Aristotelianism, 
 and therefore the influence of symbolism on it, and particu- 
 larly of the number 3, is not on the whole very great. See 
 supra, p. 7, n. 2 . 
 
IN EOMAN LAW 67 
 
 wholly vague, suspects it to be due to Greek 
 influence and refers to Aristotle (Ehet. i. 13). 1 
 Other illustrations of this kind of trichotomy, 
 most probably borrowed from the Greek philo- 
 sophers, will be found in most of Cicero's writings, 
 other than his letters and orations, in immense 
 abundance. 2 It would, of course, be too much to 
 charge Cicero, who, as I said, was not an adherent 
 of any of the Schools, with being consciously domi- 
 nated by the influence of trichotomies. All that 
 can be said is that he shows a strange partiality 
 for them. Take, for instance, the Ad Brutum 
 Orator. A great part of the scheme of the book 
 rests on trifariae divisiones. Thus the perfect 
 orator is he who has these * tria dicendi genera ' 
 in his power, viz. 'humilia subtiliter, modica 
 temperate, magna graviter dicere.' 3 Then he 
 
 1 Ztschr. d. Sav.-Stift. xxii, p. 62, note \ It is to be re- 
 gretted that no attempt has hitherto been made to examine 
 the trifariae divisions of Cicero, to trace their source and to 
 estimate their value. I commend the subject to our younger 
 scholars. 
 
 8 See e. g. Cic. de Nat. Deor. ii. 30. 75 ; de Off. i. 3. 10 
 and ii. 9. 31 ; de Orat. i 48. Even the orations are not 
 wholly free from the charge. See e. g. pro Murena 9, 
 with which cf. de Orat. 1. c. The comparison of these two 
 last-cited passages, in so far as they characterize the functions 
 of a jurist, is instructive. 
 
 3 Cic. Orat. ad Brut. 20, 100 ; cf. Rhetor, ad Herenn. 
 iv. 8-11 'Sunt igitur tria genera, quae genera nos figuras 
 
68 TEICHOTOMY 
 
 points out that there are three things that must 
 be looked to by the orator, viz. ' quid, quo loco, 
 et quomodo dicat,' of which the first depends on 
 inventiOj the second on dispositio, and the third on 
 elocutio. 1 Then he makes a tripartite division 
 (tripartUa varietas) of the objects of perfect elo- 
 quence, viz. ' ad probandum, delectandum, flecten- 
 dum,' 2 and then further treating of the ' forma 
 et character orationis ', he makes the ' collocatio 
 verborum in eloquentia' of three different kinds. 3 
 Leaving Cicero I will take Seneca. Seneca 
 was an out-and-out Stoic, 4 and supplies us with 
 good illustrations of the habit I am discussing. 
 Thus, in the Second Book of his Naturales 
 
 appellamus, in quibus omiiis oratio non vitiosa consumitur ; 
 unam gravem, alteram mediocrem, tertiam extenuatam 
 vocamus.' 
 
 1 Orat. ad Brut. 43-57. Despite his tripartite division 
 ( 43) Cicero, ignoring it, actually makes four parts, 54, 55. 
 Cf. de Off. i. 3. 10. 
 
 2 Ibid. 69. 
 
 3 Ibid. 149, 175 ; cf. Rhetor, ad Herenn. iv. 12 sq., 27. 
 Other trichotomies that I have noticed in the Rhetor, are 
 i. 8. 12 (' narrationum tria sunt genera ', &c.) ; i. 9. 14 (' tres 
 res convenit habere narrationem ', &c.) ; i. 11. 18 ('constitu- 
 tiones itaque . . . tres sunt conjecturalis, legitima, juridi- 
 cialis ') ; i. 2. 2 ; iv. 25. 35 (cf. ii. 26, 41). 
 
 * On Seneca's relation to Stoicism see Dill, Roman Society 
 from Nero to M. Aureliits, pp. 306-9, and authorities there 
 cited. 
 
IN KOMAN LAW 69 
 
 Quaestiones, he deals with a topic which readily 
 lends itself to divisions and subdivisions. He 
 professes to treat of lightning and thunder, 
 and he begins ( 1) by saying of the universe, 
 'Omnis de universo quaestio in coelestia, sub- 
 limia, terrena dividitur' (meaning astronomy, 
 meteorology, and geography), and he then proceeds 
 to make, in his subsequent chapters, almost every 
 classification of his subject-matter tripartite. This 
 may be seen, for instance, in sections 10, 12, 33, 
 39, 40, 47, 50. Probably all of them are tradi- 
 tional. Once or twice he expresses himself as 
 dubious about the propriety of a particular divi- 
 sion, but not with much confidence. 1 Most, if 
 not all, of the trichotomies are in any view open 
 to criticism as strained and unscientific. 2 
 
 Another writer whose writings seem to show 
 the influence of the ternarius numerus, both as re- 
 gards words and ideas, is Apuleius. As is well 
 known, he was a Phoenician who flourished 
 during the reigns of Antonius Pius and Marcus 
 Aurelius, and was thoroughly imbued with the 
 current ideas of philosophy both of Greece (where 
 
 1 See especially Nat. Quaest. ii. 39, 47, 48. 
 
 2 See also Seneca, Epist. 75, 9-15, where he says of 'pro- 
 ficientes in philosophia et studio sapientiae ' that * in tres 
 partes, ut quibusdam placet, dividuntur ', &c. 
 
70 TEICHOTOMY 
 
 he studied as a young man) and of Kome, and 
 borrowed from his predecessors freely. 1 
 
 It is needless to insist here upon the influence 
 exercised upon the jurists by the Stoic School (not 
 to speak of other schools) of philosophy; it has 
 been often and abundantly discussed. 2 Such philo- 
 sophers as Panaetius and Posidonius were espe- 
 cially influential. At the present day lawyers are 
 not much influenced, either in theory or practice, 
 by the doctrines of psychology or metaphysics. 
 On the contrary, they as a rule regard abstract 
 speculation in relation to law with suspicion if 
 not dislike. We are living in an age of intense 
 commercial competition, both individual and in- 
 
 1 The following passages may be referred to: Apul. de 
 dogm. Plat. i. 584 (' deorum trinas species '), i. 591, ii. 597 
 (' tria genera ingeniorum '), 606 (' justitiam vero, quod trinis 
 animae regionibus sparsa sit '), iii. 638 ; de Deo Socrat. 664, 
 684 ; de Mundo, 755 ('tria fata sunt ') ; Apolog. 461, 484; 
 Florid, ad init. ; Metam. iii. 55, iv. 82. 
 
 2 Several illustrations of this have been given in the pre- 
 ceding pages. See also Dig. 1. 16. 124, attributed by Cujas 
 (viii, p. 622) to the Stoics, and Dig. xli. 3. 30, where the 
 influence of Stoic principles of realism on the division of 
 things and the law of usucapio is well shown in the tripartite 
 division given by Pomponius. Cf. Seneca, Epist. 102. 6 ; 
 Sokolowski, Philosophie im Privatrecht, i, pp. 48 sq. and 119. 
 Pomponius's three illustrations df corpus ex distantibus, 
 viz. grex, populus, legio, may be compared with Seneca's 
 exercitus, populus, senatus, which Sokolowski apparently 
 overlooks (p. 119, I.e.). 
 
 
IN EOMAN LAW 71 
 
 ternational, and empirical rules of practical con- 
 venience govern the development of the law. 
 But far otherwise was it at the close of the Eoman 
 Eepublic and the early period of the Empire. 
 Graecia capta, having no longer any national life 
 and few commercial interests, exerted her still 
 keen intellectual forces in the fields of philosophy 
 and speculation. And with these she dominated 
 the Koman intellect not only in art and philosophy 
 but to a large extent in jurisprudence, which, 
 after the establishment of the pax Romano, under 
 Augustus, became the favourite study of the 
 Eomans and the securest avenue to fame. That 
 Ulpian and Paul, among others, did not escape 
 this influence is not open to doubt. 1 Savigny, in 
 dealing with the tripartite division of pupils into 
 infantes, wfantiae proximi and pubertati proximi 
 (see Inst. iii. 19 $ 9, 10), the infantes being those 
 who have not completed seven years, makes the 
 
 1 On the influence of the schools of philosophy generally 
 on the jurists see Pernice, Labeo, i, pp. 30 sq. ; Voigt, Jus 
 Naturak, &c. i, pp. 250 sq. On their influence on the 
 doctrines of property and possession in particular, see Soko- 
 lowski, Philosophic im Privatrecht, vols. i and ii. Puchta 
 (Inst. 102) seems to quite underestimate the Stoic in- 
 fluence, though before his time there was perhaps a tendency 
 to exaggerate it. Marcus Aurelius subsidised chairs at Athens 
 for the four principal schools Platonists, Peripatetics, 
 Stoics, and Epicureans. Dio Cass. Ixxxi. 31. 
 
72 TRICHOTOMY 
 
 following observations : ; Nun wurde den Bomern 
 eine uralte Lehre griechischer Philosophie be- 
 kannt, welche der Zahl Sieben geheimnissvolle 
 Krafte, und den siebenjahrigen Lebensperioden 
 eine besondere Wichtigkeit beilegte. Diese Lehre 
 kam jenem praktischen Bediirfniss auf die will- 
 kommenste Weise entgegen, und so geschah es, 
 dass die Granze der Kindheit gerade auf das 
 Ende des siebenten Jahres allgemein angesetzt 
 wurde, anstatt dass wohl auch Sechs oder Acht 
 Jahre dafiir angenommen werden konnten.' * 
 
 On the influence of Greek speculations generally 
 on the jurists of the early Empire, Sokolowski, 
 with much truth and cogency in my opinion, 
 remarks : ' Sicher ist der elementare, produktive 
 Eechtsinstinkt in der Kaiserzeit gesunken, zu- 
 gleich aber uiiter dem Einfluss der griechischen 
 Geistesbildung die Neigung fiir allgemeine Speku- 
 lationen ganz bedeutend gestiegen. Das Kecht 
 tritt aus dem Gebiet der praktischen Bedurfniss- 
 befriedigung eines politisch eminent geschulten 
 Volks in die Welt wissenschaftlicher Gedanken. 
 . . . Der griechische, spekulative Genius unter 
 warf sich im zweiten und dritten Jahrhundert 
 nach Christo das gesamte Geistes- und Sitten- 
 
 1 System, iii, p. 32. See also Quinctil, i. 1, 18, and Dig. 
 xxvi. 7, 1 2, commented on by Savigny (1. c.). 
 
IN KOMAN-LAW 73 
 
 leben, er umspann die einfachen und elemen- 
 taren Lehren Christi mit einer Ftille schon friiher 
 ausgereifter, philosophischer Gedanken, er dureh- 
 drang vielfach auch die schlichten, praktischen 
 Kechtsbegriffe des alien Komertums.' l 
 
 The Greek Schools of Philosophy also exercised 
 an undoubted influence upon the works of the 
 early Christian writers, similar to what they 
 exercised upon those of the jurists. From the 
 influence of numbers, and of the number 8 in 
 particular, they did not emancipate themselves. 
 It will be found in the works of Augustine, Ter- 
 tullian, and others. 2 Nor can it well be doubted 
 that the tripartite divisions both of the Deity into 
 Father, Son, and Holy Ghost and of man into 
 body, soul, and spirit the soul or psyche being 
 further subdivided by Philo, following Aristotle, 
 into three principles (nutritive, emotional, and 
 
 1 Philosophic im Privatrecht, ii, p. 126. 
 
 2 See Augustin. Civ. xi, c. 25 (' de tripertita totius philoso- 
 phiae disciplina '), where he says : ' Tria sunt quae in uno 
 quoque homine artifice spectantur . . . natura, doctrina. 
 usus ' ; Augustin. Epist. ad Januarium, ii. 18. 33, where 
 he says : ' Tenarius numerus in multis sacramentis maxime 
 excellit.' Cf. Cassiodor. as cited by Teuffel, Gesch. d. rom. 
 Litt. 483, notes 6, 7, 9, 10, 11 ; Tertull. de Baptismo, VI 
 in fine ; Usener, Ehein. Mus. f. Phil. Iviii, pp. 41 sq. 
 Curious illustrations of the mystic value of the number 3 
 in theological writings are given by Hugo de Palma, Theo- 
 logia Mystica, ed. de Montes, Amsterdam, 1647. 
 
 1211 K 
 
74 TRICHOTOMY 
 
 rational) 1 were due to that influence. 2 Many 
 other illustrations might easily be given. 3 
 
 It may be objected to the foregoing thesis that 
 tripartite classifications of subjects are in them- 
 selves natural and convenient and very common 
 (the latter part of the proposition no one who has 
 been accustomed to listen to sermons will dis- 
 pute), and that there are in the Roman texts 
 many duads and quadrads and other classifi- 
 cations, of which fact also there is abundant 
 evidence. In every system of law, moreover, 
 including our own, there are plenty of tripartite 
 divisions. 4 They assist the memory and are thus 
 an aid to composition. Such objectors must then, 
 
 1 Eth. Nic. i. 13. 
 
 2 Cf. Plato's division of man's nature into intellectual, 
 appetitive, and passionate (TO flv/xoaSc's), which however is 
 perhaps a natural consequence of his view of his subject- 
 matter. To what confusion of thought and extent of con- 
 troversy the trichotomy of body, soul, and spirit may lead, 
 see J. B. Heard, Tripartite Nature of Man, Edinburgh, 1870. 
 
 3 On the Pauline triad of virtues Faith, Hope and 
 Charity see Lobeck, Aglaophanus, p. 388 note ; Epist. 
 St. John, i. 7, 8. 
 
 4 Some of them traditional and illogical, e. g. in English 
 law the old divisions of Crimes into (1) treasons, (2) felonies, 
 and (3) misdemeanours (treason being just a felony, though 
 with special rules of procedure), and of Contracts into (1) 
 contracts of record, (2) contracts by specialty, and (3) simple 
 contracts. See Broom, Com. Law, 9th ed., pp. 255-6 ; 
 Blackstone, Commentaries, ii, p. 465 ; Kenny, Outlines of 
 
IN KOMAN LAW 75 
 
 however, hold that the Koman trichotomies above 
 criticized are either defensible in themselves on 
 grounds of simplicity, convenience, logic, or the 
 like, or that they were made casually, not to say 
 ' perperam et sine ratione '. But this I have shown 
 cannot be maintained. Nor can it be disputed 
 that the traditional dominance of the number 3 
 was potent at the time of the classical jurists. It 
 is not necessary for me to hold, nor, as I have 
 said, do I hold, that all the various trichotomies of 
 Paul, Ulpian, and others that I have given were 
 deliberately invented by them under the influence 
 of symbolism or the traditional force of numbers. 
 On the contrary it is almost certain that many of 
 them were tralatitious, and it is probable that when 
 the jurists ventured to make divisions of their own 
 they were often less illogical. 1 Gaius, at least, who 
 was not apparently much influenced by Stoic or 
 other philosophy, seems entitled to this credit. 
 And the same credit may doubtless be given to 
 some other jurists. But Ulpian and Paul cannot be 
 acquitted in my opinion of the charge of adopting, 
 and to some extent inventing, more or less con- 
 sciously, symbolic tripartite categories, with the 
 
 Criminal Law, p. 83 ; Stephen, Hist, of Crim. Law, ii, p. 187. 
 Of. Kussell on Crimes, 6th ed., p. 192. 
 1 See, e. g., Gai. iv 1. 
 
76 TKICHOTOMY 
 
 result of obscuring the subject and producing fre- 
 quently fanciful and misleading doctrines. 
 
 Trichotomy, unlike dichotomy (e.g. A and not 
 A), has no apparent universality. It depends 
 on the subject-matter which is being discussed 
 whether it can in any given case be justified. 
 There is no principle in logic which should, for 
 instance, lead to a given genus being divided into 
 three species rather than any other number ; the 
 special subject-matter must in each case decide 
 what the number of co-ordinate species should be. 
 
 If my theory be accepted, it will perhaps put 
 an end to the vain attempts that have been made, 
 and are still being made, by commentators on 
 the Eoman law to find reasons to justify classifi- 
 cations which their authors themselves could not 
 have attempted to justify as logical or natural, 
 and to which they, in fact, attached no practical 
 importance. How much labour would have been 
 saved, and may yet be saved, to jurists, philo- 
 sophers, and theologians by a little attention to 
 anthropological research ! 
 
 Of course, nothing that I have said can materially 
 detract from the merits of the Eoman jurists 
 from their lucidity and logical acumen and power 
 of philosophic analysis to which all who have 
 studied them must ever pay profound reverence. 
 
IN EOMAN LAW 77 
 
 It merely shows that classification and definition 
 of legal doctrines and rules were not their strong 
 points. 1 They were in fact in these matters the 
 children of their age, and were unable to emanci- 
 pate themselves from the traditional influences of 
 their predecessors and of the schools of philosophy. 
 
 1 Savigny, System, ii, Beilage vi, p. 508, observes : l . . . die 
 allermeisten Definitionen der alten Juristen tiberaus mangel- 
 haft sind.' 
 
 Oxford : Printed at the Clarendon Press by Horace Hart, M.A. 
 
CLARENDON PRESS BOOKS 
 
 LAW 
 
 Jurisprudence 
 
 Bentham's Fragment on Government. Edited by F. C. MONTAGUE. 
 8vo. 7s. 6d. 
 
 Bentham's Introduction to the Principles of Morals and Legisla- 
 tion. Second edition. Crown 8vo. 6s. 6d. 
 
 Studies in History and Jurisprudence. By the Right Hon. JAMES 
 BRYCE. 1901. Two volumes. 8vo. 1 5s. net. 
 
 The Elements of Jurisprudence. By T. E. HOLLAND. Eleventh 
 
 edition. 1910. 8vo. 10s. 6d. net. 
 Elements of Law, considered with reference to Principles of General 
 
 Jurisprudence. By Sir W. MARKBY, K.C.I.E. Sixth edition revised, 
 
 1905. 8vo. 12s. 6d. 
 
 Roman Law 
 
 Imperatoris lustiniani Institutionum Libri Quattuor ; with 
 introductions, commentary, and translation, by J. B. MOYLE. Two 
 volumes. 8vo. Vol. I (fourth edition, 1903), 16s. ; Vol. II, Translation 
 (fourth edition, 1906), 6s. 
 
 The Institutes of Justinian, edited as a recension of the Institutes of 
 Gaius. By T. E. HOLLAND. Second edition. Extra fcap. 8vo. 5s. 
 
 Select Titles from the Digest of Justinian. By T. E. HOLLAND and 
 C. L. SHADWELL. 8vo. 14s. 
 
 Also, sold in parts, in paper covers : Part I. Introductory Titles. 2s. 6d. 
 Part II. Family Law. Is. Part III. Property Law. 2s. 6d. Part IV. 
 Law of Obligations. No. 1. 3s. 6d. No. 2. 4s. 6d. 
 
 Gai Institutionum luris Civilis Commentarii Quattuor: with 
 a translation and commentary by the late E. POSTE. Fourth edition. 
 Revised and enlarged by E. A. WHITTUCK, with an historical introduction 
 by A. H. J. GREENIDGE. 8vo. 16s. net. 
 
 Institutes of Roman Law, by R. SOHM. Translated by J. C. LEDLIE : 
 with an introductory essay by E. GRUEBER. Third edition. 1907. 8vo. 
 16s. net. 
 
 Infamia : its place in Roman Public and Private Law. By A. H. J. 
 GREENIDGE. 8vo. 10s. 6d. 
 
 Legal Procedure in Cicero's Time. By A. H. J. GREENIDGE. 8vo. 
 25s. net. 
 
 The Roman Law of Damage to Property : being a commentatory on 
 the title of the Digest ' Ad Legem Aquiliam ' (ix. 2), with an introduction 
 to the study of the Corpus luris Civilis. By E. GRUEBER. 8vo. 10s. 6d. 
 
 Contract of Sale in the Civil Law. By J. B. MOYL.E. 8vo. 10s. 6d. 
 Trichotomy in Roman Law. By H. GOUDY. 8vo. 4s. net, 
 
 The Principles of German Civil Law. By ERNEST J. SCHUSTER. 1907. 
 8vo. 12s. 6d. net. 
 
CLARENDON PRESS BOOKS 
 
 English Law 
 
 Principles of the English Law of Contract, and of Agency in its 
 relation to Contract. By Sir W. R. ANSON. Twelfth edition, 1910, revised 
 by M. L. GWYER. 8vo. 10s. net. 
 
 Law and Custom of the Constitution. By Sir W. R. ANSON. In two 
 volumes. 8vo. 
 
 Vol. I. Parliament. Fourth edition. 1909. 12s. 6d. net. 
 Vol. II. The Crown. Third edition. Part 1, 1907. 10s.6d.net. Part II, 
 1908. 8s. 6d. net. 
 
 Introduction to the History of the Law of Real Property. By 
 Sir K. E. DIGBY. Fifth edition. 8vo. 12s. 6d. 
 
 Legislative Methods and Forms. By Sir C. P. ILBERT, K.C.S.I. 1901. 
 8vo. 16s. 
 
 Modern Land Law. By E. JENKS. 8vo. 15s. 
 
 Essay on Possession in the Common Law. By Sir F. POLLOCK and 
 Sir R. S. WRIGHT. 8vo. 8s. 6d. 
 
 Outline of the Law of Property. By T. RALEIGH. 8vo. 7s. 6d. 
 
 Law in Daily Life. By RUD. VON JHERING. Translated with Notes and 
 Additions by H. GOUDY. Crown 8vo. 3s. 6d. net. 
 
 Cases illustrating the Principles of the Law of Torts, with table 
 of all Cases cited. By F. R. Y. RADCLIFFE and J. C. MILES. 8vo. 1904. 
 12s. 6d. net. 
 
 The Management of Private Affairs. By JOSEPH KING, F. T. R. 
 BIGHAM, M. L. GWYER, EDWIN CANNAN, J. S. C. BRIDGE, A. M. LATTER. 
 Crown 8vo. 2s. 6d. net. 
 
 Calendar of Charters and Rolls, containing those preserved in the 
 Bodleian Library. 8vo. 1 11s. 6d. net. 
 
 Handbook to the Land- Charters, and other Saxonic Documents. By 
 J. EARLE. Crown 8vo. 16s. 
 
 Fortescue's Difference between an Absolute and a Limited 
 Monarchy. Text revised and edited, with introduction, etc., by 
 C. PLUMMER. 8vo, leather back. 12s. 6d. net. 
 
 Villainage in England. By P. VINOGRADOFF. 8vo. 16s. net. 
 
 Welsh Mediaeval Law : the Laws of Howel the Good. Text, trans- 
 lation, etc., by A. W. WADE EVANS. Crown 8vo. 8s. 6d. net. 
 
 Constitutional Documents 
 
 Select Charters and other Illustrations of English Constitutional History, 
 from the earliest times to Edward I. Arranged and edited by W. STUBBS. 
 Eighth edition. 1900. Crown 8vo. 8s. 6d. 
 
 Select Statutes and other Constitutional Documents, illustrative of 
 the reigns of Elizabeth and James I. Edited by G. W. PROTHERO. Third 
 edition. Crown 8vo. 10s. 6d. 
 
 Constitutional Documents of the Puritan Revolution, selected and edited 
 by S. R. GARDINER. Third edition. Crown 8vo. 10s. 6d. 
 
CLARENDON PRESS BOOKS 
 
 International Law 
 
 International Law. By W. E. HALL. Sixth edition by J. B. ATLAY. 
 1909. 8vo. 1 Is. net. 
 
 Treatise on the Foreign Powers and Jurisdiction of the British 
 Crown. By W. E. HALL. 8vo. 10s. 6d. 
 
 The European Concert in the Eastern Question, a collection of 
 treaties and other public acts. Edited, with introductions and notes, by 
 T. E. HOLLAND. 1885. 8vo. 12s. 6d. 
 
 Studies in International Law. ByT. E. HOLLAND. 1898. 8vo. 10s. 6d. 
 The Laws of War on Land. By T. E. HOLLAND. 1908. 8vo. 6s. net. 
 
 Gentilis Alberici de lure Belli Libri Tres edidit T. E. HOLLAND. 
 1877. Small quarto, half -morocco. 1 Is. 
 
 The Law of Nations. By Sir T. Twiss. Part I. In time of peace. New 
 edition, revised and enlarged. 8vo. 15s. 
 
 Pacific Blockade. By A. E. HOGAN. 1908. 8vo. 6s. net. 
 
 Colonial and Indian Law 
 
 The Government of India, being a Digest of the Statute Law relating 
 thereto, with historical introduction and illustrative documents. By Sir 
 C. P. ILBERT, K. C.S.I. Second edition. 1907. 8vo, cloth. 10s. 6d. net. 
 
 British Rule and Jurisdiction beyond the Sea. By the late Sir H. 
 JENKYNS, K.C.B., with a preface by Sir C. P. ILBERT, and a portrait of 
 the author. 1902. 8vo, leather back. 15s. net. 
 
 Cornewall-Lewis's Essay on the Government of Dependencies. 
 Edited by Sir C. P. LUCAS, K.C.M.G. 8vo, leather back. 14s. 
 
 An Introduction to Hindu and Mahommedan Law for the use of 
 students. 1906. By Sir W. MARKBY, K.C.I.E. 6s. net. 
 
 Land-Revenue and Tenure in British India. By B. H. BADEN- 
 POWELL, C.I.E. With map. Second edition, revised by T. W. HOLDER- 
 NESS, C.S.I. (1907). Crown 8vo. 5s. net. 
 
 Land-Systems of British India, being a manual of the Land-Tenures, 
 and of the systems of Land-Revenue administration. By the same. 
 Three volumes. 8vo, with map. 3 3s. 
 
 Anglo-Indian Codes, by WHITLEY STOKES. 8vo. 
 
 Vol. I. Substantive Law. 1 10s. Vol. II. Adjective Law. 1 15s. 
 1st supplement, 2s. 6d. 2nd supplement, to 1891, 4s. 6d. In one vol., 6s. 6d. 
 
 The Indian Evidence Act, with notes by Sir W. MAHKBY, K.C.I.E. 
 8vo. 3s. 6d. net (published by Mr. Frowde). 
 
 Corps de Droit Ottoman : un Recueil des Codes, Lois, Reglements, 
 Ordonnances et Actes les plus importants du Droit Inte"rieur, et deludes 
 sur le Droit Coutumier de 1'Empire Ottoman. Par GEORGE YOUNG. 1905. 
 Seven vols. 8vo. Cloth, 4 14s. 6d. net; paper covers, 4 4s. net. 
 Parts I (Vols. I- III) and II (Vols. IV to VII) can be obtained separately; 
 price per part, in cloth, 2 17s. 6d. net, in paper covers, 2 12s. 6d. net. 
 
YC 36554 
 
 M105768 
 
 
 THE UNIVERSITY OF CALIFORNIA LIBRARY