¹ Howe, L. L., The Pretorian Prefect from Commodus to Diocletian (Chicago, 1942), pp. 119–20.Google Scholar

² The two constitutions where Krueger unhesitatingly retained the reading praefectus praetorio are CJ 9.51.1 (Oclatinius Adventus and Opellius Macrinus) and CJ 1.50.1 (Domitius). A third (CJ 7.33.1) marks its recipient Julianus as praetorian prefect, and Krueger leaves the appellation in the inscription, though suggesting in a note that it is to be deleted. Howe, op. cit., p. 70 no. 20 would let it stand.

³ For a chronological list of prefects that contains all those possibly labelled as such by the CJ manuscript tradition or early editors, see Cuq, E. in Oeuvres complètes de Bartolomeo Borghesi, x (Paris, 1897), pp. 803–7Google Scholar. Howe, op. cit., pp. 87ff., provides thorough discussion of these. The following are classed as dubious: Decimus in 228 (CJ 1.54.2); Ammonius in 240 (CJ 6.45.2); Pompeianus in 293 (CJ 8.1.3); Tryphonius c. 293 (CJ 9.51.12). Those rejected are: Metrodorus in 207 (CJ 4.24.1); Severus in 223 (CJ 4.56.2); Herodotus in 238 (CJ 5.11.2); Fabianus under Gordian III (CJ 9.51.6); Celer in 239 (CJ 1.54.3); Claudius in 240 (CJ 5.11.3); Achillinus under Valerian and Gallienus (CJ 10.65.1).

⁴ Howe, op. cit., p. 120 for the supposition of other cases. In six instances he identified unlabelled recipients of rescripts with otherwise known prefects: CJ 2.11.9, probably Q. Maecius Laetus (Howe, op. cit., p. 71 no. 21); CJ 8.30.2, probably Domitius (Howe, op. cit., p. 78 no. 44); CJ 5.30.2, 5.31.19, 5.70.4, 8.17.9, all probably Julius Asclepiodotus (Howe, op. cit., pp. 84–5 no. 60).

⁵ These have received attention: Huchthausen, L., ‘Herkunft und ökonomische Stellung weiblicher Adressaten von Reskripten des Codex Iustinianus (2. und 3. Jhdt. u. Z.)’, Klio 56 (1974), 199–228CrossRefGoogle Scholar and eadem, ‘Soldaten des 3. Jahrhunderts u. Z. als Korrespondenten der kaiserlichen Kanzlei’, in Scheel, H. (ed.), Altertumswissenschaft mit Zukunft. Dem Wirken Werner Hartkes gewidmet (Berlin, 1973), pp. 19–51Google Scholar. For a general treatment of women at law, Marshall, A. J., ‘Ladies at Law: the Role of Women in the Roman Civil Courts’, in Deroux, C. (ed.), Studies in Latin Literature and Roman History, v (Brussels, 1989), pp. 35–54Google Scholar. Paul, Dig. 5.1.12.2, in a discussion of those eligible to serve as judges, states the principle regarding women plainly – they were excluded, non quia non habent indicium, sed quia receptum est, ut civitibus officiis non fungantur. See also Paul, Dig. 16.1.1.1 and Ulp. Dig. 50.17.2 on the inability of women to serve as judges. As for soldiers, while they might serve in certain instances as judges of civil matters, in all the instances where imperial constitutions are directed to men called soldiers, these men are clearly litigants in the case at hand. For a brief discussion of soldiers as judges, cf. MacMullen, R., Soldier and Civilian in the Later Roman Empire (Cambridge, Mass., 1967), pp. 54ff.Google Scholar

⁶ It is worth noting that naming officials as such seems to begin in earnest with the tetrarchs. Aside from the praetorian prefects mentioned above (nn. 2 and 3), there are only the following before 284: CJ 3.31.1 (Augurino proconsuli Africae, 170); CJ 9.43.1 (Rutiliano consulari Ciliciae, 215); CJ 7.54.1 (procuratoribus hereditatium, Caracalla?); CJ 10.5.1 (ad rationales, 228); CJ 9.9.4 (Iuliano proconsuli Narbonensis, Severus Alexander); CJ 11.40.1 (quattuorviris et decurionibus Fabraternorum, Severus Alexander). It is hard to say with absolute certainty whether common practice before the time of Diocletian or the work of editors has resulted in this paucity of titles in the earlier constitutions. Unfortunately, inscriptions and papyri seem to be of rather little help in resolving the problem. The Greek rescripts on papyrus and stone are now conveniently collected by Oliver, J. H., Greek Constitutions of Early Roman Emperors from Inscriptions and Papyri (Philadelphia, 1989)Google Scholar. Only one document recorded there goes to an official of the Roman state. Number 289 (Stud. Pal. 5.119) is a letter from Gallienus to the equestrian procurator Aurelius Plutio (PIR ² A 1576), who is addressed without a title. Latin documents, though I cannot claim to have seen them all, appear to give similar results. From Aezanis in Asia, for example, comes a dossier containing various official correspondence (CIL iiii.355). Though the original address formula was not copied, a letter from Hadrian to the proconsul Avidius Quietus (PIR ² A 1409) is there described as an Exempl(um) epistulae [Cae]saris scriptae ad Quietum. Note also that in this letter the emperor refers to a previous governor of the province (Mettius Modestus, PIR ² M 568) simply by name. The documentary sources, though they certainly cannot have the last word here, at least do not show that titles are to be expected. On the other hand, it should be remarked that editorial deletion of recipients' titles, at least from fourth-century constitutions, is certainly known in the CJ. For example, CJ 4.61.4 is inscribed, Imp. Constantinus A. ad Iunium Rufum. At CTh 4.13.4, the same rescript goes to the same recipient, who is now called consularis Aemiliae. I should suspect, in fact, that the lack of titles for officials in the pre-Diocletianic constitutions of the CJ probably came about largely because of the editorial practices of the Codex Gregorianus.

⁷ Borghesi long ago pointed out that lack of an official title in a constitution's inscription does not at all prove the recipient to be a private person rather than a magistrate. See the note by Cuq, in Oeuvres complètes de Bartolomeo Borghesi, pp. 117–18Google Scholar. There are also instances where unlabelled addressees were recognised as officials already by Krueger. For example: CJ 7.45.1, to a Quintilianus, whom Krueger suggests to have been the praefectus vigilum (cf. PIR ² I 511); CJ 8.50.1, to Ovinius, probably Ovinius Tertullus the governor of Moesia Inferior (cf. PIR ² O 191). See further below (n. 42) for more on this.

⁸ For example, CJ 9.41.4 (Pars ex rescripto imp. Antonini A.) or CJ 10.5.1 (Pars epistulae imp. Alexandri A). Note as well a document such as the rescript, now chopped into three pieces, from Severus Alexander to Syrus (CJ 3.42.2, 9.2.2, 9.35.1). The remaining fragments may or may not represent the complete original. A list of such constitutions is provided by Honoré, T., Emperors and Lawyers (London, 1981), p. 35 n. 78.Google Scholar

⁹ I list just a few examples of such documents: CJ 4.5.1, 5.12.2, 5.71.1, 7.75.1.

¹⁰ Again, a few examples as representative: CJ 2.12.5, 4.29.1, 8.15.3, 8.42.1.

¹¹ Regarding interest on sales, the seller's legal claim to such and the obligations of the purchaser, Bechmann, A., Der Kauf nach gemeinem Recht, iii. 1, 2 (Leipzig, 1905), pp. 138–9Google Scholar, and de Zulueta, F., The Roman Law of Sale (Oxford, 1945), p. 51Google Scholar. Note also Kaser, M., ‘Zum heutigen Stand der Interpolationenforschung’, ZRG 69 (1952), 75–6Google Scholar, where it is remarked that pretium in a sale is to be understood as totum pretium, i.e. including interest. The stipulation here may have involved novatio, in which case the novum could well have been the addition of interest. On novation generally, Kaser, M., Das römische Privatrecht² (Munich, 1971), p. 647Google Scholar. Cf. also Marcellus, Dig. 13.5.24.

¹² The rubric of CJ 4.28 is Ad Senatus Consultum Macedonianum; hence, there can be no doubt as to which SC is here intended. I should like to thank Dennis Kehoe for bringing my attention to the important role of the action's title.

¹³ On this SC generally: Berger, A., Encyclopedic Dictionary of Roman Law (Philadelphia, 1953), p. 698Google Scholar; Kaser, op. cit., p. 532; Talbert, R. J. A., The Senate of Imperial Rome (Princeton, 1984), p. 443 no. 70.Google Scholar

¹⁴ Ulp. Dig. 12.1.24, Si quis certum stipulatus fuerit, ex stipulatu actionem non habet, sed illa condicticia actione id persequi debet, per quam certum petitur. See also Kaser, op. cit., p. 542.

¹⁵ On the near equivalence of the two actions, Kaser, op. cit., p. 531. Cf. also Berger, op. cit., p. 591 s.v. ‘mutuum’.

¹⁶ On the importance of employing the right type of action under the formulary procedure and on the continued use of formulae with the cognitio extra ordinem, Wenger, L., Institutes of the Roman Law of Civil Procedure (New York, 1955), pp. 15–16 and 262ffGoogle Scholar. respectively. Note also Ulp. Dig. 2.13.1 pr., where it is said that the plaintiff must make known the action by means of which he brings suit so that the defendant might contest it.

¹⁷ This interpretation is lent further weight by a passage from Ulpian's ad edictum (Dig. 14.6.3.3). We are there told that loans to sons-in-power are not allowed because they are detrimental to the parents, et ideo etsi in creditum abii filio familias vel ex causa emptionis vel ex alio contractu, in quo pecuniam non numeravi, etsi stipulatus sim: licet coeperit esse mutua pecunia, tamen quia pecuniae numeratio non concurrit, cessat senatus consultum. It is not clear from the context just why Ulpian thinks the initial contractual debt becomes a loan if the creditor attempts to collect on the basis of stipulation. Might it be that rather than stating a generally held rule, Ulpian is in fact thinking of the very case here under discussion?

¹⁸ Most significant for what follows is Turpin, W., ‘Imperial Subscriptions and the Administration of Justice’, JRS 81 (1991), 101–18Google Scholar. He argues persuasively that private individuals usually petitioned the emperor in order to elicit favours or direct action regarding their litigation. They wanted ‘practical assistance rather than intellectual guidance’ (p. 119).Google Scholar

¹⁹ Note (e.g.) CJ 4.30.2, where Caracalla explicitly informs the petitioning litigant of the judge's eventual decision, given the parameters of the case. This type of response is rather common. It should here be mentioned that Levy, E., Die Konkurrenz der Aktionen und Personen im klassischen römischen Recht, ii (Berlin 1922), p. 19 n. 7Google Scholar takes the last clause of CJ 4.28.3 to be a Justinianic interpolation, reflecting the later situation when the formulary procedure was inoperative. Levy supposes a similar interpolation regarding an action's title (though the legal question is different) in another place, Tryph. Dig. 46.1.69 – et non titulus actionis, sed debiti causa respicienda est. While both of these passages might be interpolations, I should think that they could just as likely reflect the attitude of Severan jurisprudence in this regard. Indeed in another area, i.e. where the voluntas of a testator was concerned, it seems that the Severans preferred to let their perception of equity supersede the strict letter of the law. Cf. Kehoe, D. and Peachin, M., ‘Testamentary Trouble and an Imperial Rescript from Bithynia’, ZPE 86 (1991), 159–60.Google Scholar

²⁰ On the praetor, cf. Lenel, O., Das Edictum perpetuum³ (Leipzig, 1927), p. 299Google Scholar (§ 110). We know, at least, that the urban praetor might try an actio venditi. Ulpian, in his monograph de officio praefecti urbi (Dig. 1.12.1), reproduces a letter from Severus to Fabius Cilo, outlining the duties of the urban prefect. The law of sale does not figure in it. However, Paul, Dig. 4.4.38 pr. mentions a sale of a fundus, and the praefectus urbi handling litigation concerning the sale. Generally on the judicial duties of the urban prefect in civil matters, Sachers, E., RE xxii.2 (1954), 2522–3Google Scholar. On the curule aedile, Honoré, A. M., ‘The History of the Aedilitian Actions from Roman to Roman-Dutch Law’, in Daube, D. (ed.), Studies in the Roman Law of Sale Dedicated to the Memory of Francis de Zulueta (Oxford, 1959), pp. 132–4Google Scholar. Also Ulp. Dig. 21.1.23.4 on aedilician actions for sales by sons-in-power. As regards the praetorian prefect and his jurisdiction over Italy, see below n. 34. For iuridici, Eck, W., Die staatliche Organisation Italiens in der hohen Kaiserzeit (Munich, 1979), pp. 256ff.Google Scholar, and Simshäuser, W., ‘Untersuchungen zur Entstehung der Provinzialverfassung Italiens’, ANRW ii.13 (Berlin and New York, 1980), pp. 429ffGoogle Scholar. On the provincial governor (e.g.), Crook, J. A., Law and Life of Rome (London, 1967), pp. 70–3Google Scholar. Also Mierow, H. E., The Roman Provincial Governor as He Appears in the Digest and Code of Justinian (Colorado Springs, 1926), pp. 30ffGoogle Scholar. Delegates of these officials (at least of the praetorian or urban prefect or provincial governor) are, of course, also possible.

²¹ Leunissen, Paul M. M., Konsuln und Konsulare in der Zeit von Commodus bis Severus Alexander (Amsterdam, 1989), pp. 308–9Google Scholar. It may be that Manilius Cerealis should be added to the list during the years 200/201. See Peachin, M., ‘Prosopographic Notes from the Law Codes’, ZPE 84 (1990), 108–9.Google Scholar

²² Alföldy, G., ‘Septimius Severus und der Senat’, BJ 168 (1968), 136–7 and 154Google Scholar. Also Alföldy, G., ‘Eine Proskriptionsliste in der Historia Augusta’, BHAC 1968/1969 (Bonn, 1970), 4Google Scholar = Die Krise des Römischen Reiches. Geschichte, Geschichtsschreibung und Geschichtsbetrachtung (Stuttgart, 1989), p. 167Google Scholar. See also Leunissen, op. cit., pp. 400–2. As Alföldy points out, Marcianus ought to have been a praetorius or a young consularis in about 197. Otherwise, next to nothing can be said of his career.

²³ Stein, A., Der römische Ritterstand (Munich, 1927), p. 120Google Scholar; von Petrikovits, H., RE xviii.1 (1942), 542Google Scholar; Pflaum, H.-G., Les carrières procuratoriennes équestres sous le Haut-Empire romain (Paris, 1960), p. 668Google Scholar; Millar, F., The Emperor in the Roman World (Ithaca, 1977), p. 126Google Scholar; PIR ² O 108. I have not been able to consult Cavuoto, P., Macrino (Naples, 1983).Google Scholar

²⁴ Another vague possibility might be the father-in-law of Severus Alexander, apparently named Sallustius Macrinus; however, there are numerous problems with this man, and he is, in the end, an unconvincing candidate. On him, Stein, A., RE i A 2 (1920), 1910–13Google Scholar [Sallustius no. ]4. The HA (Sev. Alex. 58.1) also mentions a Varius Macrinus, active in Illyricum and a relative of Severus Alexander. This man is another invention of the HA author. See Syme, R., Ammianus and the Historia Augusta (Oxford, 1968), pp. 45–6Google Scholar. Also now, Maria Angustias, Villacampa Rubio, El valor histórico de la Vita Alexandri Severi en los Scriptores Historiae Augustae (Zaragoza, 1988), pp. 345–6Google Scholar, and Leunissen, op. cit., pp. 209–10.

²⁵ Bastianini, G., ‘Lista dei prefetti d'Egitto dal 30^a al 299^p’, ZPE 17 (1975), 300.Google Scholar

²⁶ HA Did. Jul. 7.4.

²⁷ On his career generally: Howe, op. cit., pp. 68–9; Stiglitz, R., RE viii A 2 (1958), 1901Google Scholar; Pflaum, op. cit., no. 179bis. Cf. also Birley, A., Septimius Severus² (Plymouth, 1988), p. 99.Google Scholar

²⁸ Howe, op. cit., p. 69 and PIR ² F 300.

²⁹ Howe, op. cit., p. 69 and PIR ² F 554, based on the evidence of CIL vi.224, where Plautianus is attested (his name is restored, but with some certainty) as praetorian prefect on 9 June 197.

³⁰ Howe, op. cit., p. 16; Ensslin, W., RE xxii.2 (1954), 2408Google Scholar; Millar, op. cit., pp. 127–8; Halfmann, H., Itinera principum. Geschichte und Typologie der Kaiserreisen im Römischen Reich (Stuttgart, 1986), pp. 103–4.Google Scholar

³¹ We know, on the other hand, that the other prefect, Plautianus, was with Severus and Caracalla in the East. See Halfmann, op. cit., p. 104 with the relevant evidence. It could be, of course, that the emperors chose to communicate in written form with Macrinus even though he was near by. I am unable, however, to locate any direct evidence for such procedure.

³² On the movements of Severus, Halfmann, op. cit., pp. 217 and 220.

³³ This interpretation might seem to cause trouble because of the subscription with PP. The usual assumption is that the remark proposita indicates a response to a libellus posted up at the emperor's residence, and which the petitioner was responsible to copy. See e.g. Honoré, op. cit., pp. 27–8. If my interpretation of the present constitution is correct, this can hardly have been the procedure followed; rather, the rescript must have been sent directly to Macrinus. The fact is, however, that PP is not an indication of much at all. On this, Nörr, D., ‘Zur Reskriptenpraxis in der hohen Prinzipatszeit’, ZRG 98 (1981), 9 n. 24Google Scholar (citing Krueger's Praefatio to the editio maior of the CJ) and 14ff. on the problems generally of the ‘Propositionsvermerk’.

³⁴ Ulp. Coll. 14.3.2. See also Karlowa, O., Römische Rechtsgeschichte, i (Leipzig, 1885), p. 548Google Scholar, and Howe, op. cit., p. 34 n. 7. Note further Ulp. Dig. 1.12.1.4 or Dio 52.21.2, remarking that cases within the hundredth milestone belong to the praefectus urbi. Wenger, op. cit., p. 46 n. 55 argues that the jurisdiction of the praetorian prefect was unrestricted territorially; however, I am aware of no evidence that indicates this for the period of the Principate. It is clear that the praetorian prefect could handle appeals from judgements of provincial governors, or cases sent him by such officials. See (e.g.), CJ 4.65.4 (222) or 9.2.6 (243). However, I do not think that we should expect to find prefects outside Italy and administering justice, unless in the company of the emperor, or unless specially delegated to do so. When travelling, of course, the emperor might well delegate a case to the prefect. See (e.g.), P. Col. vi, lines 45–9. Millar, op. cit., pp. 129–30 argues that the prefect was generally at the emperor's side when he gave justice. See also the discussion by Howe, op. cit., pp. 32–7, which gives no indication of a jurisdiction unlimited territorially for the prefect.

³⁵ Note Ulp. Dig. 32.1.4, a praefectis vero praetorio vel eo, qui vice praefecti ex mandatis principis cognoscet, etc. On vice-praetorian prefects generally, Stein, A., ‘Stellvertreter der Praefecti praetorio’, Hermes 60 (1925), 94–103Google Scholar. See also Chastagnol, A., ‘Deux chevaliers de l'époque de la Tétrarchie’, AS 3 (1972), 223–31Google Scholar. Interestingly, the praefectus vigilum and the praefectus annonae seem to have been the men generally chosen to represent the praetorian prefects. The mere fact of the existence of such substitutes indicates that there was a need always to have someone of this stature on hand in Rome.

³⁶ Note that Birley, op. cit., p. 175 assumes that of the two praetorian prefects, Papinian and Maecius Laetus, only the former accompanied Septimius Severus on the British campaign. That is supported by CJ 2.11.9 (18 February 208), directed to a Laetus who is almost certainly the prefect (cf. Howe, op. cit., p. 71). The emperor (N.B. oddly Caracalla alone) allows that anyone who has defended the public interest of his patria will not resultantly suffer infamia. We may assume that someone from an Italian town was here involved, and that the case had come before the prefect in his capacity as chief appellate judge for Italy. We may also presume that Laetus received written communication about the case because the imperial entourage had already headed north. For the departure date (late 207 or early 208), Halfmann, op. cit., p. 219.

³⁷ Paul, Dig. 12.1.40 (loan), Papin. Dig. 22.1.3.3 (fideicommissum). Ensslin, W., RE xxii.2 (1954), 2415Google Scholar suggests that by the Severan period civil jurisdiction was regularly a part of the duties of the praetorian prefect. He also adds two possible instances of such: Paul, Dig. 29.2.97 (Papinian is thought here to be already praefectus praetorio); CJ 1.26.1. (A.d. 230), where it is said that a libellus praefecto praetorio datus pro contestatione haberi non potest. Note as well Paul, Dig. 3.2.21, where a theoretical suit for iniuria seems to take place in the first instance before the praetorian prefect.

³⁸ Howe, op. cit., p. 34.

³⁹ Still probably the best treatment of the rescript system as a whole is that of Wilcken, U., ‘Zu den Kaiserreskripten’, Hermes 55 (1920), 1–42Google Scholar. This may now be supplemented by: Honoré, op. cit., pp. 24–53; Nörr, , ZRG 98 (1981), 1–46Google Scholar; Williams, W., ‘The Libellus Procedure and the Severan Papyri’, JRS 64 (1974), 86–103Google Scholar; idem, ‘The Publication of Imperial Subscripts’, ZPE 40 (1980), 283–94Google Scholar. Note as well the important articles of Liebs, D., ‘Juristen als Sekretäre des römischen Kaisers’, ZRG 100 (1983), 485–509Google Scholar and idem, ‘OM 13,1 und das Reskriptwesen in der Historia Augusta’, BHAC 1982/1983 (Bonn, 1985), pp. 221–37.Google Scholar

⁴⁰ Honoré, op. cit., p. 57 n. 13.

⁴¹ Honoré, op. cit., p. 34. See further on this Nörr, , ZRG 98 (1981), esp. 9ff.Google Scholar

⁴² Cf. above n. 7. Note also that Honoré's reviewers, for example, do not question the private nature of the constitutions examined by him. E.g. Burton, G., CR 34 (1984), 66Google Scholar or Millar, F., JRS 36 (1986), 277Google Scholar. By way of illustration, I adduce another example of a constitution directed to a magistrate, but employed by Honoré (op. cit., p. 57 nn. 13 and 49) as a private rescript – CJ 2.50.1. The inscription reads, Impp. Severus et Antoninus AA. Chiloni, and the constitution is subscribed, PP. k. Nov. Laterano et Rufino conss. (1 November 197). Settlement of a soldier's will is in question, and the constitution clearly went to L. Fabius Cilo as leg. Aug. pr. pr. Pannoniae superioris. This was first recognised by Ritterling, E., ‘Die Statthalter der pannonischen Provinzen’, AEM 20 (1897), 34–5Google Scholar. See now, PIR ² F 27 (p. 99) and Thomasson, B. E., Laterculi Praesidum (Göteborg, 1984), p. 18 no. 45Google Scholar. Other examples of this phenomenon are noted by Peachin, , ZPE 84 (1990), 105–12Google Scholar and Turpin, , JRS 81 (1991), 104 n. 20Google Scholar, although the latter generally regards the CJ rescripts as private.

⁴³ I have conducted an initial examination of the dated constitutions in the CJ from the reigns of Septimius Severus and Caracalla, using the method described at the beginning of this article. For the purposes of the present calculation, only officials actually labelled as such in Krueger's edition are listed as certainly official. The group called ‘possibly official’ includes various constitutions whose addressees are not labelled with official titles, but who have been identified as magistrates, e.g. CJ 2.50.1 (cf. above n. 42): total rescripts [leges geminatae are counted together as one], 355; certainly private, 268 = 75.1%; certainly official, 2 = 0.56%; possibly official, 85 = 23.9%. By way of comparison, in his study of all the rescripts (N.B. the numbers given above consider only dated documents in the CJ) from the period between A.d. 194 and 305, Honoré, op. cit., p. 34 finds that 54 of a total 2,639 rescripts (0.02 %) are official.

⁴⁴ For example, neque enim aequitas patitur is a hapax in the CJ (21 April 200) and non enim aequitas hoc probare patitur appears only once in the Dig., and comes from Papinian's pen. One is inclined, with Honoré, to see the same man at work in both places. See Honoré, op. cit., p. 58, with this and other such examples. It further seems to me that a reading of these rescripts in chronological order indeed gives at least a rough impression of periodic changes in style. However, fixing exact dates for the changes, as well as determining anything like real consistency of style within a given period of time, is not so neat as Honoré makes it out to be.

⁴⁵ For the dates, Honoré, op. cit., p. 144.

⁴⁶ The opinion expressed here as to how exactly private rescripts were composed is very close to that of Millar, , JRS 36 (1986), 278Google Scholar, though he does not consider the distinction between rescripts to private individuals and communications to officials, which seems to me essential to the problem. I should like to remark that this paper grew from a graduate tutorial during the spring of 1991 at New York University with Sean Redmond, whom I thank for initial discussion of this law. Thanks go also to Géza Alföldy and this journal's referee for helpful comments on an initial draft. Bruce Frier and Dennis Kehoe have provided me with invaluable advice about things legal, and I am especially grateful to them.