IV THE LITERATURE OF THE CLASSICAL AGE: ITS FORMS AND ITS TRANSMISSION

The present chapter is perhaps the most important of this book. The classical juristic literature represents indeed the core of our sources of Roman law and every methodical inquiry depends on the true valuation of these texts as they have come down to us, a valuation which can only be achieved by a clear insight into the form and the fate of this literature.

1. Leaving aside, for the present, the law-creating and declaring acts of State,¹ Gaius’ Institutiones are the only classical work that has reached us anything like complete. Of the rest we possess only post-classical abridgements and fragments: these, however, are so extensive that it is possible to discern the structure of many of the works and to reconstruct large portions of text. This work of reconstruction, which goes by the name of palingenesia, was begun in the humanistic period, but after the first, relatively immature essays it was long before further progress was made? It was left to Otto Lenel, in his Edictum Perpetuum (1883) and, building on that, in his Palingenesia Juris Civilis (1889), to make a real advance. His Palingenesia is a serviceable instrument for further research, a basis from which fresh advance can be made,³ but, as Lenel himself recognized, it does not exhaust the results that are obtainable from the evidence. Future progress will, however, depend on the detailed study of the whole of the surviv­ing fragments of this or that individual work considered together. What has been accomplished by researches conducted on these lines since 1889 will be noted in our accounts of the individual works.

2. The immediate goal of this palingenetic research is necessarily the reconstruction of the works as they appeared in the editions from which our fragments were extracted. Now these editions were one and all post-classical, so that the further question arises

¹ Below, p.

² Lenel’s work does not include sacral or constitutional law, for which one must therefore use the collections of fragments in Bremer, ii. i and ii. 2, and in Seckel- Kubler, i. But so far as Lenel is available, Bremer’s work, which is seldom helpful, is better left aside. whether and how far they were identical with those current in classical times or, for that matter, with the work as originally published. The critical school of the last quarter of the nineteenth century began by assuming as self-evident that the post-classical editions were, apart from copyists’ errors and the intrusion of, unimportant glosses, faithful copies of the classical. In particular it was believed that the editions from which the texts of Justinian’s Digest were excerpted gave the classical text, and that, con­sequently, to detect the interpolations of Justinian’s compilers was to re-establish the classical text. Conversely it was argued: 'this text is unclassical and therefore is due to the compilers.’¹ Consistently it was believed that the post-classical edition of Gaius’ Institutes contained in the Veronese manuscript gave, apart from scribal errors and a few insignificant glosses, the classical text; that the Epitome Ulpiani was an abridgement of Ulpian’s Regulae giving Ulpian’s authentic text; that the abridge­ment of Paul’s Sententiae contained in the Lex Romana Visigo- thorum, apart from certain Visigothic interpolations, presented the true Pauline text, and so on. These beliefs remained unshaken till the second decade of the twentieth century. But thereafter every serious research has led again and again to the same con­clusion, namely that in post-classical times the classical texts were subjected to alteration, sometimes superficial, sometimes profound. To the later age the classical works seemed too long; they were therefore abridged, in some cases by combining two or more works into one. Besides this, additions of many kinds were made—of rubrics, supporting arguments, abstract summaries of the case law, and corrections of substance.² This work of re-editing continued right through the post-classical period, but it was most radical in the earlier part of it, at the end of the third and the beginning of the fourth centuries.³ At that time the classical originals had not become consecrated as ius, and no scruple was felt in adapting them with the necessary freedom to present needs.

¹ So expressly Gradenwitz, Interpolaiionen, 43.

¹ Galen., Hepl τών ιδίων βιβλίων (Scripta min. ii, 1891, Teubner), Praef. 9: Πολνηδώ; Ιλωβησαντο πολλοί τοΐς ίμΰΐί βιβλίοις, άλλοι κατ' άλλα των ίάνων αναγιγνώσκοντας ώς ίδια μβτά τού τα μαν άφαιραΐν, τά δλ προστιθάναι, τα δ’ ύτταλλάτταιν.

³ See belpw, ρ. ί8ο.

⁴ Dionysius of Corinth (about A.D. 171) complains (Euseb. Hist. eccl. iv. 23. 12): ‘The apostles of the devil have filled them’ (scil. Dionysius’ works) ‘with tares by leaving out some things and putting in others ’ (Engl, transl. taken from the edition in the Loeb Class.). Galenus, l.c. See further Jachmann, Naehrithten der Gesellschaft obvious parallels may be found in the history of the transmission of medieval legal literature.¹ In the Middle Ages, as in the last centuries of antiquity, juristic literature was treated with no more respect than we nowadays show for cookery books,² travellers’ guides, song-books, and similar literature. The revision of such books is hampered by no scruples; there is no question, or expecta­tion, of fidelity to the original; no one regards their alteration as falsification.

* See, for example, Ksmtorowicz, Z xliii (1922), 21 ff., on the unauthentic layers of text in Gandinus. By comparing the various editions everybody can see how Baldus’ Concilia have been tampered with.

³ On antique cookery books see Bilabel, PW ii. 932 ff.

. ³ See H. Kruger, Die HersteUung der Dig. Justinians (1922), s. 10; Schulz, Epit. Ulp. pp. 9 and 18 ff.; * Ueberlieferungsgesch. d. Responsa des Cervidius Scaevola ’, Symb. Friburg., 216 ff.; Niedermeyer, ‘ Vorjustinianische Glossen u. Interpolationen ’, &c., AC I 1933, Roma, i. 353 ff.; De Francisci, Conjerenze (1931), 29. For the rest see the literature cited below on the individual classical works. the post-classical editor will have abridged the original in his own language, but without altering the law stated: in such a case it is clearly impossible to reconstruct the original with any sort of certainty.¹ Palingenesia will in future have to be limited to the reconstruction of the post-classical but pre-Justinian version. From these texts we may hope to recover, not indeed the classical originals, but still their essential contents, i.e. the classical legal doctrine, and so to obtain a picture of classical jurisprudence both as a whole and with many details.^{[213] [214]} This after all is the essential point for the legal historian.

Reconstructions of the classical texts must be received with scepti­cism. Such value as they have is as restitutions of the classical author’s substantial meaning, not of his words. The linguistic researches of recent years also retain value, but their results must to some extent be applied from a different point of view. A text shown on linguistic grounds to be post-classical, but prior to Justinian, must not simply be thrown aside. It may in substance be reproducing classical doctrine; in fact, unless its doctrine can be argued undassical, the presumption is in favour of its being in substance classical in spite of its being expressed in unclassical form.

3. A document of general importance for the literary history of the classical jurisprudence is the so-called Index Florentinus. Justinian had ordered that the compilers of the Digest should compose an index of all those books from which they had inserted extracts in the Digest, and that this index should be prefixed to the Digest. His order was carried out, and our main manuscript

145 of the Digest, the Codex Florentinus, has preserved this index,¹ whereas the medieval Vulgate MSS. have omitted it.

The index is written in Greek, the Latin version, which must needs have existed, is lost. It begins with a Greek title: Έξ όσων αρχαίων και των νπ αυτών γενομένων βιβλίων συνκεΐται το “παρόν των digeston ήτοι πανδέκτου του ευσεβέστατου βασιλέως ’Ιουστινιανού σύνταγμα.² This title makes it quite clear that the Index is a list of those books from which fragments have been inserted in the Digest, not of those which the compilers had also consulted but rejected as unfit for their purpose. The list begins with Julian and Papinian, i.e. with the two greatest jurists according to Justinian’s evaluation. Then follow the other jurists in what the compilers believed to be their chronological order, beginning with Q. Mucius Scaevola and ending with Hermogenianus. The books of each jurist are arranged according to their size, the most bulky books coming first and the least voluminous last.³ But the order is also influenced by the order in which the books followed within the four masses,⁴ with which we shall deal later.³ The names of the jurists are invariably given in the genitive case, formed after the Greek flexion even when the lettering is Roman. The titles of the books are mostly written with Roman letters, but sometimes Greek inflexions are used. Sometimes the original title has been abbreviated or completely transformed, probably according to the usage of the Byzantine law-school. The number of the libri of each book is always given in Greek words (not figures). This mixture of Greek and Latin is a well-known peculiarity of Byzantine law-Greek. It has long been observed that the contents of the Index do not quite agree with its title: books are men­tioned in the Index of which no fragments are inserted in the Digest; on the other hand, the Digest contains fragments of books which are not mentioned in the Index. Of these discrepancies plausible explanations can be given.

These explanations cannot be adequately discussed here. Hence

¹ Ed. Mommsen, Digesta, i (1870), p. Hi. Literature on the Index: Spangenberg, Einleitung in das Romisch-Justinianeische Rechtsbuch (1817), 24 ff.; Puchta, Kleine Schriften (1851), 2r6; Mommsen, Digesta, i (1870), Praefatio, p. xi; Lintelo de Geer, Verslagen en Mededeelingen der Koninklijke Akademie van Wetenschappen 2. Reeks 6. Deel (Amsterdam, Ã877), 334 fr.; Buonamici, Annali delie Univ. Toscane, xxiii (1901); Joers, PW v (7905), 492; H. Peters, Leipzig SB Ixv (1913), offprint, pp. 75 ff.; Ebrard, Z xl (1919), 124 ff.; G. Rotondi, Scritti, i (1922), 298 ff.

³ = ‘Ex quibus veteribus iurisconsultis librisque ab ipsis conscriptis constet praesens digestorum sive pandectarum piissimi imperatoris lustiniani corpus.’

’ There are, however, some unaccountable exceptions to this rule.

⁴ Rotondi, l.c. 323. ^s See below, p. 319.

4497.1 _L

we give only the two lists of books* with references to the pages on which they are discussed below.

I. Books of which no fragments are inserted in the Digest but which are mentioned in the Index.

1. Alfenus Varus, Digestorum libri XL, below, p. 205.

2. Massurius Sabinus, Iuris civilis libri III, below, p. 156.

3. Cervidius Scaevola, De quaestione familiae liber singularis, below, P· 233·

4. Gaius, Dotalicion liber singularis, below, p. 253.

5. Ulpianus, Pandectarum liber singularis libri X, below, p. 122.

6. Paulus, De officio praetoris tutelaris liber singularis, below, p..

7. Paulus, De extraordinariis criminibus liber singularis, below, p. 257.

8. Paulus, Ad formulam hypothecariam liber singularis, below, p. 202.

9. Paulus, Ad municipalem liber singularis, below, p. 196.

10 Paulus, Ad legem Velleam liber singularis, below, p. 189.

11. Paulus, De iure patronatus quod ex lege lulia et Papia venit liber singularis, below, p. 188.

12. Paulus, De actionibus liber singularis, below, p. 255.

13. Paulus, De donationibus inter virum et uxorem liber singularis, below, p. 253.

14. Paulus, De legibus liber singularis, below, p. 188.

15. Paulus, De legitimis hereditatibus liber singularis, below, p. 255.

16. Modestinus, De legatis et fideicommissis liber singularis, below, p· 255·

17. Modestinus, De testamentis liber singularis, below, p. 255.

II. Books of which fragments are inserted in the Digest but which are not mentioned in the Index.

1. Aelius Gallus, De verborum quae ad ius pertinent significatione, be­low, p. 283 n. 8.

2. Alfenus Varus, Digestorum a Paulo epitomatorum libri, below, p. 205.

3. Alfenus Varus, Digestorum ab anonymo epitomatorum libri, below, p. 206.

4. Gaius, Ad legem Glitiam liber singularis, below, p. 187.

5. Gaius, Regularum libri III, below, p. 174.

6. Gaius, Ad s.c. Orfitianum liber singularis, below, p. 189.

7. Gaius, Ad s.c. Tertullianum liber singularis, below, p. 189.

8. Gaius, De tacitis fideicommissis liber singularis, below, p. 255.

9. Maecianus, Ex lege Rhodia, below, p. 255.

10. Modestinus, De praescriptionibus libri IV, below, p. 256.

11. Paulus, De adsignatione libertorum liber singularis, below, p. 253.

12. Paulus, De articulis liberalis causae liber singularis, below, p. 253.

13. Paulus, De cognitionibus liber singularis, below, p. 256.

* The lists given by P. Krüger, Joers, Peters, and others are not quite exact.

14. Paulus, De conceptione formularum liber singularis, below, p. 255.

15. Paulus, De dotis repetitione liber singularis, below, p. 253.

16. Paulus, Ad legem Fufiam Caniniam liber singularis, below, p. 189.

17. Paulus, De liberali causa liber singularis, below, pp. 196, 253.

18. Paulus, De officio adsessorum liber singularis, below, p. 246.

19. Paulus, Ad s.c. Turpillianum liber singularis, below, p. 189.

20. Paulus, De variis lectionibus liber singularis, below, p. 222.

21. Pomponius, Enchiridii liber singularis, below, p. 170.

22. Proculus, Ex posterioribus Labeonis, below, p. 210.

23. Ulpianus, Excusationum liber singularis, below, p. 249.

24. Ulpianus, Ad legem Aeliam Sentiam libri IV, below, p. 18g.

25. Ulpianus, De officio consularium liber singularis, below, p. 247.

26. Ulpianus, Pandectarum liber singularis, below, p. 222.

(Ü)

Our next business is with the various forms taken by juristic literature and their specimens. Let us begin with considering the forms and transmission of the Roman acts of State which create new or declare existing law.¹

1. Leges (leges rogatae, plebiscita, leges datae). There was no change in their literary form nor was any official or private collection published even in this period? In the course of the first century lex as a form for creating new law receded into the back­ground and practically disappeared.³

2.Senatusconsulta. Here, too, the literary form of the decree remained unaltered,⁴ but we have now to take account of the fact that when a senatusconsultum was passed on the proposition of the Princeps* it was the introductory oratio principis that the jurists quoted from, not the senatusconsultum itself. As of leges, so of senatusconsulta and orationes no collection was made;⁶

¹ Above, p. 87. See for what follows v. Schwind, ‘Zur Frage der Publikation im rom. Recht ’, Münchener Beiträge zur Papyrusforschung u. antiken RG. 1940 (inaccessible).

* Cass. Dio, 57.16. 2, reports that Tiberius charged a committee of three senators to collect some old public documents (δημόσια γράμματα). This hardly means (as Geizer, PW x. 525 believes) that Tiberius planned a collection of leges.

³ Mommsen, Staatsr. iii. 345; Sehr. ³. 285. For the leges of this period see the literature cited above, p. 87. The two Spanish leges municipales {leges datae) from the time of Domitian show a similar stratification to that of the lex Rubria and the lex Ursonensis. Text: Bruns, no. 30. ILS 6088/9; FIRA i. 202ff. Literature: Mommsen, Sehr. ³. 280 ff. and above, p. 88. ⁴ Above, p. 87.

³ Mommsen, Staatsr. ii. 898; Radin, PW xviii. 869 ff.

* There is no collection of known senatusconsulta similar to Rotondi’s of leges. Wlassak, Prozessgesetze, ii. 173 ff., gives a list of those affecting private law under the Empire; Cuq, Consilium principis, 424, and Radin, PW xviii. 871 ff., give a list of orationes principum. See further, Stella Maranca, ‘ Di alcuni senatus consultis nelle iscrizioni latine ’, Rend. Lincei (Class, di Scienze Mor., Stor, e Philosoph. Ser. VI, vol. i, 1925), 504 ff., and above, p. 88.

Pomponius’ libri V de senatusconsuUis and Paul’s liber singularis de seis, were only treatises on certain selected senatusconsulta.¹ It is from copies obtained from the aerarium² that the texts preserved in juristic literature ultimately derived. Such copies could natur­ally be obtained by a metropolitan jurist, but others might at times have difficulty in obtaining a proper text and have to rely on second-hand information.³

3. Edicta. The only feature known to us which is common to the Edicts of the Emperor and of the magistrates in general is the heading: name and office of the edicens, sometimes the date,⁴ followed by dicit (in Greek versions Aeyei). There was no literary collection of edicts, even imperial.⁵ Only of the jurisdictional edicts, that is those of the praetors, governors of provinces, and curule aediles, is there anything more to say. As already men­tioned,⁶ these edicts were by Hadrian’s order stabilized by a senatusconsuttum in the forms settled by Julian. These codifica­tions were transmitted in book form ;⁷ Justinian still knows of a 'little book'⁸ containing the Edict, by which no doubt is meant that of the praetor urbanus along with that of the aediles. Our knowledge of this Edict depends entirely on the surviving frag­ments of the classical commentaries Ad Edictum. From these a reconstruction of the Edicts of the praetor urbanus and the aediles is possible and has been carried out in all essentials, finally by Lenel.’ The edictal system is so important in the history of juristic systematization that we must at least exhibit it in outline¹⁰ by giving the edictal titles in their order, as reconstructed by Lenel.¹¹

¹ Lenel, Pal. ii. 148; i. 1294.

I On the archives: Mommsen, Schr. iii. 295 ff.; Staatsr. iii. 1010. On copies: ibid. 1013. Cf. the SC. de nundinis (Bruns, no. 61): * SC. de nundmis saltus Beguensis in temtorio Casensi, descriptum et recognitum ex libro sententiarum in senatu dictarum Kari luni Nigri, C. Pomponi Camerini cos.

³ Gaius, i. 32b: ‘postea dicitur (!) factum esse senatusconsultum....’

⁴ In the Edict of Claudius, FIRA i. 417, and the Cyrenean Edicts of Augustus, ibid. 403, the date is in the heading, whereas in the Edicts of provincial governors it is at the end: e.g. Mitteis, Chrest. no. 192; Wilcken, Chrest. no. 19.

⁵ Some imperial edicts: Haenel, Corpus Legum (1857); Cuq, Consil. princ. 456;

Orestano, BuU. xliv (N.S. iii, 1936-7), 241 ff.; provincial governors’ edicts: Weiss, St. z. d. òîò. Rechtsquellen, 71 ff.; Wilcken, Z xlii (1921), 137. Also Voc. ii. 425, 10 f.; Haberleitner, Phil. Ixviii (1909), 271 ff. ⁶ Above, p. 127.

⁷ The jurisdictional edicts were current in book-form earlier: Gellius (rr. Ã7; cf. Weiss, Z 1 (1930), 256) had read them in the Library.

⁸ Const. JcSuKev, s. 18. Cf. ILS 8987: ‘praetoris volumen’.

⁹ Ed. 1, 1883; French ed. i (1901); ii (1903); ed. 2,1907; ed. 3,1927, to which our

citations refer. ¹⁰ Lenel, Ed. pp. xvi ff., 3 ff.

II We give the titles as in Lenel, though some are not authentic and some uncertain

or disputed. We are not concerned here with their exact formulation. See Riccobono, FIRA i. 335 ff. _>

Pars I,

De his qui in municipio colonia foro iure dicundo praesunt.

De iurisdictione.

De edendo.

De pactis conventis.¹

De in ius vocando.

De postulando.

De vadimoniis.

De cognitoribus et procuratoribus et- defeiisoribus.

De calumniatoribus.

De in integrum restitutionibus.

De receptis.

De satisdando.

Quibus causis praeiudicium fieri non oportet.

Pars lia

De iudiciis.

De his quae cuiusque in bonis sunt.

De religiosis et sumptibus funerum.

De rebus creditis.

Quod cum magistro navis institore eove qui in aliena potestate erit negotium gestum erit.

De bonae fidei iudiciis.

De re uxoria.

De liberis et de ventre.

De tutelis.

De furtis.

De iure patronatus.

Pars Ilb

De bonorum possessionibus.

De testamentis.

De legatis.

De operis novi nuntiatione.

De damno infecto.

De aqua et aquae pluviae arcendae.

De liberali causa.

De publicanis.

De praediatoribus.

De vi turba incendio ruina naufragio rate nave expugnata.

De iniuriis.

Pars III

De re iudicata.

De confessis et indefensis.

’ Riccobono, Bull. xliv (N.S. iii, 1936-7), 9 ff. Lenel: ‘ De pactis et conventionibus.

Not quite correct Koschaker, Festschrifi Hanausek (1925), 156·

Qui neque sequantur neque ducantur.

Quibus ex causis in possessionem eatur.

De bonis possidendis proscribendis vendundis.

Quemadmodum a bonorum emptore vel contra eum agatur.

De curatore bonis dando.

De sententia in duplum revocanda.

Pars IV

De interdictis.

De exceptionibus.

De stipulationibus praetoriis.

Appendix: Edictum aedilicium

De mancipiis vendundis.

De ramentis vendundis.

De feris.

Stipulatio ab aedilibus proposita.

The contents of the praetorian Edict’ can be summed up as constituting the praetor’s programme of office: he is announcing to the public, at the beginning of his term, how he intends to exercise his office. Its arrangement, which is in four main parts, is based on processual considerations. Part I orders and guarantees procedure in an action up to joinder of issue (litis contestatio), Part III regulates execution of judgment, Part IV is a collection of official formulae, interdicts, special defences (exceptiones), and praetorian stipulations. Part II, which is in two subdivisions, deals with the remedies not disposed of in Parts I and III, each remedy being accompanied by its appropriate formula, where this is not reserved for inclusion in the collection in Part IV.

Part I. An introductory part deals with the safeguards of the juris­diction of praetor and municipal magistrates. Then follows the title De edendo. Editio actionis (notice of claim) is put first as being a duty incumbent on the plaintiff even before the issue of summons. To it is attracted the special duty of bankers to allow inspection of documents (also edere). De pactis, which comes next, owes its position to the pact of compromise (transactio), placed here owing to the consideration that where there has been compromise there will be no in ius vocatio. This particular pact attracts the subject of pacts in general.² Next come the titles on summons (in ius vocatio), demand before the magistrate (postu­latio), securities for reappearance (vadimonia), and representatives be­fore the magistrate (cognitores, &c.). This last title ends with a subhead De negotiis gestis, which is attracted by the case of a procurator appear­ing in iure on behalf of an absent party:³ proceedings in iure are thus

’ Cf. Lenel, Ed. pp. 14 ff. * Lend, ibid., p. 32.

³ Partsch, * St. z. negotiorum gestio i (Heidelberg SB, Abh. 12), 13.

151 a negotium. The position of the title De calumniatoribus is determined similarly by its relevance to the prosecution of an action. Next comes the title De in integrum restitutionibus, because i. i. r. involves that the praetor reopens a remedy which has been barred iure civili. The title consisted originally of the rubric De minoribus xxv annis followed by Ex quibus causis maiores·, later, it is not certain when,¹ in integrum restitutio metus causa was placed at the head of the title, though the rubric Ex quibus causis maiores was thereby rendered illogical;² i. i. r. propter metum attracted the actio quod metus causa, and this in turn the actio de dolo, in spite of there being no i. i. r. propter dolum.³ Next comes the title De receptis, its position being due to receptum arbitri, i.e. agreement to act as an arbiter; receptum cauponum, &c., and receptum argentarii are attracted.

Part II. We can distinguish two main subdivisions. The first (Ila) deals with the ordinary, the second (116) with the summary, remedies; the latter are not confined to the ordinary terms of court-sittings (actus rerum, in the provinces conventus). Ila divides into three main heads: property in bonis, property extra bona (De religiosis), and personal claims. The treatment of personal claims opens with the title De rebus creditis·. the leading topic, which attracts the others, is loan for consumption, this suggesting loan for use (commodatum). Next come the bonae fidei iudicia, the actio rei uxoriae attracting the title De liberis. It is curious to find De furtis interposed between De tuteUs and De iure patrona­tus.* II&, on summary remedies, deals with bonorum possessio, which introduces De testamentis and De legatis. With the title De liberali causa we reach the iudicia before recuperatores.

For further information we must refer the reader to Lend and to Riccobono’s footnotes in his edition of the edictum, FIRA i. 335 ff.

Mommsen called this edictal order a disorder,⁵ and certainly it is anything but a masterpiece of systematization. An unsatisfactory feature is that some of the formulae are assembled in Part IV, while the rest are scattered over the Edict. The primitive practice of grouping topics by association still plays an important part.⁶

* Schulz, Z xliii (1923), 222 ff.

² Because restitutio of maiores would include restitutio propter metum.

³ Originally rest, propter fraudem creditorum (Lenel, Ed. s. 225) came next after

rest, propter metum and was itself followed by the actio de dolo, which it attracted. The plairing of rest. p. fr. cr. in Part 3 seems to have come from Julian: Schulz, Z xliii (1923), 237, n. 5. ⁴ Lenel, Ed. p. 36, attempts an explanation.

¹ Sckr. i. 164.

⁶ Specially clear in the title De receptis, where the word recipere is the only. connexion between rec. arbitri, cauponum, &c., and argentarii (Lenel, Ed. p. 33). The actio arborum furtim caesarum is in the title De furtis merely because of the word furtim, since it is not a case of true furtum (Lenel, p. 42). The system recalls the connexion by catchwords in Theognis: Christ-Schmid-Stahlin, Gesch. d. griech. Lit. i (1929), 376, and E. Diehl’s edition (Teubner). On the order in the Epistle of James·. Christ-StaJalin, ii (ed. 6,1924), 1155.

The scheme is visibly one that has grown up gradually from one generation to another. How far Julian’s final redaction departs from the hitherto traditional arrangement we have not the means of judging save in some exceptional cases.¹ Before him all the formulae had been collected at the end; it was he who first reduced the last part to interdicts, special defences, and praetorian stipulations.^{[215] [216] He seems also to have been responsible for moving in integrum restitutio creditorum far away from the title De i. i. r.[217] [218] [219] Neither of these changes is happy. But on the whole his alterations do not seem to have been important: any serious change of order would have made the older commentaries on the Edict difficult to use. No doubt this consideration ought not to have been decisive, but, in spite of Q. Mudus, the interest of the jurists in logical, Hellenistic systematization remained very mild. We shall speak below of the adoption of the edictal order of topics in juristic works.}

4. Imperial rescripts* i.e. the Emperor’s answers to questions, memorials, and petitions addressed to him, took the form of either epistula or subscript™. The former is a rescript in the form of a separate letter. It was headed in the usual epistolary style by a greeting (e.g. ‘Imp. Caesar Vespasianus magistratibus et senatori­bus Vanacinorum salutem didt’), and ended with vale or the like, written in the Emperor’s own hand, and the date and place.® Such rescripts issued from the office ab epistulis.^[220] A subscript™ is a rescript written at the foot of the petition (libellus) itself. It began with the names of the Emperor (nominative) and the addressee (dative), but without greeting; at the end the head of the chancery added recognovi and the Emperor, in his own hand, scripsi or rescripsi, with date and place.^{[221] [222] Subscriptiones issued from the office a libellis* and were publicly posted (propositio) with date and place noted on the libellus.[223] Rescripts were filed in their two classes by the Roman Chancery; literary collections giving full and exact texts were· unknown in the classical period. The collec­tion made by the younger Pliny of the letters addressed to him by Trajan had no juristic purpose nor, so far as we know, was it used by the jurists. Also non-juristic is the collection known as Divi Hadriani sententiae et epistulae.1 There was, however, a collection of Marcus Aurelius’ rescripts entitled Semestria, which seems to have been made anonymously from the imperial archives; it is cited four times in our sources.2 Papirius Justus’ bulky Constitu­tionum libri xx,3 though never cited by the jurists, is certainly classical, indeed pre-Severan. Its author may have been an official of the archives.4 It does not give the rescripts verbatim, but only the archival minutes, which, however, at times preserve the wording of the originals.® The collector added no comments. No other collections are known and probably none existed. It was only after the complete victory of bureaucracy under Diocletian that rescripts began to be codified.6 Faithful to its policy of compromise with the republican tradition7 the Principate left the task of informing the public of rescripts to the jurists, whose leaders were from the time of Hadrian members of the imperial consilium and thus in a position to keep abreast with the rescripts issued. So far as they had not direct access to the archives the jurists must have relied on copies.8 But at times they seem to have contented themselves with oral information.9}

¹ Edited by Becking in Corpus iur. Rom. anteiustiniani (1841), col. 201, and by Gotz, Corp, gloss, lai. iii (1892), p. 30, 14; Krüger, 285, n. 5.

¹ Tryph. D. (18. 7) 10; Tryph. D. (2. 14) 46; Ulp. D. (29. 2) 12; Inst. 1. 25. 1 (from an indeterminable classical source). Cf. Krüger, 119.

¹ Lenel, Pal. i. 947; Μ. Scarlata Fazio, SD. v (1939), 414 ff.

⁴ He is probably identical with Μ. Aurelius Veranius Papirius Dionysius who was a libellis and a cognitionibus and lived in the period in which Papirius Justus must have lived. See on Papirius Dionysius: CIL x. 6662 = ILS 1455; Prosopogr. (2nd ed.), no. 1567. A new inscription was published by Segrfe, Bulletin de la Sociiti Royale d'Archiologie d’Alexandrie, xxxii (1938), 138; see L’Annde ipigr. 1938, no. 60. Papirius Dionysius was killed by Commodus and therefore possibly was called ‘lustus’ after his death. Before the discovery of the new inscription it was unknown that he had the name ‘Veranius’. ¹ D. (18.1) 71; (48. 12) 3.

• Below, p. 287. ⁷ Niedermeyer, AC I Roma, i. 364-66.

• The inscription from Skaptopara (Bruns, no. 90; Wilcken, Hermes, Iv (1920), 31, 37) is a copy from a rescript as posted up; the inscription from Smyrna (Bruns,. no. 84; FIRA i. 435, Wilcken, 37) is a copy of a rescript kept in the archives, made by imperial permission.

• Gaius, 2. 221: ‘ quae sententia dicitur (!) divi Hadriani constitutione confirmata esse.’ Marcellus, D. (23. 2) 50: Proxime constitutum dicitur. Paul, D. (35. 2) 1. 14: divus Antoninus iudicasse dicitur. D. (41. 4) 2. 8: a divo Traiano constitutum dicitur. Marcian, D. (49. 14) 18. 9: ut et constitutum esse refertur. D. (1. 22) 2: ‘ut aliquo quoque decreto principali refertur constitutum ’ (probably interpolated: Gradenwitz, Z xxvi (1905), 347 ff.).

5. As in the republican period,¹ collections of decisions of the courts were unknown. None existed even of the decisions of the Emperor’s high court, but minutes of its sittings were kept in the imperial archives, where, like the records of rescripts and edicts, they were open to the jurists. The jurists made occasional use of these decisions in their books,² and sometimes reproduced the minutes more or less verbatim.³ But works dealing solely with such decisions were rare. An example must have been Titus Aristo’s Decreta Frontiniana* but the outstanding example is Paul’s Libri vi imperialium sententiarum in cognitionibus prola- tarum.^s This, however, was no mere collection of decisions but a selection of cases in the imperial court at which Paul had assisted. A classical jurist would have considered the simple reproduction of official minutes as mere hack-work. Paul’s work followed the order of the Edict (or of the Digesta).⁶ Justinian’s compilers did not possess the original work, but only two post-classical abridge­ments, one entitled Imperialium sententiarum in cognitionibus prolatarum ex libris sex and the other Decretorum libri tres.¹

6. Naturally local collections of such materials as imperial edicts, rescripts, decrees, and mandala (to the provincial governor),⁸ and of the governors’ own edicts, rescripts, and decrees,⁹ were kept in the offices of regional officials,¹⁰ particularly in that of the governor, in Egypt that of the praejectus Aegypti and the I dialo­gue ;^u provincial practitioners also may well have made collections of their own.¹² We have some remains of this class of provincial literature, the authors being presumably provincial officials or

* Above, p. 92. ¹ Evidence: Voc. ii. 107. 38 f.

³ D. (28. 4) 3; (48. 7) 7. The minutes were made up from a shorthand note: v. Premerstein, PW iv. 743.

⁴ Only D. (29. 2) 99. Cf. Mommsen, Sehr. ii. 22.

’ Lenel, Pal. i. 959, n. 1. Berger, PW x. 722, 725; Sanfilippo, Pauli Decretorum libri Ires (1938), containing a valuable commentary on the fragments.

⁶ On the Digesta see below, p. 226. ⁷ See Note X, p. 340.

⁸ Mandata: Finkelstein, T xiii (1934), 150 ff. See further the κΐφαλαϊον Ικ των Kaiaapos Ιντολων, CIL iii, Supplem. 7086, and Mommsen’s commentary, no. 25. On Greek models see P. Tebt. 703. Justinian’s Nov. 17 imitates a liber mandatorum.

⁹ Governors’ edicts: Weiss, St. z. ram. Rechtsquellen, 81 ff.; Wilcken, Z xlii (1921), 137 ff. Governors’ rescripts: Wilcken, Hermes, Iv (1920), 27 ff. Decree of the governor of Sardinia (CIL x. 7852; ILS 5947; Bruns, no. 71a; FIRA iv. 322): Mommsen, Sehr. v. 325 ff.; v. Premerstein, PW iv. 733.

¹⁰ See P. Krüger, 315; Mommsen, Sehr. ii. 363.

” Plin. Ad Trai. 65; v. Premerstein, PW iv. 756. P. Oxy. xvii. 2104 is a copy of an imperial rescript. There is a note on the foot of the document: * inserted in die commentarii ai the prefect of Egypt by Annianus* (obviously a secretary of the • prefect). See P. Μ. Meyer, St. Bonfante, ii. 341 ff.

¹² In the papyri we find in the minutes of court-proceedings repeated citations of enactments of the emperors and governors by the advocates. practitioners, namely a work on the Augustan forma idiologi,¹ frag­ments of collections of authorities on soldiers’ marriages^{[224] [225]} and on longi tempons praescriptio,^[226] and of a collection of Caracalla’s Edicts and the like.^{[227] [228] [229]} There was thus in the provinces a minor juristic literature, which must have had its importance in provincial practice, but was, of course, ignored by the metropolitan jurists.

(iii)

Pure formularies, i.e. collections of precedents of contracts, testaments, actions, and defences, were no longer made by the classical jurisconsults; indeed, even in the last century of the Republic this class of work had already been left to the lower orders of lawyers.® Owing to the ample official collection of pro­cedural formulae supplied by the fully developed Edict such collections were no longer needed for litigation in the classical period, but they were still required by contractors and testators. Nor were they lacking, but since they belong to the minor litera­ture already mentioned their authors are unknown to us; they were perhaps anonymous. Evidence of the existence of such works is provided by the formula BaeticaP and again by the Transyl­vanian triptych of a.d. 139,^[230] in which, as Mommsen^{[231] [232] has pointed out, the draftsman of the record of a mancipation of a female slave twice slips into the masculine gender, thus betraying that he is using a precedent in which the hypothetical sale was of a male.’}

Though the jurists comment, as occasion arises, on documents or their individual clauses,^[233] it is doubtful whether there were any classical works dealing exclusively or mainly with such matters. Venuleius’ Libri x actionum may have been such a work, supposing the word actio to have been used in the old sense of the Manilianae actiones.¹ The treatment devoted to the procedural formulae was not merely incidental to more general themes, as we see from Gaius, book 4, and from such monographs as Paul’s De conceptions formularum, his De actionibus, and his De concurrentibus actioni­bus.^{[234] [235] We know little about this literature for the simple reason that it had lost all interest for Justinian’s compilers.[236]}

(iv)

Under the Republic, as we have seen,^[237] there had been the first beginnings of an isagogic literature, but it was only in the classical period, especially from Hadrian onwards, when legal education became more academic,® that a true elementary legal literature came into existence. The growing activities of the bureaucracy created a demand for more thorough legal education, of the kind that could be imparted only in a law school. A series of school­books speedily resulted.

1. The earliest known to us comes from the first century of our era—the Libri tres iuris civilis of the celebrated law teacher Massurius Sabinus. Significantly its author abstained from label­ling it as elementary (institutiones), but so bare an outline, couched in terse axiomatic sentences,^[238] can have been nothing else.^[239] Clear as is its derivation from the basic systematic work of Roman jurisprudence, Q. Mucius’ Ius civile* comparison of the sizes of the two works (Mucius 18 books, Sabinus 3) reveals at once their fundamental difference. The republican pontifex wrote a com­prehensive work for ripe lawyers; the composition of an elementary text-book was as remote from his mind as teaching in a law school.’ The imperial law teacher wrote a text-book for his pupils. Here and there he modified Mucius’ arrangement of topics, but he accepted his conception of the matter to be dealt with: no more than Mucius did he confine himself to ius civile in the narrower sense, i.e. to the exclusion of ius honorarium;¹⁰ but he left unex­amined important topics which a comprehensive treatise of the

157 time of Tiberius ought not to have passed over. Of the real con­tracts none, of the consensual only sale and partnership, were dealt with. No rationed explanation of these omissions can be given, and it is out of the question that Sabinus himself should have published as a systematic work anything so fragmentary. Obviously what we are dealing with is not such a work, but a collection of lecture-notes made by the famous professor for his pupils and first published after his death.¹ It may be that he had not yet worked out his scheme in all its parts, or again that many of his notes were not forthcoming when the posthumous publica­tion was prepared. But, for all these defects, the work remained the text-book of the Sabinian school* up to the reign of Hadrian and perhaps later. We find Pomponius, Paul, and Ulpian still devoting extensive commentaries to it.^{[240] [241] [242] After Ulpian all trace of it is lost; the Interpretatio of the Law of Citations[243] affirms its disappearance; Justinian’s compilers possessed no copy. Though it is registered in the Index Librorum,[244] there is not one fragment from it in the Digest, and the compilers, had they possessed a copy, could hardly have failed to select one or two excerpts, if only reverentiae antiquitaiis causa, from a work of such ancient renown. A few fragments are given textually by Gellius,[245] and a few more can be gleaned from the above-mentioned classical commentaries.[246] The frequent quotations from Sabinus elsewhere never cite his lus civile as the source.[247] Its arrangement of topics can be reconstructed in outline from the classical commentaries Ad Sabinum:[248]}

I. Law of inheritance. (1) Testaments: (a) execution, (ft) institution of heres, (c) exheredation, (if) acceptance and refusal of hereditas. (2) Intestate succession. (3) Legaia.

II. Law of persons. The various forms of power over free men and slaves. Emancipation and manumission.

III. Law of obligations. (1) Sale, including mancipation. (2) Societas. (3) Actio rei uxoriae. (4) Actio tutelae. (5) Obligations ex delicto·. (a) furtum, (b) damnum iniuria datum (I. Aquilia), including damnum infectum, (c) iniuria. (6) Unjustified enrichment. (7) The aedilician Edict. (8) The literal contract. (9) The verbal contract.

IV. Law of things. (1) Acquisition of ownership, including donatio. (2) Servitudes. (3) Fiducia. (4) Postliminium.

This arrangement, as comparison shows, is derived from Mucius.¹ In part I Sabinus’ separation of legata from testamentary law is hardly an improvement. Another change is the placing of the law of things last; the subhead 'servitudes’ is better placed than, as by Mucius, in connexion with locatio conductio. Under obligations the actio rei uxoriae (dotal law) is placed by Sabinus after societas, perhaps in conformity with the edictal order² or, it may be, because the form of societas known as consortium attracts matri­monial consortium omnis vitae. As in the Edict,³ the actio tutelae (law of guardianship) follows the actio rei uxoriae and is itself followed⁴ by furtum', in contrast to the Edict, furtum is followed by the other delicts. That obligations ex delicto should have led up to unjustified enrichment and the aedilician actions is under­standable. The real contracts are absent, and so are locatio conductio and mandatum. Lenel’s favourable judgment of this system taken as a whole is not readily intelligible.⁵

2. We know of no further elementary books till the second century. The first to be mentioned is Florentinus’ Institutiones, an extensive work, in twelve books, the only known work of this author, of whom we know nothing more. He must, like Gaius, have been simply a law teacher ;⁶ like him he is never cited by the classical jurists. The position of his book in the Index Librorumf after the works of Marcellus and before those of Gaius, in the absence of other evidence, fixes his date. At any rate there is no valid argument for putting him in the Severan period.⁸ The work was thorough, well arranged, and written in classical style.

¹ Above, p. 95. ² Above, p. 149. ³ Above, p. 149.

⁴ Above, p. 149. Lenel, Ed. 36, n. 1.

^s Sabinussystem, 104 (offprint). ⁶ Above, p. 107.

⁷ ‘XIX. Φλωροτίνου instituton βιβλία δίκαδυο.’

* Krüger, 215; Brassloff, PW vi. 2755. The only jurist cited in our fragments is Trebatius, the only Emperor dims Pius, both in D. (41.1) 16. The Florentinus in Alexander’s rescript, C. (3. 28) 8, is not our jurist, as C. (6. 30) 2 shows. The notice in SHA, Alex. Sev. 68, is valueless: Mommsen, Sehr. ii. 66. But Patzig, Byz. Z. xiii (1904), 44 ff., seems still to be misled. Cf. Hohl, Klio, xiii (1913), 420.

Its disposition was as follows :* I. Sources (book 1); II. Marriage and Tutela (books 3-5); III. Property (book 6); IV. Obligations (books 7-8); V. De statu hominum (book 9); VI. Inheritance (books 10-n).

This work survived for centuries² in spite of the competition of Gaius and, though not mentioned in the Law of Citations, was used in the compilation of the Digest and the Institutes. The excerpts from it in the Institutes can, however, be identified with complete certainty only when they recur in the Digest, which unfortunately is not often the case.³ But these duplications suffice to reveal that the manuscript of the work used for the Digest was not the same as that used for the Institutes, and that in both manuscripts the work was in a post-classical state.⁴ A notable point in our fragments is that with one exception all citations of jurists and imperial constitutions have been cut out: their excision must be the work of the post-classical editor, and the exception is due to his oversight.

3. Two elementary works by Gaius are known—the four books of Institutiones and the seven of Res cottidianae.

A. The Institutiones.³ This is of the greatest importance, as being the only classical work which has reached us nearly complete. In its definition of its subject and its order of topics it shows once again the influence of Mucius’ Ius civile.⁶ It is not restricted to ius civile in the narrower sense, but takes account of ius honorarium also, though only in so far as this is closely implicated with ius civile. Thus bonorum possession is dealt with because of its con­nexion with hereditas, the actio vi bonorum raptorum* because allied to furtum, the actio iniuriarum in its praetorian form’ because of iniuria in the Twelve Tables, while, on the other hand, the actio de dolo, for example, and the actio quod metus causa are simply omitted. The order of topics in the Institutes is too familiar

¹ Pal. i. 171, n. 1. We do not possess a fragment from books 2 and 12.

³ Cited in Schol. Sin. (below, p. 325), xiii. 35.

³ Brassloff, PW vi. 2758.

⁴ See the duplicate passages D. (46. 4) 18 and Inst. (3. 29) 2. Small stylistic and other differences indicate the use of different manuscripts. In both the oral formula of the novating stipulate has been confused with the narrative form in which it would be recorded in writing (cf. D. 45.1.122. 2). This contamination is due to the post-classical editor who, moreover, added the formula of the acceptilatio. See Wlassak, Z xlii (1921), 394 ff.; Perozzi, ii. 404; Solazzi, L’estinzione della obbligazione (1931), ²5^x 5 Index Interp.

³ See in general Kubler, PW vii. 494; Kruger, QueUen, 206,276; Bizoukides (below, ' p. 166), i. 34 ff. ⁶ Above, pp. 72, 156.

⁷ Bonorum possessio secundum and contra tabulas: 2. 119-21, 125, 126, 129, 135-7, 147-50, B, p. ab intestate: 3, 25 f. · 3. 209. ’ 3. 220.

160 THE CLASSICAL PERIOD

to need to be set out here. Its departures from the Mucian scheme, from which it is obviously derived, are not always happy. In­heritance is removed from the first place and put after property and before obligations, the three subjects being combined as ius quod ad res pertinet—a decidedly heterogeneous assemblage. There are many omissions, the most remarkable being the absence of dotal law and of the real contracts of commodatum, depositum, and pignus. These are points to which we shall return.

We must in the first place determine the literary genus of the work. The title, Commentarii,¹ applied to a work obviously intended for academic use, can only mean ‘lecture-notes’,^{[249] [250] but we cannot decide whether we have before us the notes used by Gaius himself when lecturing or notes taken by pupils at his lectures ano pwvfjs ;[251] [252] this, however, is of no importance. The implication of the title is confirmed by the whole style of exposition. Though the audience is not addressed in the second person,♦ the ever­recurring admonitions (admonere)[253] for the purpose of emphasis, the constant references to the scheme of topics,[254] and the fact that the author supposes himself to be speaking (loqui)[255] are in true lecturer’s style.[256] [257] A further, more important question is whether the work was published by Gaius himself or only by some pupil.’ Its}

161 numerous defects, some of which cut deep, suggest the hypothesis that Gaius left his lectures imperfect and unfinished and that they were first published after his death. Gaius’ Institutes, like Sabinus’ lus civile,¹ resemble Aristotle’s Metaphysics^{[258] [259]· in this respect.[260]}

For nearly a century Romanistic scholarship treated the text of the Institutes as sacrosanct,^[261] admitting only some relatively few and unimportant glosses. But in the twentieth century criticism has become ever more radical. Kniep^[262] tried to distin­guish four strata of text—a pre-Gaian work used by Gaius, Gaius’ own elaborations of it, post-Gaian additions, and, within the pre- Gaian work, an original text and later modifications and additions. Such a thesis was doomed to failure, even if it had been carried out with greater insight and caution than it was. No doubt Gaius made use of older literature, but whether he took one work as a model is a question that cannot be settled. In any case, the mysterious ‘Sabinian school-book’ which Gaius is supposed to have used and elaborated is pure fantasy. The only such book that ever existed was Sabinus’ luris civilis libri Hi. It is possible, of course, that Gaius made use of lecture-notes taken by himself in his own student days. Be that as it may, Gaius did not simply copy and adapt a previous work; and it is therefore impossible by analysis to distinguish such a work from Gaius’ own additions. The still more radical view that the Institutes are not classical at all, but belong to the fourth or fifth century,^[263] has been taken seriously by no one of any judgment and has been disproved^[264] by the modem discoveries of manuscript remains. For the rest research has been focused on individual passages. Attempts have been made to prove, on linguistic or substantial grounds, that this or that passage of our text is unauthentic, non-Gaian, post-Gaian, or post- classical.^[265] The inherent defect of this procedure has been that it

studies passages in isolation instead of applying critical analysis to the work as a whole.¹ The ultimate result of bringing each section and each sentence one by one under the microscope, of striking out every imperfection, of regarding every word that is not strictly necessary sis superfluous, everything that is superfluous as objectionable, and everything that is objectionable as unauthentic,² is that hardly anything is left. Hence the conclusion to which certain critics have been driven is that our actual text was com­posed in the main in the course of the third and at the beginning of the fourth century, this being the only period in which such radical alterations of a classical text would have been ventured.³ This conclusion has not been refuted, as has sometimes been thought, by the newly discovered manuscript remains, but it is nevertheless improbable. It is improbable that a later editor would have deformed an originally faultless classical work into that which we have. The impression one has is rather of a work which its author had revised again and again, but had never brought to a final form.

This impression could only be verified by a minute analytical interpretation of the whole work. Here we must be content to draw attention to a few significant points.

(i) Of the real contracts only mutuum is mentioned; commodatum, depositum, and pignus are omitted.⁴ The explanation has been proposed that Gaius was following an ancient book written before these contracts had become actionable.⁵ But even granted that Gaius followed so old a model, he would necessarily have amplified it in this matter, since these contracts had become actionable, and that iure civili, long before his time, and he himself mentions depositum and commodatum under the law of actions.⁶ He simply cannot have omitted them in his lec­tures, because if in giving an outline of the Roman law of contract he had mentioned all four of the consensual, but omitted three of the real, contracts, he would simply have been misleading the beginner. Nor can the omission of these three contracts be due to their having been deleted by a later editor; they must have been absent from the beginning. Thus the only conclusion left is that Gaius for some reason had not yet written up this topic in his own manuscript or else that this part of his lecture could not be found when it came to posthumous publication, (ii) Pre­cisely the same considerations apply to his silence as to the law of dos.¹ (in) There are places where Gaius promises future treatment of some

’ D. (r. 3) 24.

* v. Wilamowitz-Mollendorff, EM. in d. griech. Tragodie, 250.

¹Above, p. 142. ⁴ Gaius, 3. 90.

⁵e.g. by Krüger, 210, n. 57. ⁶ Gaius, 3. 207; 4. 47, 60, 62,182.

⁷ Ibid. 1.178,180; 2.63; 3.95a, 125; 4.44,62,151. question (yidebimys), a promise that is not kept,¹ and others where he raises a question (quaeritur, quaesitum est), but contrary to his custom of proceeding to discussion, with citation of jurists,² leaves it unex­plored.³ Here again we cannot hold a later editor responsible; the impression left is that of a book not finished and published by its author, (iv) He introduces the work with a conspectus of the sources of Roman law in general,⁴ and then passes abruptly to ius privatum, as though there were no such thing as ius publicum. A transitional passage re­mained to be supplied, (v) In the introduction ius civile is contrasted, as being the law peculiar to Roman citizens, with ius gentium, or law common to all peoples,⁵ and the author promises that the exposition which follows will tell us to which of the two the several legal institu­tions belong.⁶ He keeps his promise in book 1 and the first part of book 2, but seems to forget it when he reaches the law of inheritance; here ius civile becomes antithetical to ius honorarium.⁷ He returns to his original intention in the treatment of obligations, but carries it out carelessly:⁸ one would have expected to be told at least that sale, barter, and hire belonged to the ius gentium. In book 4 the promise is once more completely forgotten; ius civile is simply private law which is not ius honorarium.⁹ Once again one has the impression of a book not finally revised by its author.¹⁰

It follows that the supposed authentic text, perfect in expression, absolutely correct in law, and conforming to the strict classical standards, is pure fantasy; it never existed any more than a well- ordered, carefully expressed and consistent Aristotelian Meta­physics. Criticisms, correct in themselves, have led in part to false conclusions. It is true enough that there are post-classical altera­tions in our text—this can be and has been proved—but merely formal defects, such as departures from classical juristic idiom, bad arrangement, unskilful transitions, inexact formulations, and the like, do not properly justify the conclusion that here is post- classical work. In rough lecture-notes, not written npos ckSooiv,¹¹ one cannot count on complete legal accuracy or the highest

^I Ibid. 2.120,121; 3.116,202. ³ e.g. ibid. 2. 79, 200, 244.

» Ibid. 2. 90,94, 95; 3.119,143, ^T44,145» ³72,189; 4· 125.

⁴ Ibid. 1. 2: ‘ Constant autem iura populi Romani....’

⁵ Ibid. 1.1; cf. above, p. 137.

⁴ Ibid.: ‘quae singula qualia sint, suis locis proponemus.’

⁷ Ibid. 2. 114, 115, 118, 135,136,149,151,170,197,198, 206, 218, 220, 241, 253, 255; 3· 34, 36, 37, 66, 71· ’ Ibid. 3. 93,132.

• Ibid. 4. 38, 45,107, 116.

¹⁰ Its frequent repetitions have long been noted: Demburg, op. cit. 40 ff.; Kruger, 209, n. 55. Moreover, the book has no preface.

^II Galenus, Ilf pl tu>v iStav /3i/JXian> (Scripta minora, ii, 1891, Teubner), Praef. 10: piXois yap /xafhyrais eSlSoro yaipis fmypaprjs d>s av ovSiv irpos exSomv, aAA* aurocs (Klims yeyov&ra Seqdfioiv v rjKovaav Ixav vnofnajpuira. linguistic polish. Also some responsibility may be attributed to whoever published the notes. It is proper to expose these defects, but to do so only completes in detail the proof of our thesis that the Institutes are a work which was left unfinished by its author and was first put together and published by some pupil.

That publication must have been soon after the death of the Emperor Pius.¹ A second edition, but withput considerable changes,³ may have been produced in the third century; in essen­tials the text was stabilized early, at the end of that century.³ Commentaries on it took the form of separate works,⁴ not that of a continuous gloss.⁵ Of course marginal glosses were entered by readers,⁶ and some of these eventually got into the textin one case, which seems to be exceptional, a passage was struck out.⁸ But in essentials the text was left unaltered,’ for the law schools of the fourth and fifth centuries, unlike the bureaucracy, tended to classicize the law.¹⁰ Even book 4, which the disappearance of the formulary system had rendered completely out of date, was laboriously preservedeven the polytheism which Justinian ex­punged from his compilations¹² was left untouched.¹³ Though used

¹ No constitutions of Marcus Aurelius are cited; Pius is cited as living except in

2. 195 (divi Pit constituiione). But on various grounds this passage (‘sed hodie* reB.) cannot be Gaian. But neither can it be post-classical, since the victory of the Sabinian doctrine in the full classical period is certain. It must therefore be an addition made by the first editor, and thus is important evidence as to the date of publication. On Gaius, 2.195: Beseler, Beitr. ii. 105; Z xliii (1922), 536; Ciapessoni, St. Bonfante, iii. 664; Voci, Teoria dell’ acquisto del legato (Milan, 1936), 49 ff., 62 ff.

^I It is significant that Marcus’ constitution (Epit. Ulp. 22. 34) was not quoted against the severe, formalistic decision in Gaius, 2.177.

³ So Schulz, Epit. Ulp. p. 13, before the discovery of the new texts of Gaius, which confirm the early stabilization of the text.

⁴ e.g. the Autun lecture-notes, on which see below, p. 301.

³ Below, p. 184.

⁶ Thus the Greek marginal' and interlinear glosses in the Antinoite Gaius: below, p. 166.

⁷ In three cases—Gaius, 1.53; 3.113,126—a reader had noted regula in the margin, and the word has passed into the text. Schulz, Z 1 (1930), 227. Again, in 4. 16, * adversarius quoque dicebat similiter: et ego te ’ is a gloss which has got into the Veronese text, as its absence from the Antinoite text shows: Arangio-Ruiz, Bull, ³ (N.S.), 607.

⁸ That the Veronese version has shortened the text of 3. 154 is now proved by the Antinoite version.

’ Even in the late post-classical period Marcus’ constitution (Epit. Ulp. 23. 34) was not quoted against Gaius, 2. 177.

¹⁰ Below, p. 281.

^II Preserved by both the Veronese and the Antinoite manuscripts and lectured on still, as the Autun lecture-notes show.

¹² Voc. s.v. ‘deus’, 2. 204. 49; D. (1. 3) 2 (Otov for 9ñøê); Inst. 2. 1. 8 and 1.8. 2, compared with the corresponding passages of Gaius; Joers, PW v. 538,1. 65.

¹³ Gaius, 1. 53; 2. 4; 1.112. ' for centuries as a school manual the text, like that of Euclid’s Elements, remained relatively pure. It was not superseded either by an epitome or a Greek translation: our chief manuscript dates from the fifth or sixth century. Justinian’s compilers possessed the work and made it the basis of their own Institutes, besides taking some excerpts from it for the Digest. After Justinian the existing copies gradually perished; thus the Veronese manuscript was overwritten, and thereby practically destroyed, as early as the seventh century. Gaius’ Institutes were superseded by Justinian’s in the East and in Italy, and by the Epitome Gai of the Visigothic Breviarium in Spain and France.

A scientific edition of Gaius’ Institutes ought to establish, as far as may be, the text of the first edition in all its incomplete­ness and with all its defects. It ought further to exhibit the transformations through which the text passed up to the time of Justinian, since the work is an important authority for post- classical as well as classical jurisprudence. The existing materials for such an edition are: I. remains of ancient copies of the Institutes, and II. ancient excerpts from the works derived from the Institutes.

I. (1) Codex Veronensis.¹ This our chief manuscript is a parchment codex written in Italy in the fifth or early sixth century, and over­written partly in the seventh and partly in the eighth centuries? Only one of the surviving leaves is not palimpsest. The manuscript is almost complete, only four folios (the last blank) having been lost. It was discovered at Verona in 1816 by Niebuhr³ and has been gradually deciphered, in the greater part, but not completely, by various scholars. The standard reading is that of Studemund and P. Kriiger, which is to be found in Studemund’s Apographum? the photographic reproduc­tion³ is of little service. (2) In 1927 A. S. Hunt published (P. Oxy. xvii. 2103) three fragments, discovered at Oxyrhynchus in Egypt, of a copy written on papyrus rolls in the third century, perhaps even as early as the middle of that century. It is thus our earliest evidence. It has been re-edited several times, partly with the aid of further study of the papy­rus by Hunt.⁶ (3) Two and a half folios of a parchment codex written at the end of the fourth or the beginning of the fifth century. The

■ Described by Studemund, Apographum, 5 f.

²On the superimposed text of St. Jerome: Studemund, l.c. in last note.

³Savigny, Vermischte Sehr. iü. 155 ff. Studemund, Apogr. p. viii.

⁴ Gai institutionum commentarii IV. Codicis Veronensis denuo collati apographum, 1874; additions in the edition of Kriiger and Studemund (below, p. 166), p. xvii f., and by Capocci, Bull, xxxvi (1928), 139 ff.

⁵ Gai Codex rescriptus... phototypice expressus, Leipzig, 1909; Kriiger, Z xxxi (1910), 2 ff.

⁶ Collinet, RH vii (s. 4, 1928), 92 ff.; Levy, Z xlviii (1928), 532 ff.; St. Bonfante, ii (1930), 277 ff.; De Zulueta, LQR 1928, 198 ff. fragments were bought by Medea Korsa from a dealer at Cairo in 1933, but where they were discovered is not known; the presumption is in favour of Antinoe. They were identified as coming from a copy of Gaius’ Institutes and first edited by Arangio-Ruiz (PSI xi. 1182, Florence, 1933) with photographs of the two complete folios. It has been re-edited by him and others.¹

II. (1) Excerpts from ancient copies are found chiefly in Justinian’s Digest and the Collatio.¹ (2) Derived works: (a) Gaius’Res Cottidianae,¹ (6) the so-called Epitome Ulpiani,⁴ (c) the ‘Autun Lecture-notes’,⁵ (d) the Visigothic Epitome Gai,⁶ (e) the Greek translation on which Theophilus’ so-called Paraphrase of the Institutes'¹ was based, (/) Jus­tinian’s Institutes.

The best edition is still that by P. Krüger and Studemund in the Collectio librorum iuris anteiustiniani.⁶ It is out of date, if for no other reason because it was made before the new discoveries of texts, but it remains indispensable to the researcher because it is the only edition which gives full information as to the readings of the Codex Veronensis. Kübler’s edition⁹ incorporates all the new fragments and has a full and valuable apparatus of parallel passages, but owing to the insufficiency of its critical apparatus it is scarcely an advance on Krüger and Stude­mund. F. Kniep’s uncompleted edition¹⁰ must be condemned, its text being on wrong principles and its commentary too capricious and un­satisfactory. Bizoukides’s edition" is valuable on account of its notes and other accessories, but its text is open to objection.

The other modem editions¹¹ are simply school-books. The student should accustom himself to checking the text by continual reference at least to Krüger and Studemund, and, where possible, to Studemund’s Apographum or, when the occasion arises, to the photographs of the new fragments. Zanzucchi’s Vocabolario di Gaio¹³ is indispensable.

¹ Bull, i (N.S., 1935), 571 ff.; Collinet, RH xiii (1934), 96 ff.; Levy, Z liv (1934), 258 ff.; Monier, Les.Nouveaux Fragments des Institutes de Gaius, 1935; Zulueta, JRS xxiv (1934), 168 ff.; xxv. 21 ff.; xxvi. 174 ff.; Albertario, Studi, v. 463.

² Rudorff, Die lexicalen Exzerpte aus d. Inst, des Gaius, Abh. Berlin. Ak. 1865 (1866). Bizoukides, ii. 195 ff., gives the passages.

³ Below, p. 167. ⁴ Below, p. 180. · Below, p. 30z.

⁶ Below, p. 302. ⁷ Below, p. 305.

^I Vol. i, ed. 6,1912.

⁹ Teubner, ed. 7,1935.

¹⁰ Gai instituiionum commentarius primus, Jena, 1911; comm, secundus, ss. 1-96, 1912; comm, secundus, ss. 97-289, 1913; comm, tertius, ss. 1-87,1914; comm, tertius, ss. 88-225, 1917.

^II Gaius, I. Prolegomena and Text of the Institutes, Salonica, 1937; II. Adnotationes. Indices, 1938; III. 1 Fragmenta Gaiana (Gaius ad edictum provinciale), 1939;

III. 2 Fragmenta Gaiana, 1939.

¹² English ed.: Gai Inst, with a translation and commentary by the late E. Poste, ed. 4, E. Whittuck, 1904,1925. Supplements to same: De Zulueta, 1935. French ed.: in Girard's Textes de dr. rom., ed. 6,1937, by F. Senn. Italian ed.: in FIRA ii (1940).

¹³ Milan, n.d. Elvers’s Promptuarium Gaianum (Gottingen, 1824) is out of date. Bizoukides’s edition is to include a vocabulary.

B. The Res cottidianae.¹ Our knowledge of this second elemen­tary work ascribed to Gaius is much smaller, depending entirely on the excerpts taken from it by the compilers of the Digest and Institutes. In the copies used by them the work was entitled Gai rerum cottidianarum sive aureorum libri vii, which was naturally shortened in citations? The description aurea (dicta, no doubt) must have been coined by some editor ; it cannot come from Gaius himself.³ Res cottidianae does not, it seems, mean ‘law in everyday life’ ;⁴ cottidianus means here rather‘‘usual’, ‘familiar’, ‘common’,⁵ and res cottidianae are thus the elementary topics of the traditional academic curriculum. The work comprises what is obviously a version or elaboration of Gaius’ Institutes, though the order of topics is altered. Our fragments show that book 1 dealt with manumission, from which it may be inferred that it covered much the same ground as book 1 of the Institutes, namely persons in potestate manu mancipiove and tutela. Book 2 began, like Inst. 2, with property, but went on to contracts. Book 3 began with delicts (Inst. 3. 88-225) >' inheritance (Inst. 2. 97-3. 87) must have followed. The text of the Institutes is sometimes reduced, some­times amplified; thus, under the real contract, commodatum, depositum, and pignus, which, as already observed,⁶ are absent from the Institutes, follow mutuum.⁷ It may be that the whole work, by which we mean the first three books,⁸ belongs to the beginning of the post-classical period (the end of the third cen­tury),⁹ but it may be that it represents lecture-notes¹⁰ composed, if not by Gaius himself, at any rate by some law teacher of the classical period, though not by a jurist of the front rank, and in places revised in post-classical times.” Our evidence makes it impossible to be more precise.

* On what follows : Goschen, Z./. Geschieht!.. R.W. i (1815), 54 ff. ; Buchholtz, Hugos

Civilist. Magazin, vi (1830), 228 ff. ; Dirksen, Hinterlassene Sehr. ii. 397 ; Krüger, Quellen, 203,210 ; Albertario, Rend. Lomb. lix (1926), 409 ff. (Studi, iii. 95 ff.) ; Arangio- Ruiz, Mèi. Cornil, i (1926), 93; St. Bonfante, i (1930), 493 ff., 509; Felgenträger, Symb. Frib. 365; Lenel, Pal. i. 251. ¹ See Note Z, p. 341.

³ The verses containing the Pythagorean ethics were called χρυσά ίπη by Jamblichus. Cf. Anthol. lyr., ed. Diehl (ed. 2), ii. 87; Schmid-Stählin, Gesch. d. griech. Lit. i (1929), 741. Lucret. 3. 12 : aurea dicta (Epicuri). Thes. 1. 1491. 61 f.

* Still held by Kniep, Der Rechtsgelekrte Gaius, 104. But the specificalo of another’s

materials, or writing and painting on his property (D. 4r. 1.7.7 ; 9.1), were not daily occurrences even at Rome. ⁵ Thes. 4. 1090, 48 f.

⁶ Above, p. 162. ⁷ D. (44. 7) r.

* On books 4-7 see immediately below.

* And in the western Empire, not, as is often hastily assumed, at Byzantium. See above, p. 142.

³⁰ Justinian calls the work commentarli. See above, p. 160, n. 1.

³³ In the western Empire.

We must explain why so far we have accounted for only three of the seven books of the Res cottidianae. Book 2 cannot like Inst. 2 have reached the law of inheritance, for it contained, besides the law of pro­perty (at somewhat greater length than in Inst. 2), the law of contracts. Delicts must have come next; all the Digest fragments from book 3 are on that subject. But delicts cannot have occupied the whole of book 3; hence at least part of the law of inheritance must have been in that book, after the law of delicts, in contrast to the order of the Institutes. Now part of book 3 and the whole of books 4-7 would be far too much for inheritance and actions, assuming the exposition to have been on the same scale as in the rest of the work. One is thus driven to the con­clusion, already suggested by Mommsen,¹ that the Res cottidianae were a composite work, of which the first three books were a version of Gaius’ Institutes carried up to the end of the law of inheritance, but omitting the law of actions. The remaining four books were just Gaius' Institutes. The Digest excerpts from books 1-3 are inscribed as coming from this composite work, but those from books 4-7 are inscribed as coming from the Institutes, because copies of that work independent of the composite work were still in use in the law schools and were in the hands of the compilers.

Further research into the questions of origin and transmission must take the form of a critical examination of the fragments considered as a whole; the study of individual passages can throw no light. Three problems must be distinguished more clearly than they have been hitherto, namely (1) how much of our texts is due to the compilers, (2) how much is in all probability post-classical (western) work, and (3) what passages can, in spite of their falling below the highest classical standards, be supposed to represent the lecture-notes of a classical law teacher of the second rank ?

4.Pomjtonius’² elementary work, called by the compilers briefly Enchiridion (the exact title is unknown),³ is absolutely unique. Our knowledge of it depends entirely on the few excerpts in the Digest. It appears to have been in only two books and to have given a sketch, necessarily summary and elementary, of private law. Pomponius evidently shared with his contemporary and fellow law teacher Gaius⁴ an interest in legal history, for the

¹ In Kruger’s edition of Gaius, p. Ixviii; but the view cannot be accepted as there expressed.

² Osann, Pomponii de origine iuris fragm. (1848); Sanio, Varroniana in den Schriften d. rom. Juristen (1867); Schulin, Ad Pandectarum tit. de orig. iur. comment. (1876); Joers, 8 ff.; Lenel, Pal. ii. 44; Ebrard, Z xlvi (1925), 117 ff.; Felgentrager, Symh. Frib. 369.

’ The full title must have been enchiridion iuris civilis or the like. On enchiridion as a book-title (i.e. brevis libellus, manuals) see Thes. 5. 2. 557. 52; Liddell and Scott, s.v. exx«p/a>; Geld.praef. s. 7. ⁴ Above, p. 134. work began with an historical introduction in three parts: I. De origine el process» iuris,¹ on the sources from prehistoric times to that of Pomponius, II. De magistratuum nominibus et origine,² and III. De auctorum succession#,³ a survey of the jurists from the beginning of the Republic up to Julian, excluding mere orators and jurists whose literary activity had been confined to ius sacrum and ius publicum. Literary precursors of parts I and II are the republican works on public law, Varro’s antiquarian works, in particular his lost Libri xv de iure civili* and Capito’s Conie- ctanea.⁵ Part III belongs to another literary genus, that of the SiaSo^ai, which were lists of the heads and members of the various schools, particularly the philosophical, accompanied by more or less extensive accounts of their lives and writings.⁶ Of this type of literature the outstanding Roman examples are Cicero’s Brutus, which gives a survey of the Roman orators, and Suetonius’ De rhetoribus et grammaiicis ;⁷ similar sketches are given by Vitru­vius, Cornelius Celsus, and Quintilian.⁸ No exact opinion can be formed of Pomponius’ sources of information. Possibly, indeed probably, he drew on Varro’s De iure civilidirectly or indirectly he also used Cicero’s Brutus, De oratore, and De re publica.¹⁰ This historical introduction belongs to a literary genus previously unknown in Roman legal literature, Roman legal science being held not to include legal history.¹¹ It is therefore written not in the true juristic style affected by the classical jurists,¹² but rather in that of Cicero and of historical works,¹³ especially those on literary history.¹⁴

* D. (1. 2) 2 pr. 13. ¹ D. (1.2) 2.13.

³ D. (1. 2) 2. 13, 35 f. * Schanz-Hosius, i, ss. 182 ff., 189.

⁵ Above, p. 158; below, p. 227.

⁶ Schmid-Stahlin, Gesch. d. griech. Lit. i (1929), 724; F. Leo, Die griech.-rdm. Biographic (1901), 35 ff.; P. Oxy. x. 1241 (list of Alexandrian librarians); Schubart, Papyruskunde, 168.

⁷ In general the literature De illustribus viris: Schanz-Hosius, i, s. 126; Leo, op. cit. ii ff.

* Vitruv. 2. 1. 5 f.; Celsus, De medicina, praef.; Quint. Inst. 3. 1.

* So Sanio, Varroniana, but his arguments are not convincing. But more recently Taubler, Vntersuch. z. Gesch. des Decemvirats u. Zwolftajeln (Hist. St. herausg. v. Ebering, 148,1921), p. 40, takes the same view.

¹⁰ A careful comparison of Cic. De re pub. 2. 36. 61 with Pomp. D. (1. 2) 2. 24 shows that Pomponius did not obtain the idea of making the power of legislation a part of the summum imperium from Cicero: Taubler (op. cit., last note) 40.

¹¹ Above, p. 134. ¹¹ Beseler, SD i (1935), 280.

¹³ D. (1. 2) 2. 24: ‘... captumque amore virginis omne fas ac nefas miscuisse....’ Again, s. 46: ‘Tubero, qui Ofilio operam dedit’ compared with Cic. Brut. 42. 154: ‘Cumque discendi causa duobus peritissimis operam dedisset, L. Lucilio Balbo C. Aquilio Gallo....’

¹⁴ The nearest parallel is the style of Suetonius, De rhet. et gramm. Leo, op. cit. 14,

After Pomponius, perhaps before the end of the classical-period, this already brief work was epitomized in a single book; later still this epitome was enlarged, perhaps from the original complete work. The compilers possessed both the original and the epitome,¹ and took a few excerpts from each.^{[266] [267] Unfortunately it was from the epitome that they took the historical introduction; the fuller version probably seemed too prolix. The extensive fragment in the second title of the Digest (D. 1. 2, De origine iuris et omnium magistratuum et successione prudentium, L. 2) is one of the most corrupt texts in the Digest, being full of scribal errors, careless abbreviations, and intruded glosses.[268] The chief blame perhaps lies with the author of the epitome, since the compilers, though they may have made matters worse by further abridgements, cannot conceivably be responsible for the additions which dis­figure the text. Clearly the copyists of classical or post-classical times took very little trouble about thè text of this historical part, which for jurists was an άλλότριον.}

Pomponius found no imitators, no one even to carry his list of jurists beyond Julian. Yet this excerpt, for all its faults, remains of great value, provided that one bears these faults constantly in mind.^{[269] [270] Much of it is based on excellent sources. One writer has indeed denied that Pomponius wrote the Enchiridion or, at any rate, its historical introduction ; he would attribute the work to some post-classical law teacher.3 But this view (a veritable will 0’ the wisp)[271] is disposed of by the simple fact that the list of jurists ends with Julian, that is at Pomponius’ own date. A post- classical writer would have been bound to make some mention of Gaius, Scaevola, Papinian, Paul, and Ulpian, and the compilers would have been only too pleased to preserve any information about the men from whose works the bulk of the Digest extracts are derived. The conclusion is inescapable that our fragments, corrupt as they are, must come ultimately from Pomponius. Of the merits of his original work we have not the materials for judging, but so· industrious a man as Pomponius cannot have produced anything thoroughly worthless. It was in any case a pioneer work in juristic literature, but it remained an isolated phenomenon, legal history being a meal that Roman jurisprudence could not swallow.1}

The Liber singularis enchiridii must be an epitome made by a post-classical anonymus?· Not that it never happened in antiquity that an author abridged or expanded his own earlier work, but simply because the work is too bad to be either an epitome executed by Pomponius or a first edition of the book. Moreover, it is hardly conceivable that he should have abridged what was already a very brief outline. It is, however, possible that the libri duo enchiridii were, like the liber singularis, only an epitome of a larger work, and that Pomponius’ original Enchiridion was in more than two books; if so, it did not reach the compilers.

A fresh edition and a critical analysis of this interesting fragment (D. 1. 2. 2) would be welcome. The edition could aim only at establish­ing the text which the compilers had before them, obvious scribal errors being corrected; the text of the fuller original is no longer recoverable. The aim of the critical analysis would be to distinguish the strata of the text,³ and to determine what sources were directly or indirectly used and (equally important) what sources were not used.⁴ In dealing with the questions of authenticity and interpolation the fashionable linguistic tests will be out of place, because this historical opusculum does not belong to the juristic literary genus to which these tests are applicable: to each genus its own style.³

5.Coming to the Severan period, we must mention first three short works, the Institutiones of Paul, Callistratus, and Ulpian, in two, three, and two books respectively. The authenticity of the Institutiones of Paul and Ulpian may well be doubted, but our evidence does not enable the doubts to be resolved. It is certainly difficult to believe that these two great and high-placed jurists wrote short and necessarily very elementary school-books of this character, or that, if they did so, their books would not

¹ Above, p. 134.

¹ Pemice (at p. 15 of a lecture programme, summer 1899): ‘The sections are of various value; the second contains a number of gross errors along with undoubtedly correct information. Implicit trust therefore cannot be placed in the information given by the other sections, where we are not in a position in all cases to check it. The exposition is often, though by no means always, confused and in very bad Latin. The fragment is perhaps only a defective extract from a work by Pomponius.’

’ An example above, p. 115.

⁴ See, e.g., Taubler, op. cit. 40.

» Above, p. 169.

have displaced Gaius’ out-of-date and defective work in the law schools.¹

Paul’s Institutiones. Our remains consist of (i) excerpts, of which there are three in the Digest ;* any that there may be in Justinian’s Institutes are unidentifiable; (2) a small fragment in Boethius’ commen­tary on Cicero’s Topica; and (3) three fragments in an unpublished com­mentary on Cicero’s De inventione, from which perhaps more may prove to be obtainable.³

Of Callistratus’ Institutiones we have only the five short excerpts in the Digest.⁴

Of Ulpian’s⁵ we have (1) a few fragments in the Digest, some of which recur in the Institutes; (2) those preserved in Coll. 16. 5-9; (3) Vienna fragments of a fifth- or sixth-century parchment codex ;⁶ and (4) a frag­ment in Boethius’ above-mentioned commentary, which, however, does not preserve the original wording.

6.The Institutiones of Marcian⁷ are a strange work. We know it only from the fragments incorporated in the Digest and Insti­tutes, but these are sufficiently numerous to allow of a clear picture of the work being drawn. It embodies two distinct writings: Marcian’s uncompleted preparatory studies for a book of Institutes, and his preparatory studies for a systematic Digest. They were composed under Caracalla or shortly after his death ;⁸ the date of publication (not, of course, by Marcian himself) is unknown. The work is never cited, not even by Ulpian.’ It is characterized by the abundance of its citations of imperial rescripts, showing that its author must have had access to the imperial archives.

¹ The conjecture that Ulpian’s Institutes are a later compilation from other works was already advanced by Hugo, Lehrb. d. Gesch. d. rom. Rechts (ed. rr, 1832), ii. 849. This would mean western post-classical work, not Byzantine, as Steinwenter, St. Bonfante, ii. 432 ff., and others assume. See further Solazzi, ‘ Glosse a Gaio I St. Riccobono, i. 93. ¹ Pal. i. 111-14; Berger, PW x. 726.

³ The fragment from Boethius will be found in Collect, libr. iuris anteiust. ii. 160; the same and no. 3 in Seckel-Kübler, ii. 162, and in Girard-Senn, Textes, 453; also in FIRA ii (1940), 421. On the Digest passages: Index Interp.

⁴ Lenel, Pal. i. 97; v. Kotz, PW Suppl. iii. 228.

⁵ Pal. ii. 928; Joers, PW v. 1447; Krüger, Kritische Versuche (1870), should be rejected.

⁶ These can be found in the collections cited in the last note but two. Facsimile in Krüger, Krit. Versuche. On the Digest passages: Index Interp.; also Steinwenter, St. Bonfante, ii. 433. On the Fragm. Vindobon.: Solazzi, St. Bonfante, i. 93.

⁷ Lenel, Pal. i. 652 ff.; Pemice, Festg.f. Dernburg (1900), 3 ff.; Ferrini, ii. 277 ff., 285 ff.; Buckland, St. Riccobono, i. 275 ff.; Krüger, Quellen, 251; Joers, PW i. 524.

⁸ Fitting, Alter u. Folge, 122. The rescript in D. (37.14) 5.1 may be a later addition. * Marcian’s Institutes seem to be cited by Paul, D. (j. 9) 8 pr., and by Ulpian,

D. (28.1) 5. But either these citations have been added later, or the name is corrupt. Krüger, 251; Fitting, Alter, rr8; Ferrini, ii. 317, n. 1; Albertario, Bull, xxxiii (1923), 26, n. 1.

Books I and 2 dealt with the sources and the law of persons, including marriage and tutela; book 3 with the law of property; the next six books gave a relatively full exposition of the law of inheritance; books 10-14 treated of various leges and the iudicia publica, precisely like the second part of the so-called Digesta, the system of which we shall explain below. The contents of the remaining books 15 and 16 cannot be determined. There is no trace of any treatment of the law of obligations. It is clear that this opus is neither ‘something halfway between a compendium for novices and a commentary’¹ nor ‘a book which junior imperial officials might learn their law from and use for reference’,² but is just a literary monstrosity. If we are right to credit the classical writers with a sense of language and style, we can safely claim that their sense of form would not have permitted the publication of such a work as this; in fact, Marcian himself cannot have published it. Books 1-9 seem clearly to have been composed in preparation for an Institutiones; it is unfinished work, since the law of obligations is absent and the treatment of the law of inheritance is far too long and would have had to be compressed later. Books 10-14 ^3X6 beyond doubt the second part of a systematic Digesta, unaccompanied by the first part.

‘The style of the exposition is always elegant and sometimes elevated; the composition strikes me as exceptionally pure and carefully executed’: so wrote Pemice (1900). Only a close study of the whole of the remains could decide whether this verdict is justified.

(v)

Closely akin to the isagogic works we have just described is a group of writings entitled regulae, deftnitiones, differentiae, or sententiae. In juristic usage regulae and definitiones are synony­mous, meaning, like όροι,³ abstract statements of law, juristic principles, in antithesis to case law.⁴ Differentia means the same as διαιρί'σις.⁵ By sententiae, as the title of a book, it is indicated that its contents are in the nature of maxims.⁶ The common

’ So Krüger, 251, a thoroughly opportunist way out of the difficulty.

¹ So Pemice, Festg. f. Dernburg, 3 ff. Ferrini’s conjecture, that we are dealing with a work intended for jurists in the provinces, is unproven and unprovable.

³ See, e.g., the pseudo-Platonic 5/»», Plat. Opp., ed. Burnet, vol. v; Pseudo-Galen. (ed. Kühn), xix. 346: όροι Ιατρικοί; v. Arnim, Fragm. Stoic. Veter, i. 140.8; ii. 8.33-9. 6; iii. 247. 28, on book-titles nepl όρων. We have already (above, p. 94) spoken of the liber sing, όρων doubtfully attributed to Q. Mucius.

⁴ Krüger, Quellen, 141, n. ii, and above, p. 66.

⁵ Book-titles irepl Siaipioauv, see v. Arnim, Fragm. Stoic. Veter, ii. 9, and above, p. 63. On rhetor, differentiae Goetz, PW v. 481.

⁶ Ad Herenn. 4.17. 24; Quint. 8. 5.1; Isid. Etym. 2. it. Cf., e.g., Appius Claudius’ Sententiae (Schanz-Hosius, i, s. 20), Cato the Elder’s Ad filium (Schanz-Hosius, i, s. 66), and the sententiae Varronis (Schanz-Hosius, i, s. 194, p. 577). characteristic of works of this group is that they contain abstract statements of principle, approximating at times to the rules of school grammar. Case law is absent or very much in the back­ground; they are not commentaries; they embrace the whole field of private law, in part also that of criminal law. Many, but not all, of them are isagogic in character and resemble the Institu­tiones in form and contents. Not all of them are authentic, for the post-classical age sorely needed works of this kind and in some cases manufactured them out of classical materials and labelled them with the names of classical authors. But so little remains of many of these writings that conclusions cannot be drawn with certainty.¹

1. Neratius Priscus, Regularum libri xv.^2, The Digest yields only seven fragments. The work seems authentic, but the copy used by the compilers contained post-classical glosses.

2. Pomponius, Regularum liber singularis. Authenticity doubtful. One passage of the Regulae is taken almost word for word from Pom­ponius Ad Plautium (D. 44. 7. 24 compared with 12.1.12).^{[272] [273] [274] [275] It may be that the Regulae were collected from Pomponius’ writings by a later writer. The fact that the work contains notae by Marcellus does not exclude this possibility, for these notae may also have been inserted later from Marcellus’ writings. Our remains are unfortunately no more than the seven fragments in the Digest*}

3. Gaius, Regularum libri Hi, and Regularum liber singularis. The compilers, who possessed both works,^[276] took two fragments from the first and only one from the second. We find neither anywhere else; they were not used for the l. Romana Burgundionum.^[277] Their authenticity is doubtful, especially that of the liber singularis. The single fragment from it (D. i. 7. 21) is certainly spurious, though not invented by the compilers.^[278]

4. Cervidius Scaevola, Regularum libri iv. Only the Digest extracts survive. If the work is authentic, it seems to contain later additions. Did any classical liber regularum really contain the text of the I. Itdia maiestatisl¹

5.The so-called Fragmentum Dositheanum.²· In a late Roman collection of passages for translation from Latin into Greek and from Greek into Latin we find amongst other things a juristic passage dealing with manwmissio. The Latin text, as commonly in such cases,³ is written one word, below the other, in a vertical column, and the Greek translation by its side in a similar column. The Latin is in part not the original text but a retranslation from the Greek. The collection has nothing to do with the fourth­century grammarian, magister Dositheus, and the traditional name of the fragment is therefore misleading. The Latin text is very corrupt and its reconstruction uncertain. The anonymous collector drew on some classical elementary work (regulae* or institutiones) or on an epitome of such a work; the classical work itself seems to have come from as early as the second century; its text had already been corrupted by glosses.⁵ The modem juristic editions of this fragment⁶ are unsatisfactory because they do not give the Greek text.⁷

6. Papinianus, Definitionum libri ii.⁸ The work is known only from excerpts in the Digest and a passage of the Collatio, the latter a later addition.⁹ If Papinian really wrote a book of Definitiones, it has been defaced by a later hand, for many of the passages exhibit the same bombastic, rhetorical, and thoroughly unlawyer­like style which we find repeatedly in Papinian’s Quaestiones,¹⁰ a style for which neither Papinian nor Justinian’s compilers can be responsible. The post-classical parentage of the work is shown by other signs also.¹¹

’ D. (48.4) 4 pr. ³ Goetz, PW v. 1606; Joers, PW v. 1603.

³ Vergil also underwent similar revision: Publications de la Soc. Fouad I de

Papyrologie. Textes et documents, III. Les Papyrus Fouad I (1939), no. 5, and similar texts, p. 8. ♦ Fr. Dos. 3: ‘ regulas enirn exsequenti mihi....’

’ Schulz, Z xlviii (1928), 283.

⁴ Kruger, Collect. Libr. ii. 149; Seckel-Kubler, i. 419; Girard-Senn, Textes, 505; FIRA ii (1940), 615.

⁷ The Greek text is in Bocking’s, &c., Corpus iur. Rom. anteiust. i (1841), 213, and in Gotz, Corpus glossariorum lat. iii (1882).

⁸ Costa, Papiniano, i (1894), 233; Joers, PW i. 574.

• Coll. 2. 3; Schulz, AC I Roma, ii. 13. ¹⁰ Below, p. 234.

’* (1) D. (23. 4) 27: cf. Levy, Ehescheidung (1925), 14, n. 4, contemplating, as was then the practice, only Byzantine interpolations. (2) D. 2. 15. 5: cf. Index Interp. (3) D. (23. 2) 63: the sentence quae species... est is too compressed, and the usage of potentatus unexampled in juristic Latin. Cf. Beseler, Beitr. v. 39. (4) Coll. 2. 3, the only extract in that work, is unclassical: Solazzi, * Per la data della Collatio ’, Atti Napoli, Ivii (1936), 13, n. 5 (offprint).

7. Paulus, Regularum libri vii and Regularum liber singularis.¹ Only Digest fragments. The liber singularis must be a post-classical abridge­ment of the larger work.

8. Paulus, Sententiarum ad filium libri v.² Of all the works in the present group this was the most influential. It is arranged on the system of the classical Digesta,¹ that is, in two parts, of which the first follows the Edictal order, with appendixes and excursus added in the traditional manner, and the second deals with leges, senatusconsulta, and imperial constitutions. But it was a pocket-Digesta, so short relatively to the matters treated of that, though not pronouncedly educational, it was necessarily elemen­tary and for that very reason acquired speedy popularity with practitioners.⁴ Whether Paul was its author is doubtful. The prevailing modem opinion is that he was not, but that it was composed chiefly from Paul’s writings by some post-classical jurist.® This is likely enough, but in the present state of the evidence unprovable. It is, in any case, of no great importance, because, even if Paul did write some such work, it was radically revised early in the post-classical period (in the third century), and the revised work superseded the original so completely that it alone survived. It can no longer be disputed that our present text exhibits, both in form and substance, clear signs of a post- classical, though not Byzantine or Visigothic, origin.

As early as the third century it was used and valued as a con­venient handbook by practitioners, though even then there were

¹ Beiger, PW x. 729.

¹ See in general: ibid. 731 ff. Also: Max Conrat, Der westgothische Paulus (offprint Amsterdam Ak. viii, no. 4, 1907); Beseler, Das Edictum de eo quod certo loco (1907), 2, n. i; Beitr. ³ (1910), 99; Schulz, Einführung (1916), 38; Z xxxviii (1917), 118; xliii (1922), 203; xlvii (1927), 39; Lauria, Ricerche su Pauli Sententiarum Libri (Annali Macerata, vi, 1930); Levy, Z 1 (1930), 272; Scherillo, ‘ L’ordinamento delle Sentt. di Paolo ’, St. Riccobono, ³ (1931), 41; Niedermeyer, St. Bonfante, ii. 399; AC I Roma, ³ (1934), 367; Buckland, LQR lx (1944), 361; Ixi (1945), 34; Volterra, Indice delle Glosse, &c., 1 (extr. RSDI viii, 1935); AC I 1933, Bologna, i. 35 ff.

E. Levy, Pauli Sententiae (1945).

³See below, p. 226.

⁴ The title included the words ad filium: Krüger, Collect, libr. ii. 45; their omission by the Index Flor, xxv is insignificant. Inside the books there was division into titles, with headings, and at least in some editions titles and sententiae were numbered, as we see from Consult. 6. 5 and 6.6, and the Antinoite gloss on Gaius, 3. 173: * d>s o Paul. ßißXu» ß' ò³òÕø xS cvrg v’ sententiarum ’, i.e. bk. 2, tit. 19, sent. 50.

⁵ Even the cautious Mittels wrote (Antike Rechtsgesch. u. romanistisches Rechts­studium, Vienna, 1917, offprint p. 16; Ital. trans. Riccobono, Ann. Palermo, xii (1928), 491): * To-day no one doubts that Paul’s Sententiae are the work of a late and unskilful compiler, containing a veritable flood of interpolations.’ Recently Volterra, AC I1933, Bologna, i. 161 ff., has dissented, but his arguments are unconvincing. On Paul’s supposed self-citation in the Sententiae (D. 38. 10. 9) see below, p. 254.

177 jurists who rejected it as spurious. But the central bureaucracy declared in favour of its authenticity and value. In an enactment of 327 or 328¹ Constantine extolled its clarity and legal correcti- tude and set at rest all doubt as to its authenticity and adducibility before the law courts. Moreover, the Law of Citations of 426, not content with confirming the authority of scripta universa Pauli, affirmed once more specifically the validity of the Sententiae²·— evidently for the confutation of past and present doubts. The authority of the work being thus finally established, we can see that it was widely used. It is cited by a marginal gloss in the Antinoite Gaius³ and by the interpretatio of the Codex Theodo­sianus.⁴ It was used by the Byzantines for the Digest, though not, it seems, for the Institutes.* In what had formerly been the western Empire the compilers of the Visigothic Breviarium made an epitome of it, which they appended, along with an interpretatio, to the Breviarium* In France, where the Breviarium counted as an authority, though not as statute,⁷ later editors added further passages from the complete Sententiae by inserting them both into the text of the epitome⁸ and in appendix.’ The Sententiae also contributed to the Lex Romana Burgundionum and to the Edictum Theodorici.¹⁰ Besides the epitome in the Breviarium, which goes by the name of the Visigothic Paul, there survive only the frag­ments contained in the above-mentioned collections.“

At the beginning of the twentieth century the attitude of scholars in regard to our text of the Sententiae was still highly conservative. It was not doubted that substantially it was a genuine Pauline work. However, P. Kruger’s edition of 1878 had recognized that it contained isolated Visigothic interpolations,’* a view accepted, though grudgingly, by Mommsen.¹³ In 1895

* C. Th. (1. 4) 2. Date: Levy, Z1 (1930), 293.

* C. Th. (1.4) 3, from which Conrat (Westgoth. Paulus, 7, n. 16) inferred that doubts as to the authenticity of the text existed.

’ Above, p. 176, n. 4. ⁴ Interpretatio ad C. Th. (3.13) 2.

’ Digest: Lenel, Pal. i. 1297. Institutes: Zocco-Rosa, Ann. Catania, xi/xii (1911),

279. ⁶ On the Breviarium: below, p. 302.

⁷ Conrat, Gesch. d. Quellen u. Lit. d. rom. Rechts im fruheren Mittelalter, i (1891), 41 ff. * Ibid. 143.

’ Ibid. 141; Kruger, Collect, libr. iii. 249 ff.

¹⁰ List of the passages in the Lex Burg, in De Salis’s edition, p. 169; on the Edictum.

see Bluhme’s edition. Cf. FIRA 2. 683. 713. " See Note AA, p. 341.

¹¹ Praef.42.

¹³ Strqfr. 480, n. 1; 497, n. 2; 8ox, n. 1. These passages bear out Mitteis’s account (Antihe Rechtsgesch. &c., above, p. 176, n. 5) of a conversation in which Mommsen pronounced: ‘ Interpolations in Paul’s Sentences? No, there are none.* Our citations show that Mommsen’s view was not quite so absolute, but he certainly considered the text to be in all essentials classical.

4497.1 _N

Pemice¹ stated, as a fact long since observed and no longer dis­putable, that our epitome of the Sententiae no longer recognized the formulary procedure, from which he inferred that the classical text had undergone a thorough revision; but he left all further questions at large. It was in 1907 that Max Comat propounded the correct view,^{[279] [280] that the Visigoths, though they cut down their original severely, made no other substantial change,[281] but that this original was itself ‘not infrequently’ of post-classical manu­facture. At first Conrat’s book passed unnoticed.[282] Thus Seckel and Kiibler[283] [284] were apparently unaware of it when, in their edition of 1911, they maintained that the Visigothic epitome was far more interpolated than had previously been supposed, but attri­buted the interpolations to the Visigoths. Modem criticism of the Sententiae really begins with Gerhard Beseler, whose view that the work was a post-classical collection from Paul’s opera has met with ever-growing acceptance. Again and again studies of special questions have confirmed the unclassical character of our Sen­tentiae.b Criticism has not, however, said its last word,[285] since it has not been applied to the evidence as a whole. Conrat’s book, apart from its being primarily concerned with the interpretatio, is out of date. Future inquirers will do well to leave the question of authorship aside. Whether Paul wrote Sententiae which later, in the third century, underwent a thorough revision in the western Empire, or whether the Sententiae were a compilation made in the western Empire out of Pauline materials by some post-classical writer, does not greatly matter to us.[286] Similarly, pseudo-philo­logical reconstruction of the ‘authentic’ text ought to be aban­doned.’ Either Paul wrote the original Sententiae, in which case the post-classical revision of his text has rendered it impossible to reconstruct his work with any sort of certainty, or else the Sen­tentiae were from the beginning a post-classical compilation of Pauline materials, in which case, though the attempt may reason­ably be made to reconstruct the (post-classical) text of the work,}

179 we shall not thereby recover the original Pauline texts. The following are the questions on which research should be con­centrated: (1) how serious are the Byzantine interpolations? and (2) how far was the version of the work upon which the Byzantines and Visigoths worked unclassical in form and substance, and what light do the unclassical elements in it throw on the legal history of the western or eastern Empire in the fourth and fifth centuries? From the Visigoths the.only sort of alterations that one should expect are abbreviations.

Neither P. Kriiger’s¹ nor Seckel and Kübler’s² edition is satisfactory. The edition that is needed should confine itself to presenting the text in its various phases of development: if a passage exists in more than one version, the various versions should be exhibited side by side? Nor should passages coming from Byzantine sources be omitted, since any light on the methods of the compilers is of value. Lastly, the edition must include the Visigothic interpretatio, which Krüger omitted, as, less excusably in view of the example set by Mommsen in his edition of the Codex Theodosianus, did Seckel and Kubier? What we demand from an edition is a clear conspectus of the textual tradition. All further questions can be left to subsequent criticism, including such reconstruc­tion as may be possible of the text upon which the Visigoths and Byzan­tines worked, in other words the text current in the fourth and fifth centuries.

9. Paulus, Manualium libri Hi.⁵ This work closely resembled the Sententiae; like it, it was a pocket Digesta, only sensibly shorter—by two books. Part 1 was based on the Edictal order; part 2 dealt with individual leges, senatusconsulta, and imperial constitutions. Fortunately fifteen fragments of it are preserved in the Fragmenta Vaticana,⁶ independently of those in the Digest, and these show that by the date of the composition of the F.V. numerous unclassical elements had been added to the work. The basis, however, is classical. Whether Paul never wrote Manualia, but the work is just a post-classical compilation made from Paul’s works,⁷ or whether an authentic Manualia by Paul was merely revised in post-classical times, can be decided no more than the similar question as to the Sententiae, and is of equally little interest.

¹ Collect, libr. 2. ² Jurispr. anteiust. 2.

³ Mommsen’s edition of the Theodosianus is a model, but it has not yet been followed by the lawyers.

⁴ At present the interpretatio must be sought in Haenel’s old and clumsy edition of the Breviarium: Lex Romana Visigotharum (1849).

³ Berger, PW x. 726. On manuale as the title of a book see Thes. 8. 335; Beseler, Bull, xlv (1938), 184. ⁶ F.V. 45-58.

⁷ Beseler, l.c., thinks that the libri manuales were a post-classical epitome of the libri ad edictum.

Very clear examples of post-classical revision can be seen in F.V.¹ They afford an instructive parallel to the Senteniiae.

10. Licinnius Rufinus, Regularum libri xii or õ³³³.^ã We possess only a few Digest fragments, the unclassical elements in which may well be due to the Byzantines.³

11. Ulpianus, Regularum libri vii* and Regularum liber singu­laris. The few Digest fragments,⁵ which are all that we possess of the libri vii, do not enable a judgment of the work as a whole to be formed. But the famous definition,⁶ 'Lata culpa est nimia neglegentia, id est non intellegere quod omnes intellegunt’, can hardly have been devised by the compilers of the Digest. Since it is certainly unclassical,⁷ it proves that our text of this work contains post-classical elements.

The liber singularis.⁸ More is known of this: a Vatican manu­script preserves a large portion of an epitome of it made in the first half of the fourth century. It is a mere epitome which abbreviates, but does not otherwise alter, the text ; consequently it gives a clear picture of the general scheme of the liber singular is, which was els follows:

Introduction: The sources of law.

I.Law of persons: (i) Manumission. (2) Persons in potestate manu

mancipio. The discussion of the law of marriage included the topics of dos and donatio inter virum et uxorem. (3) Tutela and cura. (4) Lex dulia et Papia Poppaea.

II. Law of property: (1) Kinds of things. (2) Ownership.

¹ A few notes must suffice here. (1) F.V. 47 [non-transferri], (2) F.V. 47a [nee-

tradatur], (3) F.V. 50. See Riccobono, St. Peroni, 367, n. 1 ; Ann. Palermo, xii (1928), 569. (4) F.V. 55. See Index Interp. ad D. (45. 3) 26; Beseler, T x (1930), 234; Bull, xlv (1938), 184. (s)F.V.57[si-legatis], (6) F.V. 49. See Riccobono, St. Peroni, 3⁶7·..

² Index Flor.: ‘XXXV. regularion βιβλία δ.’ But the inscription of D. (42.1) 34 mentions a book 13.

³ D. (5. 1) 38 for certain, 40. 5. 16 probably—see Index Interp.

♦ Joers, PW v. 1448.

⁵ Whether they were used by the compilers of the Institutes is doubtful ; below, p. 305·

⁴ D. (50. 16) 213. 2.

⁷ Kunkel, Z xlv (1925), 315; Ròm. Rechi, s. no, 3; Lenel, Z xxxviii (1917), 288; Mitteis, RP i. 333; Arangio-Ruiz, Responsabilità contrattuale (1933), 251 ff.

* Mommsen, Schr. ii. 47 ff. ; Conrat, Gesch. d. Quellen, &c., ³ (1891), 85 ; Pal. ii. 1016; Joers, PW v. 1448; Arangio-Ruiz, Bull, xxx (1920/1), 178ff.; xxxv (1927), 19ã ff. ; PSI 1182, p. 34 (Frammenti di Gaio) ; Rev. Al Qanoun wal Iqtisad, iv (1934), 65 ff.; Albertario, Bull, xxxii (1922), 73 ff. (Studi, v. 493 ff.); Buckland, LQR xxxviii (1922), 38ff.; xl (1924), 185; liii (1937), 508 ff.; Niedermeyer, Z xlvi (1926), 486; AttiCongr. internai. 1933, Roma, i. 369; Schulz, Die Epitome Ulpiani des Cod. Vat. Reg. 1128 (1926) ; Lenel, Z xlvii (1927), 414; Felgentràger, Symb. Frib. 372 ; Volterra, Indice delle glosse, &c., ii, RSDI viii (1935).

181

III. Law of inheritance: (i) Wills, (a) Hereditas and bonorum pos sessio. (b) Legata and fideicommissa. (2) Intestacy.¹

IV. Law of obligations?

V. Law of actions?

The author of the liber singularis was certainly not Ulpian but an unknown lawyer of the third or the beginning of the fourth century. His main source was the Institutes of Gaius in the form in which they have reached us.⁴ He used other classical sources also, but our evidence only seldom enables us to identify them? We cannot say whether the epitome superseded the full liber singularis, nor whether the three fragments in the Collatio and the two in the Digest? come from the full work or only from the epitome. The Institutes of Justinian appear to have made use of neither work.⁷

The epitome, which at my suggestion has come to be known as Epi­tome Ulpiani, figures in the unique manuscript. Vat. Reg. 1128,⁸ as an appendix to the Visigothic Breviarium. Whoever first made this con­nexion possessed only an incomplete copy of the Epitome, from which the beginning, the law of obligations, and the law of actions were missing. What we have is thus only a fragment of the Epitome. It bears no title, but at the beginning is a list of the chapter-titles, with the superscrip­tion: Incipiuni tituli ex cor pore Ulpiani, which means 'titles from the following work by Ulpian’, corpus here having the literary sense of the whole as opposed to its parts? The title of the Epitome cannot have been Corpus Ulpiani, still less Tituli ex corpore Ulpiani. Whoever first connected the Epitome with the Breviarium knew that it was ascribed to Ulpian, but either did not know, or did not trouble to give, its title. In any case, this Epitome is an extract from the Liber singularis regula­rum which was attributed to Ulpian. Of the five fragments (three in the Collatio, two in the Digest) which, according to their inscriptions, are taken from the liber singularis, three recur in our Epitome; the other two, being concerned with the law of obligations and the law of actions, do not recur in the Epitome, which is deficient in these topics.

The primary source of the liber singularis was the Institutes of Gaius, whose order of topics it follows, though with occasional alterations of

¹ The Cod. Vat. ends here. ² Coll. 2. 2 is from the section on delicts.

³D- (44· 7) 25·

⁴ Arangio-Ruiz dissents, still holding that the version of the Institutes used was a second edition prepared by Gaius himself. Such an edition never existed; there was only one version, which became stabilized relatively early (above, p. 164). A second edition would have superseded the first, as always happens with works serving a practical purpose. Only bibliophiles preserve the first edition.

⁵ Schulz, Epit. Ulp. p. 17. ⁶ Coll. 2. 2; 6. 2; 16. 4; D. (22. 5) 17; (44. 7) 25.

⁷ The cases where Justinian’s Inst, agree with Epit. Ulp. against Gaius can be otherwise explained. · Description and history: Schulz, Epit. Ulp. 1 ff.

» Schulz, p. 20.

detail; also, from other sources, it inserts various topics omitted by the Institutes, such as dos and the Lex lulia et Papia. The text is at times taken word for word from the Institutes, and that slavishly, with all the faults of the original; but frequently it is recast: there are not very happy condensations and purposeless paraphrases. One cannot credit Ulpian with so inferior a revision of Gaius’ text.¹ He was not the author of the Liber singularis.

The standard edition is still that given by P. Kruger in the second volume of the Collect, libr. anteiust. (1878), where alone the readings of the manuscript are fully registered. The editions in the other collections are suitable only for scholastic use. Schulz’s edition shows in detail that the author cannot be Ulpian. In any future edition the process of de­formation by the post-classical author ought to be exhibited by con­tinuous juxtaposition of the independent texts.

12. Ulpianus, Opinionum libri vi.^z This work also is wrongly ascribed to Ulpian. Its form and contents point to a post-classical author,³ who no doubt worked on Ulpian’s libri ad edictum. The order is obviously that of the Edict. Book 6 deals with Tit. xv of the Edict (in Lenel’s reconstruction), which is reached by book 24 of Ulpian’s commentary on the Edict. Justinian’s compilers possessed no more than the first six books of the Opiniones: either its unknown author never finished the work or the later part had been lost. Assuming the work to have been finished, it must have comprised at least twenty books, and more than that if it was a Digesta.

13.Marcianus, Regularum libri v. We possess only the Digest frag­ments, but these are fairly numerous. Although a comprehensive study of them has still to be made, it is already clear that the text used by the compilers contained post-classical passages.⁴

14. Modestinus, Regularum libri x.^s Here again we possess only the Digest fragments, and these need a comprehensive critical study. The edition used by the compilers contained an interpolated text.⁶

¹ Schulz, pp. 12 ff. Volterra’s Indice, ii (above, p. 180, n. 8), cites the literature on these faults at the individual passages.

² Joers, PW v. 1450; Rotondi, Scritti, i. 453, giving a full vocabulary.

³ Already recognized by Jacob. Gothofredus, Novus in tit. Pandect, de diversis regulis iuris antiquis commentarius (Genevae, 1553) ad 1. 61, p. 259. Now the generally accepted view: Lenel, Pal. ii. 1001; G. Rotondi, Scritti, i. 453; Felgentrager, Symb. Frib. 371.

⁴ D. (15.1) 40. [Quomodo... nascitur peculium.] Beseler’s criticism, Z liii (1933),

25, goes much too far. ³ Brassloff, PW viii. 670.

⁶ Did Ulpian’s pupil really write ‘Licet “capitalis" Latine loquentibus (!) omnis causa existimationis videatur’ (D 50. 16. 103)? Cf. Levy, Die rom. Kapital­strafe (Heidelb. SB, 1930/1), 44. It was not the compilers who wrote Latine loquentibus. See further, Beseler, Z lii (1932), 61; St. Albertoni, i. 436; Solazzi, Rend. Lomb. Ixix (1936), 986, n.; Bull, xliv (1937), 4°4> n. 1.

15.Modestinus, Differentiarum libri ix. De differentia dotis Uber sin- gularis.¹ Of the libri ix we fortunately possess two fragments in the CoUatio¹ in addition to those in the Digest. The latter contain much that is objectionable from the linguistic and substantial points of view,¹ but the fragments in the Collatio also show undeniable post-classical characteristics.⁴ Hence, either the Differentiae were not written by Ulpian’s famous pupil, or else his authentic work has been thoroughly overlaid by later work.

Of the liber singularis all that we possess is one Digest passage,⁵ which is certainly unclassical, but hardly Byzantine.

(vi)

The most important category of legal literature in the classical period is the commentary, including under that term the com­menting epitome. The form taken by the commentary in non- juristic literature during the first three centuries of our era is well known.⁶ Regularly, one may perhaps say universally, it is that of the lemmatic commentary. The text commented on and the commentary are separate works, written on separate rolls, and the reader of the commentary is informed of what particular passage of the text is being commented on by means of lemmata, that is to say by words of the text being used as headings or captions. The lemma may be the passage in question, or its initial words, and it is made easy to find by being written outside the text (åê#ñSt. Riccobono, i. 305; Z liii (1933), 48.

* On Coll. 1.12 see Beseler, T x (1930), 208; Z li (1931), 198. Coll. 10. 2 is a passage of ill repute which has frequently been dealt with in modem discussion. See Volterra’s Indice. ^s D. (23. 3) 13. See Index Interp.

⁶ On what follows see especially G. Zuntz, ‘Die Aristophanes-Scholien der Papyri ’, Byzantion, xiii (1938), 631 ff.; xiv (1939), 545 ff. Faulhaber, BZ xviii. (1909) 383. ’See Note BB, p. 341.

⁸ P. Fay. 3 contains an edition of Aristotle’s Topica with a paraphrase following each paragraph of the text: Cronert, AP ii (1903), 367. See further Galen’s commen­taries on Hippocrates, e.g. Corp. Med. Grate, v. 10. 1 (1934), 178.

The marginal commentary, on the other hand, was as yet unknown. The margins were used for noting variant readings and for signifying by means of symbols doubts as to the soundness of the text—in short for what we should call a critical apparatus ; also, of course, for readers’ own notes, but these belong to another category, since they are not part of the original' copying. Unknown in the classical period were the extensive catenae of stereotyped marginal glosses which we find in the Byzantine period.

The epitome is found in all branches of non-juristic literature.¹ No fixed form was developed. The purpose of an epitomisi is to shorten a more extensive work by extracting from it what seems to him essential. This can be done in ways varying from simple reproduction of select passages to independent summarization of contents. When, as may well happen, the epitomisi adds observa­tions of his own, we have what we may call the commenting epitome.^{[287] [288]}

Classical juristic commentaries took the same forms as those of contemporary non-juristic literature.^[289] There were no marginal commentaries, no editions of statutes, the Edict, or Sabinus’ lus civile, whether in the style of older editions of the Corpus luris, in which the text was surrounded by the Accursian Gloss, or in that of modem texts with footnotes; readers’ marginal notes are another matter.^[290] Commentaries were regularly, perhaps always, in the separate lemmatic form described above. This form naturally admitted of many variations. A commentator on the Edict might use as lemmata simply the titles and rubrics of the Edict, each of which would permit of a considerable range of discussion. Or one might use for the purpose not merely the edictal titles and rubrics, but also the clauses of particular edicts and formulae, interpreting clause by clause and word by word. Or one might take a middle course.

It is possible that the form of commentary combined with text ■was also used in classical times. Of such, a juristic work combined with the notae of a later jurist would be an example, and we do in fact possess fragments from a fourth- or fifth-century edition of Papinian’s ResponSa combined with notae by Paul and Ulpian,^[291]

literature : forms and transmission 185 in which the notae are not in the margin but inserted into the text immediately after the relevant responsa. It is possible, though an edition of so late a date does not prove the fact, that such combinations of commentaries with text were already known in the classical period. But in any case they presuppose a previous stage in which the commentary contained only selected lemmata Let us put a case. Cervidius Scaevola is writing notae on Julian’s Digesta. He may begin by entering them in the margin of his own copy of Julian, and then order the issue of an edition of Julian with his own notae inserted in the text. This would be a possible proceeding, but most unpractical. The many lawyers who already possessed a copy of the famous Digesta would hardly be tempted to buy another copy of this extensive and therefore costly work for the sake of the comparatively unimportant notae. A far more practical way of reaching a wide public would be to publish the notae separately, as a lemmatic commentary. If and when a new edition of the Digesta was called for, Scaevola’s notae could be inserted into the text, but this would never happen at all except where there was a sufficient demand for a new edition of the work commented on. Otherwise the notae would remain in the form of a separate lemmatic commentary, as happened in the case of Julian’s commentary Ad Urseium Ferocem.¹ Another possibility is that the notes incorporated in the text of a new edition of a work might be derived from sources other than a lemmatic com­mentary on it. For instance, an editor of Papinian’s Responsa might construct a note out of a discussion of one of Papinian’s responsa occurring in Ulpian’s commentary on the Edict and incorporate this note in the text of his edition. Such a procedure would explain why Justinian refers to the various notes on Papinian as 'ea quae in notis Aemilii Papiniani ex Ulpiano et Paulo nec non Marciano adscripta sunt’.²

Bare epitomes do not figure among the classical works: they were obviously regarded as hack-work. But we do meet with the commenting epitome, in which the epitomist, after quoting some passage more or less literally, adds his own observations. Here also the passages excerpted from the original work were taken as lemmata for the epitomist’s notes, but nevertheless the basic difference between a lemmatic commentary and a commenting epitome is obvious: in the former only the words to be commented are quoted; in the latter also other passages of the original which seemed remarkable to the epitomist. Since we only possess

* D. (23. 3) 48.1; below, p. 216. ² Const, Deo Auct. s. 6.

fragments, it is obviously impossible in some cases to determine whether a given fragment is from a lemmatic commentary or a commenting epitome.- Consequently in what follows we cannot treat these two forms apart.

Works of either kind were specially exposed to corruption and suffered greatly in the course of being transmitted. Sometimes, as the result of later abbreviation, the distinction between lemma and comment disappeared; short notes simply assenting to the lemma might be cut out; notes completing or correcting the lemma might be fused with the lemma. Or the confusion of lemma and commentary may sometimes be due to a copyist.¹ On the other hand, notes gave rise to further notes, and citations of literature to fuller citations. Corruptions of this kind come from the third and early fourth centuries? Doubtless the excisions and insertions of Justinian’s compilers made things considerably worse, but on the whole they injured the texts far less than scholars in the last thirty years have been apt to assume.

We proceed now to the individual specimens of this class of literature, grouping them according to the work commented on; for example, we shall keep the libri ad edictum together. The catalogues of works attached by the handbooks to the biographies of their authors are of no value to the historian of legal science.³

A. Commentaries on leges and senatusconsulta.

The classical writers frequently comment on the leges and senatusconsulta in the course of their great general works, especially in the second part of their Digesta.* Such comments sometimes developed into large treatises occupying several libri, and ended by becoming distinct literary entities. We shall speak only of com­mentaries which were or developed into separate books, though this means that we cannot give a picture of the full extent of classi­cal commentative work on leges and senatusconsulta. Some of the works to be mentioned appear to be only detached portions of the larger works.

i. The Twelve Tables. At the beginning of our period Labeo wrote a commentary on them,³ of which all that we know is the three fragments preserved by Gellius.⁶ Often as Labeo is cited, this work is never cited.

¹ A case in the commentary on the Theaetetus, Berliner Klassikertexte, ii, p. x.

²Above, p. 142.

³Our Index gives the works arranged according to authors.

♦ Below, p. 226.

³ Pal. i. jor; Pemice, Labeo, i. 51; Joers, PW i. 2250.

⁶ Gell. r. 12.18; 6.15.1; 20. 1.13.

No extract from it was included in the Digest; doubtless it did not come down to the Byzantine compilers. To the unhistorically minded classical jurists¹ it was, indeed, of no interest, the Twelve Tables being for the most part obsolete. After Labeo the only writer who dedicated an independent commentary to them was Gaius; it was in six books² and illustrates his personal interest in history. As he puts it in his highly rhetorical praefaiio, which the compilers have preserved (D. 1. 2. i),³ a man who comes to the task of inierpretatio without knowing the origins and beginnings is like a man who comes to dinner with unwashed hands. The passage is of interest to the historian of science and should be read; the idea that one ought to begin at the beginning is expressed by others,⁴ and the phrase illotis manibus is proverbial.³ The survival of this work in the post-classical period is due to Gaius’ great popularity in that period and to the historical leanings of the post-classical law school.⁶

2. Classical commentaries on republican leges.¹ Of such we know very little. From Paul we have a small fragment of his Ad I. Cinciam liber sing.³ and some fragments of his Ad I. Falcidiam, the latter containing unclassical matter.’ Of Rutilius Maximus’ Ad I. Falcidiam liber sing.¹⁰ we have only one small fragment, while the one fragment of Gaius’ Ad

I. Glitiam liber sing, is certainly not authentic.¹¹

3. Commentaries on leges of the Augustan period. The I. lulia et Papia Poppaea, in particular, provoked a number of separate commentaries¹³ —thus Gaius* in fifteen books¹³ and Mauricianus’ in six. Marcellus’ five or six books appear to be merely a separate edition of the portions of the author’s Digesta (books 26-30; probably also 31) treating of the I. lulia et Papia, since he can hardly have also written an independent

* Above, p. 135.

² Pal. i. 242; Index Flor.·. ‘XX. δυο& Thes. vii. 1, col. 400, 28; Otto, Spriclmdrter, 212, 274 ff. ⁶ Below, p. 281.

⁷ On the following works see in general Lenel’s Palingenesia. ' ⁸ D. (1.3) 29.

⁹ In D. (35. 2) 3 pr. vix... rations is not authentic, but certainly does not come from the compilers: Schulz, Z xlviii (1928), 214; Beseler, Z 1 (1930), 20.

«° D. (30) 125.

¹¹ D. (5. 2) 4. On the interpolation: Pal. i. 246; Index Interp. There is no question of the delicate psychological annotation of which Kubler, Gesek. 193, speaks.

¹² In the Index Flor, the titles of the works about to be mentioned are abbreviated to libri ad leges, luliam et Papiam Poppaeam being understood.

¹³ Ferrini, ii. 261 ff.

commentary on the same scale. Besides these, there are Terentius Clemens’ commentary in 20 books¹ and Paul’s in 10? The latter’s Liber sing, de legibus (viz. lulia et Papia Poppaea) was doubtless a post- classical extract from the larger work, as likewise his Liber sing, de iure paironatus quod ex lege lulia et Papia venit* Ulpian’s commentary in 20 books⁴ gave a painstaking interpretation of the wording of the lex, after the manner of his commentary on the Edict?

Of the leges belonging to the province of the ordinary criminal courts (quaestiones) the only one commented on separately⁶ was the I. lulia de adulterUs. Papinian’s De adulteriis libri ii were, in a wide sense, a commentary on this lex,’’ but the Liber sing, de adulteriis* attributed to him was merely a collection of quaestiones.⁹ If such a work ever existed, it has come down to us in an altered form. The six fragments of it preserved in the Collatio¹⁰ are characterized by the same rhetorical style¹¹ that we have already remarked in Papinian’s Definitiones.¹¹ Of Paul’s De adulteriis libri Hi, also a commentary in a wide sense on the I. lulia, we possess only a few Digest fragments.¹³ The Liber singularis de adulteriis ascribed to Paul, of which the Digest preserves one frag­ment¹⁴ and the Collatio three,¹³ was doubtless a post-classical produc­tion.¹⁶ Ulpian, Ad I. luliam de adulteriis libri v,¹¹ is an authentic work; of the Digest fragments, which are all we have, we lack a comprehensive study.¹⁸

^I Ferrini, ii. 251. ⁸ Ibid. 237; Berger, PW x. 708.

³ The titles are given by Index Flor. xxv. 63 and 68. But nothing survives of either work. Ferrini, ii. 237. ⁴ Joers, PW v. 1445.

³ The small fragment, P. Oxy. xvii. 2089, certainly is part of a discussion of the I. lulia et Papia, but whether it comes from a separate commentary on that statute or from the Digesta of some jurist is naturally an unanswerable question. Levy, Z xlviii (1928), 555, is mistaken.

⁶ Works de iudiciis publicis: below, p. 256.

⁷ Joers, PW ³. 57ã; Costa, Papiniano, i. 234.

⁸ Joers, PW, l.c.; Krüger, 224.

• Coll. 4. 8 is at any rate not a responsum in the technical sense, but an answer to a purely theoretical question.. ¹⁰ Coll. 4. 7-ix; 6.6.

^II eg. CoU. 4. 10: ‘quare aperte contra legem fecisse eum non ambigitur. sed si de poena tractas, non inique aliquid eius honestissimo calori permittitur...’; cf. Beseler, Beitr. ii. 21. On Coll. 4. 8: Beseler, Z xlv (1925), 453; Ivi (1936), 58. On Coll. 4. 9: H. Krüger, Z xlviii (1928), 668; Beseler, Z li (T931), 65; liii (Ã933), rr. On Coll. 4. ro: Beseler, Beitr. v. 13; ² x (1930), 195. On CoU. 6. 6: Albertario, Rend. Lomb. Iviii (1925). For the Digest passages see Index Interp. Also in favour of a post-classical abridgement: Solazzi, Bull, xxxvii (1929), 96; AG civ (1930), p. 22 offprint; Beseler, Z liii (1933), 9; ACII i. 341.

^,a Above, p. 175. ³³ Berger, PW x. 716.

¹⁴ D. (48. r6) r6. ¹⁵ CoU. 4. 2-4; 4. 6. Pal. i. 593; Berger, PW x. 715.

¹⁶ So also Solazzi, BuU. xxxvii. 96; AG civ (Ã930), p. 22 offprint. Literature on the passages of the Collatio: Volterra, Indice III (RS DI ix, 1936).

¹⁷ Joers, PW v. 1446.

¹⁸ Ulpian’s text appears to have been shortened in post-classical times. Its citations of literature are few; the non-mention of Papinian’s work is remarkable. From the citations a commentary by Sex. Caedlius Africanus can be inferred: Joers, PW v. 1446.

In addition we know of commentaries on the I. lulia vicesimaria, the l. Fufia Caninia, and above all on the I. Aelia Sentia. We possess just a few fragments¹ of Aemilius Macer, Ad l. vicesimam hereditatium, of Paul, Liber singularis ad l. Fufiam Caniniam and Ad l. Aeliam Sentiam libri iv, and of Ulpian, Ad l. Aeliam Sentiam libri iv.

4. Among commentaries on post-Augustan leges we know of Paul’s on the I. VeUaea and on the I. lunia. Of his Liber singularis ad I. VeUaeam nothing survives ;² of his Ad I. luniam libri ii only one fragment.³

5. The only known separate commentaries on senatusconsulta are by Gaius, Paul, and Marcian; the Pauline commentaries appear to be merely separate editions of the relevant parts of his commentary on the Edict. Gaius’ known commentaries are a Liber singularis ad SC. Orfi- tianum and a Liber singularis ad SC. Tertullianum, a single fragment of both surviving ;⁴ Paul’s are libri singulares on the following senatuscon­sulta, or rather orationes principum:* Ad orationem divorum Marci Antonini et Commodi (two fragments showing signs of post-classical revision),⁶ Ad orationem divi Seven (three fragments, two showing post- classical revision),⁷ Ad SC. Claudianum, Ad SC. Libonianum, Ad SC. Orfitianum, Ad SC. Silanianum (probably merely a separate edition of book 46 of his AdEdictum), AdSC. Tertullianum, AdSC. TurpilUanum, Ad SC. Vdlaeanum (doubtless identical with his Liber singularis de intercessionibus feminarum). The two last-mentioned are probably only separate editions of book 30 of his Ad Edictum. The few surviving frag­ments of these works can be found in Lenel’s Palingenesia. Of Marcian’s Liber singularis ad SC. Turpillianum we have but one fragment, though that a rather long one: D. 48. 16. 1, which has a post-classical appear­ance.⁸

B. Commentaries on the Edicts of the praetors, the aediles, and the provincial governors

Cicero’s early complaint⁹ that most lawyers derived their di­sciplina iuris from the praetorian Edict was at the time it was uttered a rhetorical exaggeration, since the practice of commenting on the Edict was only just beginning ;¹⁰ but it came to represent the truth in classical times. Not only in their edictal commentaries, but also in other works, particularly in their Digesta and in their collections of responsa, quaestiones, and disputationes, the jurists

¹ They can be read in Lenel’s Palingenesia.

* Mentioned by Index Flor, xxv. 51. Date of lex: Rotondi, Leges pub. 466.

³ D. (40. 9) 15, with luliam by mistake for luniam in the inscription: Mitteis, Zxxi (1900), 204; Berger, PW x. 708. ♦ D. (38. 17) 9 and 8.

³ On these works see Lenel’s Palingenesia. -

⁶ D. (23. 2) 60; cf. Index Interp.

⁷D. (ay. 9) 2,13; cf. Index Interp.

• Beseler, Beitr. iii. 165, iv. 191; Index Interp.

’ De leg. 1. 5. 7. ¹⁰ Above, p. 91. based themselves on the Edict by adopting its order of topics., Hence all such works may be termed, more or less strictly, 'com­mentaries on the Edict*. Nevertheless here we must confine our­selves to those works which expressly styled themselves libri ad edictum, since the other works adopting the edictal order possess special characteristics which require separate treatment.

1. First in ddte is Labeo’s commentary,¹ in at least thirty libri (book-rolls)/ on the urban praetor’s Edict, with that of the peregrine praetor in appendix. We have evidence of its being in use right through the classical period, as late as Ulpian.³ But the compilers took no extracts from it; doubtless they did not possess it. There survives only one short fragment of text; for the rest we have only citations. On the aedilician Edict Labeo commented in a separate work, of which we have only citations.⁴

2. Massurius Sabinus’ Ad edictum pr. urb., like his lus civile, was very brief: by book 5 he had already reached the title De operis libertorum* which Paul reached only in book 40 and Gaius in book 14. We possess only one certain citation of the work;⁶ it was not used by the compilers.

3. Caelius Sabinus wrote a commentary on the aedilician Edict,⁷ which the compilers did not use and of which we possess but two fragments of text.⁸ We have also a number of citations, in which the writer’s name has frequently been shortened by the compilers to Caelius or Sabinus alone, with the result that some of them have been wrongly referred to Massurius instead of to Caelius, Sabinus. For example, the interpretation of morbus in the aedilician Edict attributed in the Digest to Sabinus is proved by Gellius to come from Caelius, and not, as is generally supposed,⁹ from Massurius.

4. Vivianus must also have written on the praetorian and aedilician Edicts, but we have only citations, and these never give the title of the work.¹⁰

5. On the same Edicts Sex. Pedius wrote an extensive com-

¹ Pemice, Labeo, i. 55; Joers, PW i. 2550. ¹ D. (4. 3) 9. 4a.

’ Joers, PW v. 1479.

⁴ Fabius Mela, a contemporary of Labeo, seems also to have written a commentary on the edictum. We possess only some citations in which the title of the work is never mentioned. See Ferrini, ii. rr if.; Brassloff, PW vi. 1830.

^J D. (38.1) 18. ⁶ The collection in Bremer, ii. r. 568 If., is unsupported.

⁷ Joers, PW iii. 1272, v. 1484.

* Gell. 4. 2. 3-5; 6. 4. 1-3. Also 4. 2. 1-13, though probably not taken verbally from Caelius. See Seckel-Kubler, i. 92; Dirksen, Hinterlass. Schr. i. 39 ff.

• So Pal. ii. 200, no. 98; Bremer, ii. 1, 545; Joers, PW v. 1481; Dirksen, l.c. The texts are D. (2t. 1) r. 7; Gell. 4. 2.

¹⁰ Pal. ii. 1225. In the second century there was an epitome ex Viviano: Joers, PW v. 1485.

191 mentary, not utilized by the compilers, of which we possess (through the Digest) some citations.¹ Unfortunately it cannot be determined whether it was written before or after Hadrian’s codification of the Edict.’

6. The compilers possessed two distinct edictal commentaries by Gaius.

(a). The first was a collection of commentaries on particular

titles of the urban praetor’s Edict? It ran to ten books and covered the edictal titles 26-33 36-37* The libri possessed by

the compilers⁵ were not numbered progressively in one series;⁶ consequently they were not fragments of a single complete com­mentary on the whole Edict, but rather distinct commentaries on the ten edictal titles, which were later combined into a collection. It may be that Gaius never wrote a complete Ad edictum praetoris urbani.⁷

(b) The second commentary was entitled Ad edictum provinciate libri xxxiiP The two last books were on the aedilician Edict and must therefore have been joined to the commentary on the pro­vincial Edict, as its books 31 and 32, by some post-classical editor; for in the provinces there was in general no aedilician Edict.⁹ The

¹ Ferrini, Open, ii. 391; Girard, MA. i (1912), 214, 299; Berger, PW xix. 4r. The scale of the work can be judged from the fact that the subject of Pedius, book 25 (D. 37. i. 6. 2), namely Ed. Tit. xxv (Lenel), is the same as that of Paul, Ad ed. 41, and Ülpian, 39, and of Julian, Dig. 23 (order of Edict followed).

² Girard, MA. i. 2r4, 299, thinks before.

³ Pal. i. 182; Bizoukides, Gaius, ii. i; Kniep, Der Rechtsgelehrte Gaius (1910), 314 ff.; Kübler, PW vii. 492. Index Flor.: ‘ XX. ad edictum urbicum τά μόνα ΐύρΐθίντα βιβλία &1κα.’ The Digest inscriptions vary between ad ed. pr. urbani, ad ed. pr., and ad ed. urbicum.

* According to Lenel’s reckoning in the 3rd ed. of his Edictum perp.

³ Two de testamentis, three de legatis, one on operis novi nuntiatio, damnum injectum and de aqua (Ed. xxviii-xxx); two de liberali causa (xxxi), one de publicanis and de praedialoribus (xxxii, xxxiii), one de re iudicata and on the title qui neque sequantur (xxxvi, xxxvii). In all io libri, which happen to cover io edictal titles. Krüger, 202.

⁶ If they had been, the inscriptions would invariably have given the number in question. In fact they give simply the title, and when the title was commented on in more than one book they added the number: e.g. ‘ libro secundo de testamentis ‘libro ad ed. pr. urb. tit. de damno infecto’.

⁷ Possibly a complete commentary may have existed at one time, in which case the editor must have expunged the serial number, as being meaningless after the loss of a part of the work. He did not renumber. Some of even Gaius’ works disappeared, e.g. his commentary on Q. Mucius (Gaius, 1.188). His reference (ibid.) to his edicti interpretatio may be to lectures, since the word libri is avoided.

⁸ Bizoukides, Gaius, ii. 1; Kniep, Der Rechtsgelehrte Gaius, 149 ff.; Kübler, PW

vii. 492; Beseler, St. Albertoni, i. 428. Recent discoveries have disposed of v. Velsen, Z xxi (1900), 73, and Beitr. z. Gesch. d. Edictum pr. urbani (1909), 105.

⁹ In the senatorial provinces, with which alone Gaius’ commentary was concerned, the governor’s quaestor had the aedilician jurisdiction (Gaius, 1. 6). The aediles at

first thirty books were a commentary on the edictum provincial«, in the stereotyped form, which every governor was obliged by Hadrian to issue.¹ Gaius’ motives for commenting at such length on the provincial Edict, which in all essentials agreed with the praetorian, are a matter of pure conjecture; perhaps the work represents lectures given by Gaius in some provincial city. Throughout the commentary the magistrate named as having jurisdiction is the proconsul;⁸ the occasional mentions³ of the praetor are due to corruption⁴ or to interpolation.⁵ Possibly the work, like the same author’s Institutes? was no more than lecture-notes, in which case many things in it can be accepted as authentic which, if found in a finished work by a great classical writer, or even by Gaius himself, would have to be regarded as later depravations. Our text is, however, not free from post-classical pre-Justinian elements.⁷ What is needed is a critical study not merely of particular passages but of all our fragments as a whole. The complete work was in the hands of the compilers.

7.The longest commentary on the praetorian and aedilician Edicts was that of Pomponius. It was in far more than eighty- three books,⁸ which sufficed for Ulpian, amounting in fact to some 150 libri (book-rolls). It certainly contained an imposing assem­blage of the existing literature; it was a sort of codification, a

Cirta (Numidia) are anomalous: CIL viii. 7986, ILS 6862: aedüis habens iurisdictionem quaesioris. Cf. Mommsen, Sehr. v. 484, 490; Weiss, Z 37 (1916) 167.

¹ Above, p. 127.

² Passages with ‘proconsul’: Rudorff, Z.f. RG. iii (1864), 18, n. 15; Kniep, Der

Rechtsgel. p. vi. Thus Gaius wrote for the senatorial provinces or one of them. In the imperial the Emperor was proconsul, but not the propounder of the Edict. Egypt was an imperial province, but the Edict propounded was that of the praefeclus Aegypti, as P. Giess. 40 (Z xxxii. 378; Meyer, Jur. Papyri, no. 27) shows beyond doubt. Thus Ermann-Krüger, in Kroll, Zur Gaius-Frage (Diss. Münster, 1917), 13, are wrong. ³ Passages in Kniep, Der Rechisgel. pp. vii and 245 ff.

♦ Some copyist, or even the compilers, may have misread the siglum for proconsul as praetor.

’ The passages have till now been handled with astonishingly little critical sense. Thus: D. (2.11) 1—when putting the fragment at the head of the title the compilers must have substituted praetor. D. (4. 7) 3. 1—quia... libertatibus has long been recognized as interpolated: see Index Interp. D. (6. 2) 8—unde... pretium is cer­tainly interpolated: Pringsheim, Z 1 (1930), 416, gives the literature, but himself goes astray. D. (15.1) 27. 2—text suspicious: Beseler, Z1 (193°), 62. D. (29. 5) 25. a —not genuine: Index Interp. D. (29.3) 1 pr.—same procedure as in the case of D. (2.11) 1, mentioned above. In the same title, D. (29.3) 7, the original proconsul has been allowed to stand. Cf. Endrich, St. Cagliari, ix (1917), 155· ⁶ Above, p. 160.

⁷ Take D. (35. 2) 73. 5: one cannot credit Gaius with its pointless and indeed misleading arithmetic: Jensius, Stricturae ad Pandd. 343 ff. The text was used in a strange manner for Inst. (2. 22) 3. On D. (4. 7) 3. 1: Index Inierp., it is certainly not compilers’ work. D. (4. 4) 27 pr.: Beseler Beitr. i. 57. D. (29. 5) 25. 2: Lenel, Ed. 365; Beseler, T x (1930), 185. · See Pal. ii. 44. supplement to Hadrian’s codification of the Edict itself. But it came too soon: the Severan period had still much to say. Hence it was already out of date in the third century. After the appear­ance of the more modem and sensibly shorter commentaries of Paul and Ulpian the impressive work of our industrious author fell into oblivion. The composer of the Fragmenta Vaticana djd not use it; nor did the compilers take excerpts from it, clearly because they possessed no copy. We know it only through the citations of Paul and Ulpian; its loss is one of the heaviest for our science.

Book 83, as is shown by D. 38. 5.1,14, and 27, dealt with the edictal title Si quid in fraudem patroni (Lenel, Ed. xxv, s. 151), which Ulpian had reached in his book 44 and Paul in his book 42.

8. A commentary by Q. Saturninus is once cited by Ulpian. The citation is taken from the tenth book of the work, in which the author apparently dealt with the aedilician Edict.¹

9. The compilers possessed an edictal commentary by Calli­stratus in only six books,² entitled apparently Edicti monitorii libri vi.³ N0 certain explanation of the title can be given; the work in all probability was a commentary on the provincial Edict. What the compilers possessed was perhaps a post-classical epitome of a larger work; at any rate our fragments (all from the Digest) betray post-classical workmanship.

The latest explanation of the title, by H. Krüger,⁴ is the most un­likely that has been proposed. He believes that the so-called referential Edicts (i.e. the Edicts which referred to other sources—leges, senatus­consulta, and imperial constitutions)⁵ were termed edicta monitoria, and that Callistratus commented on these. But there is no proof that the classical jurists ever treated these Edicts as a special group or termed them edicta monitoria. One does not see any motive that could have led them to comment on them separately. Moreover, if the explana­tion were correct, the title would have been Edicta monitoria or Ad edicta monitoria. True the Index Florentinus has the plural edicta, but that can carry no weight against the constant singular of the inscriptions

¹ D. (34. 2) 19. 7. Cf. D. (50. 16) 74; Pal. ii. 1178 f. On the personality of Q. Saturninus see Krüger, 200.

³ v. Kotz, PW Suppi, iii. 226.

³ Index Flor.·. ‘XXVII. edicton monitorion βιβλία ΐξ.’ The inscriptions of the Digest fragments are as a rule: libro... edicti monitorii; once only (D. 2. 6. 2): libro primo ad edictum monitorium.

⁴ Z xxxvii (1916), 230 ff., 301 ff. Older views: Rudorff, Z. f. RG. iii (1864), 28 {monitorium edictum means edictum perpetuum); H. Pernice, Misc. z. Rechtsgesch. u. Texteskritik (1870), 102 {monitorium (substantive) edicti, i.e. monitorium ad edictum), followed by Karlowa, RRG i. 635; Buonamici, AG Ixv (1900), 68 ff.

⁵ e.g. Ed. s. 10 De pactis.

44OT.I _O of the Digest fragments.¹ Probably monitorium means commonito­rium,² and edictum (com)monitorium is just an artificial and isolated description of the Edict, 'coined before the expression edictum perpetuum had become established.³ Since its codification by Hadrian the Edict had become ‘of precept’, i.e. a form which, by the instructions of the Emperor and Senate, the praetors, aediles, and provincial governors were obliged to publish and to act under.⁴ In favour of the work having been on the provincial Edict is the only surviving fragment from 4, which cannot refer to missio legatorum servandorum causa, but must be on the Edict de sumptibus funerum.¹ This means that Callistra­tus dealt with this Edict in the same place as Gaius in his Ad edictum provinciale, i.e. after legata, which must have been its position in the pro­vincial Edict, though not in the urban.* Some of the fragments show post-classical revision.⁷

io. Paul’s Libri ad Edictum ran to eighty book-rolls, the last two of which were on the aedilician Edict.⁸ Besides numerous Digest fragments we have thirteen extracts in the Fragm. Vaticana⁹ and a fragment from an Egyptian parchment codex of the fourth to sixth centuries.¹⁰ In all probability we can also claim for the work the two considerable fragments which go by the unfortunate names of Fragm. de formula Fabiana¹¹ and Fragm. Berolinense de bonorum possessione.¹³· The question cannot be dis-

¹ In principle one should trust the inscriptions of the Digest fragments against the Index Flor.: see, e.g., the plural in Index, v, xxxii, xxxiii, where the title clearly had the singular.

² On commonitorium meaning ‘ litterae quibus ab imperatore magistratibus pecu­liaria mandantur’: Thes. iii. 1934, 8r; Seeck, PW iv. 775.

. ³ Cf. Pringsheim, ‘Zur Bezeichnung des Hadrianischen Edikts als edictum perpetuum ’, Symb. Frib. 1 ff. ⁴ Above, p. 127.

³ So already Lenel, Pal. i. 96, n. 4. 6 Lenel, Ed. pp. 9 ff.

⁷ e.g. D. (4. 6) 9, where vel potentiore vi refers to the potentiores (the socially powerful: D. 48. 19. 28. 7; Mitteis, Mil. Girard, ii. 225 ff.): interpolated, but not by the compilers. D. (11. 1) 1, where the interpolation comes at most only in part from the compilers: Index Interp. With minus frequentantur compare minus frequentatur in D. (4. 6) 2; the expression cannot come from even a provincial jurist of the Severan age.

⁸ Berger, PW x. 705. · F.V. 298-309, 319.

¹⁰ The so-called Oxford Fragment, ed. princeps by Grenfell and Hunt, Neto Class. Frs. ii (1897), no. cvii. Re-edited by Scialoja, Rend. Lincei, 1897, 236, and (with photograph) Kruger, Z xviii (1897), 224. But now one should use Girard-Senn, Textes, 460, or Seckel-Kübler, ii. 163, which rest on a new collation by Seymour de Ricci. FIRA ii, 423. Wrong: ColHnet, Conferenze 42; Albertario, St. 1, 304.

¹¹ Ed. princeps by Pfaff and Hofmann, Mitteil, aus d. Papyrus Rainer, iv (1888); later (with apograph) by P. Krüger, Z ix (1888), 144 ff. Also Collect, libr. iii. 299; Seckel-Kübler, ii. 165. Scholastic editions: Girard-Senn, Textes, 457; FIRA ii (ed. 2), 427. Photographs in the ed. princeps, Wessely’s Schrifttafeln z. älteren lat. Palaeogr., Tab. 19, no. 42, and Steffens’s Lal. Palaeogr. r4.

¹² Berlin Museum Pap. 1VJ53. Ed. princeps·. P. Μ. Meyer, Z xlii (1921), 42 ff.; thereafter Girard-Senn, Textes, 454, and FIRA ii, 427. Photo.: Mallon 48. cussed here. In the texts outside the Digest the work of one or several post-classical hands is particularly clear; in the Digest fragments the original has been considerably shortened, especially in the citations of literature.¹

Besides the eighty libri ad Edictum the compilers possessed a shorter commentary (23 books) on the Edict bearing Paul’s name? Its title is uncertain, since the Index Flor, and the inscriptions of the fragments vary. In fact the work was not a simple edictal commentary, but a short Digesta,³ the second part of which occupied only the last book,* the Edict occupying the first twenty- two. It seems to be a post-classical epitome of the larger com­mentary. Besides the scanty Digest fragments there are two in the Fragm. Vat.⁵ The work is cited in the Collectio definitionum.⁶

As to the title: Index Flor. xxv. 4 has 'brebion βιβλία βϊκοσι τρία’, meaning brevium libri xxiii. The Digest, up to book 26, calls it libri brevium, and from book 27 libri brevis edicti. F.V. has libri ad edictum de brevibus; the Collectio definitionum, libri brevium. Aitet brevium, brevi­bus, we should understand HbeUorum, HbeUis.⁷ H. Kriiger⁸ would understand a work commenting on the so-called referential Edicts only; he gives edicta brevia the same meaning as edicta monitoria. It cannot be proved, even with a show of probability, that edicta brevia was a classical name for these edicts, and still less that they were collected into a separate group and commented on together. In D. (50.16) 55 qui... creditori is post-classical,’ but not from the compilers.

Some of Paul’s minor works have all the appearance of being portions of his edictal commentary which have become distinct by having been separately edited in post-classical times and, incidentally, more or less thoroughly revised. The materials are, however, too slight to permit of more than conjectures. Only of the Liber sing, de iniuriis can it be stated positively that it is a post-classical extract from book 55 of his Ad edictum or from a post-classical edition of that book.

In his Ad edictum Paul devoted the whole of book 55 to the title De iniuriis (Ed. xxxv). He is not likely to have written another special work of exactly the same length on the same title. The two passages of the Liber sing, de iniuriis which we possess (Coll. 2. 5; 2. 6) show it to

¹ Cf. the Oxford Fragment.

² Berger, PW x. 714; Beseler, Bull, xlv (1938), 167.

³ Like Paul’s Sententiae: above, p. 176; on the scheme of Digesta: below, p. 226.

⁴ According to F.V. 310, 311, the I. Cincia came in book 23, and the second part

of Digesta regularly begins with this lex: Schulz, Z xlvii (1927), 52, n. 6, and below, p. 226. ^s F.V. 310, 311. ⁶ Below, p. 308.

⁷ Thes. ii. 2179, 14 f. ⁸ Z xxxvii (1916), 231, 301 ff.

’ Beseler, Beitr. ii. 62; Z liii (1933), 45. have been a commentary on the said title. The smaller work was not used by the compilers of the Digest, but the makers of Justinian’s Insti­tutes (4. 4 pr.) probably'drew on it, perhaps only at second hand. The texts in the Collaiio show unmistakably that Paul’s text had been re­vised in post-classical times. On the edictal clause (Ed. s. 190): 'certum dicat quid iniuriae factum sit*, CoU. 2. 6 has: ‘demonstrat autem hoc loco praetor non vocem agentis, sed qualem fonnulam edat... illud non cogitur dicere, dextra an sinistra, nec qua manu percussa sit.’ Quite true, but too trivial to come from Paul; dextra an sinistra may be a marginal gloss on qua manu, which it ought to follow. The whole treat­ment of certum dicere (CoU. 2. 6. 2-5) smacks of the post-classical law school.¹ The other passage, CoU. 2. 5, was composed by some post- classical writer out of classical materials?

Liber sing, de liberali causa. This is probably a post-classical extract from books 50 and 51 of the Ad edictum. All we have of it is D. 40.12. 33; it is absent from the Index Flor., and is perhaps identical with the Liber sing, de articulis liberalis causae (only D. 40. 12. 41)? which is similarly absent.

Liber sing, ad municipalem. This is probably an extract from book 1 Ad edictum; we have F.V. 237 and 243 only.

Liber sing, de inofficioso testamento: Index Flor. xxv. 45; fragments (few) in Pal. 1. 1113. Doubtless an extract from book 13 Ad edictum.

11. Ulpian’s Ad Edictum was in eighty-three books,^{[292] [293] about the same length as Paul’s, but only about half as long as Pomponius’. Evidence of its wide diffusion in post-classical times is abundant: the composers of the Fragm. Vai. and the Collatio drew on it, the Sinai Scholia cite it ;s it was read in Egypt ;[294] and finally it served the compilers of the Digest as the leading commentary, not merely because it was the latest of the great commentaries, but assuredly also on account of its intrinsic merits. Consequently the remains that have reached us are particularly extensive. We have, besides the numerous Digest fragments, (i) F.V. 120, 266, 318, 320-4, 339-46; (ii) Coll. 2. 4; 7. 3; 12. 7; (iii) P. Ryl. iii. 474 ;[295] (iv) two}

197 small pieces in Priscian and Pacatus ;* (v) a citation by Justinian in C. 6. 28. 4. 3-5, of 531; (vi) probably the so-called Fragm. de iudiciis ;² and (vii) possibly the very fragmentary P. Fay. x?

Following classical tradition, Ulpian treated first of the Edict of the praetor urbanus and in the last two books of that of the aediles. The commentary keeps strictly to the order of the Edict, except that the rubrics inside the titles de iudiciis and de his quae cuiusque in bonis sunt (Lenel xiv and xv) are, after the example of Julian’s Digesta, arranged in a different and more practical order.⁴ We find some excursus: thus, the commentary on the edictal title Ad legem Aquiliam gives the text of the statute, with a careful interpretation;⁵ in connexion with hereditatis petitio the SC. luventianum is reproduced and thoroughly examined;⁶ similarly the text of important rescripts is given and interpreted.⁷ Nevertheless, more seriously than his predecessors, Ulpian attempted to disregard any law that was not ius honorarium, but ius civile in the narrower sense. He deals with the latter only so far as was unavoidable in expounding the praetorian law. For example, it was impossible to give an account of the actiones legis Aquiliae utiles without setting out the statute itself and its actiones directae. A strict scheme of exposition is adhered to. The commentary on each title begins with a general consideration of its heading which provides an introductory orientation in regard to the individual Edicts of the title.⁸ The commentary on an individual Edict gives (1) the text of the Edict; (2) a close inter­pretation of its clauses, in which the clauses serve as lemmata or captions, and the commentary follows;’ (3) the text of the formula offered by the Edict; (4) any necessary interpretation of the

¹ Collect, libr. ii. 160; iii. 298; Seckel-Kübler, i. 502; Giiard-Senn, Textes, 497; Riccobono-Baviera, Fontes, ii (ed. 2), 313.

² Ed. princeps by Mommsen and Kruger, Monatsberichte Berlin Ak., phil.-hist. KI., 1879, 501 ff. (Mommsen, Sehr. ii. 68) with apograph. Photographs in Mommsen- Krüger and in Wessely, Schrifttafeln z. älteren lat. Palaeogr., tab. xix, no. 43. Collect, libr. iii. 298; Seckel-Kübler, ii. 171; Girard-Senn, Textes, 498; FIRA ii, 625. Lenel, Ed. 144, attributes the fragment to Ulp. Ad ed. 16.

³ Ed. princeps P. Fay. (1900), no. X, with photograph. Corrections: Plassberg, Wochensckr. f. klass. Philol. xviii (1901), 141. The attribution to Ulp. Ad ed. is unfortunately only a possibility. Ferrini, i. 454.

♦ See the conspectus in Lenel, Ed. p. xvii; cf. pp. 11 ff.

s Lenel, Pal. ii. 522.

⁶ Pal. ii. 501; cf. ii. 640, no. 981.

⁷ e.g. Lenel, Pal. ii. 650, no. 983..

⁸ Many of the introductions are unauthentic: below, p. 200. But that to the title De iniuriis (Ed. Tit. xxxv), in which the I. Cornelia de iniuriis is also interpreted, is authentic: Pal. ii. 766, nos. 1335-8.

⁹ e.g. Lenel, Pal. ii. 765, no. 1330 f., or ii. 771, no. 1350. formula, also in the lemmatic form.¹ Of course where the Edict contained nothing but a formula,^{[296] [297]} there was nothing but this and its commentary. It is not possible to say whether the earlier commentators used so consistent and perspicuous a method, but it may be that Ulpian’s thorough-going application of the scheme above described constituted a considerable superiority of his commentary.

Ulpian’s general purpose is plain. Himself a member of the central ministry of justice,^[298] he shared the bureaucratic tendency towards codification.^[299] This end had not been attained by Hadrian’s codification of the Edict :^s the classical interpretation of the Edict remained to be standardized. What Pomponius had essayed Ulpian was to achieve, and achieve without departing from the classical tradition, that is, not by virtue of a senatusconsult or imperial constitution, but by his private enterprise as an authorita­tive jurist, a codification in the form of a restatement. By means of a copious, but judicious, selection from the classical literature, accompanied by precise citation of the works referred to, the classical interpretation was to be resumed so completely that the practitioner would have no further need to look up the literature for himself, or at any rate not that of the Republic and the first century of the Empire.

The basis of Ulpian’s work was naturally Pomponius’ colossal commentary. This may be assumed as self-evident, though it must be admitted that it is not always demonstrable from our evidence, which in this respect is untrustworthy. No doubt before embarking on his commentary Ulpian must have read the literature of the second century and the most important works of the first for himself. The assessor of the praefectus praetorio, Papinian, can have done no less. But to what extent Ulpian in the execution of his work simply took over Pomponius’ citations, and to what extent in that case he checked them in the originals,⁶ or to what extent he derived them from materials amassed during his own preliminary studies, is a question which our evidence enables us to decide only in exceptional cases; happily the answer is a matter of complete indifference. The only important point is that Ulpian had mastered the whole intricate complex of problems raised by the classical edictal interpretation down to the smallest detail,

199 and that his selection of literature was thus based on a complete knowledge of the subject. A whole world separates him from the facile compilers and epitomists, from the half-learned sciolists, of post-classical times. He had in view not legal reform but the well-justified aim of codification. His object was to sift the existing materials and to make them cognoscible, to crystallize accepted opinions, and by his authority to lay still outstanding controversies to rest. His guarded quietism preserves the true classical spirit ;* his style, equally, breathes the classical tradition. He does not, like Papinian, affect the manner of an old-time Roman, but expounds his enormous materials in an easy, but never slovenly style, in language that is at once clear, unaffected, businesslike, and completely unrhetorical. His commentary, if he ever suc­ceeded in properly finishing it (which is uncertain), must have been one of the great works of Roman jurisprudence. It stands in the same rank as Q. Mucius’ lus civile, except that Mucins is the beginning and Ulpian the end.

During the early post-classical period, at the end of the third and the beginning of the fourth centuries, this commentary, like so many other classical works,² underwent revision in many respects. We are not in a position to determine how far the revisers shortened the work, in particular by cutting out literary citations and especially the exact references, but we can recognize the numerous post-classical additions. Where an introduction to a title was lacking in Ulpian it was supplied; where Ulpian had neglected to abstract a general principle from the case law, rules or maxims were formulated; objections to views stated by Ulpian were noted and divergent solutions suggested; literary citations and other matters were added. Even in the Digest fragments we find passages which neither Ulpian nor the compilers can have written. The pre-Justinian remains complete the proof of the fact that even before Justinian Ulpian’s text had undergone far- reaching alterations. There are passages in the Fragm. Vat. and the CoUaiio which cannot possibly be authentic; these, since it is common ground that the compilers of these collections did not themselves indulge in interpolation, prove beyond contradiction that Ulpian’s text came to these compilers already altered. The Fragm. Vai. were composed at latest between 372 and 438, but probably much earlier ;³ thus the corruption of Ulpian’s text must have taken place in the first half of the fourth century or at the end of the third. Justinian’s compilers in making their excerpts ¹ Above, p. 129. ² Above, p. 142. ³ About AJ>. 320. Below, p. 3x1.

and arranging them under their titles naturally deformed the text still further, especially by abridgement and merciless excision and telescoping of the literary citations,¹ which in the eyes of men devoid of any real interest in the historical development of legal doctrine were purely ornamental.

The pioneer in the literary criticism of Ulpian’s Ad Edictum was Per­nice (1885). But his work, though meritorious, came too soon, before the necessary preparatory studies and lexicographical apparatus existed. His acceptance of the substantial authenticity of our text led him to take a false view of Ulpian and his methods. Joers (1903) dissented, but even he put too much faith in the traditional text. The most important result reached by later research is that Ulpian’s commentary was sub­jected to a very thorough revision in early post-classical times, the extent of which will only be measurable after a complete critical examina­tion of the work. Such an examination is urgently required. Up to the present interpolations due to Justinian have not been distinguished sufficiently from those made before his time; also, interpolations which have been recognized to be pre-Justinian have often been attributed to the Byzantine law school, whereas in reality they originate from the western Empire.³ On such questions the texts preserved by pre-Jus­tinian sources are naturally of decisive weight. To-day there should be no further doubt as to the interpolation of those in the Fragm. Vai. and the CoUaiio, but the texts are too difficult for a demonstration to be given here. The literature for the CoUaiio is given by Volterra³ and Niedermeyer.⁴ Whoever still denies the fact of a pre-Justinian revision of the texts in the CoUaiio must prove the authenticity of the words in Coll. 12. 7. 8: sed non... cautionem, which is simply impossible.⁵ As to the Fragm. Vai., F.V. 266 cannot be Ulpian’s text;⁶ we have already shown that the words ut Proculeiani contra Sabinianos puiani are out of place.⁷ In F.V. 321 quamquam... legerit is clearly spurious: the comment on the opinion of Papinian cited, quod nescio ubi legerit, is really delicious. Of pre-Justinian interpolation in the Digest fragments particularly good examples are furnished by the numerous introductions to the edictal titles.⁸ A specially clear case is D. (4. 8) 3.1; this loose, rhetorical passage, which says nothing of juristic import, cannot be the work of either Ulpian or the compilers.’ In D. (21. 2) 55 the style of 'quid ergo’ rell. recalls the Autun commentary.¹⁰ A case of post-classical addition of citations occurs in D. (17.2) 52.6-10. These examples could

¹ e.g. D. (9. 2) 27. 10, where the compilers have changed puiat to puto. But such proceedings on the part of the compilers stand out more clearly in Ulp. Ad Sabinum. Below, p. 214. ² Above, p. 143. ³ ESDI ix (1936).

♦ ACI, 1933, Roma, i. 353 ff., 371 ff. ³ Schulz, Einf. 22.

⁶ Beseler, Juristische Miniaturen (1929), 124; Z Ivii (1937), 124; Albertario, St. iv. 10.

⁷ Above, p. 123. ⁸ Schulz, Einf. 35 ff. Steinwenter, Festschr. Koschaker 1.97.

⁹ So rightly Beseler, Beitr. ii. 91; iii. 35.

¹⁰ Schulz, Einf. 24; Autun Gaius, s. 34. be multiplied¹ by any reader of tecent literature. In such circumstances it is erroneous to claim that P. Ryl. iii. 474 disproves the post-classical authorship of D. (12.1) 1?

Joers made the correct observation³ that the literary citations from book 53 onwards differ from those of the previous books in omitting the title of the work and number of the book cited and in giving less pro­minence to the jurists of the second than to those of the first century and the Republic. Joers’s explanation was that Ulpian had revised and completed the citations of books 1-52, but that his revision never reached the last thirty books. This will not do. The procedure attributed to Ulpian is highly improbable. One would have expected, on the con­trary, that he would have begun by citing the jurists of the second cen­tury, with exact references, and that the addition of citations of the earlier jurists and the simplification of references would have been made later; but neither is this probable. Nor are we entitled to ascribe the difference between the citations of the earlier and the later books to post-classical revision. We must fall back on Justinian’s compilers. Joers can see no motive for them to have treated the citations differently from book 53 onwards. But it is possible that at book 53 another com­piler took over the work, or again that by then the compilers had begun to realize that their Digest would be too long, if they continued to repro­duce citations so fully. One can understand their allowing the second- century jurists to fall into the background in the citations, because they were precisely the authors who were specially well represented in the Digest by actual excerpts. Also, it suited contemporary taste to make a brave show of citations from the ‘jurists of olden times’, that is of the Republic and the first century a.d. : ‘ut altius videantur iura callere, Trebatium loquuntur et Cascellium et Alfenum.’⁴

12. Ulpian is the last classical writer on the Edict. The com­mentary of Furius Anthianus⁵ is a post-classical work, of which the compilers, on their own showing,⁶ possessed only a part, namely five books (including the first), the rest being lost. Book 1 deals with topics discussed by Ulpian only in his book 14, so that Anthianus’ work cannot have comprised more than six or seven books. The work attained no importance in practice; it is never cited, and the compilers themselves took only three excerpts from it.⁷ All three are substantially unclassical, but only in part Byzantine.⁸

¹ Another good example is D. (2. 11) 2. 8, certainly pre-Justinian on account of nonne. Cf. Beseler, Beitr. iii. nr.

² As Zulueta, St. Besta, i. 139 fif., thinks. ³ Joers, PW v. 1501.

* Ammian. Marc. 30. 4.12. ⁵ Pal. i. 179; Brassloff, PW vii. 319.

⁶ Index Flor.: ‘ XXXVI, Άνθον ήτοι Φαιρίου ’Avthavov μίρος edictu βιβλία rthnt’

1 D. (2. 14) 62; (4. 3) 4°; (6· ι) 8ο.

⁸ On the fragments cited in the last note see Index Interp. The end of the first fragment is not Byzantine: extorquere is not in Longo’s Vocabulary of Justinian’s

13. We have, finally, to mention the commentaries on particular edictal titles. Some of Paul’s libri singulares fall into this category; we have already conjectured¹ that all or some of them are merely post-classical editions or abridgements of portions of Paul’s full commentary. Besides these, three commentaries ad formulam hypothecariam are known to us which require to be discussed together. The compilers possessed three libri singulares·. (1) Gaius, De formula hypothecaria or Ad formulam hypothecariam ;² (2) Paul, a work with similar title,³ (3) Martian, Ad formulam hypothe­cariam.⁴ Paul’s work, though included in the Index Florentinus, was not used in the Digest. The two other works are represented by Digest fragments, and Martian's is cited once in the Schol. Sinaitica* Too little survives of Gaius’ work for a judgment to be formed as to its structure, but we can see that Martian's commented closely, probably in lemmatic form, on the clauses of the formula, one by one, and that it ended⁶ with general questions of the law of hypotheca. It may be that Gaius’ and Paul’s treatises were only post-classical editions or abridgements taken from their general edictal commentaries, but this cannot be true of Martian's, so that there is no doubt that he did compose such a work. All three works were, however, revised in post- classical times and the compilers possessed only the post-classical texts. One of the revisers (supposing there to have been several, which is a point that we cannot decide) introduced the word hypotheca into the title and text. The classical title had been Ad formulam Servianam or De formula Serviana. The person who made this change must have belonged to the eastern Empire, since the post-classical west-Roman sources never use the word hypotheca or its adjective,⁷ except in a single passage of the lex Romana Visigothorum which is a constitution from the (east­Roman) Codex Theodosianus.³ Whether the other post-classical

Latin (Bull, x (1987/8), 186). On the word: Beseler, Z Ivii (1937), 39. Also in D. (4. 3) 40 the clause nisi rell. is probably of western origin.

¹ Above, p. 195.

² Index Flor. xx. 13, gives the title popular in the Byzantine law school: ‘ vTroftj- Kaplas fiipMov tv.’ The Digest inscriptions have generally de formula hypoth.·, only D. (20. 6) 7 has ad formulam hypoth.

³ Index Flor. xxv. 42, records among Paul’s /xovd/k/iAa: ‘ vmih)Kdpia ’, from which no safe inference as to the Latin title can be drawn.

⁴ Index Flor.'. ‘XXIX, viro&r]Kapias /wvopiP^ov.’

⁵ Schol. Sin. v. 11: ‘tovto k6riKaplq.'

⁶ Cf. Pal. i. 649.

⁷ As can now be proved simply and surely. Look up hypotheca in Levy, Ergan- zungsindex zu ius u. leges (1930); it is a blank.

⁸ C. Th. (4.14) 1 pr. (Brev. 4.12.1): qui pignus vel hypothecam...; law of 424. elements in our texts are due to this eastern reviser or to some west-Roman cannot be determined; the latter is the more probable. In classical speech pledge is always pignus, whether or not the creditor obtained possession. Even if, as is not to be assumed, the classical writers here and there used the word hypotheca, they would never have spoken in the book-titles of the formula hypoth., but would have used the technical edictal name formula Serviana.¹ It is now certain that the word hypothecaria did not occur in the Edict.² The titles De formula hypoth. and Ad form, hypoth. cannot have been invented by Justinian’s compilers—such a title was already known to the author of the Schol. Sinaitica—and must therefore come from a post-classical editor. That being the case, it is also certain that this editor would not confine this interpolation to the title. But seeing that the interpolations of hypo­theca in the text were hastily and inconsistently carried out, and that the compilers also, as is notorious, introduced the word at times, it is impossible in individual cases to pronounce whether the word is pre­Justinian or not. The only thing certain is that the word hypotheca, which as we have said never occurs in the post-classical texts of the western Empire, was not in the classical original. The following are specially clear examples:

D. (20. 4) 11 pr. Gaius: 'Potior est in pignore, qui prius credidit pecuniam et accepit hypothecam....'

D. (20. 6) 7. 4 Gaius: ’Illud tenendum est, si quis communis rei par­tem pro indiviso dederit hypothecae, divisions facta cum socio non utique earn partem creditori obligatam esse, quae ei obtingit, qui pignori dedit....’

D. (41. 2) 37 Marcian: 'Bs pignoris nomine data et possessione tradita deinde a creditore conducta convenit, ut is qui hypothecam dedis­set....’

The pioneer in this matter was Martin Fehr in his admirable Beitr. z. Lehre vom rom. Pfandrecht in d. klass. Zeit (Upsala, 1910). Naturally Fehr could only follow the then universal view³ that all interpolations came from Justinian. So far his critics were right, but in other respects the attacks on his book⁴ were mistaken and are now quite out of date. His one mistake was as to the origin of the interpolations: that part of them (in particular those in the book-titles) do not come from the compilers, but from the post-classical revision, is to-day quite beyond dispute.⁵

¹ The name still used by s. 5 of the pre-Justinian tract De actionibus (Z xiv (1893), 89). ² Lenel, Ed. s. 267, p. 493, n. 13. ³ Above, p. 142.

♦ Erman, Mil. Girard, i (1912), 419 ff.; Manigk, PW ix. 343, 364; Kunkel, s. 94; Berger, KVJ xvi (1914), 101.

⁵ The right view is already taken by Beseler, Beitr. iii (1913), 48. The commentary on the two works of Gaius and Marcian given by Ebrard, Die Digestenfragmente ad formulam hypoth. (1917), 75 ff., is serviceable, though it carries its conclusions too far. Recently Rabel, Seminar, i (1943), 32.

C. Commentaries, Epitomae, and Notae on juristic works.

I. Works on Q. Mucins’ lus civile. This, the fundamental work of Roman jurisprudence, continued to be read throughout the classical period; the first commentary on it is of the second century.

(a) A quotation by Gellius (15. 27. 1) makes us certain of the existence of a work Ad Q. Mucium by one Laelius Felix¹ in at least two books, but it remains doubtful whether it was juristic in character or antiquarian and anecdotal. The jurists appear hardly to have noticed it. Two passages from Paul’s Ad Plautium (D. 5. 3. 43 and 5. 4. 3) citing a certain Laelius are commonly taken to refer to it, and in fact the work referred to by fr. 3 does appear to be that quoted by Gellius. But the reference occurs in the middle of a long passage which certainly does not come from the pen of a classical writer.® Laelius is there appealed to, not on a point of law, but as recording a case of five children being bom at one birth. But in fr. 43 Laelius is perhaps a mis-writing of Caelius. The commentary is not mentioned elsewhere by the jurists..

(ά) In his Institutes³ Gaius mentions libri quos ex Q. Mucio fecimus, which must have been a lemmatic commentary on Mucins’ lus civile. There is no other mention of the work, nor was it used by the compilers; probably it disappeared early.

(c) Pomponius⁴ wrote a commentary entitled, apparently, Ad Q. Mucium lectionum libri xxxix.⁵ Its arrangement was the same as Mucius’.⁶ Its method was lemmatic: the passage to be com­mented on was given in full, introduced by Q. Mucius ait or scribit or perhaps just Q. Mucius; the comment was headed Pomponius. In the Digest fragments, which are all we have, the distinction between lemma and comment has become blurred,⁷ perhaps only by Justinian’s compilers. The texts contain many things which

¹ Berger, PW xii. 416; Seckel-Kiibler, i. 94.

² Cf. Index Interp. and below, p. 216.

³ i. 188. Beseler, T x (1930), 180, but see above p. 163.

⁴ Di Marzo, Saggi critici sui libri di Pomponio ad Q. Murium (1899).

³ Index Flor.: ‘ XI, ad Quintum Mucium lectionum βιβλία τριάκοντα avia.’ The Digest inscriptions omit lectionum, obviously for brevity.

⁶ Above, p. 95.

⁷ We still have Q. Mucius scribit (or the like)... Pomponius in the following fragments: D. (9. 2) 39; (19. 1) 40; (33. 1) 7; (34. 2) 10; (34. 2) 34. In contrast, in

D. (40. 7) 29.1 the signature Pomponius has been excised before Labeo. In D. I#]. 2) 77. 1 the signatures of Mucius and Pomponius are both lacking. Here the classical text ram: ‘ Q. Mucius: Si quis... esset. Pomponius: Haec vera sunt: nam ’ reU. The reviser cut out both signatures and was thus led to rewrite the first clause: ‘ Haec [Q. Mucius refert et] vera sunt.’

205 can come neither from classical times nor the compilers,¹ showing that the work was revised in the intermediate period.

2.Works on Alfenus Varus’ Digesta.² The Index Florentinus registers a large collection of responsa by Alfenus Varus in forty books,³ but it is doubtful whether the compilers really possessed a copy of it. At any rate they were content to take their excerpts from two abridgements, from which, however, they could probably see how many books the original work contained. If they did possess the original work, which is not impossible, they may have preferred to avail themselves of the abridgements instead of working through the extensive original.

(a) One of the abridgements stands under the name of Paul. We know neither its exact title nor the number of its books. It is not in the Index Florentinus, and in the inscriptions of the frag­ments in the Digest the title varies.⁴ The compilers took excerpts from it only as far as book 8; but whether that was the last we cannot say. Paul evidently kept to Alfenus’ order and, following classical practice, seems to have quoted passages from Alfenus more or less word for word and added any observations of his own that he deemed necessary, for Paul is the very last man whom one would credit with a work of pure epitomization. It must be admitted, however, that in our Digest fragments the line between Servius’ responsa and the notes thereon of Alfenus and Paul has become obliterated, with the result that a unitary text devoid of authors’ names has been produced; only twice, obviously by oversight, Servius respondit has not been deleted.⁵ Whether this revision is due to the compilers or to some pre-Justinian post- classical editor must remain doubtful; the extracts are so few that

¹ For example, in D. (24. 1) 51 the misogynistic ratio given for the celebrated praesumptio Muciana is not classical, at any rate not Mudan, but also was not inserted by the compilers. The words et verius... habeat are evidently outside the construction of the sentence. On evitandi...probasse (before evitandi there had been Pomponius) see Beseler, Beitr. iii. 50. Mucius was reporting a court practice, and Pomponius observed that Mucius seemed to have approved it.

¹ Joers, PW i. 1472; Berger, PW x. 723; Ferrini, Opere, ii. 169; Peters, Z xxxii (191T), 464.

³ On Alfenus see above, p. 92.

⁴ Down to D. (19. 1) always ‘ Paulus libro... epitomarum Alfeni digestorum ’. From 19. 2 onwards either the same or ‘Alfenus’ (or ‘Alfenus Varus’) ‘libro... digestorum a Paulo epitomatorum ’. But this is not due to two different abridge­ments having been used, but to pure caprice and carelessness of the compilers or their clerks..

⁵ D. (28. 5) 46 and (33. 7) 16. 1. It is established that Alfenus did not simply report Servius’ decisions: above, p. 92. The variation between respondi and respondit means nothing; both were expressed by a siglum, which the copyists extended at their pleasure. the former is quite likely. But in any case there appear to be pre-Justinian interpolations in our texts.¹

(6) There is in the Digest a second set of excerpts invariably inscribed Alfenus (or Alfenus Varus) libro... digestorum. This does not disclose that the excerpts come from an abridgement; on the contrary, taken in conjunction with the entry in the Index Florentinus, it indicates that they are taken from the original work. Yet in truth they too come from an abridgement, as the following considerations show conclusively.^{[300] [301] Alfenus’ original was in forty books. The work from which the Digest fragments were taken followed the edictal order and in its book 7 dealt with the title De liberali causa (Tit. xxxi of the Edict, according to Lenel). A writer who disposed of thirty-one titles in seven books cannot possibly have devoted thirty-three books to the remaining fourteen (Titt. xxxii-xlv). Thus the work extracted from must have been in far less than forty books and have been an abridge­ment of Alfenus’ original. The epitomist’s name is unknown. He must have lived later than Paul, doubtless in early post-classical times. In this series of fragments Servius’ name does not occur, so that one cannot tell where it is Servius and where Alfenus that is speaking. Nor can we say whether this feature is due to the epitomist or to Justinian’s compilers.[302] The edictal order in any case is due to the former; the order of the original was different, as is shown by Paul’s epitome. This group of fragments once more shows pre-Justinian interpolations.[303]}

3. Works on Labeo’s Pithana* Here, too, the Index Florentinus would lead one to suppose that the compilers excerpted from an

. 207 original work by Labeo in eight books. But the Digest fragments prove to be excerpts from an epitome by Paul¹ in eight books. The textual extracts began: Labeo libro... or Labeo eodem libro, and were separated from Paul’s comments by Paulus. Some editor has cut out Labeo everywhere, except by oversight in one case,^{[304] [305] but has left Paulus. He may have been either some pre­Justinian lawyer or one of the compilers. The text contains pre­Justinian interpolations.[306] [307]}

4. Works on Labeo’s Libri posteriores.* (a) Once more the Index Plorentinus gives the impression that the compilers extracted from Labeo’s genuine ‘posthumous works’, but so small a number of books as ten indicates that what they actually used was not the original work, which is known to have comprised at least forty.^[308] The inscriptions of the Digest fragments show that they derived their extracts from an epitome by lavolenus Priscus, a work which conformed with the traditional classical design as exhibited by Paul’s epitome of Labeo’s Pithana described above. Labeo’s own words were given and lavolenus’ comments followed, the two being carefully distinguished by the authors’ names. Of this epitome, which seems to have been in ten books, the compilers possessed two versions (here referred to as A and B), from both of which they took excerpts. This should cause no surprise, in spite of the fact that, in the organization of the work of excerpting, both versions were placed in the same group (the ‘Appendix group’),^[309] for we have just seen^[310] that excerpts were taken from two epitomes of Alfenus’ Digesta, though both were in the ‘Sabinus group’. The texts of versions A and B are very dissimilar.

From A came all those Digest fragments the inscriptions of which mention Labeo’s Posteriores first, e.g. ‘Labeo libro...

208 the classical period

posteriorum a lavoleno epitomatorum.’¹ In this version, as the fragments show, the general design of lavolenus’ epitome had not been tampered with: Labeo speaks in the first person and is not reported in the third. But in other respects the classical text has been revised: the signatures Labeo and lavolenus have dis­appeared, except in one case where, by oversight, lavolenus has been left;^{[311] [312] [313] lavolenus’ comments have been excised or rendered unrecognizable; the numerous citations, which version B shows the original epitome to have contained, have been expunged? There can be no certainty as to the authorship of this revision, but it is so thoroughly in the manner of the compilers that we may assign it provisionally to them, and assume that version A gave them the text substantially in its classical form. Pre-Justinian additions appear to be present in these fragments only to a small degree.[314] Since the fragments come from books 1-6 only, it is possible that the compilers did not possess books 7-10.}

Version B, from which are to be presumed to come all fragments whose inscriptions mention lavolenus first and Labeo second (e.g. ‘lavolenus libro... posteriorum Labeonis’),^[315] was in a quite different state. Labeo does not speak in the first person, but is reported by lavolenus, except only^[316] in D. (7. 4) 24. 2, where the intended alteration was overlooked. lavolenus’ notes are ap­pended, but without the signature which, of course, they no longer needed. The citations of literature sire fuller than those of version A. It is thus clear that in this version the original classical text of the epitome had been revised, a piece of work which we cannot ascribe to the compilers, since it would have taken longer to do than the insignificant resulting reduction of text would have been worth. Hence B was a post-classical version of lavolenus’ Epitome,

. 209 in which the classical text had been altered in the ways mentioned. This revision was unskilful: in particular, it has made it at times impossible to decide whether an opinion is Labeo's or lavolenus’. Pre-Justinian interpolations are present to a greater extent¹ than in version A. The compilers possessed all its ten books.

The existence of two versions is disputed: some writers hold that all the fragments the inscriptions of which mention Labeo’s Libri posteriores come from a single source, the authentic Epitome by lavolenus. This view, which unfortunately Lenel shared and applied in his Palingenesia, has recently been defended by Berger, but unsuccessfully. All that he shows is that one argument, which was previously taken to be decisive in favour of two versions, cannot be proved. The probability now is that the two versions did not, for purposes of excerption, belong A to the 'Appendix group’ and  to the ‘Sabinus group’, but were both in the 'Appendix group’. But this does not prove that they do not come from two sources: we have seen that excerpts were taken from both abridge­ments of Alfenus’ Digesta, though both were in the 'Sabinus group’. The decisive arguments against the existence of only one version are the following: (1) The difference in the inscriptions of the two sets of frag­ments, though not in itself decisive, corresponds to internal differences in the texts. In A Labeo speaks in the first person, in  he is reported by lavolenus. Berger’s explanation is that the compilers varied their inscriptions precisely in accordance with this difference, a procedure which, though not inconceivable, is not exactly probable. But why should lavolenus himself have thus varied his method of giving Labeo’s opinions ? (2) No fragment from books later than book 6 makes Labeo speak in the first person. Why is this ? If the compilers possessed only one version, how is it that no such fragments occur in books 7-10 ? The only possible explanation is that the compilers possessed two ver­sions, but the version in which Labeo was made to speak in the first person only as far as its book 6. (3) In the fragments in which Labeo so speaks (version A) we find neither citations of other jurists nor notae by lavolenus, whereas in the other group of fragments (version B) both are found. Here again the simple explanation is that the two groups of fragments come from two different versions.

(&) In just one fragment from version  we find a nota by Paul.¹ This does not justify the inference, which has been drawn, that there were Notae Pauli on Labeo’s Libri posteriores; any reader might make a marginal entry from any of Paul’s works in his copy. The evidence for Notae Aristonis on the Libri Posteriores^{[317] [318] [319] is likewise bad.1 Again, Proculus wrote neither Notae1 on nor an Epitome[320] [321] [322] of that work.}

5. Works ad Vitellium. (a) Massurius Sabinus’ Libri ad Vitel­lium^[323] are known only by a few citations, which give no indication of the character of the work; the title must mean 'on (some work of) Vitellius’, not 'dedicated to Vitellius’.^[324] No jurist of the name is known.

(6) Paulus, Ad Vitellium libri iv.^[325] We possess a number of fragments in the Digest. They show that the work was really a lemmatic commentary on Sabinus’ Ad Vitellium, but the lemmatic form has been so spoilt that it is only occasionally recognizable.^{[326] [327]} It is a case of post-classical revision, in the course of which quite ' a number of citations from Cervidius Scaevola’s Responsa,* and much besides, were inserted. A good sample is D. (32) 78. 4, which is in the style of neither Paul nor the compilers,^[328] but might well come from Cassiodorus’ Variae.

(c) For Notae by Cassius and Aristo ad Vitellium the evidence is poor.^{[329] [330]}

6.Works on Massurius Sabinus’ lus civile. Sabinus’ luris civilis libri Hi were read and cited by the jurists throughout the classical period. Works dealing exclusively with it are first met with in the second century, probably not before its second half.

(a) Notae ad Sabinum (doubtless his lus civile) by Aristo are mentioned, but we cannot tell whether the work referred to was written in the form of a lemmatic commentary.¹¹

(b) The earliest extensive commentary comes from Pomponius. It was a counterpart, though on a smaller scale (only 35 books), to his Ad edictum, its main purpose being to expound the ius civile as distinct from the ius honorarium. As in his Ad edictum Pom­ponius here industriously collected the older literature. It was certainly a strange idea to choose as the groundwork of such an exposition Sabinus’ opusculum. The Edict, now codified, had become lex perpeiua and a proper subject for a detailed com­mentary, but it was simply grotesque to expound the literature of the civil, non-praetorian, law in thirty-five books in the form of a commentary on Sabinus’ work, which was brief—indeed that was its chief merit—unfinished, glaringly faulty in arrangement,¹ and by no means exclusively confined to the non-praetorian civil law.³ This choice shows Pomponius to have been of the epigoni and to have possessed a truly classical insensibility to faults of arrange­ment. In this respect and in its industrious assemblage of litera­ture Pomponius’ work provokes comparison with the voluminous Ausführliche Erläuterung der Pandekten of the laborious Glück who ‘without the least need and one may fairly say unfortunately’³ adopted the order of topics of Hellfeld’s lurisprudentia forensis (1764).

The Index Florentinus and the Digest inscriptions give the title as Libri ad Sabinum, whereas Ulpian ordinarily cites it as Libri ex Sabino and only exceptionally as Libri ad Sabinum. The varia­tion is readily comprehensible, since the work was a lemmatic commentary, in which the lemmata were ex Sabino and the com­mentary ad Sabinum. In the Digest the line of division between lemmata and commentary has been everywhere erased, but can sometimes still be discerned,⁴ for example in D. (17. 2) 59 pr., (34. 2) ³. i, also (41. 3) 29 and (41. 4) 6. 2.

Although Pomponius’ commentary, being overshadowed by Paul’s and Ulpian’s, is never mentioned in the post-classical period, it nevertheless was added to then. Our knowledge of it depends entirely on the excerpts taken by Justinian’s compilers from the copy which they possessed, and on citations of it by Paul and Ulpian, mostly to be found in Digest fragments. A clear example of post-classical, but pre-Justinian, interpolation is furnished by

* Above, p. 156.

² It dealt, in particular, with the aedilician Edict: above, p. 158.

³ So Arndts, Die Lehre von d. Vermächtnissen, ³ (1869), s. 1517 a init. (Glücks Pandekten, vol. xlvi).

♦ Scialoja, Bull, ii (1889), 176 ff.; Riccobono, Bull. vi (1893), 153, n. 2; Schulz, Sabinus-Fragmente in Ulpians Sabinus-Commentar (1906), 93 f.

D. (28. 5) 29, of which we have spoken above.’ Another example is D. (33. 5) 6 ^{[331] [332] [333]}

(c) Paul’s Ad Sabinum was considerably shorter than Pom­ponius’, being in only sixteen books? It too was lemmatic in plan, but this is only occasionally discernible in the Digest frag­ments, which are our only source of information.^{[334] [335] Thus D. (13. 7) 36s clearly shows that D. (47. 2) 20 pr. is a lemma ex Sabino, and D. (45. 1) 22 preserves a part of Paul’s commentary on it.}

The Scholia Sinaitica⁰ show that Paul’s commentary was still used in post-classical times, but it was overshadowed by Ulpian’s: it is the latter, not the former, that is commented on by the Schol. Sinaitica, while the Fragm. Vaticana use only the latter. Moreover, Justinian’s compilers took Ulpian’s as the leading commentary.

(d) Thus we know most about Ulpian’s commentary ;^[336] we possess not only numerous and extensive fragments of it in the Digest, but also considerable passages in the Fragm. Vaticana; the latter throw very important light on the history of the text. Outside the Digest we have (1) F.V. 59-64,70-2, 74-89, 269; (2) a small piece in Priscian,^[337] (3) the citations of the Schol. Sinaitica^[338] and the Collectio definitionum,^{[339] [340] and (4) a citation by C. 6.40.3.2 (a.d. 531).}

The work, like Ulpian’s Ad edictum, was consistently lemmatic. In the Digest the signatures Sabinus and Ulpianus have been cancelled, so that the distinction between lemmata -and com­mentary is blurred, but at times one can make out the classical form of the texta clear example is D. (47. 2) 43. 4, where we have the assistance of Gellius (11. 18. 21). In F.V. 269 the signature Ulpianus survives.

The purpose of this commentary was the same as that of the commentary on the Edict:^[341] Ulpian intended in it to restate the

. 213 interpretation of the ius civile, as in his Ad edictum he restated the interpretation of the Edict. Here again he took Pomponius as his basis, but amplified considerably. As our remains show, he re­ported the older literature so exhaustively that his commentary was more than sufficient for the purposes of the practitioner and dispensed with any need to go back to the older books. We do not know the exact length of the commentary. The compilers pos­sessed an edition in fifty-one books, but this does not cover the whole of Sabinus’ system, the entire law of property being absent.¹ Book 51 treats of the same matters as Pomponius’ book 20 and Paul’s book 13. If Ulpian completed his work on the scale thus indicated, it must have comprised about sixty-two books. Thus, either Ulpian never finished his work, or its later books did not reach the compilers.

Early in the post-classical period Ulpian’s work underwent radical changes, and this revision completely displaced the classical original. Tribonian was well aware that his own copy was a second edition, a revision;² that is why, as a bibliophile and classicist, on the occasion of the publication of the Codex repetitae praelectionis he recalls Ulpian’s Libri ad Sabinum as a classical precedent for a repetita praelectio.³ It is antecedently improbable that this second edition was made by Ulpian himself; his literary production was too great and was accomplished in too short a time to allow of this. But the fact that Ulpian A d Sabinum was indeed revised in post-classical times is beyond question; it is flagrant in the Fragm. Vaticana. There the fragments contain sentences which can on no supposition be authentic; they cannot have been penned by Ulpian even as a rough draft. They are not only unclassical in expression, but, what is decisive, they exhibit that uncertainty, that ignorance and half-knowledge in matters of law, which, while characteristic of the post-classical law school, are simply impossible in one who was assessor to Papinian and later a libellis and praefectus praetorio. The text was in this depraved condition as early as the fourth century; such was its secunda editio repetitae praelectionis. If from its beginning the modem study of interpolations had been conducted methodically, as unfortunately it was not, it would have started from these

¹ Above, p. 158.

» He learnt this no doubt from its preface.

³ Const. Cordi, s. 3: * In antiquis etenim libris non solum primas editiones, sed etiam secundas quas “repetitae praelectionis” veteres nominabant, subsecutas esse invenimus, quod ex libris Ulpiani viri prudentissimi ad Sabinum scriptis promptum erat quaerentibus reperire.’ Cf. Joers, PW v. 1441.

Vatican texts. Instead, they were overlooked, and even now no complete study of them exists. The present work is, however, not the place in which to provide one.¹

Justinian’s compilers made the already depraved text of their copy even worse, but mostly by compression and excision, less often by other kinds of interpolation. In this respect also the passages in the Fragm. Vaticana, which recur in part in the Digest, are instructive. Comparison of the two versions³ shows the kind of alteration that may or may not be attributed to Justinian’s compilers. They were specially ruthless in cutting out citations and running several citations into one, with the result that the views of one jurist are without scruple attributed to another.³

7. Works on Cassius’ lus civile.

(à) The Notae of Aristo survive only in a few citations, from which no conclusions as to their literary form⁴ can be drawn.

(á) lavolenus, Ex Cassio libri xv.^s This was either a lemmatic commentary or a commenting epitome.⁶ lavolenus quoted por­tions of Cassius more or less textually and appended his own remarks, text and comment being distinguished by the signatures Cassius or Idem and lavolenus. This lemmatic scheme has been obliterated in the Digest fragments, which are all that survive. Not only are the signatures erased, but Cassius no longer speaks in the first person, but is reported by lavolenus ;⁷ there is only one passage⁸ in which by an oversight the signature Idem (meaning Cassius) has been allowed to stand, so that the original scheme is revealed. Thus the revision was on the same lines as the post- classical revision of lavolenus’ epitome of Labeo’s Posteriora.⁹ Possibly the same hand was at work in both cases. In any-case the surviving fragments, scanty as they are, give other indications of pre-Justinian revision.¹⁰

8. An abridgement of Vivianus is referred to by a single citation in Ulpian’s commentary on the Edict.¹¹

¹ A few references must suffice here: Schulz, Einf. 28; Beseler, Beitr. iv. 170, v. 9; Z xliii (1922), 538; xlv (1925), 442; xlvi (1926), 270; 1 (1930), 73; liii (1933), xi; Ivii (i937)> 15; T x (1930), 222; St. Riccobono, i. 311; Wolff, Z liii (1933), 297, 301; Index Interp. ad D. (23. 3) 34· ² Joers, PW v. 1459, is far too superficial.

³ Compare F.V. 71 and D. (7. 1) 12 pr.: the compilers have put a doctrine of Neratius into Labeo’s mouth.

♦ D. (7. 1) 7. 3,17.1; (39. 2) 28. Below, p. 228. ³ Berger, PW xvii. 1833.

⁴ Above, p. 186. 1 D. (35.1) 54; (40. 7) 28.1; (46. 3) 78.

⁸ D. (4. 8) 39 pr. ’ Above, p. 207.

¹⁰ In D. (17.1) 36.1 quid enimfiet... cognoveris cannot be authentic; the remark is trivial and its reference to the maxim mandatum gratuitum esse debet misplaced. The style is that of the post-classical law school. Beseler, Beitr. ii. 86, iii. 69; Z liii (1933), 25; Ivii (1937), 10. ¹¹ Coll. 12. 7. 8; Joers, PW v. 1485.

9. Plautius was more than once commented on, but we do not know whether the subject of comment was a definite work by Plautius or a selection from a number of his writings.¹

(a) A single citation by Ulpian² informs us of Libri ex Plautio by Neratius.

(δ) The title of lavolenus’ commentary³ is given by the Index Florentinus* as Ad Plautium βιβλία ema, by the inscriptions of the Digest fragments as libri ex Plautio.* This leaves us once more doubting whether we have to deal with a commenting epitome or a lemmatic commentary. In the few Digest fragments the lem­matic scheme has been completely obliterated.⁶ Paul cites the work once in his Libri ad Plautium¹

(c) The commentary of Pomponius is described by the Index Florentinus* as Ad Plautium βιβλία embook 5 that Julian deals with stipulations praetoriae.

² Whence the title libri ex Minicio, the lemmata being ex Minicio; above, p. 211.

³ Riccobono, Bull. vii. 226, viii. 225, and Krüger, 175, assume an edition of Minicius with notes by Julian; the idea of a lemmatic commentary did not occur to them ; cf. above, p. 216.

⁴ D. (33. 3) 1, with Riccobono, Bull. vii. 228.

³ D. (3. 3) 76; (17. 1) 33; (46. 8) 23. Similar reviser’s work: above, p. 216.

⁶ Riccobono, l.c. ⁷ On which see ibid. 228 ff.; viii. 248 ff.

⁸ The words commodissimum... continget cannot be classical; neither are they the compilers’, whose own solution is commodius... praestare at the end. Cf. Lenel, Ed. s. 17, n. 6; Beseler, Beitr. iii. 153; Z Ivii (1937), 12; Index Inlerp.

* Landucci, St. F. Serafini 403 ff.; Ferrini, Opere, ii. 229 ff.; Berger, PW x. 709.. ¹⁰ Krüger, Quellen, 188.

¹¹ D. (7. 8) 23.

¹² Thus in D. (24.1) 63 De eo... essent cannot have been written by the compilers. On the passage: Beseler, Z Ivii (1937), 22; Index Inlerp. ¹³ Above, p. 184.

¹⁴ Whereas the Bolognese Glossators signed at the end.

218 the classical period

first appeared cannot now be determined;¹ our earliest example, an edition of Papinian’s Responsa, is of the fourth or fifth century ;² that such editions were already known in classical times is possible, but probably, as already suggested,³ the original publication of the annotations was in some other literary form, and only later were they inserted into the texts of the works commented on. Thus, in all probability, Ulpian’s notae on Marcellus’ Digesta were first published in the same form as Paul’s on Neratius and Julian’s on Minicius and on Urseius Ferox, that is as independent lem­matic commentaries, and were only later inserted into Marcellus’ text, after which the independent commentary ceased to be of interest and disappeared. Why did not the same fate befall Julian’s and Paul’s similar commentaries? No doubt simply because even in the second century there was not sufficient interest in the works of Urseius and Minicius to justify new editions of them enriched with notes taken from Julian’s and Paul’s com­mentaries, and because by the third century the same fate had overtaken the writings of Neratius. The result was that the commentaries retained an independent existence.

In considering the ultimate origin of the notae which the Byzantines, in their editions, found combined with texts, we have to reckon with three distinct sources.

(i) In part the notae are derived, as we have just indicated, from lemmatic commentaries. In the process of their transforma­tion from independent commentaries into annotations appended to texts, their wording was changed, especially by compression, and now and then there was interpolation.

(ii)In part, however, the notae were derived from other classical writings of all kinds. Thus, an editor of Papinian’s Responsa might charge some subordinate with the task of reading through Ulpian, or maybe his Ad edictum, and extracting the passages in which Ulpian had pronounced on this or that responsum of Papinian. The extracts would then be worked up into notae. Such was the exact process by which the summarizing notes were composed by the editors of the Corpus luris from the commentaries of Bartolus and Paul de Castro. During the process the expres­sions of the classical texts were frequently altered, and at times their substance as well, especially when the adapter did not fully understand his original.

(iii) In part, finally, the notae come from marginal glosses made by readers which the editor for some reason took as coming from

’ Above, p. 184. ¹ Below, p. 220. ³ Above, p. 185.

219 some classical jurist and therefore included in his edition. Such notes are substantially as well as formally unauthentic. The result is that, though we are sometimes unable to say from which of these three sources a given nota is derived, notae must always be read with a specially critical eye. This, however, is not our present affair. Our only task is to assemble what survives of the signed notae found by the Byzantines in their copies of the texts.

(à) Notae on Julian’s Digesta. As our fragments show, the Byzantine edition of the Digesta contained notae by Marcellus, Scaevola, and Paul.¹ That Marcellus wrote notae on Julian’s Digesta^{[342] [343] is rendered certain by their being cited by Paul and Ulpian. But such citations as inveni Marcellum apud lulianum adnotasse[344] do not justify the conclusion that Paul had read the note in question in an edition of Julian’s Digesta; he may have taken it from an apparatus of notes in lemmatic form. And, as has just been pointed out, the fact that Marcellus wrote such notes does not entitle us to infer that every one of the notes that has reached us is authentic. Quite apart from the possibility of inter­polations by the compilers, some of them may be entirely spurious. We have only two notae by Cervidius Scaevola ;[345] they may have been extracted from any of Scaevola’s writings.[346] The same is true of notae by Paul,[347] of which we have not a large number.}

(á) Notae on Pomponius' Liber singularis regularum. The Byzantine edition contained notae by Marcellus.^[348]

(c) Notae on Marcellus’ Digesta. The Byzantine edition con­tained notae by Cervidius Scaevola and Ulpian.^[349] That Scaevola and Ulpian wrote notae on Marcellus is proved by the fact that Ulpian cites them both.^[350]

(d)Notae on the Digesta and the Responsa of Cervidius Scaevola. In the Byzantine editions there were notae of Tryphoninus^[351] on the Digesta, and of Tryphoninus and Paul¹¹ on the Responsa. Both apparatus have been shown to be unauthentic.¹²

(e) Notae on writings of Papinian.¹³ Notae by Paul and Ulpian

figured in the Byzantine edition of Papinian’s Responsa, but hardly notae by Marcian.¹ The former, but not the latter, are found in the Berlin and Paris fragments (fourth or fifth century) of the Responsa, and the same is true of the Digest fragments. It is certain that Paul wrote notae on the Responsa: he cites them himself.² The combination of his notae with those of Ulpian must be the work of some later editor, since the latent opposition between the two great men precludes the supposition that they co-operated in an edition of Papinian’s Responsa, or that, when editing Papinian, one of them combined the notae of the other with his own.

The Byzantine editions of Papinian’s Quaestiones contained notae by Paul,³ but their authenticity is doubtful.⁴

The Byzantine edition of the Libri ii de adulteriis contained notae by Marcian.³ Possessing only two of them, we cannot pro­nounce on their authenticity..

The notae of Paul and Ulpian on Papinian were declared invalid by a constitution of Constantine, which was confirmed by the Law of Citations of 426. Both laws were included in the Codex Theo- dosianus⁶ and remained in force till the issue of the Digest, except that for the purposes of the compilation of the Digest they had been repealed by Justinian in 530.⁷ Martian's notae remained valid at first, not being mentioned in either Constantine’s con­stitution or the Law of Citations, both of which were inserted unaltered in the Codex Theodosianus. But Const. Deo Auctore of 530 says expressly that Martian’s notae had also been invalidated,⁷ and this must be true, since Tribonian cannot have made the Emperor state what to the knowledge of contemporary jurists was manifestly untrue.⁸ Therefore between 438 and 530 some imperial constitution must have added Martian's notae to those condemned we should have possessed it, had the Codex Justinianus of 529

¹ H. Krüger, l.c. 313, is in error. For the notae of Paul and Ulpian see Pal. i. 1143; ii. 950. ³ D. (27. 9) 13. i.

³ Pal. i. 813,1143. * H. Krüger, St. Bonfante, ii. 311.

» Pal. i. 803. ⁶ C.Th. (1. 4) i; 3.

’ Const. Deo auctore, s. 6..

⁸ H. Krüger, St. Bonfante, ii. 312, is in error.

⁹ Krüger, 299, explains that Constantine’s ban was applied by implication to Marcian’s notae, but this seems hardly possible. Balog, Pt. Girard, ii. 524, and H. Krüger, l.c. 312, suppose that the Law of Citations invalidated all Marcian’s writings, including his notae, but this is improbable. True, the Law of Citations does not mention Marcian’s works, but he seems to be cited (though only owing to a copyist’s mistake) by Ulpian and Paul (D. 28.1. 5; 7. 9. 8), which would suffice to make him citable. Schol. Sin. (5,11) cite Marcian’s Ad formulam hypoth., but this proves nothing: below, p. 282, n. 4.

221 survived;¹ from the Codex repetitae praelectionis of 534 it was naturally excluded, just as was the Law of Citations,^{[352] [353] because the issue of the Digest had made it inapplicable. The condemnation of these various notae did not cause the ancient editors of Papinian to expunge them, for the compilers found them still in their copies.[354] One need not be surprised, seeing that the condemnation of polytheism did not result in the alteration of the classical texts.[355] Constantine’s motive for condemning the notae cannot have been the outstanding authority of Papinian.[356] Such an assumption is forbidden not only by the terms of his constitution[357] but also by the fact that the criticisms of Papinian which were to be found in Paul’s and Ulpian’s works generally were not invalidated. That leaves only one possible motive, namely that the imperial chancery knew that the notae of Paul and Ulpian commonly read in the editions of Papinian were in part apocryphal and in part depraved reproductions of what the classical writers had really written. Thus Constantine’s constitution is a further proof of the deprava­tion of classical texts in early post-classical times.[358] [359]}

14. We come lastly to works which contained a number of excerpts from the writings of various jurists.

(a) The most extensive work of the kind was Pomponius’ Lectiones or Variae lectiones,* a title which reveals the nature of the contents. In the first and second centuries lectiones was the title bestowed on books containing the fruits of reading, collections

of excerpts or florilegia? Thus Pomponius’ work gave extracts from legal literature, which he may have collected in the course of preparing his large commentaries Ad edictum and Ad Sabinum; naturally he added remarks of his own. Since it is cited by Paul, Ulpian, and Martian, there can be no doubt that Pomponius did write such a work,² but whether or not the copy of it used by the compilers contained pre-Justinian interpolations cannot be estab­lished from our scanty fragments. It was in at least forty-one books,³ but the compilers possessed only the first fifteen. In post- classical times an epitome was made of the Lectiones and the Epistulae of Pomponius, to which we shall attend later.⁴

(b) Paul, De variis lectionibus liber singularis? This was a work of the same literary character. Whether it was authentic cannot be determined from the three short fragments that we possess.

(c) Ulpian, Pandectarum libri x.⁶ This, to judge by the title,⁷ was a similar collection of excerpts, with comments by Ulpian. Though the compilers knew of and perhaps possessed it, they did not draw on it for their Digest. They possessed a small abridge­ment of the full work, from which they inserted two short excerpts in the Digest?

(d)Modestinus, Pandectarum libri xii. This was perhaps yet another collection of extracts, possibly merely a new edition of Ulpian’s Pandectae by his pupil Modestinus.⁹ Its arrangement is peculiar.

(e) Hermogenianus, luris epitomarum libri vi. As the title indicates, this work also contained excerpts from older works, which, however, are never expressly named in the surviving frag­ments. We learn from the author himself¹⁰ that the topics were arranged in the edictal order, that there was an introduction on the sources and some appendixes at the end. The style of our Digest fragments is post-classical throughout.” Possibly the com-

¹ Gell. praef. 6.

² Pringsheim’s doubt (l.c. in n. 8, p. 221) is misplaced in view of such a citation

as Ulpian’s in D. (8. 5) 8. 6. ³ Lenel, Pal. 2. 154. ♦ Below, p. 231.

³ Pal. i. 1301 (title wrong); Berger, PW x. 722. Not in Index Flor. A post-classical

work according to Guarino, SD v (1939), 468.. ⁶ Joers, PW v. 1447.

⁷ On iravSIicrai as book-title: Gell. praef. 7; Pliny, Hist, nat., proof. 24.

⁸ The inscriptions of both give the title as liber sing, pandectarum. Omitted by Index Flor.

’ A number of the excerpts from Modestinus’ Pandectae agree word for word with Epit. Ulp.: Ferrini, ii. 418; Schulz, Epit. Ulp. p. 17. I took it that the author of the pseudo-Ulpian had used Modestinus’ Pandd., but it may be that both he and Modestinus used Ulpian’s Pandd. ¹⁰ D. (1. 5) 2.

¹¹ See Pringsheim, Sytnb. Frib. 31, and the literature given by Felgentrager, ibid. 365.

223

pilers had only a post-classical abridgement, but Hermogenian is so late a classic—one should perhaps rate him as post-classical— that it is also possible that our text is really his own. Here it would be a mistake to attempt to purge our actual texts and reconstruct a classical original which may never have existed.

(vii)

We proceed now to the consideration of a group of works which we will style the Problematic Literature, i.e. the works devoted exclusively to problems, to the most difficult and perplexing questions of law. Their titles vary—Digesta, Responsa, Quaestiones, Disputationes, and so on—and some are not free from affectation. The problems are discussed individually, at varying length; they are not interconnected by any text, and though they are sometimes arranged on a plan (oftenest the so-called system of the Digesto),¹ the connexion of a given problem with the rubric under which it is placed is frequently loose and at times artificial, and the discussion diverges into disparate departments of law. In contrast to the isagogic literature and the large commentaries Ad edictum and Ad Sabinum this problematic literature is definitely esoteric.² The isagogic works hardly touch upon the harder problems; the great commentaries occasionally deal with them, but as arising out of basic doctrines expounded by a continuous text. Some of the commentaries, abridgements, and notae we have just been dis­cussing fall within our present problematic group; this is, of course, specially true of epitomes of works which themselves were proble­matic.

The inspiration of this form of literature was undoubtedly Hellenistic. Since the time of Aristotle Greek literature had known books of προβλήματα or ζητήματα,³ in Latin quaestiones or disputa­tiones* No doubt the simple republican collections of responsa are native Roman products, but the casuistic collections of Servius and his school already bear another stamp: their problems are derived from juristic speculation as well as from practice, a characteristic feature of the Greek collections. Still, granted its Hellenistic inspiration, this form of literature also corresponded to profound and deeply rooted tendencies of the classical lawyers

¹ Below, p. 226.

² Cic. De fin. 5.5.12; Gell. 20. 5; Augustine, ep. 135. 1; Seckel, ‘Die Haftung de peculio ’, Festschr. BeUker (1907), 349 (offprint 27); Mommsen, Schr. ii. 8.

³ See Note DD, p. 342.

♦ Thes. v. i, col. 1437. 46 and 81, 1440. 36: disputatio meaning pervestigatio quaestio.

themselves—their truly Roman predilection for a fully com­prehensible concrete case, their professional taste for detail, and their very modest interest in systematization. It is thus no acci­dent that the most important works of the most important jurists belong precisely to the literature of problems.

We cannot, with the materials at our command, break this group of writings into sub-groups. The obvious course of classi­fying them according to the origin and character of the problems dealt with proves on examination to be impracticable.

1. It is not always possible to distinguish between problems suggested by the writer’s own speculations and those propounded to him by others. It matters nothing whether the question is introduced by quaero or quaesitum est. In the first place these stock phrases were often abbreviated in ancient manuscripts, and the abbreviations were later expanded indiscriminately. Further, even if we knew that the classical author wrote quaero, he might well have repeated in this form a question put to him by someone else. Again, no conclusion can be drawn from the fact that a question is raised in a collection entitled Responsa, since such works included questions suggested by speculation as well as those occurring in practice.^[CCCLX]

2. When a question is put to a jurist by someone else, there are still three possibilities.

(a) Question and answer may have been by letter, and the letters may later have been included by the jurist in his Problemaia, either in their original form or with modifications. The jurist’s answer was not always a responsum in the technical sense, even supposing it to have related to a case that had arisen in practice; moreover, a purely theoretical question might be raised in a letter from a friend or pupil. Many of the responsa in our collections may thus have been given by letter, though the epistolary form has been expunged..

(&) The problematic literature undoubtedly does contain a large number of responsa in the strict technical sense, but they are distinguishable by no sure criterion. Respondere is not decisive,

225

since the word might be used of a letter answering a theoretical question; indeed a jurist might use it of his answer to a question raised by himself. Even the collections entitled Responsa are not composed exclusively of responsa in the technical sense. Whether a text has responded, respondi, or respondit is completely insignifi­cant, since these words were all represented by a symbol (R), which later would be expanded at pleasure.

(c) The question may have been put to and answered by the jurist in oral disputation. Here too respondere was, applicable. Many of our responsa may have come from this source; they may even be presumed to have done so when the jurist introduces his answer by 'dixi’.¹ The words disputaiio and disputare have been pronounced spurious in classical legal literature? This is an error, but the point is devoid of interest for the historian of legal science, seeing that it cannot be doubted that the classical jurists did in fact take part in legal disputations and record them in their problematic works. The use of the word dixi is proof enough, since it would have been sheer affectation for a jurist to use the past tense in reference to an opinion which he was reaching at the moment of writing. Indeed, even in the absence of positive proofs,³ disputations of one kind or another would have to be presumed in the classical law schools, for otherwise teachers and pupils would simply not have been lawyers! Even leading jurists must have taken part in such disputations, though doubtless only occasionally and before select audiences. In another category of disputations they certainly took part, those namely which occurred in the consilium of a praetor or judge, or especially in that of the Emperor or the praejectus praeiorio. Paul in his Decreta mentions such disputations.⁴ A disputaiio described as having taken place in auditorio* is sometimes one which had taken place in court, not in the lecture-room.⁶ Such disputations also were recorded in the problematic literature; many a so-called responsum, the origin of which is undisclosed, may thus have been given in the course of a disputation.

For these reasons it is impossible to subdivide the group of problematic works, and we are obliged to take them simply in their historical order, without regard to their titles. But before doing this we must say a few words on two general questions, namely

¹ See Vocdb. ii. 212. 50 f. Dicere = declamare·. This. v. 970. 26.

² Beseler, T x (1930), 190. ³ Krüger, 151 ff.

♦ D. (29. 2) 97; (49. 14) 50; Marcellus, D. (28. 4) 3.

⁵ Vocdb. lur. Rom. i. 520. 21; Thes. ii. 1295. 78,1296. 54; Kubitschek, PW ii. 2278.

⁶ e.g. D. (12. 1) 40; Berger, PW x. 691, is right.

4497.1 Q that of the order of topics in these works and that of the trans­mission (speaking generally) of their texts.

As to system: in many of the works none is now discernible and none perhaps ever existed. But from Celsus and Julian onwards it became usual to follow a customary order, which we call that of the Digesta·.¹ there was a first part following the edictal order and a second treating of a traditional series of leges, senatusconsulta, and imperial constitutions in a traditional order.

In the matter of transmission the problematic literature fared particularly ill in the post-classical period. On the one hand abridgement and epitomization led to the statements of the facts of cases being pruned of the colourful actual details which enlivened the classical original; they were stripped of all that was legally irrelevant and made merely typical; the epistolary form was expunged ;* sometimes the statement of facts was even struck out altogether and the discussion thus reduced to naked abstract rules. This process, begun by the classical writers themselves, was energetically prosecuted by their successors and carried to its conclusion by the compilers. On the other hand, this very same literature lent itself very readily to amplification. Its problems were the materials of the post-classical scholastic disputations, and manifold depravations and interpolations were the conse­quence. The result is that the critical study of this group of works presents special difficulties: the Quaestiones of Africanus, Paul, and Papinian are among the most perplexing texts that we possess. They still await a critical analysis. Each work needs to be studied critically from beginning to end as a whole, a task which, since Cujas, has been neglected.

We have now to treat of the specimens of this literary group one by one.

i. Labeo

(a) Epistolae.^{[361] [362] [363] Only one certain citation, by Pomponius.[364] [365] (b) Re- sponsa,3 in at least fifteen libri. Again, only one certain citation, by Ulpian.[366] (c) Pithana.'1 Apart from a couple of citations by Pomponius and Ulpian, our knowledge of this work is confined to the Epitome by 227 Paul mentioned above.1 The title, a piece of preciosity inspired by Greek philosophy,[367] [368] is no guide to the contents. The surviving frag­ments show it to have been casuistic in character.[369] (d) Libri posteriores[370] At least forty libri, posthumously published by an unknown editor. In addition to some citations we have the epitomes already mentioned.[371] Another collection of quaestiones.}

2. Capita

(a) Coniectanea.^[372] A collection of problems chiefly concerned with ius publicum in at least nine libri. We have, besides citations, some frag­ments in Gellius.^[373] (&) Gellius gives a letter of Capito’s,^[374] which is pro­bably from a collection, but whether a juristic collection is not certain.

3. Massurius Sabinus

(a) Responsa, in a least two libri. All that we have is a single certain citation by Callistratus,^[375] but many citations of Sabinus not specifying the work referred to may be references to this work, (b) Memorialia.^{[376] [377]} A collection of problems in at least eleven libri, the subjects being sacral law and antiquarian questions of public law. A few fragments and citations.¹¹ (c) The libri ad Vitellium mentioned above^{[378] [379] also belong to the problematic literature.}

4. Proculus™

The compilers possessed an edition of Proculus’ Epistulae and took a few fragments from it for the Digest.^[380] In some of the fragments the epistolary forms (letter raising the question and that answering it, with the customary initial greetings) are preserved, but in others they have been expunged, probably by the compilers and not before them.^[381] But the Byzantine edition already contained post-classical depravations and interpolations.^[382]

5. Fufidius

Of his Quaestiones in at least two libri we have nothing but a single sure citation by Africanus.¹

6. Plauiius

His work, which survives only in the fragments of the above-men­tioned commentary by Paul,² appears to have been a collection of problemata.

7. Urseius Ferox

His work, in at least ten libri, vns a casuistic collection. Apart from a few citations all that survives is the above-mentioned commentary by Julian.³

8. Minicius

A collection of problemata, surviving only in the above-mentioned commentary by Julian.⁴

9. Aristo

He seems to have written Digesta in at least five libri, which, like all works so entitled, was casuistic in contents.⁵

10. lavolenus Priscus

The compilers possessed a version of his Epistolae, in fourteen libri, from which numerous extracts passed into the Digest. In these the epistolary forms have been radically impaired, so that only occasional glimpses of them are left.⁶ Whether this is due to some post-classical editor or only to the compilers cannot be said. The compilers’ copy already contained post-classical deprava­tions and interpolations.⁷

lavolenus’ commentaries on Labeo, Cassius, and Plautius, of which accounts have already⁸ been given, all come under the category of problematic literature. '

11. Neratius Priscus

(a) Of his Epistolae, in at least four libri, all that we have are two cer­tain citations, by Ulpian.’ (&) From his Responsorum libri Hi the com­pilers took just a few excerpts, but not enough to show how the work was arranged. Besides the fragments there are a few citations by Paul and Ulpian.¹⁰ (c) His Membranarum libri vii were a collection of quae­stiones and responsa.¹¹ The stilted and unique title (membranae=parch­ments) does not indicate a book in the form of a parchment codex, but

¹ D. (34. 2) 5. ² Above, p. 215. ² Above, p. 216.

♦ Above, p. 216. s Krüger, 180; Mommsen, Sehr. ii. 23.

⁶ D. (41. 3) 21; (39. 5) 25; Pal. i. 285.

⁷D. (38. 2) 36 is interpolated from quia solvendo onwards: Beseler, Beitr. iii. 32.

See also D. (45. 1) 108, with Beseler, Beitr. v. 12; D. (50. 17) 200, with Beseler, St. Bonfante, ii. 72. ⁸ Above, p. 206 ff., 214!.

» Pal. i. 763. ¹⁰ Ped. i. 775. ” Pal. i. 765.

229 simply ‘sketches’—the sort of thing commonly written on parchment.¹ The Byzantine edition contained post-classical additions?

12. Celsus

His Digest was a large collection, in thirty-nine libri, oi proble­mata, arranged in the accustomed order of works so entitled.³ It incorporated collections of problemata previously made by Celsus under the titles Commentarii, Epistolae, and Quaestiones. It appears that the Digesta noted the earlier work from which a quaestio was derived, since in Ulpian’s citations of the Digesta we sometimes find, e.g., ‘Celsuslibro... digestorum, commentariorum (epistularum, quaestionum) libro.. (D. 34. 2. 19. 6; 4. 4. 3.1; 28. 5. 9. 2; 12. i. i. 1). The ordinary meaning, in antiquity, of such double references⁴ would be that book and title of a work as a whole and book and title of a subordinate part were being given side by side.⁵ But this cannot be the meaning here, for Celsus’ older collections must have been broken up and distributed according to the scheme of the Digesta,⁶ either by Celsus himself or some later writer. That the double citations are not more numerous is due to the compilers. It is hardly probable that Ulpian himself verified his citations both in the Digesta and the earlier work; he must have found the latter already noted in the Digesta. The compilers excerpted relatively few passages from this important work; these fragments show that their copy was seriously corrupt.⁷

13. Julian

(a) Digestorum libri xc.^s This outstanding work also belongs to the category of problematic literature.’ It contains a collection

¹ This. viii. 630. 50 f.; Birt, Das antike Buckwesen (1882), 57 ff., 93; Kritik u. Hermeneutik nebst Abriss des antiken Buchwesens (1913), 289, 345; W. Schubart, Das Buch bei den Griechen u. Romern (ed. 2,1921), 115, 185, is incorrect.

² Thus in D. (41.1) 14 pr. nam...fiunt is not genuine, but not from the compilers: Perozzi, Ist. ³. 599, n. ³. Cf. Beseler, Z liii (1933), 16; Index Interp.

» F. Stella Maranca, Intorno ai frammenti di Celso (1915).

♦ It is not true that they are entirely unknown in ancient literature, as Mommsen, Sehr. ii. 91, says.

⁸ Thus in the conspectus of the titles of Justinian’s Digest one reads, e.g.: ‘ex ordine digestorum liber quintus, de iudiciis liber primus.’ Again in Didymus on Demosthenes: *4«Çijîè ïåð³ Ar]p>oAfricanus

(à) At least twenty libri Epistularum, of which we have nothing except a single sure citation by Ulpian.⁴

(á) Quaestionum libri ix.^s A collection of responsa of various kinds, the arrangement of which cannot be made out. Its con­nexion with Julian and Julian’s Digesta is very close. Julian is sometimes named, and in some cases where we have simply ait, respondit, and the like, it can be shown that we should understand lulianus. The Byzantine jurist Dorotheus attributes to Julian the opinions stated in a number of passages in the Digest which do not themselves mention him.⁶ This does not prove that Dorotheus possessed a copy of Africanus’ Quaestiones in which Julian’s name still stood.⁷ In short, though much of what we find in the Quae­stiones indubitably comes from Julian, we lack the means of dis-

¹ See Note EE, p. 342. ² Himmelschein, Synth. Frib. 409 S.

³ Above, p. 216. ♦ D. (30) 39 pr.; Pal. i. 1.

⁵ Ibid. 1. 2 should be discarded for Lenel’s new edition in Z li (1931), 1 ff. Older literature: Mommsen, Schr. ii. 14 ff.; Buhl, Z ii (1881), 180 ff.; Salvius lulianus, i. 78 ff.; W. Kalb, Roms Jurisien nach ihrer Sprache dargestdlt (1880), 66; Schulze, Z xii (1892), 114. The once celebrated commentaries of Cujas and Scipio Gentilis (Spangenberg, Einl. in das romisch-justinianeische Rechtsbuch, 1817, p. 204) are now out of date, but still useful. The later literature is given by Lenel, Z li (1931), 1 ff·, and, for the individual passages, by the Index Inlerp. and by the Index to Z i-1. The discussion is not yet closed.

⁶ Mommsen, Schr. ii. 16; Buhl, Z ii 197.

⁷ Mommsen, Schr. ii. 17; above, p. 206, n. 3.

231 tinguishing between his and Africanus’ own contributions.¹ Clearly the text used by the compilers was already in a deplorable state and the disorder of our present text is due to them in only a small degree. In particular, they had no motive for deleting the name of the conditor edicti whom they esteemed so highly.® The text obviously contains a quantity of pre-Justinian corruptions and interpolations, and if Africanus really did write Quaestiones, his work must have undergone far-reaching transformation in the post-classical period.³ But the whole work may be spurious; it may be some sort of post-classical hotchpotch composed from Africanus’ Epistulae and Julian’s Digesta.⁴ The older attempts to distinguish with certainty between the work of the two men are now out of date, because not sufficiently critical; even Lenel’s revised edition of his palingenesia of the Quaestiones, from which all future research must start, is far too conservative. But without fresh evidence a definitive analysis, a complete clarification of the genesis of the text, is impossible. It is one of those cases in which no more can be attempted than to recover the text as the compilers found it. Line by line the signs of post-classical workmanship must be marked down, and the question must be considered how far matter which is clothed in post-classical forms is nevertheless classical in substance.

15. Pomponius

(α) Epistulae* Citations by later classical writers show the existence of a collection of letters in at least twelve books. The Index Florentinus registers such a collection, in twenty libri.⁶ The inscriptions of four Digest fragments⁷ give the title as Epistularum et variarum lectionum libri·, this must be the correct title, the Index and the remaining inscriptions having merely abbreviated. It is thus clear that what the compilers possessed was a post- classical work combining extracts from the Epistulae and the Variae lectiones.⁶ That in the authentic Epistulae the epistolary forms were invariably preserved is shown by those of our fragments’

¹ No stress should be laid on the variation respondi respondit: above, p. 225.

² The Index Flor, puts him first, in defiance of the historical order.

³ First remarked by Kalb, Roms Juristen, &c., 66.

♦ The work is never cited by the later classics.

³ Pal. ii. 52, 53, n. 3; for older literature see Kruger, 193.

⁶ Index Flor.: ‘ XI, Ιτηστολων βιβλία ΐίκοσι.' ·> D. (4. 4) 50; (4. 8) 18; (40. 13) 3; (50.12) 14.

⁸ Above, p. 22Ϊ.

• Specially clearly by D. (40. 5) 20, where an old lawyer, at the age of 78, ‘with one foot in the grave ’, as he himself, quoting a Greek saying (cf. Lucian, Hermotimus,

from which the compilers or the pre-Justinian. editor have failed to expunge them completely. The text available to the compilers contained post-classical interpolations.¹

(b) Fideicommissorum libri v. Also in the nature of a collection of quaestiones.

16. Gaius

De casibus liber singularis. We possess only seven short frag­ments, in Justinian’s Digest.^3,

17. Marcellus

(à) Digestorum libri xxxi. A collection of problemata derived from various sources, arranged in the order of Digesta, and obviously thoroughly in Julian’s manner. The compilers’ copy contained post-classical interpolations.³

(á) Responsorum liber singularis. This may have been merely a post-classical abridgement of Marcellus’ Digesta·, at any rate our fragments contain a number of post-classical interpolations.⁴

18. Maecianus

To judge by its title, Maecianus’ Quaestionum de fideicommissis libri* xvi belonged to the category of problematic literature, though in our Digest fragments its character has been in part obscured: a pre-Justinian hapd is unmistakable.⁶ Further study is needed.

19. Cervidius Scaevola

(a) Digestorum libri xl. Responsorum libri vi.¹ The transmission of the text is in both cases complicated. Both works are collections of responsa, all of which apparently deal with cases that had arisen in practice. These responsa, to all appearance, were never published by Scaevola himself, but only in the third centuiy, later 78), puts it, raises a question of law out of pure love of his science. The greetings of the two letters have been struck out by some editor.

¹ A clear illustration: in D. (46. 3) 9a pr. haec manumissio... tenearis cannot have been written by Pomponius: Beseler, Z liii (1933), 44. Nor had the compilers time for such child’s-play. Beseler’s reconstruction cannot be accepted. Probably the whole pr. is a marginal summary (whence sit and tenearis), which has got into the text. ³ Pal. i. 181.

³ Of considerable extent. e.g.: D. (9. 2) 36 nam sane... exstitit (Beseler, Beitr. iii. 27; Index Interp.) is interpolated, but not by the compilers; D. (34. 5) 24, the whole interpolated, but not by the compilers (Beseler, Z Ivii (1937), 22; Index Interp.).

* Schulz, Ueberliefierungsgesch. d. Responsa des Cervidius Scaevola, Symb. Frib. 236.

⁵ Full title: Ulpian, D. (q. 1) 72; abbreviated to libro quaestionum in Papinian, D. (29. 2) 86 pr., probably by the compilers. Index Flor.: ‘XIII fideicommisson ßißXla ScKtdf ’; in the Digest inscriptions: libri fideicommissorum.

⁶ e.g, in the two long fragments D. (35. 2) 30,32; cf. Index Interp.

⁷ R. Samter, ‘Das Verhältnis zwischen Scaevolas Digesten u. Responsen’, Z xxvii (1906), 151; Schulz, Symb. Frib. 143 ff.

233 than Ulpian, by an unknown editor, and presumably under the title Digesta. Later still, but not later than the first half of the fourth century, this editio princeps was abridged: the textin the compilers’ copy was abbreviated, though only moderately. But they also possessed a far more radical abridgement of the Digesta made by an unknown epitomist: Responsorum libri vi. Among our fragments we find eighteen passages in which one and the same responsum appears in both versions, that of the abbreviated Digesta and that of the Responsa.^[383] Comparing the two one obtains an exceptionally good picture of the process of abbreviation; we see that the abbreviators had no intention of altering the law stated. There is nothing to show that they belonged to the eastern Empire. A few glosses have got into the text, especially the apocryphal notae of Tryphoninus and Paul of which we have spoken above? These are the premisses on which a comprehensive critical study of this large and perplexing mass of materials must proceed?

(&) Quaestionum libri xxA A work arranged in the order of the Digesta and authenticated by Ulpian and Martian? The compilers’ copy contained serious alterations and interpolations?

(c) Quaestionum publice tractaiarum liber singularis.¹ A work known only from a few extracts in the Digest; these seem to come from a post- classical epitome, the author of which has added remarks of his own? They certainly contain considerable post-classical, pre-Justinian, work?

(d) De quaesiione familiae liber singularis. This little work, of which we possess only the title as given by the Index Florentinus (xviii. 6), may also be presumed to have been a post-classical abridgement.

20. Tryphoninus

Disputationum libri xxi, a collection of quaestiones arranged in the edictal order, the quaestiones being derived¹ from disputations in school or consilium.^3. Throughout, the remains show post- classical workmanship ;³ indeed one may well doubt the authen­ticity of the whole work⁴—it is never cited by the later classical writers, not even by Ulpian. But if the basis is a real work by Tryphoninus, that work was certainly seriously overlaid and enlarged in the post-classical period.

21. Papinianus

(a) Quaestionum libri xxxvii.⁵ This, too, is a collection of proble­mata in the style of Julian’s Digesta, with which its contents show it to be closely connected. There are letters, answers given in disputations in lecture-room or consilium? (always recognizable by dixi)³ and responsa elicited by actual or supposed cases. The order is that of the Digesta. There are many excerpts in the Digest owing to Papinian’s great reputation at Byzantium, and there are also citations both in the Digest and the Fragm. Vaticana and by Justinian himself, as well as an isolated citation by Julian of Ascalon ;⁸ lastly there are three important fragments in the Fragm. Vaticana.⁹ Careful study of this difficult work shows that what the compilers had before them was not the true classical text but the original text intensively worked upon, altered, abbreviated, and added to in the early post-classical period. Almost every sur­viving fragment exhibits the ravages of the post-classical editor, who is constantly recognizable by his uncertain grasp of classical law, and often betrays himself by a pompous and sentimental rhetoric which is quite unsuitable for precise legal statement and* serves

¹ The decision is sometimes introduced by dixi (above, p. 225): D. (20. 4) 20; (28. 2) 28 pr.; (29.1) 41. 3; (34. 5) 9 pr.; (37. 4) 20; (46. 1) 69. Once we have dixi in auditorio, i.e. in the lecture-room or in court (above, p. 225): D. (23. 3) 78. 5.

² Above, p. 225. /

³ See on the passages Index Interp. and Z i-1 Index. Specially clear signs in D. (34. 3) 27; (26. 7) 55. 4 (eandem faciem patrimonii, ex temporis intercapedine); (49.17) 19.5. A thorough critical study of the whole evidence has not yet been made.

⁴ So Beseler on many occasions, e.g. Z xlv (1925), 255; T x (1930), 190. Cf. Felgen- trager, Syrnb. Frib. 370. But the word disputatio in the title is no ground for suspicion (above, p. 225); to that extent Beseler, T l.c., is wrong and Lenel, Z 1 (1930), 15, right.

’ Costa, Papiniano, i (1894), 222. Cujas’s commentary (Ppp. T. 4) is still a valuable aid, though naturally out of date.

⁶ Above, p. 225. 7 Above, p. 225.

⁸ Harmenopulos, Manuale, ii. 4. 51; Zachariae v. Lingenthal, Z x (1899), 252; Ferrini, Opere, i. 444, 446; Riccobono, ‘La citazione del 1. Ill quaest. di Papiniano in Armenopulo’,St. Fadda, i (1906), 289; Scheltema, T XVII, 424L ’ F.V. 224-6. only to darken counsel—in short the sort of thing that cannot possibly have come from Papinian’s pen. This verdict is based on a close study of the whole evidence.¹ What follows is intended merely as illustration.

D. (5. 2) 15 pr.: 'Nam etsi parentibus non debetur filiorum hereditas propter votum parentium et naturalem erga filios caritatem, turbato tamen ordine mortalitatis non minus parentibus quam liberis pie relin­qui debet.’ This piece of sentimentalism is given as a justification for allowing the querdla inofficiosi testamenti to be brought by parents. The sense is this: 'although in view of their desire (that their children should survive them) and of their natural love for their children parents are not owed the inheritance to their children, nevertheless, if the natural. order of mortality is broken (i.e. if their children predecease them), piety requires that something be left to parents as much as that (in the con­verse case) something be left to children.’ This is the language of a late Roman rhetorician, not a classical lawyer. The clause etsi... caritatem is not only irrelevant from the legal point of view, but misleading ;² the next phrase turbato ordine mortalitatis, reads like a sepulchral inscrip­tion erected by sorrowing parents.³ The style is that of neither Papin- ian nor the compilers.

An even better illustration is D. (35. 1) 72 pr.: 'Cum tale legatum esset relictum Titiae “si a liberis non discesserit”, negaverunt eam recte cavere, quia vel mortuis liberis legati condicio possit existere. Sed displi­cuit sententia: non enim voto matris opponi tam ominosa non inter­ponendae cautionis interpretatio debuit.’ The argument is: the cautio Muciana⁴ is not applicable, because the condition might be realized in the legatee’s lifetime, i.e. if the children predeceased her. But, says the closing sentence, this interpretation, that no cautio can be entered into must not be opposed to the mother’s desire (to enter into it), because it is of evil omen (for the mother’s wish is to die first): it is of evil omen to mention the possibility that the children may die before their mother. This again is obviously late rhetorical twaddle, not the utterance of the greatest of the Severan jurists.

D. (28. 7) 15: ‘Filius, qui firit in potestate, sub condicione scriptus heres, quam senatus aut princeps improbant, testamentum infirmet patris, ac si condicio non esset in eius potestate: nam quae facta laedunt pietatem existimationem verecundiam nostram et, ut generaliter dixe­rim, contra bonos mores fiunt, nec facere nos posse credendum est.’ Let

¹ Only present unpropitious circumstances have prevented me from publishing

my critical commentary on Papinian’s Quaestiones. To derive one’s conception of Papinian’s style from the traditional text of the Quaestiones (Kubier, Z xlii. 528) is to court error. '

² On caritas: Albertario, Rend. Lomb. bciv (1931), — Studi, v. 23 ff.

³ See the collection of such inscriptions given by Lier, Phil. Ixii (1903), 456 ff.; cf. Münzer, Rom. Adelsparteien, 381.

⁴ H. Krüger, Mil. Girard, ii (1912), 1 ff.; Buckland, Textbook, 289. us pass over the interpretation of this difficult text as a whole and consider its last sentence, which ranks as Papinian’s most celebrated pronouncement, Cujas observing: vox est Christiano digna. (In fact it is un-Christian, since human sinfulness is a basic principle of Christianity.) Papinian never made it. The grotesque presumption that what ought not to be cannot be destroys the whole point of the decision: it would mean that a condition which the son cannot satisfy is an impossible condition, and therefore must be struck out as a nullity. The ancient schools of rhetoric taught the following argument for the case where the interpre­tation of a document was disputed: 'according to the interpretation of the other side this document offends against statute or morals. Now it is not to be presumed that the party from whom the document comes intended anything illegal or immoral. Therefore the interpretation is false.’¹ Thus this famous dictum is a mere reminiscence of a lesson in rhetoric.²

We cannot discuss here the important, but difficult, F.V. 224-6?

(J) Responsorwm libri xixf This famous work also was a collec­tion of problemata arranged in the Digesta order. In it Papinian seems to have included principally cases that had arisen in practice. We cannot say whether this is true of every case, for Papinian sub­jected his decisions to a thorough-going revision when he incor­porated them into this collection. The cases are reduced to their juristic minimum, the facts, the question, and the answer being no longer kept separate? Unlike Scaevola’s Digesta, Papinian’s Responsa are case law reduced to abstract terms. Everything extraneous is excluded and the bare legal problem is isolated from the manifold and legally irrelevant details of the actual case. Problem and answer are formulated ■with great, perhaps excessive, elegance and the utmost compression. At times the brevity of expression borders on the baroque.⁶ Even a contemporary student

¹ Auct. ad Heren, ii. 10,14: ‘Deinde id quod scriptum sit, aut non posse fieri aut non lege, non more, non natura, non aequo et bono posse fieri; quae omnia noluisse scriptorem quam rectissime fieri nemo dicet.’ The words of the pseudo-Papinian are probably a reminiscence of the well-known utterance of Solon: ‘is cum interro­garetur, cur nullum supplicium constituisset in eum, qui parentem necasset, respondit se id neminem facturum putasse* (Cic. p. Roscio Amer. 25. 70).

² In 1889 Ferrini, Teoria gen. dei legati, 706, rightly pronounced the principle as extremely elevated, but hardly juridical. Cf. De Ruggiero, Bull, xvi (1904), 168.

’ Cf. Leist-Glück, Serie der Bücher, xxxvii/xxxviii. 5 (1879), s. 1622, no. 154, pp. 197 ff.; Bonfante, Corso, i (1925), 176, n. 12; Albertario, AG c (1928), 236; H. Krüger, St. Bonfante, ii. 307; Beseler, Z xliii (1922), 539; liii (1933), n; T x (1930), 222.

⁴ Costa, Papiniano, i. 196; Joers, PW i. 573.

¹ Pemice’s remark (Labeo, i. 62, n. 9), that from book 8 onwards the decision is regularly introduced by respondi or respondit, is incorrect: Krüger, 222, n. 74.

⁶ ‘ For me Papinian is too subtle ’ said Pemice in a lecture in the summer of 1899; toward Mommsen he has obviously uttered an even sharper verdict: cf. Mommsen’s 237 must have found the work difficult reading: often one is obliged to begin by making up one’s mind as to the facts underlying the decision. They are correctly and completely stated by the author, but with extreme brevity. Papinian is evidently aiming at the laconic lapidary style of the jurists of old, at which long ago Horace had laughed,¹ and which in the time of the Seven was something of an archaistic affectation. The legend portraying Papinian as meeting death like a true Roman¹ is faithful to an essential characteristic of this exceptional man.

Our information as to the Responsa is relatively good. Besides the numerous fragments in the Digest and some citations we have fragments in pre-Justinian sources.

(1) F.V. 2.17, 6441-66,121,122, 250-65, 294, 296, 327-33.

(2) Fragments from a parchment codex of the fourth or fifth century, which contained an edition of the Responsa. They were found in Egypt, and some passed to Berlin, others to Paris.³ Both are in a very bad state.

(3) A small fragment standing at the end of the Lex Romana Visi- gothorumA In its place a tenth-century manuscript of the Brevia­rium has another fragment from the Responsa.*

The work did not escape thorough-going revision in early post- classical times. The reviser was either the same man as the reviser of the Quaestiones or was closely connected with and imitated him. For we find in the Responsa, in places, the same inflated rhetoric and the same signs of lack of juristic competence. How far this revision was carried will only be determinable when all surviving texts have been thoroughly examined,⁶ and as a whole. Naturally the post-classical, but pre-Justinian, elements are most clearly identifiable in those fragments which we have independently of the Digest, or not through the Digest alone, because in such cases obituary notice of Pemice (Schr. iii. 579): * to him Papinianism was repugnant, and so he caught at Labeo ’. ¹ Above, p. 62.

² SHA, Carat., 4 and 8; Costa, Papiniano, i (1894), 14, 34.

³ Berlin fragments·, ed. princeps by P. Kriiger, Monaisber. Ak. Berlin, 1879, 509; 1880,363, with an apograph. Numerous addenda: Kruger, Z i (1880), 99; ii (i88r), 83. Definitive edition: Collect, libr. iii. 287, followed by Seckel-Kubler, i. 430; Girard- Senn, Textes, 372; and FIRA, ii (ed. 2), 437. Paris fragments·, ed. princeps by Dareste, NRH vii (1883), 361, with photograph. Fresh collation by Esmein, NRH x (1886), 219. Apograph: Kriiger, Z v (1884), 166. Definitive edition: Collect, libr.

iii. 291, followed in the other collections.

* Collect, ii. 157; Seckel-Kfibler, i. 429; Girard-Senn, Textes, 372; FIRA ii (ed. 2), 437·

³ Collect, libr. iii. 296; below, p. 291, n. 9.

⁴ Cujas’s commentary {Opp. T. 4), though naturally quite out of date, is still a valuable aid.

there is no question of interpolation by the compilers. But we cannot discuss the difficult texts here.*,

(c) De adulteriis liber singularis. A collection of quaestiones. Coll. 4. 8 shows that the cases are derived by no means solely from practice. As demonstrated above,^{[384] [385] the classical text has been overlaid with post-classical work.}

22. CaUistraius

Quaestionum libri ii: only Digest fragments, showing signs of post- classical work.^[386]

23. Tertullianus

Quaestionum libri viii: cited by Ulpian; two fragments in the Digest.

24. Papirius Fronto

Responsa: in at least three books. Only a few citations, by Callistra­tus and Marcian.

25. Paulus

(a) Quaestionum libri xxvi.^[387] This was a collection of problemata in the usual order of Digesta, the cases coming from practice, disputations, and letters to Nasennius Apollinaris,^[388] Latinus Lar­gus,^[389] Licinnius Rufinus,^[390] Nymphidius,^[391] and some anonymi.^[392] The text was heavily edited in the post-classical period,^{[393] [394] but how far altered could be determined only by a critical study of the whole evidence.11 The fact of alteration is clearly shown by F.V. 227, our only fragment outside the Digest.}

F.V'. 227.*² ‘Apollinaris Paulo. Duo sunt Titii, pater et filius; datus est tutor "Titius” nec apparet de quo sensit testator: quaero, quid sit )ris. Respondit: [is datus est, quem dare se testator sensit; si id non

239 apparet, non ins deficit, sed probatio, ergo] neuter est tutor. [Hoc rescriptum est in Sticho manumisso, si duo sint Stichi et incertum de quo testator senserit, vel si Erotem legaverit qui plures eodem nomine habuit servos. Quod in nummis legatis non ita placuit: si non adparet voluntas, id acceptum est quod minus est.]'

In Paul’s day, if it were clear which Titius was meant by the testator, his appointment might well be held valid. But if it were not clear, the appointment must necessarily have been void, and void iure civili. But the editor, obsessed by the doctrine of voluntas, declares for is quern dare se testator sensit, which is false, or true only st apparet de quo sensit. But our editor is consistent: he goes on to say that s» id non apparel, it is not the law, but the evidence of fact, that is defective. The parallel cases subjoined are very badly formulated: one can only guess the meaning of ‘Quod in nummis’ rell.' Even non-juristic details throw light on the history of the text. Our editor found still preserved in his text the greet­ing of the questioner’s letter, but in corrupt form; in other letters we find Nesennius (properly Nasennius) Apollinaris. But the greeting of the answer had already been reduced to a bare Respondit. As far as neuter est tutor we have the same text in the Digest, but with the remains of the greeting excised and igitur instead of ergo. The compilers doubt­less found igitur in their copy. In any case the text shows how mistaken it is to condemn a passage solely on account of initial igitur.^{[395] [396]}

(6) Responsorum libri xxiii.^[397] A collection, in the Digesta order, of numerous responsa, largely, though probably not entirely, derived from practice. Outside the Digest we have: (1) F.V. 94-112, 114-18, (2) Coll. 10. 9, (3) two citations in the Schol. Sinaitica'. 2. 4 and 11. 31. In the original each case was probably presented according to a fixed scheme: facts, question (introduced by quaero or quaesitum est),^[398] answer (introduced by Paulus re­spondit). This work, as usual, underwent post-classical revision, including probably considerable abbreviation. Sometimes the question has been struck out, sometimes the statement of facts as well, and the responsum is thus left to stand alone.^[399] The text has been condensed in other ways also, and it has been added to.

A comparison of F.V. 94 and D. (24.3) 49.1, representing the same passage of Paul, is highly instructive.¹

26. Ulpianus.

(a) Disputationum libri x* a collection of disputations in school and court,³ and possibly of quaestiones derived from other sources. At any rate the author often refers to his decision in the past tense: dicebam or dixi.⁴ The existence of such a work is vouched for by a citation in a rescript of Diocletian,³ and there is no sufficient ground for doubting its authenticity;⁶ but the classical text reached the compilers much altered by post-classical editing. The extensive Digest fragments teem with signs of post-classical work, which deserve a comprehensive critical study. We must be con­tent to illustrate from D. (15. 1) 32 pr.,⁷ which is duplicated by a fragment of a fifth-century parchment codex, discovered in Egypt and acquired by Strasbourg:⁸ '[sed licet hoc iure con]tingat, tarnen [aequitas] dictat resc[issorium iudi]cium in [eos dari].’ Except that the Digest omits rescissorium the texts are identical, so far as the Strasbourg fragment goes.’ The passage cannot be genuine: apart from the sentimental aequitas dictat¹⁰ the whole antithesis of ius and aequitas belongs to Aristotelian rhetoric,¹¹ not to Roman jurisprudence. Aequitas, if accepted by the law, is the law and can no longer be contraposed to ius.¹* A substantially

¹ Beseler, Z xliii (1922), 538; xlvii (1927), 360. F.V. 102 also shows obvious post- classical workmanship: see Albertario, St. v. 561; Beseler, Z xliii (1922), 539; xlv (1925), 457·

² Joers, PW v. 1446. ³ Above, p. 225.

* Dicebam: e.g. D. (27. 8) 2; (28. 4) 2; (28. 5) 35; (33. 4) 2; (35. 2) 82; (36.1) 23 pr.; (44· 3) 51 (49· 17) 9· Dicebamus: D. (29. 2) 42. 3. Dixi: D. (26. 1) 7; (28. 5) 35. 5; (36.1) 23. 4. See above, p. 225.

³ C. (9. 41) 11.1. Beseler’s doubt as to the authenticity of the citation (Z 1 (1930), 45) is unjustified.

⁶ This in spite of Beseler’s dogmatic pronouncements: St. Riccobono, i. 313, ‘we know now that Ulpian’s Disputaiiones are not by Ulpian’; cf. T x (1930), 190; also Z xlv (1925), 255, n. 1; 1 (1930), 45. The fact is that Beseler’s thesis is unproven and, with the existing evidence, unprovable; disputaiiones is no argument against authenticity—neither the word nor the thing: above, p. 225. What is true is that the work, like many other classical writings, was heavily edited in post-classical times, and this for practical purposes comes to much the same for the legal historian.

⁷ Beseler, Z 1 (1930), 45, giving the literature on this passage.

⁸ Ed. princeps, with photographs, by Lenel, SB. Berlin. Ak., phil.-hist. KI. 1903, 922; 1156. Later, Lenel, Z xxiv (1903), 416; xxv (1904), 368. Followed by Seckel- Kubler, i. 496; Girard-Senn, Textes, 494; FIRA ii (ed. 2), 308.

⁹ Its lacunae, supplied from the Digest, are given in square brackets.

¹⁰ Cf. Ammian. Marcell. 22. 6. 5: ‘unde velut aequitate ipsa dictante lex est promulgata.’ According to Thes. v. ion. 38, 1012. 84, this is the only text with aequitas dictat besides D. (15.1) 32.

241 correct statement would have been: 'licet hoc iure civili¹ contingat, tamen aequitas dictat rescissorium iudicium dari’, i.e. relief must be sought from the ius praetorium. Lenel regarded the Strasbourg text as disposing of previous doubts as to the authenticity of the Digest version. That was because, when he wrote (1904), scholars contemplated interpolation only by Justinian’s compilers: what did not come from them was necessarily classical.²

(6) Responsorum libri ii.³ Just a poor post-classical abridge­ment of Ulpian’s Responsa, but one which is shown by F.V. 44, the only fragment preserved independently of the Digest, to have been already in existence in the fourth century. The abridgement was radical: the statement of facts and the question were cut out, and only the responsum proper was left. Each responsum began with the name of the addressee followed by respondit governing the decision in the accusative and infinitive, from which it may be inferred that originally the work had been a collection of letters on legal subjects. Where our fragments lack this beginning it must have been excised by the compilers. The editor seems not to have altered the substance of the law stated.

27. lulius Aquila

Responsa, of which we have only two short fragments.⁴

28. Modestinus

(a) Responsorum libri xix. This collection, arranged on the Digesta plan, dealt chiefly, if not exclusively, with real cases from practice. The Digest fragments, which are all we have, show a relatively pure classical text. Clearly the original arrangement— facts, question, responsum—was preserved in the compilers’ copy. The question was given in full and the answer, which followed, was no doubt always introduced by the stock Herennius Modestinus respondit, with the decision in direct or oblique speech indifferently. The abridgements to be found in the Digest fragments are probably the work of the compilers. In other respects also the text is comparatively genuine, though not entirely free from pre-Justinian interpolation.®

(&) De enucleatis casibus liber singularis. To judge by its title⁶

¹ Although sed licet hoc iure contingat cannot be read in the Strasbourg fragments, there would not be room in the line for iure civili.

² Above, p. 142. ç Joers, PW v. 1446. ⁴ Berger, PW x. 167.

^s Thus in D (20. 1) 26. 2 hoc est in hypothecis is interpolated: M. Fehr, Beitr. (above, p. 203), 74. Again the rhetorical pompousness of D (1. 3) 25 shows it to be spurious beyond doubt: Beseler, Beitr. v. 13; Index Interp.

⁶ What is meant by ‘shelled’ cases is obscure; ‘abridged cases’? See Thes. v. 616.

4497.1 jj

this little work also belongs to the problematic group. The five Digest fragments make an unfavourable impression:¹ we seem to be dealing with a post-classical work.

(c) De heurematicis liber singularis. The paucity of the remains allows us only to assign this work to the problematic group. Its title is an enigma. No Greek word eu/Mj/xaTweds exists, and heure- maticus is found nowhere else in Latin literature.² There is, however, a Greek literature irepl dpripu£rfourth century and it contributed to the Fragm. Vaticana and the CoUaiio·, moreover, the anonymous liber de officio proconsulis¹ referred to in the Lexicon of the pseudo-Philoxenus² is probably Ulpian’s work or an abridgement of it. Finally, there are some fairly long extracts from it in the Digest. Comparison of the pas­sages which occur both in the Collatio and in the Digest shows that the text, apart from small variants, remained unchanged in the intervening period and that the interpolations of the compilers were, as usual, confined to modest limits. But the text used for the Collatio was not in all respects the classical. Diocletian’s reorganization had abolished the distinction between imperial and senatorial provinces; though there were still ‘proconsuls’,³ this was merely a title given to governors who had the distinction of being subject directly to the Emperor.⁴ Thus by proconsul a fourth-century reader would understand praeses prownciae and it was not necessary to change proconsul in the text to praeses provinciae. But other changes were made, which marred the plan and character of the classical original.

D. (47. n) 9 speaks of the crime of scopelismus, which, as the text expressly says, was confined to the province of Arabia. Again D. (47.11) 10 speaks of the destruction of the dams of the Nile. Since Arabia and Egypt were not senatorial provinces neither text can have been written by Ulpian, who, if he had wished to mention crimes confined to parti­cular provinces, would have sought his examples in some senatorial pro­vince or, finding none, would have said nothing. Laws special to this or that imperial province would have been no subject for a work de officio proconsulis. A later hand is likewise revealed when a text of the CoUaiio has praeses for proconsul,’ and praeses in the Digest passages is equally spurious, though we cannot tell whether the change was made before Justinian or, contrary to their usual practice of allowing proconsul to stand, by Justinian’s compilers. We cannot deal here with the post- classical additions of other kinds.⁶

Though the style of the work departs considerably from the classical,⁷ severe post-classical revision is scarcely probable, since in that case proconsul would have been systematically changed to praeses. The work belongs to a literary genus apart: writing for ¹ SeeRudorffstreatise. ¹ Goetz,PWvii. 1439. ³ C.Th. 1.12',C. Just. 1.35.

♦ Karlowa, RRG 856 ff.; Kubler, Gesch. 323 ff. ³ CoU. 3.3.1; 14· 3· r and 2.

⁶ See first the fragments preserved by the Collatio in Volterra’s Indice (RSDI 9, 1936), and then the Digest passages in Index Interpol, and Z Index i-1.

r So, rightly, Beseler, Beitr. iii. 39; v. 25. Also Z li (1931), 188, and elsewhere; Felgentrager, Symb. Frib. 371.

a wide, unprofessional circle of readers Ulpian would naturally use a more popular style than the strictly traditional classical juristic style of works on civil law.

3. De officio praesidis

Macer, libri ii.^[CD] Only a few Digest fragments.

4. De officio praefedi urbi

Libri singuläres by Paul (one short fragment) and Ulpian (one fairly long fragment).²

5. De officio praefedi vigilum

Libri singuläres by Paul and Ulpian; seven fragments of the former, one small piece of the latter.

6. De officio praefedi praetorio

Of the liber singularis of Arcadius Charisius we have only one frag­ment (the beginning of the work), but that in two versions: D. (1. Ii) 1 and Joh. Lydus (age of Justinian), De magistral. 1.14. Lydus’ version, which is in Greek, is probably not derived from the Digest but from the same edition of the work as that used by the compilers.³ The language of the Digest version departs from classical juristic usage. This in itself, as has been pointed out above,⁴ is no sign of unauthenticity, even if Charisius be regarded as a classical jurist.³

7. De officio quaestoris

We have of Ulpian’s liber singularis two fragments: D. (1. 13) 1, which corresponds with Joh. Lydus, De magistral. 1. 24. 28, giving an historical introduction. The language departs from classical usage:⁶ again no proof of unauthenticity.⁷ D. (2. 1) 3 seems to be a post- classical distinction

8. De officio curatoris rei publica#

Six fragments of Ulpian’s liber singularis.

9. De officio assessorum

Ulpian once cites a book of Massurius Sabinus called assessorium and a book of Puteolanus called assessoria. The character of these works remains problematic.¹⁰

Four Digest fragments of Paul’s liber singularis de officio assessorum,^[CDI] of which D. (1.18) 21 and (3. 3) 73 show post-classical workmanship.⁸

247

10. De officio consularium¹

One small fragment from Ulpian's liber singularis.^{[402] [403]}

11. The compilers possessed a work in Greek entitled Ilamvuwov 'Aotwo/mkos povdßißXos,^{[404] [405] but took only one passage from it for the Digest.* One can hardly believe that that 'true Roman’ Papinian[406] would have written on law in Greek, still less that he would have written on the astynomi of Greek cities, a subject of not the slightest importance to a man in his position. If an authentic work by Papinian did form the basis of the Greek work, it must have been in Latin and have been concerned with Roman officials; these, as Mommsen has shown, can only have been the quattuorviri viis in urbe purgandis.[407] [408] [409] On this view, what the compilers used must have been a Greek epitome.}

12. A specially interesting group is formed by the program­matic works for the praetor tutelarius, a magistrate created by Μ. Aurelius and from the fourth century called praetor tutelaris.¹

(a) Paulus, De excusationibus tutelarum liber singularis? The title is variously given: in the Index Florentinus (xxv. 31) it is de excusationibus tutelarum, in D. (27. 1) 26 and F.V. 231 de excusationibus, in D. (27.1) 11 and F.V. 246 de excusatione tutorum, and in D. (27. 1) 7 excusationes.

(&) Paulus, De officio praetoris tutelaris liber singularis.^[410]

(c)Paulus, De iurisdictione tutelaris.¹⁹ More than one book.

Our fragments of these three works, which are few, are of special interest for the history of classical juristic texts. Of (a) we possess D. (26. 3) 4 and (27. 1) 11, and F.V. 231 and 246. Of (o), which is registered by the Index Florentinus (xxv. 40), but was not used by the compilers, we have only F.V. 244 (cited D. (27.1) 6.19) and 245. Of (c) (not in the Index Florentinus) all we know is F.V. 247.

Scanty as this evidence is, a judgment on the authenticity of the three works is nevertheless permissible. We have no reason to doubt the authenticity of (&), the classical title of which was of course de officio praetoris tutdarii. The versions of D. (27. 1) 6. 19 and F.V. 244

differ, but for this abbreviation by Justinian’s compilers is mainly responsible however, the interpolation ex epistula divorum (Hadriani et Antonini et] fratrum shows that the edition used for the Fragm. Vaticana was already interpolated in places.

Work (c) can only be another edition of work (&), as may be seen from the only surviving fragment, F.V. 247. The inscription is : ‘Paulus libro i editionis secundae de iurisdiciione tutelaris.’ The classical title would have been de iurisdiciione praeloris tutelarii, which would mean precisely the same as de officio praetoris tutelarii. We must therefore refer the inscription to a second edition of work (6). This second edition was not by Paul himself, but, like the second edition of Ulpian Ad Sabinum, by some post-classical editor,^{[411] [412] as is shown by the following consideration. If Paul had written a second edition, enlarged to more than one book, this would have entirely superseded, or at least have overshadowed, the liber singularis of the first edition, whereas it was obviously quite un­known to the compilers: the Index Florentinus mentions only the liber sing, de off. praet. tut. Moreover, the text of the second edition shows signs of post-classical workmanship. A decisive point is that it refers (F.V. 247) to Severus and Caracalla as domini nostri. This is the only juristic passage which so refers to the Emperors ;[413] during the Principato one addressed the Emperor orally[414] and by letter[415] as dominus, but otherwise scrupulously avoided a title which would have been a breach of the ‘republican’ manners affected by the Principate.[416] [417] And there are other post-classical stigmata in the passage.}

Finally, work (a), De excusationibus, was in all probability merely a post-classical abridgement of work (6), De off. praet. tut. The latter must necessarily have dealt with the law of excusationes, and it is un­likely that Paul returned to it in a separate work. Be that as it may, the text of the De excusationibus (a) used for the Fragm. Vaticana was largely of post-classical making. The post-classical equation of cura with tutda¹ occurs in one of our two passages, F.V. 231:^[418] the addition of curatio is clear from the fact that the editor has neglected to complete it at the end of the passage. Moreover, the words velpermixto... separa- las are spurious, and the final clause has been rendered unintelligible by abridgement.¹

(d) Ulpianus, De officio praetoris tutelaris.¹

(e) Ulpianus, De excusationibus liber singularis.³ The Index Florentinus (xxiv. 22) shows that the compilers possessed work

(i) ; from it they took D. (27.1) 3,5,9/ It is cited once, title given, in a post-classical addition to Modestinus’ work⁵ (D. 27.1. 6.13), and we have the long series of fragments: F. V. 173-223; 232-6; 238-42.

Work (e) is not in the Index. Apparently the compilers had no copy of it, but obtained the single Digest fragment (D. 27. 1. 7) from the post-classical additions to Modestinus, which refer to it as Ulpianus libro singulari de excusationibus⁶—an incontestable proof of the existence, though not of the authenticity, of the work.⁷ Then, too, there is F.V. 123-70: the beginning of the first fragment, which would have given author and work, has perished, but Mommsen has shown that the whole series should be ascribed to Ulpian’s liber de excusationibus.⁶ (i) F.V. 145 and 151 recur in 222 and 223, which come from Ulpian’s liber de off. praet. tut. Again, two passages ascribed in the Digest to Ulpian’s liber de excus. (D. 27.1. 7; 15.16) recur as F.V. 185, 240, and 189, which are given as coming from Ulpian’s liber de off. praet. tut. Thus it is certain that F.V. 123-70 come from a work bearing Ulpian’s name, (ii) That work can only be the liber de excus.; the extracts from the liber de off. praet. tut. begin only at no. 173.

The relation between the two works is that work (e), De excus., is a post-classical abridgement of work (4), De off. The latter must neces­sarily have dealt with the law of excusationes,⁹ and in fact we find four of our fragments of (e) in a more extended form as fragments of (i): F.V. 145 = 222; 151 = 223; D. (27.1) 7 = F.F. 185, 240; D. (27. 1) 15. 16 = F.V. 189. Thus the liber de excus. is an abridgement of the liber de off. praet. tut. Mommsen held the converse, namely that the liber de off. was an enlarged edition, by Ulpian himself, of the De excus., but this cannot be accepted. The passages from the liber de excusationi­bus are typical epitomist’s work; in their shortened form they cannot possibly have been written by Ulpian. Mommsen was led to his opinion by his having observed that in the liber de officio Septimius Severus is always referred to as dead (divus), but not in the liber de excusationibus.

¹ What the epitomist meant to say is shown by F.V. 186. On permixto modo see D. (27.1) 2. 9, with the Index Inlerp.

² Joers, PW v. 1452. » Ibid. x. 1451.

* The dominant view, that the compilers obtained these fragments from Modestinus’ work, is wrong.

⁸ Below, p. 252. ⁶ D. (27. 1) 15. 16.

r H. Kruger, St. Bonfante, ii. 323, denies its existence !

⁸ In his edition of the apographum of the Fragm. Vaticana (below, p. 344), p. 394; Joers, PW x. 1454. Kruger’s objections (Quellen, 247, n. 191) cannot be allowed; F.V. 233 and 235 are by Paul: see Mommsen in his edition of the Fragm. Vaticana (Collectio libr. iii. 72). · But not that only: Joers, PW x. 1454. But the evidence in the latter work is very slight—only five passages : F.V. 158 pars orationis imperatoris Severi; and 125,147, and 159 (pro­bably also 149) imperatores nostri. Mommsen’s observation is correct, but it does not justify us in concluding that the liber de excusationibus was written while Septimius Severus was still alive. The epitomist shirked the long title imperator noster et divus Severus and so wrote simply imperatores nostri.¹ Thus we read F.V. 246, 'Paulus libro singu­lari de excusatione tutorum. Imperatores nostri Aelio Diodoto suo salutem*, where obviously the abbreviator (see above, p. 248) is respon­sible for imperatores nostri, since Severus and Caracalla cannot have referred to themselves in these words.

The liber singularis de officio praetoris tutelarii was a genuine work of Ulpian’s, but in the version used for the Fragm. Vaticana the classical text had already undergone alteration by a later hand. In particular, as in the liber de excusationibus, cura is put on the same footing as tutela.²·.

(/) Modestinus,De excusationibus² This work, written in Greek, is of special interest. We possess a considerable number of rather long extracts in the Digest and among them, what is rare in our juristic remains,⁴ the beginning of the work.⁵ This consists of a dedicatory epistle addressed to an otherwise unknown Egnatius Dexter⁶ and giving the title of the work (Παραίτησής επιτροπής) and, as was the literary custom, an outline of its scheme. It is explained that the work is in Greek, though the author well knows that it is considered difficult to expound Roman law (τά νόμιμο.) in that language; that statutes are quoted textually in order that it may be easy for advocates to quote them in court; and that it is believed that a very useful book has resulted. This epistle is no forgery. It is written throughout in the style of Greek dedicatory epistles and is governed by their topology, as may be seen by those interested in such matters if they will compare it with the prefaces of the Pseudo-Skymnus and of the anonymous Stadias- mus.⁷ The remark on the difficulty of translating Roman jurispru­dence into Greek reminds one of Lucretius’ lines⁸ on the similar difficulty of translating Greek philosophy into Latin.

¹ Cf. Riccobono junior, Misc. 48. ² See the literature cited above, p. 248, n. 7.

³ Older literature: Ant. Augustinus, Lib. sing, ad Modestinum sive de excusationibus (Venice, 1543, and in Otto’s Thesaurus, iv. 1559). Modem: Peters, Z xxxiii (1912), Six; Ebrard, Die Digestenfragmente ad formulam hypothecariam (1917), 144; Η. KrQger, St. Bonfante, ii. 315; Brassloff, PW viii. 670.

♦ There seem to be only the beginning of Gaius on the Twelve Tables (above, p. 187), that of the work de adulteriis ascribed to Paul (above, p. 188) and that of Arcadius Charisius’ De officio praefecti praetorio (above, p. 246).

³ D. (27. 1) i pr. ⁶ See Prosopogr. Imp. Rom.

r Geogr. gr. min. 1.196, 427. ⁸ 1. 136 ff. Against him Cic. De fin. 1. 3.10.

(i) There is no doubt either that Modestinus is the author or that he himself wrote the work in Greek. It is the only classical juristic work so written,¹ but Modestinus is the latest of the classical writers. His exact motives for writing in Greek can only be conjectured. He certainly had connexions with the East:³ he writes to his teacher Ulpian from Dalmatia,³ and we find questions put in Greek among his responsa. He may have been induced by Egnatius Dexter to write in Greek, just as Eike von Repgow was induced by Graf Hoyer von Falkenstein to write the Sachsen- spiegd in German. The Constitutio Antoniniana had made the need of books on Roman law in Greek specially urgent. Gregorius Thaumaturgus, a contemporary of Modestinus, who had learnt Latin and was an advocate, declares that the marvellous Roman law, which (since the Constitutio Antoniniana) applied to all subjects of the Empire, is difficult to leam, because, for all its excellence, it is in Latin and very difficult for him to read.⁴ It is obviously erroneous to suppose that the work is a translation of Modestinus’ Latin by ‘some Byzantine’ ;^s in view of the fact that the Byzantines had long learnt how to translate Roman jurispru­dence into Greek, no Byzantine would have remarked on the difficulty of so doing. The remark suits Modestinus, but not a Byzantine.

(ii) The title is given by the dedication as Παραίτησις emrpomjs,⁶ by the Index Florentinus (xxxi. 5) as excusationum βιβλία ίξ, and by the Digest inscriptions as libri excusationum. A Greek work, however, would naturally have a Greek title.

(iii) The dedication promises to quote the actual words of the statutes (νόμοι), so far as necessary. This means the imperial constitutions.⁷ Our fragments do at times give the texts of con­stitutions, and as a rule in Latin.⁸ Where the text is given in Greek, this was presumably the language of the original constitu­tion. An example is Antoninus Pius’ epistula to the κοινόν of Asia.⁹ Otherwise the Latin text is given because, as the dedication

¹ On Papinian’s αστυνομικοί: above, p. 247; Maecian: below, p. 255.

² Petere, Z xxxiii (1912), 513. ³ D. (47. 2) 52. 20.

⁴ Paneg. ad Originem, I. 7: ‘ ol Θαυμαστοί ημών νόμοι, ots νυν τά πάντων των υπό τήν Ρωμαίων αρχήν ανθρώπων κατιυθννίται πράγματα, οντ* συγκΐίμχνοι οντν καί ικμαν- θανόμίνοι άταλαιπώρωί· Svres μίν αυτοί σοφοί τ« καί άκριβίϊί καί θαυμαστοί καί συνιλόντα ίίπτΐν Έλληνικώτατοι' ίκφρασθίντκ καί παραβοθάντΐί τβ Ρωμαίων φωνή... φορτική H όμωί Ιμοί.’ ³ So Η. Kr&ger, St. Bonfante, ii. 323.

⁶ Modestinus’ translation of excusatio tutelae; LiddeU and Scott give ‘excuse, declining’ for παραίτησίί.

i D. (27. i) 6. 9.

⁸ D. (27. 1) IO. 4; 13. 12; 15 17; (26. 6) 2. 2.

• D- (27. 1) 6. 2.

says, the intention was that the constitutions should be readable in court straight out of this book; it seems, therefore, that even in the provinces Latin constitutions were read out in Latin.¹

(iv) There are citations of Paul and Ulpian; quotations from them are usually in the original Latin. Every one of these Latin texts must have been inserted in post-classical times: it is not credible that so early as Modestinus classical juristic writings should have been put on the same footing as imperial constitutions. Modestinus had no reason for not translating into Greek any juristic text that he wished to quote. It does not seem to have been classical practice to read out in court passages from the jurists; this became established later, when the classical literature had been elevated to the rank of ius.

The quotations from Paul and Ulpian fall into two groups. In the first everything is in Latin—name of jurist, title of work, book-number, and text. The citations of the second group begin with a clause in Greek, giving the jurist’s name, but not the title of the work; then follows (though not always) the text in Latin. About the first group there is no difficulty: the quotations must, as has long been recognized, all have been inserted after Modestinus, who, if he desired to quote a text in Latin, would surely have introduced it by a Greek phrase. But they cannot have been inserted by the compilers, so that they constitute a proof that Modestinus’ work underwent a post-classical revision. The second group is less straightforward; but here too it may be claimed that the quotations in Latin are not due to Modestinus.

(v) There are other indications of a post-classical revision which cannot be dealt with here. An example is the equation throughout of tutela and cura.²

(ix)

We are left with a considerable number of works which we cannot fit into our classification, partly because we know very little about them. Of many we possess but a few fragments, of some only the title registered by the Index Florentinus. The libri singulares among them are in part merely classical or post-classical separate editions or post-classical abridgements taken from por­tions of larger works; this holds in particular of Paul’s numerous libri singulares. Here it is sufficient to give a summary conspectus of the works in question, with short remarks on notable points.³

¹ See the report: Bruns, no. 69 (Wilcken, Chrest. 462) and E. Weiss, Z xxxiii (1912), 223. It may be that there have been fresh discoveries later.

² Above p. 248, n. 7. ³ See for the fragments Lenel, Pal.

(1) Law of Persons

De manumissionibus: Gaius, libri Hi; Modestinus, lib. sing.

Paulus, De liberali causa lib. sing, and De articulis liberalis causae lib. sing, are probably one and the same: Beseler, Beitr. iii. 202; Berger, PW x. 716 contra. See above, p. 196.

Paulus, De adsignaiione libertorum lib. sing, and De libertatibus dandis lib. sing., both probably consisting of extracts from larger works.

Paulus, De iure patronatus lib. sing, and De iure patronatus quod ex lege lulia et Papia venit (known only from the Index) are probably identical and consisted of extracts from larger works. Above, p. 188.

Ulpianus, De sponsalibus lib. sing. Presumably extracted from a larger work.

Neratius, De nuptiis. Known only from a citation by Gellius 4. 4.

Modestinus, De ritu nuptiarum lib. sing.

Gaius, Dotalicion. Only in the Index.

Paulus, De dotis repetitione lib. sing. Doubtless only an extract.

Paulus, De donationibus inter virum et uxorem lib. sing. Only in the Index. Doubtless only an extract.

Tertullianus, De castrensi peculio lib. sing.

Paulus, De gradibus et adfinibus et nominibus eorum lib. sing.' This is of special interest, both as providing a good example of a post-classical forgery and as being connected with the pseudo­Pauline Sententiae. We have only one, very long, fragment (D. 38.10. 10). In Cujas’s day a complete copy of the work existed; its owner assured Cujas that the Digest fragment contained nearly the whole of the text given by the manuscript.² The work is assuredly not by Paul: it is a piece of professorial triviality, de­void of any juristic value and unmistakably post-classical in style.

The fragment begins by telling us that a jurisconsult (no less I) ought to know the grades of consanguinity and affinity; no classical writer could have penned this. Then s. 5: 'non parcimus (!) his nominibus, id est cognatorum, etiam in servis’, and s. 7: 'Parentes usque ad tritavum apud Romanos (!) proprio vocabulo nominan­tur’, and lastly s. 10: ‘Gradus autem dicti sunt a similitudine scalarum locorumve proclivium, quos ita ingredimur, ut a proximo in proximum, id est in eum, qui quasi ex eo nascitur, transeamus’ —these passages speak for themselves, especially the last, which is typical of the would-be learned post-classical schoolmaster. In

¹ Index Flor. xxv. 29; Pal. i. 1103; Pringsheim, ‘Beryt u. Bologna’ (Festschr. Lenel, 1921), 279; Scherillo, * Sul tractatus de gradibus cognationum ’ (St. Cagliari, xviii. 1931). We will not here dwell on the anonymous Tractatus de gradibus cogna­tionum (Collect, libr. ii. 166; Seckel-Kübler, ii. 182; Girard-Senn, 502; FIRA ii. 631): see Krüger,Quellen, 286. On the Stemmata cogn.: Alberti, * Lo stemma cognationum RSDI v (1934). ² Obsero, vi. 40. the classical age the term gradus vias current coin. This, we submit, is proof enough.

In the Digest our passage (D. 38. 10. 10) is preceded by a frag­ment from Paul’s Sententiae, which, after a short remark about ascending and descending lines, ends: ‘quas omnes latiore tractatu habito in librum singularem conteximus.’¹ This sentence cannot come from Paul.² Equally clearly it cannot come from the compilers, such a transition from one fragment to another being foreign to their manner. More likely the sentence was already in the post-classical version of the Sententiae from which the compilers took the extract ; it was also in the version used by the Visigoths, though they did not include the text known to us as D. (38. 10) 9 in their Breviarium, 4.10. This reference by the Sententiae to a post-classical lib. sing, de gradibus must have been inserted by a post-classical hand. It is therefore a mistake to regard it as a proof of Paul’s authorship of the Sententiae.³

(2) Law of Property

Nerva, De usucapionibus ; Pal. i. 791. Only citations.

Paulus, Iltpl bwairoandtrriav (i.e. on things difficult to separate) in at least two books. All we have is a Greek gloss. Pal. i. 966.

(3) Law of Obligations

De verborum obligationibus.

Pedius, De stipulationibus ; Pal. ii. 8, where more than one book is rightly assumed ; Ferrini, Opere, ii. 42, is wrong. Only one citation.

Gaius, De verborum obligationibus libri Hi.

Pomponius, De stipulationibus, in at least eight books. Only a citation.

Venuleius, De stipulationibus libri xix.

Paulus, De intercessionibus feminarum lib. sing. Doubtless identical with the lib. sing, ad senatusconsultum VeUaeanum. Above, p. 189.

Massurius Sabinus, De furtis. Cited only by Gellius, who is presum­ably referring to the title De furtis in Sabinus* Ius civile.*

¹ Scherillo’s view, op. cit. pp. 25 ff., is untenable.

² This is the only case where a juristic work uses ‘contexere’ metaphorically: Voc. i. 981.52 ; Thes. iv. 692.72. Incomprehensibly Volterra, ‘ Sull’ uso delle sententiae di Paolo’, Atti Congresso 1933, Bologna, i (1934), 164, takes the passage as genuinely Pauline: ‘Nelle Sententiae Paolo cita se stesso in prima persona. È questo un indizio che non trovò ancora preso in esame dall’ opinione dominante ( !).’ But the right view had already been taken by Pringsheim, Beryt u. Bologna, 279.

³ As Volterra (last note) does.

♦ Observe Gellius’ manner in citing. In 10.6 he says: ‘id factum esse dicit Capito Ateius.in commentario de iudiciis publicis...’; but he himself informs us in 4. 14 that Capito’s book ‘qui inscriptus est de iudiciis publicis’ is nothing more than book 8 of his Coniedanea.

Paulus, De iniuriis lib. sing. Just a post-classical extract. Above, P· 795·

Paulus, De usuris lib. sing.

Maecianus, Ex lege Rhodia.¹ We have only D. (14.2) 9, in Greek. Not in the Index. The contents show that either Maecianus is in no sense the author or else that the compilers derived the fragment from a Greek paraphrase of his work or somehow otherwise at second hand. It is unlikely that Maecianus wrote the work in Greek. The text runs: 'Antoninus said to Eudaemon: “I am lord of the world, but the law is lord of the sea.” ’ Impossible, from either Antoninus Pius or Μ. Aure­lius ; nor is interpolation by the compilers a possible explanation?

(4) Law of Inheritance

Paulus, De testamentis lib. sing, and De forma testamenti lib. sing, are doubtless identical.

Modestinus, De testamentis lib. sing. Only in the Index.

Paulus, De secundis tabulis lib. sing.

Paulus, De inofficioso testamento lib. sing.

Modestinus, De inofficioso testamento lib. sing.

Paulus, De septemviralibus iudiciis lib. sing. So the Index (xxv. 46) and the inscriptions of our four Digest fragments, all of which deal with the querdla inoff. test. There is a natural inclination to alter this un­known tribunal to centumviralibus throughout, Pal. i. 957, but the emendation cannot be accepted: Eisele, Z xv. 283; xxxv. 320, giving the literature.

Paulus, De iure codicillorum lib. sing.; Μ. Scarlata Fazio, La succes­sione codicillare (1939), 215.

De legatis et fideicommissis.

Valens, De fideicommissis libri vii.

Gaius, De fideicommissis libri ii; De tacitis fideicommissis lib. sing. Pomponius, De fideicommissis libri v.

Paulus, De fideicommissis libri iii. De tacitis fideicommissis lib. sing. Ulpianus, De fideicommissis libri vi.

Modestinus, De legalis et fideicommissis lib. sing. Only in Index. Paulus, Ad regulam Catonianam lib. sing.

Paulus, De instrumenti significatione lib. sing. In the Index (xxv. 58) described as de instructo et instrumento.

Paulus, De legitimis hereditatibus lib. sing. Only in Index.

(5) Law of Actions

Paulus, De actionibus lib. sing. Only in Index.

Paulus, De conceptione formularum lib. sing. ^{[419] [420]}

Paulus, De concurrentibus actionibus lib. sing.

Modestinus, De praescriptionibus lib. sing.

Venuleius, De interdictis libri vi.

Arrianus, De interdictis, at least 2 books.

(6) Private Law in General

Paulus, De iuris et facti ignorantia lib. sing. Post-classical: see Index Interp. on the one fragment, D. (22.6) 9, and Ebrard, Z xlv. (1925), 118.

Paulus, De iure singulari lib. sing. Hardly genuine, though Orestano, Ius singulare e privilegio (1937), treats it as such.

(7) Procedure

Paulus, De cognitionibus lib. sing.; Berger, PW x. 716, giving the literature. Of our seven fragments six deal with excusatio tutdae. Apparently post-classical, which would explain the variations in the titles given to the Emperors.

Callistratus, De cognitionibus libri vi.

Paulus, De iure libellorum lib. sing. Berger, PW x. 719.

Arcadius Charisius, De testibus lib. sing. Post-classical in style.

De appellationibus.

Paulus, lib. sing.

Ulpianus, libri iv. Marcianus, libri ii. Macer, libri ii. Post-dassical revision: Beseler, Beitr. ii. 142.

Paulus, Ad municipalem lib. sing.; Berger, PW x. 709. Above, p. 196.

Ulpianus, De omnibus tribunalibus libri x.^[421] The title sounds un- classical, and that of ‘Protribunalia’ given by the Index and Lydus, De magistral. 1. 48, is even less normal. Plan unrecognizable; contents post-classical in character. It may well be a post-classical collection of excerpts from Ulpian, with alterations and additions.

(8) Criminal Law

De iudiciis publicis. Maecianus, libri xiv. Venuleius, libri Hi. Marcianus, libri ii. Macer, libri ii.

Paulus, De poenis omnium legum lib. sing. Only in the Index.

Paulus, De poenis paganorum lib. sing.

Claudius Saturninus, De poenis paganorum lib. sing. One long frag­ment (D. 48.19.16). In the Index (xxi. 4) under ‘Venuleius Saturninus’. The fragment is unclassical in style: Beseler, Z li (1931), 198.

Paulus, De extraordinariis criminibus lib. sing. Only in Index. Modestinus, De poenis libri iv.

(9) Fiscal Law

Callistratus, De iure fisci et populi libri iv.

Paulus, De iure fisci libri ii.

Paulus, De portionibus quae liberis damnatorum conceduntur lib. sing. Marcianus, De delaioribus lib. sing.

Fragmentum de iure fisci (so-called).¹ The unknown source can hardly have been a special treatise de iure fisci.

Paulus, De censibus libri ii.

Ulpianus, De censibus libri vi.

Arcadius Charisius, De muneribus civilibus lib. sing.

(10) Military Law

De re militari.

Tarrutenius Patemus, libri iv. See above, p. 106.

Menander, libri iv. Macer, libri ii.

Paulus, De poenis militum lib. sing.

From this survey the important fact stands out that classical jurisprudence produced hardly any monographs. Those we have assembled (to which the works ad formulam hypothecariam men­tioned earlier should be added), so far as they are in any sense classical, are either very short or else deal with matters for which there was not room enough in the ordinary systematic works, such as fideicommissa, appeals, and criminal, fiscal, and military law. The one exception is that the law of stipulaiio received mono­graphic treatment from at any rate Pomponius and Venuleius. It could not have been otherwise. A legal science which eschewed legal history, law reform, and legal philosophy, which laid stress mainly on case law and problems and was only very mildly interested in system and abstraction, contained no place for a monographic literature of the modem type. And yet one may see in the lack of such a literature one reason why the stream of classical literature eventually ran dry. No systematic work could hope to outdo Ulpian’s great commentaries, and though problema­tic literature might have been further spun out, since casuistry is in its nature inexhaustible, one can well understand that satiety was at length reached. Only by monographic literature could new paths have been opened and explored.

¹ Collect, libr. ii. 162; Seckel-Kubler, ii. 172; Girard-Senn, Textes, 499; FIRA ii, ed. 2, 627.

4497.1 _s

(x)

We close our survey of classical juristic literature with a brief characterization of the legal language of the period.¹ Legal. language was no more uniform than in the preceding period:^{[422] [423] it was different in each of the various genera of legal literature.}

1. About the language of the leges and senatusconsulta there is nothing to add to what has been said of the previous period.^[424] As to the style of the orationes principum it is impossible to generalize, since it would depend on the personal taste of the emperor.^[425]

2. As we have related,^[426] the praetorian and aedilician Edict was revised by Julian at Hadrian’s order and stereotyped by means of a senatusconsult. Here and there, no doubt, Julian made stylistic alterations in the traditional text, but he was much too conserva­tive to recast the whole Edict into the style of a second-century jurist. This means that even his stereotyped Edict lacks linguistic uniformity. Side by side with terse clauses in the regular juristic phraseology stand passages in the antiquated style proper to the magistracy. This is equally true of the individual edicts (edicts in the narrow sense) and of the edictal formulae. Combinations of actio, exceptio, and replicatio produce highly complicated and stylistically obscure formulations. No linguistic study of the Edict and its formulae exists.

For example, the edict on restitutio in integrum of maiores xxv annis is a single long and involved sentence ending: ‘in integrum restituam quod eius per leges plebis scita senatus consulta edicta decreta principum licebit.’ Here quod is used in the sense of si, which is characteristic of the ancient official style: e.g. in the L. Cornelia de xx quaestoribus (Bruns i. 89, 1. 4-5) we read: ‘quod ( — si) sine malo pequlatuu fiat’, and again in the formulary demonstratio (Gaius 4. 40) we find quod used in the same sense.⁶ The genitive eius in the above-quoted edict is a genitive of respect or relation, which also is characteristic of the official style.⁷ Thus it is found in the common clause of leges·, ‘eius hac lege nihil rogatur’, and again in the eius... condemna of a formulary condemnaiio (Gaius 4. 40). The edict de noxalibus actionibus has vel deierare iubebo; now deierare ( = deiurare) is not used by the classical jurists,¹ but belongs to the ancient official phraseology:* see, e.g., the Lex repetundarum, s. 19 (Bruns, no. 10).

3. In general the style of the writings of the classical jurists conforms strictly to the republican tradition.³ It is a professional form of speech⁴ and therefore a thing apart, diverging in many respects from common usage.⁵ It falls under the genus tenue⁶—a plain, unadorned style, which disdains all rhetorical artifice and aims solely at simplicity and exactitude.⁷ Things are called by their technical names and by them alone,⁸ even though such terminological strictness produces a certain monotony. Neologisms and metaphors are sternly eschewedunusual words, archaisms especially, are shunned like the plague ;¹⁰ so also any expression savouring of sentiment or pomposity.¹¹ Passion, pathos, and emotion in expression are taboo; the tempo of the exposition is a serene andante.™ It goes without saying that rhythmic clausulae were not affected.¹³ For so severely professional a science as that of the classical jurists such a style was, according to the canons of antiquity, becoming (npenov, decorum).¹⁴ Further, it harmonized with the whole intellectual attitude of the jurists, especially with their decided distaste for rhetoric.¹⁵ Its clear and impressive

^I Voc. ii. 142. ¹ See Thes. v. 403. ³ Above, p. 97.

⁴ Legal Latin is not considered by J. Svennung, Untersuchungen zu Palladius und zur lateinischen Fach- und Volkssprache (1935).

³ Quint. Inst. ii. 2. 41 i.f.: ‘... magis ab usu dicendi remota, qualia sunt iuris- consultorum. ’

⁶ Cic. Or. 5. 20; Quint. Inst. 12. 10. 21, 59.

⁷ Cic. Or. 23. 78 f. set aside all rhetorical curling-tongs and rouge-pots; ‘elegantia’ (on this expression see below, p. 335) ‘modo et munditia remanebit! Sermo purus erit et Latinus, dilucide planeque dicetur... unum aberit: ornatum. ’

⁸ Quint. Inst. 5.14. 34: ‘ iurisconsulti, quorum summus circa verborum proprieta­tem labor est. ’

⁹ Cic. Or. 23. 81: ‘Ergo ille tenuis orator, modo sit elegans, nec in faciendis verbis erit audax et in transferendis verecundus. ’

¹⁰ Julius Caesar (Geli. 1.1.10) advises: ‘ ut tamquam scopulum sic fugias inauditum atque insolens verbum. ’ Cic. Or. 24. 81: parcus in priscis. Tubero was fond of archaisms; the classical age disliked his writings—D. (1. 2) 2. 46: ‘sermone etiam antiquo usus affectavit scribere, et ideo parum libri eius grati habentur. ’

^II e.g. coniux, repulsa, humanitas', cf. Schulz, 81.

¹² Cic. Or.19.63 f.: ‘ Loquuntur cum doctis... Mollis est enim oratio philosophorum et umbratilis... nihil iratum habet, nihil invidum, nihil atrox, nihil miserabile, nihil astutum. *

¹³ So expressly Quint. Inst. 11. 2. 41: ‘solutiora numeris... qualia sunt iuris consultorum.’ Cf. Cic. Or. 19. 64: nec vincta numeris·, ibid. 23. 77. Thus Rechnitz, Studien zu Salvius lulianus (1925). is fundamentally in error; cf. Ed. Fraenkel, Z xlii (1927), 396 ff., 405.

¹⁴ Cic. Or. 21. 70; 36.124; 21. 72: ‘indecorum est de stillicidiis, cum apud unum iudicem dicas, amplissimis verbis... uti.’ Here de stillicidiis is equivalent to de iure civili (pars pro toto) ;cf. Cic. De leg. 1.4.17. Quint. Inst. 2. 10. 5.

³³ Above, p. 119.

objectivity is yet another manifestation of Roman greatness and sobriety. Naturally the style of the jurists was not absolutely uniform, but in a profession which gave scanty scope even to scientific individuality¹ the stylistic idiosyncrasies were trifling. A more important point is that in certain kinds of juristic literature (legal history, or instructions as to official conduct such as Ulpian’s libri de officio proconsulis) the standard juristic style was not strictly adhered to.^{[427] [428] The same is, of course, true of uncompleted, posthumously published works like Gaius’ Institutes.[429]}

It is only by recent critical research that this characterization of the classical juristic style has been made possible. Since, as we have shown, all the surviving writings have come to us from the post-classical period with their texts more or less seriously revised, it is only by the study of the interpolations that the special characteristics of the classical style can be revealed. Less recent researches, coming from so late as the end of the nine­teenth and the first decade of the twentieth century, are now out of date.^[430] Their authors were blinded by false presuppositions as to the nature of our literary tradition.^{[431] [432] Moreover, their lexical apparatus was defective. Since 1910 the study of the linguistic usage of the classical jurists has been very active under the leader­ship of G. Beseler. Almost every modem work on Roman law contains at least some observations on the subject. We are still far from our goal; many points are still (often, it must be said, erroneously) disputed. As yet no comprehensive critical study of this widely scattered literature exists; Guameri Citati’s Indict is, however, a valuable guide. The almost finished (perhaps already finished) Vocabularium lurisprudentiae Romanae furnishes an exhaustive index of the passages in which a given word occurs. Then we have vocabularies of the Institutes of Gaius and of the fragments of Celsus and of Callistratus.[433] Dirksen’s and Seckel’s dictionaries are out of date,[434] though the latter is still indispensable.}

4. The imperial constitutions of the period are couched in yet another style,¹ though in general this is simple and unadorned and obviously seeks not to diverge far from the juristic style. But the Chancery had its own special formulae, which were alien to the jurists; at times it is more rhetorical than they. At times, also, the language is influenced by the Emperor’s personal taste. A linguistic study of the constitutions has not yet been attempted.²

5.Finally, business documents have their own linguistic pecu­liarities, so that it is erroneous to argue from their language to the juristic: it does not follow, merely because an expression is employed in business documents, that it was also employed by the jurists.³

¹Vemay, Audes Girard, ii (1913), 266; Schulz, 82.

² Materials collected in: Haenel, Corpus Legum ab imperatoribus Rom. ante lustinianum laiarum (1857)—out of date, but still useful, especially on account of its extensive indexes. A collection of the classical constitutions has been planned in Italy. The first fasc. has appeared (Acc. dei Lincei).

³ Cf., e.g., Schulz, J RS xxxii (1942), 87; xxxiii (1943).