¹ The fragment has been so widely discussed, both in general histories and in particular studies, that it would be reckless to attempt a bibliography. The following are referred to by author's name alone: Busolt-Swoboda, , Griechisches Staatsrecht (1920–1926)Google Scholar, see esp. i 252 with n. 2; Guarducci, , L'istituzione della fratria, Mem. dei Lincei vi 6 (1937), esp. 14–15Google Scholar; Hignett, , A History of the Athenian Constitution (1952), esp. 61–2, 390–1Google Scholar; Jacoby, , commentary on Philochoros, quoted by page of FGrH III b Suppl. vol. i (1954)Google Scholar; Wade-Gery, , Essays in Greek History (1958) 86–134Google Scholar (= CQ 25 1931); Wilamowitz, , Aristoteles und Athen ii (1893) 259–79.Google ScholarFerguson, , Org. = HTR 37 (1944) 61–140CrossRefGoogle Scholar, The Attic Orgeones; id., Phr. = Cl. Ph. 5 (1910) 257–84, The Athenian Phratries; id., Sal. = Hesp. 7 (1938), The Salaminioi of Heptaphylai and Somion: Sal. (by itself) = the first inscription of the Salaminioi there published, pp. 3–5.

This paper arose from a discussion of phratries in a class held in Oxford early in 1957. I am grateful to all who took part, and besides and especially to H. T. Wade-Gery and M. I. Finley: but my errors are my own, and not for lack of warning.

² Wade-Gery, Ferguson, Guarducci, Jacoby and others.

³ Busolt, Hignett and others.

⁴ The lexica are not much use: see Ferguson's analysis, Org. 62–4.

⁵ Ferguson's contention (p. 77) that the orgeones of Hesp. 11 (1942) 282 no. 55 had sometimes not enough money for an ox depends on a restoration which he abandoned in Hesp. Suppl. 8 (1949) 131.

There remains IG ii² 2355, an unbroken stone with a list of 16 names, and 2499 with its τραπέζας εἰς δύο τρίκλινα (29–30). The orgeones of Amynos who adopted Asklepios (ii² 1252, 1253, 1259) seem to be richer and might be more numerous.

⁶ Kahrstedt, , Staatsgebiet 235 n. 2Google Scholar, denied the connexion between 35a and 35b, remarking that Philochoros may have had many occasions to mention homogalaktes. But these are not just mentions, they are explanations, and in the same terms: it may be taken as certain that the lexica did not draw them from different parts of Philochoros' work.

⁷ Orgeones were mentioned in a law of Solon (Seleukos' commentary on the axones, 341 F 1), and they appear in many reconstructions of the corrupt passage of the law cited in Digest xlvii 22.4.

We had best leave out of account the wo-ro-ki-jo-ne-jo-e-re-mo (ka-ma) of Ventris and Chad wick, Documents, nos. 152.7, 171.11. Our orgeones present a specialised Attic sense of a word whose meaning elsewhere is perhaps more like ‘priest’: it is not impossible that there should be some organic connexion between the Attic and the Pylian orgeones, but it is not specially probable, and we have little idea what the latter were.

⁸ I say ‘perhaps’ because Athens is not always careful to avoid anomaly.

⁹ They are too long to summarise here again. See the translations and analyses given by Wilamowitz and Wade-Gery. The third decree, which is later, does not affect the present issue.

¹⁰ Wade-Gery (pp. 128–9) is clearly right to deduce from their oath that these are judges not advocates.

¹¹ Wilamowitz insisted that we distinguish the procedure of the special scrutiny from that of the normal, and no doubt they differ in some respects: but the whole vocabulary need not change its meaning.

¹² When the thiasos accepts, there is no appeal (and the verb ἐφιέναι is not used) but automatic reference to the phratry: appeal occurs, and the introducer has an option, when the thiasos rejects. These notions are clarified by Wade-Gery, 192–5.

¹³ His further argument from the nomenclature of the priest has not apparently been noticed except by Guarducci (p. 43), but is none the less valid.

¹⁴ There is not much agreement about the correct meaning of οἶκος: Busolt-Swoboda, 959, ‘eine wohl mehrere Thiasoi umfassende Kultgenossenschaft’; Kahrstedt, , Staatsgebiet 234Google Scholar, ‘ein οἶκος kann terminologisch nichts anders sein als ein Teil eines Genos’, for which surprising assertion he refers to Wilamowitz 266, who only said, ‘haus, οἶκος, ist ein gentilizischer begriff’, and produced examples to show that it could be; Latte, RE s.v. ‘Phratrie’ col. 750, confines himself to saying that οἶκος cannot be a name for a phratry.

The confusion is natural, for the truth is that any sense of οἶκος which is tolerable in this text is bound to be unique. There is no trouble with 41–2 ‘the priest of the Dekeleians' house’: the Dekeleians, whoever they are, might have a literal house for their hiera like the Klytidai of Chios (Syll. ³ 987: the Athenian Kerykes also have a house, IG ii² 1672.24), and their priest might easily be called the priest of such a ‘house’. The body which is to appoint synegoroi (32–3) is another matter. If the ‘house’ is not (in both passages) simply the phratry, it could perhaps in 32–3 mean the officers of the phratry, as the persons specially concerned with the cult and its house: the ‘vestry’ of nineteenth-century English parishes would be a parallel, and it would not be hard to find others. But I must repeat that the interpretation of οἶκος, which is difficult on any hypothesis, is a subsidiary point.

¹⁵ The privilege is by now very much attenuated: the law of the Demotionidai is freely amended by the assembly of phrateres, and Hierokles dilutes their jurisdiction in appeal.

Wilamowitz ascribed a position of privilege to the ‘house of the Dekeleians’, and added instances (most of those discussed immediately below) of differentiation in other phratries between ‘brothers of first and second class’. I am glad to note that my position is not in principle different from his, only reversing the parts of Dekeleies and Demotionidai. For later views see Hignett, 56 and app. ii(c), pp. 313–15.

¹⁶ One would expect the Apatouria, but Apollodoros may have feared that he would not live till then, and in fact he died before he could get Thrasyllos enrolled in his deme.

¹⁷ This phrase could be taken to mean that the law is the same for either case, a real or an adopted son (so Wyse, and this seems to read more easily); or that the law is the same for genos and phratry. Either way, the law of the genos is the law of the phratry.

¹⁸ The point seems to be that the introducer's oath was not enough by itself. It is presumably implied that others were more lax, but I doubt if (c) below is an instance: see n. 21 below.

¹⁹ Harp. (Suid.) s.v. γεννῆται remarks that in this speech Isaios which seems to mean that he is speaking throughout of relatives and not of gennetai in the technical sense. It is generally assumed without question that he really meant gennetai (Wilamowitz; Wyse; Francotte, , La polis grecque 56–7Google Scholar; Busolt-Swoboda, 957 n. 4; Latte, RE s.v. ‘Phratrie’ col. 750; etc.): Schoemann, seeing the difficulty, wished to emend to It seems incredible that Thrasyllos should say he was enrolled among his relatives (cf. not only 15–17, but 13 with 27), and I assume that Isaios meant gennetai in the precise sense of the word (Her. v 66.1 is some parallel). But Jacoby appears to accept the statement of Harpokration as it stands.

²⁰ This has been understood to mean that in this case the child was registered with the phratry before application was made to the genos (e.g. Guarducci, 25). But it does not appear from the text that the phrateres ever voted on the case, εἰσῆγεν is perhaps ambiguous: the verb can be used of effective introduction (as IG ii² 1237.18–19), but in itself it means no more than ‘introduce’, ‘present’, and I imagine, though the instances are indecisive, that it could be used of the occasion when the meion was sacrificed, when the father probably swore an oath (Dem. 57.54, below) but there was no scrutiny or vote.

²¹ This is sometimes claimed as a case where the introducer's oath was all that the law required (n. 18 above), but it seems unlikely that Kalliades' opposition and the vote of the genos were meaningless, as Wilamowitz (p. 271) alleged without argument. The proceedings would more naturally be taken as a real debate ending in a decision that Kallias should proceed according to the form prescribed by law.

²² For this, see below pp. 7–8. The alternative emendation <μ̓ > ἦγον (Blass) makes no difference for our purpose.

²³ The subject of the earlier clause is οἱ συγγενεῑς, and in spite of Ledi, Wien. Stud. 29 (1907) 213, it seems likely that there are two separate occasions. The father perhaps swore his oath when he sacrificed the meion (cf. n. 20), but died before he could sacrifice the koureion.

²⁴ Most clearly Isaios 8.19; [Dem.] 43.14. Admission to a phratry without mention of gennetai: Isaios 2.14; 3.73–6; 6.22; 10.8; 12.3; [Dem.] 39.4, 20.

²⁵ Harp., Suid. s.v. ἑρκεῑος Ζεύς; Phot. s.v. ἑρκείου Διός; EM 375·24; schol. Plato Euthyd. 302d. Some of these allege that ερκος was Attic for a house, which at least shows how they understood έρκειος.

²⁶ References in Busolt-Swoboda, 956; Ferguson, Sal. 31. The latter's objection that neither Kerykes nor Salaminioi had a separate priest of Apollo Patroos is not decisive: the archon of the genos, for instance, might discharge this function.

²⁷ It has been noted that Odysseus had an altar of Zeus Herkeios in his courtyard (χ 334), but Eumaios apparently not (ξ 5 ff., 419 ff.): perhaps he depended for this on Odysseus.

²⁸ All Spartans were descended from Herakles in the seventh century (Tyrt. fr. 8.1), but of the tribal eponyms only Hyllos was Herakles' son. (Plato, loc. cit., makes it clear that Patroos in the title of this Apollo means that he was the general ancestor of all Athenians, through his son Ion.)

²⁹ For the meaning of είσάγειν see n. 20 above.

³⁰ The subject is Atrometos throughout, so το γένος must be taken adverbially, ‘he belonged by descent to a phratry….’ If Atrometos had been a gennete, Aischines would have said so unmistakably.

³¹ The history of the genos Salaminioi and the relation of its two branches remain obscure, and none of the solutions so far proposed is wholly convincing: Ferguson, Sal.; Nilsson, , AJP 59 (1938) 385Google Scholar; Daux, , REG 54 (1941) 220Google Scholar; Guarducci, , Riv. Fil. 26 (1948) 223Google Scholar: perhaps the assumption that the seven tribes of ‘Heptaphylai’ were Kleisthenic tribes is wrong.

The inscription too often leaves us in some doubt what the position was before the two branches were reconciled. Did each then sacrifice separately to Zeus Phratrios? and if so, with a different phratry? and if so, how was the difference resolved? We can only leave this open.

³² Ferguson, Sal. 14, 28 n. 7. As he points out, Στράτων in 2345.74, three lines above Stratophon, might well be his father. It was also natural to guess that in 2345.83 might be Demon's father Demaretos, but Wade-Gery, who looked at the stone for me in 1956, assures me that this is impossible, and the squeeze in the collection of the Institute for Advanced Study, Princeton, con firms this.

³³ The view is therefore no longer tenable, that thiasoi and gene are exclusive terms, thiasoi being composed of orgeones only (e.g. Hignett, 56); nor Wade-Gery's idea that a genos might itself constitute a thiasos (p. 87, written before the document of the Salaminioi was found).

³⁴ No intelligible pattern emerges from the demotics on the stone, and the few that can be added by identification of individuals (Kirchner's note on 33–35 should be corrected in the light of Lewis', argument, BSA 50 (1955) 13–14Google Scholar). The bunching of members from Agryle in Diogenes' thiasos is something that is liable to happen to any group whose membership is hereditary.

³⁵ Harp, adds, Ἴων (fr. 32 N²) and LSJ under θίασος give instances from Euripides for the general meaning ‘company, troop’: but these have all some suggestion of or comparison with religious activity. Ion is best treated as an exception.

³⁶ The MS. reading seems improbable: Ferguson's parallel (Org. 70–1, n. 12) from Dem. 18.260, 19.199 refers to much less reputable activities, and the text of Harp. s.v. θίασος does not really prove that Harp, read the plural in Isaios. Sauppe's θιασώτας; seems the easiest correction.

³⁷ IG ii² 1261 (302/301) and many of the following. For foreigners cf. 1263.21 (301/300), 1271.15 (298/297), etc.; Poland, , Vereinswesen 20 ff.Google Scholar

³⁸ Perhaps of the same phratry, and perhaps Glaukidai and Epikleidai were gene which also belonged to it, but we cannot be sure of this. In such transactions the ‘creditors’ might be inter related, as in IG ii² 2670 (Finley no. 146) the Lykomidai had their main shrine at Phlya (Plut. Them. 1.4), a Kekropid deme.

³⁹ Cf. Ferguson, Org. 133, app. 2; Guarducci, 45.

⁴⁰ θιασῶται are listed in the Solonian law in Digest xlvii 22.4 among those associations whose internal rules are allowed to be valid so far as they do not conflict with the law of the state. Thiasoi in Solon's time might be something different again: but (as Wade-Gery reminds me) this is not an antiquarian's quotation from the axones, but a lawyer's reference, so that the law as here presented may contain later accretions and there is no certainty that Solon legislated about thiasoi.

⁴¹ It is unlikely that the headship changed annually, if a personal name was enough for identification, i.e. if the creditors of IG ii² 2723 (p. 11 above) are thiasoi (v. Premerstein, 116).

⁴² IG ii² 1241, and the reference to the Medontidai in Hesp. 10 ( 1941 ) 16, no. 1.16 ff., tell us something about the business transactions in which a phratry might be concerned.

⁴³ Jacoby's text is beyond reasonable doubt. ξένοιν, the reading of the MSS. of Harp., is implied also by Suidas’ γένοιν: the μὴ; between τις and ἐξ ἀμφοῑν, attested by Bekker for MS. A of Harp., can only be taken into the text if we are prepared to go to the length of altering ξένοιν to ἀστοῑν.

⁴⁴ Jacoby's further objection that 451/450 lies outside the limits of book iv depends on his dating of F 1, from book iii, to 446/445 or later. The authors of ATL (iii 9–12) argue that F 4 belongs to 451/450 and F 1 must be earlier, and this is doubtful for the reason given above: but they rely mainly on the probability that Doros in F 1 is the known city of Phoenicia, appropriate for 454 but most unlikely later, and this seems to me the weightier argument. That book iv could have begun before 451/450 is no positive argument for assigning F 4 to that year, but it cannot be so very much later: the fact that book ix included a document from Sept 411 prevents us from allowing a long span to book iv.

⁴⁵ I am glad to find that an explanation on these lines had also occurred to Wade-Gery, though he is more inclined than I am to refer Krateros F1 to Perikles' law of 451/450.

⁴⁶ We must not expect absolute regularity. See n. 31 for the possibility that the Salaminioi were split between two phratries; n. 38 for a possible phratry with two gene.

⁴⁷ The case between Krokonidai and Koironidai, in which both Lykourgos and Deinarchos spoke, is referred to in several entries in Harpokration, e.g. Κοιρωνίδαι. Cf. also Phrastor and the Brytidai, p. 6 above; and the arbitration between the two branches of the Salaminioi.

⁴⁸ It was a surprise to learn that Medontidai were a phratry (Hesp. 10 (1941) 16, no. 1.16 ff.), and there might be other surprises.

⁴⁹ Toepffer's Attische Genealogie lists 58 possible gene: a few more have turned up since, but at least as many of his are extremely doubtful. The demes created by Kleisthenes, primarily the city demes, should perhaps be left out of account, but there may have been city phratries. The scheme set out in Ath. Pol. fr. 3 is at variance with all we know of Athens at any later period, and if Aristotle assigned this scheme to Ion he was describing something other than the origin of the phratries and gene of his own day (Ferguson, , Phr. 257 ff.Google Scholar, defended its historicity, but abandoned it in Class. Stud. presented to E. Capps (1936) 151 ff.). See, however, Hammond, , History of Greece (1959) 153.Google Scholar

⁵⁰ Hermes 89(1961) 129–140.