[In the following excerpt, Williams examines edicts and letters of Hadrian as documentary evidence that sheds light on the emperor's personal traits.]

1. INTERNAL EVIDENCE AND THE AUTHORSHIP OF IMPERIAL CONSTITUTIONS

A considerable number of texts of official pronouncements of Roman emperors (which will be referred to, rather inaccurately,¹ as constitutions, for the sake of brevity) have been preserved on inscriptions, in papyri and in the writings of the classical jurists and the imperial Codes. Such texts provide the kind of documentary evidence which is regarded by historians of more recent periods as primary material, as narrative histories and biographies are not.² Such rigour is not possible in Roman history, but it is clearly desirable to make the best use of these documents, especially in a period such as that between Hadrian and Commodus, when they are especially plentiful but, as far as literary sources go, there is not even the doubtful light of a panegyric to supplement the glimmerings of an epitome. But the kind of use to which they are put depends on the question of their authorship: are they to be treated as the work of the individual emperors in whose names they were issued, or of a civil service which continued to operate in much the same way while emperors came and went? The latter hypothesis has commonly been taken for granted until recently, even if not explicitly stated and argued for. In 1967 Millar showed that all the external descriptions of the methods of work of the emperors took it for granted that the emperors normally dealt in person with problems presented to them and produced the constitutions issued in their names themselves: hence those who wished to uphold the second hypothesis described above would have to prove that ‘vast ranges of imperial business were handled by the bureaux in private, systematically concealed from the view of our literary sources.’³

If this hypothesis, which might be termed that of ‘universal imperial authorship’, is correct, one can hope to find each emperor revealing his own individual personality in his constitutions. The main aim of this paper is to see how far individual attitudes and idiosyncrasies of style can be detected in the considerable body of texts from the period of Hadrian and the Antonines. This period is a particularly appropriate one in which to test this hypothesis. In the first place, it is the period for which the non-juristic evidence is richest, and constantly being increased by epigraphic discoveries.⁴ Secondly, it follows immediately upon the supposed increase in ‘bureaucratization’ by Hadrian. Thirdly, Antoninus and Marcus both professed a deep devotion to the memory of their predecessors: if their pronouncements reveal a marked individuality, those of less ‘pious’ rulers could certainly be expected to do so. This investigation extends into the succeeding reigns that of Trajan's letters to Pliny by Sherwin-White.⁵ He sought to distinguish in the texts ‘the hand of the secretary’ from ‘the hand of the emperor’ by an analysis, not only of the ‘outer style’ (i.e., vocabulary and style in the ordinary sense), but also of the ‘inner style’ (i.e., principles and attitudes expressed in or underlying a text). This combination of methods will be applied to the texts discussed below, but with one major difference. It has to be admitted that only in the case of a minority of the texts can such analysis show that a document was entirely the work of the emperor in person. However, one should not assume that passages or whole texts written in routine or formulaic language or in stiff jargon are necessarily the work of a secretary, as Sherwin-White does. If the hypothesis of ‘universal imperial authorship’ is correct, the emperors would be very busy and one would expect them to use standard formulae to save thought and effort. Similarly the recapitulation of the original message to which a reply is being written need not be a ‘secretarial device’: private persons with the duty of answering family letters and with nothing in particular to say often resort to this device, and it would be especially useful in formal replies to messages of greeting to the emperor.⁶ Whether the routine passages are to be regarded as the work of the emperor or of the secretary depends therefore on the assumptions one makes on other grounds about the authorship of constitutions. The fact that it can be shown that the emperors composed some of the documents in person does not prove that the hypothesis of universal imperial authorship is correct. On the other hand, the existence of texts with no obvious signs of personal authorship is no disproof of the hypothesis either: the emperors could hardly be expected to stamp every text with some expression of emotion or some idiosyncrasy of style. Nevertheless, this investigation can supply some arguments from probability in support of this hypothesis.

The vast majority of surviving constitutions are edicts of local application, epistles or subscripts to petitions, produced in response to some enquiry or request, or mere greeting, from subordinate officials, from communities or corporations, or from private individuals. Edicts of universal application, in which emperors made public some new policy launched on their own initiative, are few in number, and there is no reason to suppose that our surviving sample gives a misleading impression of the original, complete body of constitutions.⁷ The first stage in the production of most constitutions must therefore have been the reception of incoming epistles and petitions (libelli). Now, if imperial ‘civil servants’ did have any independent responsibilities, one might expect that the ab epistulis and the a libellis would have sorted out all incoming documents into those which they thought worthy of imperial attention and those which they felt able to deal with on their own responsibility. In the case of epistles it would have been possible to do this with those from officials delivered by couriers using diplomata, but not with those from communities or corporations, because the latter were commonly delivered by ambassadors who received an audience of the emperor.⁸ However, Antoninus Pius appears to have instituted in the early years of his reign a rigorous system of control, under which cities could only send embassies to Rome when they had important problems to explain, while decrees on trivial matters or of a formal character had to be forwarded through the provincial governor or procurator.⁹ It is precisely in the period c. 140-60 that it would have been possible for an ab epistulis to have sorted out formal messages from the Greek cities and composed replies to them on his own responsibility. Yet three of the imperial epistles which end with the formula ‘the proconsul/procurator sent on the decree’ are the very epistles of Pius to Ephesos in which, it will be argued below (p. 74 f.), that emperor displayed a very personal tone of irony and sarcasm. Two of these were simply acknowledgements of the receipt of honorary decrees, the most routine form of imperial business, of which one would expect the ab epistulis to relieve an emperor, if of anything at all. And, if he did not do so in the reign of Pius when the system of control was in operation, it seems very unlikely that it was done in other reigns when embassies were the standard means of delivery.¹⁰ Since Pius' secretary did not sort through the epistles forwarded from cities, he cannot have done so for the epistles of governors and procurators, with which they were included, either; and in other reigns, if emperors found the time to deal with the often trivial messages from the cities, it is very unlikely that they let their secretaries screen the reports of governors (apart from any other reason, one might expect those of senatorial rank, at least, to take offence, if the practice were a regular one and became widely known).

Petitions were in some cases forwarded by governors, but usually they had to be handed in by the petitioner in person or by a close relative.¹¹ In the case of the former, ‘screening’ by the secretaries would only have been possible if the ab epistulis sorted through the letters of governors, which is, as we have seen, unlikely, and then passed on all petitions to the a libellis without reference to the emperor, which also seems improbable. In the case of the latter, it would have been possible for the a libellis to examine the written petitions, since the emperors can have exchanged words with few, if any, of the people who handed them in.¹² It is possible to show that in some cases the emperors had themselves read the petitions and been moved by them: Hadrian and Pius both wrote epistles to subordinates enclosing copies of petitions and expressing themselves forcibly about the appropriate action to be taken.¹³ Furthermore, if any petitions were dealt with in this way by an a libellis acting on his own responsibility, one would have to suppose that he and the ab epistulis Latinis were able to get together and send a letter to a provincial governor, giving the latter orders. The picture thus conjured up is a most unconvincing one.

The relative contributions of the emperors and of the secretaries to the production of the published replies, once the incoming documents had been examined, are impossible to assess on the basis of the internal evidence alone. Stylistic analysis can demonstrate that in some cases the emperors both made the decisions and composed the final texts of the replies. On the other hand, it could well be thought possible that there were whole categories of routine business where the latter task, at least, was entrusted to the secretaries. Three categories which suggest themselves are epistles of a formal and diplomatic character, in particular those addressed to Greek cities, edicts with the function of bringing to the public attention ‘philanthropic’ decisions taken by the emperors, and those subscripts in which the emperor acted as a kind of Citizens' Advice Bureau by explaining the terms of laws and regulations. As for the first category, Pius' letters to Ephesos show that it should not be too readily assumed that emperors would not bother with such compositions. In the case of edicts, the examples to be considered below, from the reigns of Hadrian and Marcus, point very much to regular imperial authorship of such documents. Honoré has analysed the language of the subscripts, chiefly on questions of private law, which are preserved in large numbers in the Code of Justinian from Severus' reign onwards: he seeks to show that the distinctive styles can be traced of a series of jurists who held the office of a libellis.¹⁴ This is the area in which secretaries can most plausibly be suggested to have affected both the form and the content of imperial constitutions, because the private law was one of the rare areas in which the authority of experts was recognized in Roman public life, and because eminent jurists were appointed a libellis, from L. Volusius Maecianus under Pius onwards.¹⁵ But it must be remembered that replies about matters of private law accounted for perhaps no more than half of all subscripts issued,¹⁶ and on questions other than those of private law the jurists would have no special expertise. The internal evidence does therefore support, although it can do no more, the hypothesis that the role of imperial secretaries in the production of constitutions was a limited one, except possibly in the field of private law, and that that of the emperor himself was all-important.

2. HADRIAN

Despite the considerable number of well-preserved texts from this reign, the traces of Hadrian's individual personality which can be distinguished in them are disappointingly meagre, in view of the highly original character painted by the literary sources. The clearest signs of personal temperament are occasional outbursts of impatience and anger. The clearest case is that of a reply to a petitioner who raised in his petition a matter which the emperor thought he had already settled in a judicial decision, and who tried to cast aspersions on a lady whom Hadrian knew and esteemed. The excerpt quoted in a papyrus runs: ‘and I have already made it plain that my judicial decision is of assistance to you, and I do not believe that Philotera, an excellent lady and well-known to me, will do you injustice, since she knows very well that unjust possession counts for nothing, and you wish to trouble me about matters which are not in dispute, although you have the guardian of your creditor who will restore the slaves to you.’¹⁷ The string of paratactical clauses gives the impression of a reply dictated in haste and anger, adding one ground of complaint after another. Tradition might require an emperor to attend to the petitions of his faithful subjects,¹⁸ but Hadrian could show resentment when he felt that his time was being wasted. In another case his anger was aroused by an attempt to evade a civic responsibility. A legate of Gallia Lugdunensis had reported a dodge by which one Clodius Macer sought to escape from the duty of acting as a guardian: he was a son in potestate, and his father had refused to provide security for the son's honest performance of his duties. Hadrian's advice to the legate was brief and harsh: ‘if … the father … persists with the trick, I think you will rightly counter this fraud by compelling both the son and the man himself to undertake the guardianship of Clemens' children.’¹⁹ In two further cases Hadrian directed his subordinates to hear criminal charges which had been brought to his attention, and in both he reveals the kind of impression made upon him by the parties involved. The first of these is described in an excerpt quoted by Callistratus from an epistle to a proconsul of Macedonia, but the case in question had been passed on, not to the proconsul, but to some other governor, who is referred to in the third person.²⁰ The proconsul had consulted the emperor about the relative weight to be attached to documentary and to oral evidence, and Hadrian must have referred in his reply to a recent case which was still very much on his mind, in order to drive home the principle that the evidence of witnesses was to be preferred.²¹ The fact that the plaintiff Alexander in bringing charges against one Aper had failed to produce witnesses, but had preferred to rely on documentary evidence, clearly prejudiced Hadrian against him. The emperor emphatically declared his disapproval of such evidence: ‘for which there is no place at my tribunal, for it is my practice to question the witnesses themselves.’²² Alexander had been sent to the court of his local governor and the latter was ordered to investigate the good faith of the witnesses whom Alexander was presumably expected to produce, and to sentence him to relegation if he failed to prove the charges he had made. The second case is also mentioned in an excerpt quoted by Callistratus. Hadrian again sent to the court of some subordinate a petitioner, named Iulius Tarentinus, who claimed that he had lost a lawsuit because his opponents had bribed the witnesses. The tone of Hadrian's instructions suggests that he had been favourably impressed by Tarentinus, although he leaves it to the subordinate official to investigate the question of fact: ‘if he (Tarentinus) succeeds in proving to you that he has suffered from a conspiracy by his opponents to bribe witnesses, do you punish the deed severely, and, if any decisions have been handed down by a judge misled by so wicked an act, allow him a fresh hearing.’²³ Two aspects of Hadrian's character emerge from these epistles, anger at any hint of corruption of justice, and a confidence in his own ability to assess a witness' character from face to face questioning: the latter explains the insistence on the direct interrogation of witnesses found in the texts quoted by Callistratus.

To turn to the ‘inner style’ of these documents, that is to the general attitudes and principles which can be detected in them, it is again the epistles quoted by the jurists which provide the clearest evidence. Like his predecessors and his successors, Hadrian was averse to laying down detailed rules to be rigidly applied in all circumstances.²⁴ Two more passages on the assessment of evidence collected by Callistratus illustrate this principle. Valerius Verus, probably a provincial governor, had asked for guidance on the proofs required to demonstrate the truth of a case, but he failed to elicit any hard and fast rules from Hadrian: ‘which proofs are sufficient to demonstrate any case, and in what way, cannot be adequately defined in any secure way.’²⁵ Quite often, but not always, the truth of a matter can be established even in the absence of official records, because of the large number of available witnesses, because of their rank and authority, or because of ‘consentiens fama’.²⁶ In the end Hadrian can only advise Verus not to tie himself down to any one category of evidence and to use his own judgement. In the second passage another perplexed governor, a legate of Cilicia, receives the unhelpful reply that he is in a better position than the emperor to judge the reliability of a group of witnesses, and Hadrian goes on to list all the different criteria which might be applied.²⁷ These two texts, as well as a third about the payment of witnesses' expenses, also quoted by Callistratus, illustrate another aim which Hadrian shared with other emperors before and after him, namely to get his subordinates to trust their own judgements.²⁸ Hence the despatch of cases brought to his attention by petitioners to provincial governors for investigation and trial.²⁹

A reluctance to promulgate rules with a universal application appears also in an epistle to the provincial council of Baetica about the penalties for rustling cattle. Hadrian indicated his disapproval of the imposition of a standard sentence, and especially of the maximum permissible one, for the same offence in all cases.³⁰ The harshest sentence should not be inflicted universally but only where the particular crime of cattle-stealing was rampant and special deterrents needed: ‘puniuntur autem durissime non ubique, sed ubi frequentius est hoc genus maleficii.’ This epistle must clearly have rejected a request from the landowners represented on the provincial council that some especially harsh penalties should be authorized for a crime with which this province had been repeatedly troubled.³¹ By implication the emperor told them that their situation did not justify extraordinary measures: ‘ideoque puto apud vos quoque sufficere genus poenae quod maximum huic maleficio inrogari solet.’³² The provincial council tried again with his successor, but Pius repeated what Hadrian had said, that the severest form of penalty was ‘ad gladium dare.’³³

Insofar as Hadrian, in this epistle, resisted pressure from a municipal aristocracy, it bears out Syme's claim that it provides evidence of Hadrian's hostility to ‘the distinctions of birth and class,’³⁴ although Hadrian was not unique in this, since Pius took the same line. However, the fact that in this epistle no distinction is drawn between the penalties appropriate for honestiores and humiliores is not evidence of such hostility. As Garnsey has pointed out, Hadrian was answering a specific request and it probably never crossed his mind that rustling was a crime that decurions were at all likely to commit.³⁵ Syme also cited in support of his thesis an epistle of 119, which was addressed to a consular legate conducting a census in Macedonia, and which for the first time provided criminal penalties for the removal of boundary stones.³⁶ In this epistle Hadrian did indeed distinguish between two forms of this offence: the more serious one of removing the stones in order to acquire land, and the less serious one of mere theft of the stones for use as stones. He made the social status of the principals involved the criterion for establishing an intention to acquire land: ‘nam, si splendidiores sunt personae, non dubito quin occupandorum aliorum finium causa id admiserint.’ However, in cases of the more serious form of the offence, harsher penalties are to be imposed on agents of lower social status than on principals of higher rank, relegatio for the latter, damnatio ad opus for the former; men of humbler status who had simply been stealing stones were to be beaten. In neither of these epistles can one detect any significant opposition to the established privileges of the upper classes.

The epigraphic and papyrus texts, although usually preserved in a more complete form, are in the main less illuminating for either the ‘inner’ or the ‘outer’ style of the emperor than the jurists' excerpts: most are formal replies to diplomatic greetings or honorary decrees from Greek cities, or deal with requests from or on behalf of cities in a direct manner, without any comment.³⁷ In a very brief reply to a new foundation, Hadrianopolis-Stratonicea, the city's requests are merely described as ‘just, and necessary for a newly-founded city.’³⁸ However, an epistle to Hadrian's favourite city, Athens, did open with a resounding ‘declaration of intent’: ‘you know that I take advantage of every excuse for doing good to the city as a whole and to individual Athenians.’³⁹ The public emphasis on the emperor's generosity reappears in two papyrus texts which refer to his ‘philanthropia,’ a term also applied to him by the proconsul Avidius Quietus.⁴⁰ The first of these is an epistle to Q. Rammius Martialis, Prefect of Egypt in 117-19, which was published in a Greek translation at the headquarters of the legions of Egypt on 4 August 119. It announced that in future children born to serving soldiers would be allowed to claim bonorum possessio, if their fathers died intestate, under the rubric unde cognati, that is among the third class of kinsmen to whom the praetor usually granted bonorum possessio in cases of intestacy. This epistle was probably sent in reply to an enquiry from the Prefect about the state of the law, or to a request for permission to modify its application, and in response the emperor decided to make a change which would increase his popularity among the soldiers.⁴¹ His motives are made clear in the second sentence: ‘I myself am delighted to take the opportunity to interpret in a more humane sense the rather stern rule laid down by my predecessors.’ His personal interest in the matter is stressed by the use of ‘myself’ in this sentence, and by the tone of the last sentence, in which the Prefect is directed to bring the concession to the attention of the soldiers: ‘it will be your duty to bring this concession of mine to the notice of the soldiers and the veterans, not to give the impression that I am laying claim (to their gratitude), but, in case they are unaware of it, that they may take advantage of it.’ Despite this final disclaimer, the whole epistle gives an impression of having been drafted to bring out the present emperor's generosity and consideration: in the second sentence his humaneness is placed in neat contrast to the rigour of his predecessors, and even the disclaimer itself reveals his modesty and his exclusive concern for the welfare of the soldiers. All this, and the emphatic use of the first person, suggests that Hadrian had a large part in the composition of this document in its original Latin form.

The second of these papyrus texts is that of the only surviving edict of Hadrian, preserved in three papyrus copies.⁴² It was issued in Hadrian's twentieth tribunician year, i.e. after 9 December 135, and published at Alexandria on 31 May 136.⁴³ The text falls into two parts. First, there is a preamble which explains the situation and the emperor's reasoning; second, and marked off from the preamble by the formula, ‘with good fortune,’ come the emperor's actual decisions. In the first the emperor announces that he has been informed (presumably by the Prefect of Egypt, although this is not expressly stated) that the most recent inundation of the Nile, like the previous one (they were presumably those of 134 and 135), had been a poor one, and that he has therefore resolved on a concession: ‘I thought it necessary to perform some act of human kindness towards the peasants.’ This fairly straightforward account of the background is accompanied by a repetitive and verbose insistence on the excellence of the crops produced after the inundations which immediately preceded the two defective ones: ‘even though the Nile in each of the preceding years had made not only a complete inundation but one which was greater than almost any it made before, and by covering the whole country it was responsible for the production of the greatest and finest crops.’ Then a confident prediction that in future years the present dearth will be made up for by the Nile itself and the land is toned down by the pious interjection, ‘with god let it be said,’ as if to ward off any hint of presumption in the word ‘expecting.’ Finally, the emperor's confidence in the future is explained by a philosophical commonplace about ‘the nature of things’, which is defined as a cycle of plenty, dearth and renewed plenty.

The use of the formula [tukhe de te agathe] to separate the preamble from the actual decisions is unparalleled in a Roman edict, although of course standard practice in Athenian popular decrees of the classical period, and Jouguet saw in its appearance here an attempt by the philathenian Hadrian to introduce this Attic formula into the style of the imperial secretariat.⁴⁴ This seems the most likely explanation: the Greek phrase can hardly be intended to represent the Latin formula, ‘bonum factum,’ which could be prefixed to the texts of Roman edicts.⁴⁵ If this hypothesis is accepted, the use of the formula is a strong argument for Hadrian's authorship of the text, in view of his known enthusiasm for things Athenian and of the fact that the formula is found in no other imperial edict. The actual terms of the moratorium on tax payments which follow are set out in a perfectly straightforward way, and indeed are very modest and limited, despite the loud assertions about the emperor's generosity.⁴⁶ Apart from the use of the Athenian formula, there are two reasons for treating this text as the work of Hadrian himself. First, there is the use of the edict form, in which the emperor could address his Egyptian subjects directly, instead of an epistle to the Prefect, who had presumably raised the matter with the emperor.⁴⁷ Secondly, there are the tone and the style of the preamble, with its prolixity, its emphasis on imperial ‘philanthropy’, its ostentatious piety, and, above all, its use of a commonplace of Greek philosophy. The impression the preamble gives is rather different from that conveyed by the excerpts quoted by the jurists: it is one of pomposity and vanity. This could be regarded as a reflection of a deterioration in the emperor's character which accompanied his final illness; on the other hand, similar characteristics occur in edicts of other emperors, and they can be regarded as appropriate for a public proclamation, as against an epistle to a subordinate.⁴⁸

To sum up, Hadrian's individuality is revealed mainly in his letters to his subordinates rather than in formal correspondence with cities. The most striking characteristics are impatience with time-wasting and anger at the corruption of justice or the evasion of public duties. The general principles with which he was especially concerned included getting the best possible evidence in the courts and refusing to lay down hard and fast rules, which would be rigidly applied. In documents intended for public consumption, such as the epistles to Athens and to Rammius Martialis and the edict of 136, the main stress is on the emperor's ‘philanthropy,’ and in each of these cases there is good reason for believing that the texts were the work of Hadrian himself, and not of a ‘publicity-minded’ secretary.

Notes

1. A constitution is strictly not a type of document but any utterance, oral or written, of an emperor, which can be taken as a binding precedent in law (Gaius i. 5; Ulpian, Dig. i. 4. 1. 1). Not all written documents issued in the emperor's name contained what would be regarded as constitutions, but it is a convenient term to use to cover all the different kinds of documents issued by emperors, which in fact were the main sources of constitutions in the strict sense, and it will be used here to avoid any clumsy periphrasis such as ‘official pronouncements in a written form’.

2. e.g., G. R. Elton, Political History, Principles and Practice (1970), 72-81, and especially the comments on p. 74.

3. JRS lvii (1967), 19. This view is supported with fuller arguments in Millar's book, The Emperor in the Roman World, and I am very grateful to him for allowing me to see and to refer to the relevant passages in advance of publication.

4. Most recently, an epistle of Marcus to Miletus (see nn. 90-1 below).

5. JRS lii (1962), 114-20 = The Letters of Pliny, 536-46.

6. The Letters of Pliny, 544; Millar pointed out that in a private letter (Epp. x. 7 and 10) Pliny repeated the terms of a letter from Trajan (JRS lviii (1968), 223).

7. For the distinction between edicts of universal and of local application see ZPE xvii, 43-8; cf. Millar, The Emperor, 252-9, but he is more cautious about the representative nature of our sample (p. 257, ‘we can only surmise, but cannot firmly conclude, that general edicts were in fact a relatively minor part of imperial business’).

8. See Millar, op. cit., 363-4; 375-85.

9. See Historia xvi (1967), 471-2, 474-5.

10. Only one case of the ‘forwarding’ formula survives from a later date than Pius' reign (AE 1926. 95 = IGBulg. ii. 659).

11. For cases of the former procedure, see ZPE xvii, 58-62, and for the latter, JRS lxiv (1974), 93-8.

12. Millar, op. cit., 242, cites the only evidence for discussion between the emperor and petitioners, three anecdotes in the Sententiae Hadriani (Corp. Gloss. Lat. iii, p. 31, ll. 45 ff. = p. 387, ll. 22 ff.; p. 32, ll. 33 ff. = p. 387, ll. 47 ff.; p. 34, ll. 6 ff. = p. 388, ll. 48 ff.). In general, one would have supposed that the rambling verbosity of the petitions which have survived made it impractical to have them read out in the emperor's presence (see CIL viii. 10570 = FIRA i.² 103, cols. 2-3; IGBulg. iv. 2236, ll. 8-165; CIL iii. 14191 = Abbott and Johnson, no. 141, ll. 5-34; and idem., nos. 142-4). On the other hand, it might well have taken almost as long to bring these often humble and perhaps awestruck petitioners to the point, if they were invited to express their request orally.

13. Dig. xlii. 1. 33, ‘exemplum libelli dati mihi a Iulio Tarentino mitti tibi iussi’ (the rest is quoted in n. 23); xlviii. 6. 6, ‘exemplum libelli dati mihi a Domitio Silvano nomine Domitii Silvani patrui subici iussi’ (the rest in n. 80), and P. Rendel Harris 67 (quoted in n. 81 below).

14. Aufstieg u. Niedergang d. rom. Welt II (forthcoming).

15. See W. Kunkel, Herkunft u. soziale Stellung d. rom. Juristen², 174-6, 222-4, 224-9 and 246.

16. JRS lxiv (1964), 92-3; Millar, op. cit., 240 f.

17. P. Tebtunis 286 = FIRA iii. 100. ll. 4-9. … This text is described in l. 1 as an apokrima, a term used in P. Columbia 123 to describe imperial subscripts to petitions, while the earlier imperial decision had taken the form of an epikrima, that is of a decretum pronounced at the end of a judicial hearing. See H. J. Mason, Greek Terms for Roman Institutions (Am. Stud. Pap. xiii), 126-31.

18. Dio lxix. 6. 3, and JRS lxiv (1974), 86, nn. 1-2.

19. Dig. xxvii. 1. 15. 17: ‘Si … pater autem eius … in hoc artificio perseveravit, existimo te huic fraudi recte occursurum, ut et filius et ipse ad tutelam liberorum Clementis gerendam compellantur.’

20. Dig. xxii. 5. 3. 3. The proconsul was Iunius Rufinus (PIR² I 805).

21. This fact strengthens the argument for regarding the section of the epistle quoted in the Digest, at least, as Hadrian's own work: a secretary would surely be less likely to refer in such detail to a matter which was no direct concern of the person being addressed.

22. ‘Quibus apud me locus non est (nam ipsos interrogare soleo) …’.

23. Dig. xlii. 1. 33: ‘tu, si tibi probaverit conspiratione adversariorum et testibus pecunia corruptis oppressum se, et rem severe vindica, et, si qua a iudice tam malo exemplo circumscripto iudicata sunt, in integrum restitue.’

24. cf. Trajan's remark, ‘in universum a me non potest statui’ (Plin., Epp. x. 113), and Severus' answer quoted in Dig. i. 16. 6. 3.

25. Dig. xxii. 5. 3. 2: ‘quae argumenta ad quem modum probandae cuique rei sufficiant, nullo certo modo satis definiri potest’.

26. The precise meaning of this phrase is not clear to me: does the consensus of public opinion refer to the character of the witnesses or to the matter at issue itself?

27. Dig. xxii. 5. 3. 1: ‘tu magis scire potes quanta fides habenda sit testibus. …’ For the stress on the rank and influence of witnesses in these passages, see P. Garnsey, Social Status and Legal Privilege in the Roman Empire (1970), 210-12.

28. See the remarks addressed to Gabinius Maximus: ‘alia est auctoritas praesentium testium, alia testimoniorum quae recitari solent: tecum ergo delibera, ut, si retinere eos velis, des eis impendia’ (Dig. xxii. 5. 3. 4). A very similar use of language is found in the epistle to Valerius Verus: ‘alias numerus testium, alias dignitas et auctoritas, alias veluti consentiens fama confirmat rei de qua quaeritur fidem’ (ibid. 2). It may therefore be a favourite idiom of Hadrian's own, and thus evidence that he composed both these texts himself.

29. With the cases in Dig. xxii. 5. 3. 3 and xlii. 1. 33, compare the epistle of Pius in xlviii. 6. 6, and his standard reply to petitioners, ‘eum qui provinciae praeest adire potes’ (i. 18. 8).

30. Collatio xi. 7. 1-2 (FIRA ii, pp. 571-2) and Dig. xlvii. 14. 1. pr. (= section 1 of the former), both quoted from Ulpian's de officio proconsulis.

31. Vergil, Georg. iii, 408, with Servius, ad loc.; Varro, RR i. 16. 2; and n. 33.

32. The final passage of the excerpt quoted in Coll. xi. 7. 2 puzzled Ulpian himself, with its apparent suggestion that despatch to the mines was a harsher penalty than execution. Mommsen, Röm. Strafrecht, 1040, n. 1, suggested that at this point the text had been corrupted, presumably before it reached Ulpian. Garnsey, Social Status, 131. n. 4, and 185, n. 1, apparently accepts Ulpian's own solution that by ‘ad gladium’ Hadrian was referring to gladiatorial combat, and not execution. It may be that, if Hadrian had dictated an epistle and then added a final rider without checking the whole text, he got muddled himself: I have argued that just such a procedure can be detected in the new text of Marcus from Athens (see ZPE xvii, 50-1 and 55-6).

33. Coll. xi. 6. 1.

34. Les empereurs romains d'Espagne (1965), 245.

35. Social Status, 158, n. 2.

36. Coll. xiii. 3. 1-2 and Dig. xlvii. 21. 2; for the discrepancies between the two texts, see Garnsey, op. cit. (n. 27), 156, n. 1; for the innovation of imposing criminal penalties, idem, 170; and for the addressee, D. Terentius Gentianus, RE s.v. ‘Terentius’ (48).

37. e.g., Smallwood, Documents of Nerva, Trajan and Hadrian, nos. 61, 73, 449a, 452, 454b.

38. Robert, Hellenica vi, 81-2 = Smallwood 453, ll. 8-9. …

39. IG ii.² 1102 = Smallwood 445, ll. 10-11. … The remainder of the text is very fragmentary, but it may have announced the gift of a gymnasium to the young of Athens (cf. Paus. i. 18. 9).

40. Smallwood 454b: in his covering letter to Aizanoi, enclosing a copy of Hadrian's letter to himself, Avidius Quietus wrote of Hadrian μείξας τo φιλανθρώπῳ τò δίκαιον (l. 7), and he presumably knew how the emperor liked to be pictured.

41. BGU 140, revised by Wilcken, Hermes xxxvii, 84 ff. = FIRA i.² 78 = Smallwood 333. On the legal issue, see Schulz, Classical Roman Law, 227-36, and Kaser, Röm. Privatrecht i², 700. That it was a reply to an enquiry, and not part of a programme of concessions launched by Hadrian to counteract the hostility aroused by the execution of the ‘four consulars’ (as was suggested by P. J. Alexander, HSCP xlix (1938), 144-6), is strongly suggested by (a) the fact that the decison was made public in an epistle to a governor, instead of an edict of universal application (cf. the edicts of the future Augustus and of Domitian on the privileges of veterans, FIRA i.² 56 and 76), and (b) by certain phrases in the text itself. …

42. P. Cairo 49359 and 49360 (published in REG xxxiii (1920), 375-402 = SB 6944) and P. Oslo iii. 78; see FIRA i.² 81 and Smallwood 462.

43. See the new reading of l. 24 of P. Cairo 49359 by Guéraud, reported in P. Oslo.

44. REG xxxiii (1920), 391. The formula occurs between the preamble and the main body of an Athenian decree of as late as c. a.d. 220 (Syll.³ 885, l. 9).

45. See Plaut., Poen., l. 16; FIRA i.² 53, l. 4; Suet., Div. Iul. 80. 2; Vitell. 14. 4; Tertull., De Pud. i. 7.

46. See Westermann, JEA xi (1925), 177; D'Orgeval, L'empereur Hadrien, 117.

47. … The peasants of Egypt had no political organization, such as a provincial council, through which they could make known to the emperor their difficulties; they could only submit petitions to the Prefect as individuals when he visited the nomes (and they did so in great numbers: see P. Yale 61, ll. 5-7, for 1804 handed in during two and a half days at Arsinoe). But the Prefect could not grant a moratorium on taxes on his own responsibility and would have had to consult the emperor (see G. Chalon, L'édit de Tib. Iulius Alexander, 234-5, and ll. 9 and 62-5 of the text).

48. For his illness, see Dio lxix. 17. 1-2 and 23. 2; HA, Had. 23. 1-8. For similar imperial edicts, see Plin., Epp. x. 58. 7-10 (Nerva), P. Giessen 40 and AE 1948 109 (Caracalla), P. Fayum 20 (Severus Alexander).