books on Plautius, the Digests of Celsus and Marcellus, the books on the Julian and Papian law, and others. Of the Papinian group the Quaestiones of Paul and Scaevola, the answers of Paul and Scaevola, the books on Trusts, the Receptae sententiae of Paul with the Epitome of Hermogenian, and others were respectively grouped.

Out of nearly 230 works bearing separate titles, extracts from which appear in the Digest, a small number have composed the bulk of the Digest. The works first taken in hand were large and comprehensive treatises or collections, and naturally extracts from them rendered it superfluous to make many or large extracts from most of the others, except on special points. Hommel has rearranged all the extracts under the head of the works from which they were taken, and although from the necessity of the case his book falls far short of justifying its ambitious title as a 'New birth of the Old LawBooks' (Palingenesia librorum iuris ueterum, Lipsiae, 1767) it is convenient for many purposes, and affords us a ready means of determining the proportion in which each work and each author has been made to contribute to the Digest. (See App. C.)

The extracts from Ulpian's commentary on the Edict form between a fourth and a fifth, or, if we add his Libri ad Sabinum, nearly one-third of the whole Digest. If we add to these the like treatises of Paul, Pomponius' books on Sabinus, Julian's Digesta, Gaius' commentary on the Provincial Edict, Papinian's Quaestiones and Responsa, and Scaevola's Digesta, we have one half of the Digest. Six more works, viz. Ulpian's Disputationes, Paul's Quaestiones, Responsa and books ad Plautium, African's Quaestiones and Scaevola's Responsa, making sixteen works by eight authors, raise the proportion to nearly two-thirds. About twice this number, or thirty-three works by sixteen authors give four-fifths of the whole Digest. The remaining one-fifth is contributed by nearly 200 treatises, none of which however have supplied extracts sufficient to fill much more than four pages of Mommsen's stereotype edition, and some of which have supplied only two or three lines.

The proportion of extracts supplied by each of the three groups is estimated by Bluhme approximately as the numbers 5 for the Sabinian group, 4 for the Edictal group, 3 for the Papinian group. The thirty-three important works, which, as I have stated, furnish together four-fifths of the Digest, are distributed between the groups as follows. I have added my estimate of the pages of Hommel

respectively occupied by them. The order of extraction is preserved, the works dealt with simultaneously being bracketed.

Gai Libri IX.—XVIII. ad edictum prouinciale

Ulpiani Disputationes

14

25

Gai

Libri I.-VIII.; XIX.-XXX.adedict. prouinciale 19

8

11

Modestini Excusationes

Modestini Responsa.

Iavoleni Epistulae

Pomponi Libri ad Q. Mucium

Callistrati Libri de cognitionibus

Ulpiani Libri ad legem Iuliam et Papiam

PAPINIAN GROUP:

Quaestiones

Papiniani

45

42222

9

20

20

423

N.B. The total number of extracts in the Digest occupy about 1510 of Hommel's pages.

The group called Appendix to the Papinian group, and which is found sometimes to precede the bulk of that group, sometimes to follow it, contains one very important work, Scaevola's Digest, and also some books of Labeo abridged by Javolenus, and by Paulus. It is probable that they were discovered later than the rest, and given out, probably to the Papinian Committee, separately from the other works assigned to them.

Another case of indecision is shewn by the position of some books of Ulpian's, and the others' Commentaries on the Edict. The whole of Books 52-81 of Ulpian's were given to the Edictal Committee, but they were extracted in this order: Books 56-81, then the Commentary on the Aediles' Edict, then Books 54, 55, and lastly Books 52, 53. Possibly these books 52-55 were originally assigned to the Sabinian Committee, and they being overworked eventually passed them on to the Edictal Committee. But it is very likely that the subject-matter had something to do with it. Book 52 (besides treating of the security to be given to the legatee) dealt with operis noui nuntiatio; Book 53 with damni infecti and aquae pluuiae arcendae; Books 54, 55 with claims to freedom and with the publicani. These are just matters of which the position in the Digest is somewhat difficult to account for. See what is said of Dig. XXXIX. and XL. above p. xliii. The position of the Commentary on the Aedile's Edict in the order of extracting probably indicates that that Edict was by Julian made an Appendix to the Praetor's Edict. (Hence the Florentine Index gives the number of Ulpian's books on the Edict as 83, and of Paul's as 80.)

If we turn for a minute from the works to the authors it may be noted that the extracts from Ulpian's various works are in bulk 39 per cent. of the whole Digest. Paulus has nearly 18 per cent.; Papinianus, Scaevola, Pomponius, Julianus and Gaius have together more than 24 per cent.; Modestinus, Marcianus, Africanus, Javolenus and Marcellus 9 per cent.; Tryphoninus, Callistratus, Celsus and Hermogenianus nearly 4 per cent. Thus the sixteen authors of the thirty-three works mentioned above have, when their other works are taken into account, supplied 94 per cent. of the whole. The remaining 5 per cent. was supplied by extracts from 22 writers.

If we make a chronological division, classing as older writers all down to and inclusive of Gaius, and as later writers the rest, beginning with Uenuleius and Marcellus, one-fifth of the Digest is taken from the older writers and four-fifths from the later. The Republican writers (Q. Mucius, Aelius Gallus) and the other writers before Trajan (Alfenus, Labeo, Proculus) are insignificant in bulk (not 2 per cent. of the whole): the writers after Alexander Severus, or at any rate after Gordian (Hermogenianus, Arcadius) are doubly insignificant (not we may safely say 1 per cent.). Over 95 per cent. of the Digest was originally written from cir. 100 to 230 A.D.

Such then being the authors and works used, what is the nature of the Digest as revealed by the observed order? Clearly it is not a systematic treatise on each separate head of law, composed after consultation of these authorities. Nor is it a skilful mosaic in which fragments carefully selected from these works are disposed, each in the place and order calculated to give the effect of original treatment, because permeated with consecutive thought. It is simply the result of taking the selected treatises and arranging them, partly in one line, partly in parallel lines, and then as it were squeezing them together, so as to throw out what was antiquated or superfluous, and leave only what was practical, with no more repetition than was needed for clearness. Where Ulpian, Pomponius and Paul had treated the same matter, the most comprehensive, usually Ulpian, sometimes Paul (e.g. D. XLI. 2), was made the groundwork, and the others were either dovetailed into it or clipt and appended. The larger treatises being exhausted, supplementary remarks or illustrations were taken from the multifarious string of minor works which each committee had received for examination. The same process was repeated when the contributions of the three committees were combined, but there is less fusing of the extracts together: the

largest and most important string of extracts being put first, the other two contributions were clipt and appended.

But there are, as has been said and as would naturally result from such treatment, signs of the operations of the compilers having been in a limited degree guided by the connexion of thought. First, not unusually some extracts are taken from the regular order of the works and placed at the beginning of the title as an introduction (see 11 1-6 of this title). Secondly, some extracts have been evidently mutilated so as to form part of a sentence in another extract from a different work (see e.g. 18; 1 14; 1 16; 1 69 of this title, and especially III. 3; IV. 3 passim). Thirdly, successive extracts are connected by particles, e.g. quodsi, uero, autem, enim, ergo, &c., which may or may not have been in the original work, but at any rate have, it is fair to suppose, been either left there or inserted there on purpose to connect the meaning of the two fragments. Fourthly, repeated alternation of extracts from the same books indicates the selection or compression of the one with an eye to the other. Where there is no such connexion visible, we should not be justified in imagining that the compilers intended to imply any connexion in thought between two successive extracts, unless one at least of the extracts is found in a different part of the title from that which the known order of the treatises would assign to it. And even if it is found in such an abnormal position, we cannot at once conclude that the law given in the extract is meant to be explained or qualified or supplemented by the adjoining extracts or uice uersa. The position may be merely accidental—a slip of the compilers or their copyists, a consequence of the exigencies or convenience of the work. For there is no ground for thinking that the compilers set any special value on preserving the order. The order, which we observe, is one which was natural for them to fall into, but not one thoughtfully adopted and religiously retained and protected. They struck out parts of the manuscripts before them, amended the rest where necessary, and gave them to be copied one by one or section by section as they finished with them.

But even when we can see some design in the displacement of a fragment, it is not always a design of real importance for the interpretation of the law. The connexion may be due to a much slighter Thus Bluhme (p. 293) suggests that the extract from Papinian forming 1 33 of our title has been put out of its place, merely because Julian is mentioned, and the position tus serves to introduce the

cause.