The Second Athenian Tribute Assessment Period Meritt, Benjamin D Greek, Roman and Byzantine Studies; Summer 1967; 8, 2; ProQuest pg. 121 The Second Athenian Tribute Assessment Period Benjamin D . Meritt I N THE FIRST assessment period of the Athenian tribute-quota lists there was very little, if any, systematic order in the way the names of the cities were inscribed on the first large tribute stele (454/3-451/0). There was, in the inscribed record, no separation into the «districts" which first made their formal appearance in the Lists in 443/2, five in number, and in the following order: Ionic, Helles­ pontine, Thrakian, Karic, and Island (Lists 12-16).1 Presumably the first tributes were booked very much in the order in which they were received.2 Names recorded early in List 2 were in reality in arrears for 454/3 and should normally have been recorded in List 1, but their payments were received from remote Karia too late for the helleno­ tamiai of 454/3 to take account of them. And the Karic names at the end of List 4 reflect the activity of Kimon's fleet in Karic waters during his campaign to Kypros in the late spring of 450 B.C.s On the second large tribute stele, which compassed all of the fifth and three quarters of the sixth assessment periods, there were only four «districts," and they followed in this order: Ionic, Island, Helles­ pontine and Thrakian (Lists 17-23). Karic names were merged with the Ionic and they both appeared under the heading 'IwvLKo~ cp6po~. It has been plausibly assumed that the districts within any given 1 For the Lists, see Meritt, Wade-Gery and McGregor, The Athenian Tribute Lists II (Princeton 1949) pp.8-36; IV (1953) pp. ix-xii; Phoenix 21 (1967), Article in press. 2 ATL III (Princeton 1950) pp.7, 12,30. W. K. Pritchett notes this presumption, which was made not only by the editors of ATL but also by Beloch, Busolt and Nesselhauf, and he admits some grouping of cities when "envoys who were neighbors sometimes travelled together to Athens with their money" (BCH 89 [1965J 432-3). But he finds that such group­ ings were not always the same and is disturbed by this phenomenon. Juxtaposition of names in a List does not make a group unless the cities are in fact neighbors, and not necessarily so even then. There is no Significance, as he claims, in a lack of continuity, and the order of names is best explained by the time of payment, in however random a fashion records of the hellenotamiai were handed over to the engraver. 3 ATL III pp.7-8, 298. 121 122 THE SECOND ATHENIAN TRIBUTE ASSESSMENT PERIOD assessment period were the same in the quota lists as they were in the roster of names attached to the several assessment decrees on which the quota lists depended. On the large stelai even the order of the districts was presumably the same, and after the assessments of 425/4 and 422/1 (A9 and AlO),4 even on the smaller individual stelai, it was invariably Island, Ionic, Hellespontine, [Aktaian], Thrakian and [E uxine]. 5 The order of districts changed in 430/29 to Ionic, Thrakian, Island and Hellespontine, and from 428/7 through 426/5 it was presumably Thrakian, Island, Helles­ pontine and lonic.6 These panels of names should not be taken to represent administrative districts. They were a bookkeeping conve­ nience and probably useful primarily as guides to which cities the heralds should visit when new assessments were to be announced.' Professor W. K. Pritchett quotes this judgement but apparently mis­ understands its intent.S He observes that Hif one will plot on a map the tributary cities according to the order in List IV, it will become ap­ parent that no heralds would adopt such roundabout and serpentine routes." This is beside the point. Naturally, no such meandering journeys of the heralds were ever envisaged. Pritchett does not quote the further judgement of the editors of ATL (III p.12) that "some grouping into lists that did not overlap ... was essential as soon as there was ... any occasion for the despatch of heralds." The words of the decree of 425 are significant (A9, lines 40-41): [ra]s- 8[~ 7Top]et<;,s rots KlpvX[ 0'£ rots lout xuvyypaCPUCtl Kara r ]6V h6p[ KOV r6] ~ TaKe ras hlo]s TeO 7To]pe[vlllu] ov[raJ £ htva p.~ av[roL araKro£ £00'£'-- -J. The heralds had to know what ground they were to cover, but the order of their visits to the cities was undoubtedly controlled by the convenience of travel. That it was not laid down in the assessments is shown, for example, by the great assessment of 425 B.C., in which the cities were arranged pretty much in the order of the amounts they were supposed to pay.9 The sequence of visitation was not indicated. It was natural that the panels in anyone assessment period should have a uniform order, especially when they were inscribed on the same stone (as in Lists 12-15 [16], 17-23), but variation in the order 4 Cf A TL II pp.40-4. 6 See Lists 33-40 (ATL II pp.33-9). 6 See Lists 27-29 (ATL II pp.31-2). 7 ATL III p.ll with n.25, p.12 n.29. 8 BCH 89 (1965) 432. 9 See ATL II pp.42-3 (A9). BENJAMIN D. MERITT 123 was easier when each list had its separate stele. Yet the only variation within a period of which there is evidence is that of 429/8, when the Ionic panel came last instead of first as had been the case in 430/29. There can, I think, no longer be doubt that the lists of these two years belong to one assessment period.10 There can hardly have been any sanction or rule about the order of the districts. This is well illustrated by the fact that the panels appended to the assessment decree of 425/4 display an arrangement quite different from that employed in the decree itself. The decree provides (A9, lines 5-6) that heralds were to be sent out two to Ionia and Karia, two to Thrace, two to the Islands and two to the Hellespont,ll whereas the panels of names come in the order of Island, Ionic, Hellespontine and Thrakian, with an inter­ mediate panel of Aktaian cities and a final panel of Euxine cities about which nothing whatsoever was provided in the opening decree. It might be suggested that the variation of order in 429/8 was quite incidental, the list for 430/29 not being on the same stone to serve as guide, and that the roster following the assessment of 428/7 merely followed the order of the previous list of 429/8. There is no way of knowing what order was prescribed in the decree of 428/7 (now lost), but the example of 425/4 shows that it may have had no bearing on the order of the panels of names which followed. Such order as we find in the panels of tribute cities, in the one instance where it can be tested, is completely independent of the order used in the governing decree. One can only conclude that the orators of the pertinent decrees spoke of heralds who were to go to four divisions of the Empire, no matter how many divisions, or panels, there actually were, and that the usages in the decrees and rosters were independent of each other. This can be tested only once, in A9, but this one test is sufficient. There were four divisions in the decree and six in the roster, and the order of them (so far as they exist at all) in the decree differs from the order in the roster. In the light of this circumstance, it would seem that Mattingly's assertion is unjustifiable that the similarity of order of districts in the 10 See Meritt, Athenian Financial Documents (Ann Arbor 1932) 1-12. See also a criticism of Mattingly's redating, in GRBS 8 (1967) 45-52. 11 A9, lines 5--6: 0150 [lLlV €1T' '[OV{C1:V Kal Kap{av] ~!5o Ol €['TTl @patK€V 0150 olE €'TTl N[Eao, 0150 Ol €I/J' 'EMEa1T]OVTOV. H. B. Mattingly (Historia 10 [1961] 168) dated List 25 (SEG V 28) in 426/5 and assumed that the orator Thoudippos in A9followed the order of districts in lines 5--6 as he knew them from the preceding year. In CQ 16 (1966) 179-88 (esp. 188) Mattingly again affirms the date 426/5 for List 25. But it has now been shown, in the light of evidence overlooked by Mattingly, that this is impossibly late (cf GRBS 8 [1967] 50-2). 124 THE SECOND ATHENIAN TRIBUTE ASSESSMENT PERIOD decrees of D7 and D14 with the order of districts in the roster of A9 is "striking confirmation" of his date for D7 and D14 after A9 (425/4).12 He calls the order the same, but both D7 and D14 fail to mention the Aktaian cities and the Euxine panel. The division of the Empire into five panels, rather than the later four or six, is formally attested in 443/2. But a rudimentary division suggestive of the five districts can be observed even earlier, in the order of names in cols. I-IV of List 9, in cols. I-IV of List 10, and in all of List 11.13 There is some confusion but not enough to invalidate Nesselhauf's conclusions: the districts, five in number, can be seen, one might say, as "through a glass, darkly," but they can still be seen and should be so recognized: Ionic, Hellespontine, Thrakian, Karic and Island.14 There is also a rudimentary division into district panels observable in the second assessment period in List 5 (450/49). Here the Karic and Ionic names come in cols. I and 11.15 Thrakian names occupy most of col. III and are followed by Hellespontine names at the bottom of col. III and top of col. IV. Island names occupy the bottom of col. IV.16 When Mario Segre discovered and published the Kos fragment of the monetary decree (DI4) with its Attic lettering and three-bar sigma,!7 it was a natural assumption that the two panels ofIonia and Karia had been merged into one and that the four districts of the coinage decree were the same as the four districts of the roster of names in List 5. This was the opinion held by me and suggested to Segre; it was also the opinion of the editors of ATL in 1939, in the year following the publication of Segre's article.1s The editors make no apology for 12 CQ 16 (1966) 187, 189. D7 is the decree of Kleinias about the collection of tribute, pub­ lished in ATL II pp.50-1, and dated by the editors in 448/7; D14 is the Coinage Decree (Decree of Klearchos) published in ATL II pp.61-8, and dated by the editors in 449/8. See also below, pp.126-32. 13 Herbert Nesselhauf's study of Lists 9, 10 and 11 is convincing proof of this (KUo, Beibeft 30 [1933] 39) and was so accepted by the editors of ATL I p.496). Pritchett's criticisms (BCH 89 [1965] 428-30) are invalid and are based in part on an error in identifying names. 14 Pritchett's table (BCH 89 [1965] 429), based-so he says-on "the ATL text of List 11," puts at the head of col. ill 16 Ionic names, when in fact all sixteen of these names belong to Thrace. His assertion that "there are Ionic cities in the first, second and third columns" (op.cit. 430) is not true. The reader is advised to make his own verification of the names in col. ill. 15 There are only two intruders, the Hellespontine KE{3P1JV'O' and N€avSp€,a in 5 II 33-4. 16 There can be no real quarrel with this judgement, despite Pritchett's comment (BCH 89 [1965] 430-1). 17 Clara Rhodos 9 (1938) 151-78. 18 A TL I p.496: "That Ionia and Karia formed one administrative district in the second period (450/49-447/6 B.C.) is made clear by the naming of the Ionic district alone in the BENJAMIN D. MERITT 125 having a better understanding of the evidence in 1950 than they had in 1939, and their judgement19 was more conservative. They admitted the possibility of assuming five districts (Karic, Ionic, Thrakian, Hellespontine and Island),20 but regarded the intermingling of names in col. II as indicative of a merged district ;21 yet this was considered not quite sure, because no Ionic names were mingled with the Karic names in the preserved part of col. I (fourteen lines). The question of four or five panels cannot be settled on the available evidence. If there were four districts, they were Ionic-Karic,22 Thrakian, Helles­ pontine and Island. With the evidence of Krateros that the heading KapLKos cpopos appeared in the assessment of 454/3, it is possible that there were five districts on which all the lists were based continuously down to 439/8, though the order of the second period, at least, was different from that of the third and fourth. 23 The point to remember is that, whether four districts or five, it does not matter for the date of the coinage decree, since the formulae of the decrees were independent of any sequence or number of district panels in the rosters which accompanied them and on which the lists were based. Formulae in decrees other than those of assessment had even less relevance (D7, for example) and the formula of the Congress Decree perhaps least relevance of all. 2( The dating of the coinage Athenian monetary decree, by the order of names, grouped roughly into four districts, in List 5, and by the terms of the invitation to the Pan hellenic Congress quoted in Plutarch's Pericles. " 19 Quoted by Pritchett (BCH 89 [1965] pA30). The term "administrative district" was ill- advised and misleading. See above, p.122. 20 ATL III p.3l. 21 See also ATL III p.68. 22 Later called simply Ionic except in the assessment decree of 425/4. 23 The arguments for dating the pertinent fragment from Krateros in 454/3 have been presented in ATL I p.203 and III pp.9-11. Pritchett's contrary argument that ;ragav in Thuc. 1.96.1 (when the Oelian League was first organized) means "assessed" is strained and unconvincing, and the later ragas of Aristotle (whom he quotes) is irrelevant. Arnold Gomme, in his Commentary on Thucydides I (Oxford 1959) 272, warned that the translation must be "certainly 'settled which cities were to furnish money and which ships: not as Kirchhoff and Steup 'the amount of money and the number of ships which the cities were to contribute'." Benjamin Jowett, Thucydides 12 (Oxford 1900) 65, rendered the text "they immediately fixed which of the cities should supply money and which of them ships." Jacqueline de Romilly, Thucydide (Paris 1953), renders "ils fixerent quelles villes devaient leur fournir contre Ie Barbare de l'argent ou bien des vaisseaux." See now Malcolm F. McGregor's lecture in memory of Louise Taft Semple delivered in Cincinnati on Nov. 15, 1966 (published by the University of Cincinnati). 24 Quoted as 012 in ATL II p.61. This has now been dropped from the argument also by Mattingly (CQ 16 [1966] 187 n.5). 126 THE SECOND ATHENIAN TRIBUTE ASSESSMENT PERIOD decree (D14) depends on its letter forms, especially the three-bar sigma, which Russell Meiggs, on the basis of a full study of all datable fifth-century texts, shows to belong "before 445."25 This is in confor­ mity with the date 449/8 assigned to the decree by the editors of ATL.26 The Kos fragment of the monetary decree deserves more careful study than it has hitherto received. Professor Georgiades of the University of Athens, a noted petrologist, thinks that the marble is Parian.27 He is conservative in his judgement, but quite certain that it is not Pentelic as Segre thought. I have seen the stone and, even as a layman, am equally sure that the marble is not Pentelic. I do not know that it is Parian, and can only say that it looks like many other frag­ ments of marble on Kos, of all ages from the sixth century B.C. to Roman times. Georgiades avers that it is of a different texture from those Koan samples that he has analyzed. I have collected samples of Koan marble from quarries not visited by Pritchett and Georgiades and have submitted them to Georgiades for analysis. He reports that they differ from the marble of the monetary decree. One quarry at least that I had hoped to visit in 1966 must await a later opportunity until I can return to the island. But the identity of the marble is irrelevant to the question of the Attic text and its date. The inscription is Attic, of mid fifth century, whatever the stone on which it is cut, and Meiggs has posed some interesting problems about why a Koan engraver should have used a Parian stone.28 If the marble was Parian the text is far more likely to have been cut in Athens than at Kos. Pritchett denies this,29 but in his digression from the Kos stone to the early tribute lists comes to several other conclusions that I believe to be equally in error. The statistics for dating three-bar sigmas before 445 have now been well assembled; the question of four districts instead of five in the second assessment period is still open, but of little meaningful importance; the First Tribute Stele still resists Pritchett's attempts to reinterpret it;30 and his dating of the first 25 ]HS 86 (1966) 97. 26 ATL II p.61. 27 BCH 89 (1965) 400-22. 18 ]HS 86 (1966) pp.97-8 n.44. 29 BCH 89 (1965) pA38. so See B. D. Meritt, "The Top of the First Tribute Stele," Hesperia 35 (1966) 134-40; Malcolm F. McGregor, "The Postscript of the First Attic Quota-List," above in this issue pp.103-12. What Pritchett means by rejecting IG 12 304B as evidence for sometimes starting an inscription on the reverse of a stone at some distance below the top is not clear to me BENJAMIN D. MERITT 127 tribute fragment from Krateros rests on an improbable translation of Thucydides 1.96.1. At the conclusion of his article on the Kos fragment Pritchett has an "Epigraphical Commentary," and two photographs, neither of which is as good as that published by Segre in Clara Rhodos. He questions the reading of tau (undotted) in line 2, saying that "this letter is certainly an iota." It is, on the contrary, certainly a tau. I have examined it in the museum at Kos and now call attention to the very clear photograph in Clara Rhodos (9 [1938] 156).31 One most important point Pritchett does not mention: the right edge of the stone is preserved (so described by Segre) and shows that one less letter must be restored at the right than has been the custom hitherto. Segre thought there might be one or at most two letters restored at the ends of lines 9 and 10 and chose to restore two. But this would leave no margin whatever at the very edge. No more than one letter can be there restored. D14 (The Fragment from KOS)32 a. 449/8 a. ETOIX. 41 [---------------------------~ [A€ - - - - - - - - - - - - - - - - - - -JC.:L ra y[ ... 6 ••• ] [ •••••• 1.3. • • • •• apxovT€~ €V r€aL 7T J OA€a£ € ap [XOVT€ J [s • AO€valov ...... :~ ...... hOLJ OE hdA€VOrafL[laLJ 5 [ra apyvpOK07TLa €V riaL 7TOAW£ alvaypacpovrov· €~[v oJ [E fLE opeas avaypacp€L ro €K rav 7T J OA€OV rwos, eua[yhJ [0 ho f3OAOfL€VO~ avrlKa fLCxAa €S rJEv €/-.Lalav rEv ra[v OJ [wfLOO€rov TOS €OLK€KOras· hO]L OE OWfLoO'[r]aL 7T'[VO'] [EfL€POV oovrov oLKas rot~ cpEv]auL hEKaurov· €av OE [a] 10 ['\'\o~ EXao TOY apxovrov €v r]ia£ 7TO..\W£ fLE 7Tm€L Ka[rJ [a ra €cpu€cp£a,.dva € TOY 7ToALJrov € TOY Xa'vov, [ar ]LfL[ o~J [Earo, ra OE xp'fLara avro O€]fLoaLa [Ea]ro Kat r€~ 0100 r[oJ [€7TLO'KarOv, Kat €l fJ-€ €la£v J apXOVr€S 'AO€VaLOV, €[ TTL ] [TEAwavrov hoaa €V rOL cpU€Jfff!fLan hOL apXOVT€[s hJ (BCH 89 [1965] 437 n.3). Does he imply that the upper part of the reverse of IG J2 304A (i.e., 304B) was inscribed? If so, with what text? Other evidence is found in IG J2 25 (the reverse of IG 12 24). But neither of these examples is necessary to the description of the First Tribute Stele, on which I see neither a "high curving ridge" nor an "an athyrosis" on its top. 81 The reading of tau is also confirmed by the Director of the Museum at Kos. 32 See ATL II p.63. Initial alpha of line 13 was read by Pritchett; the last omicron of line 18 was read as doubtful by Segre, but denied by Pritchett. 128 THE SECOND ATHENIAN TRIBUTE ASSESSMENT PERIOD 15 [Ot h€Kaa'T€S' TES' 7T6,\€0S" Kat] N.p. p.€ 7TOtOat KaT~ T[~ Jcp] [a€cptap.Eva, laTo KaT~ TOV apx]6vTOV TOVTOV 7T€pt [an] r .. ' <;;, , 'A(} , '<;;' \ -] , [ , \'] Lt-"taS' OLOxatS' €v€at· €V O€ TO t apyvpoK07T tot TO ap [yvptoV O€xaap.EvoS' K6cpaat p.€ l'\]cnn>[v € hEp.vav Kat] [a7ToMa(}a£ - - - - - - - - - - - - - - - - - - - - -] A new fragment of D14 has been reported recently from Odessa. It was attributed by its first editor, P. O. Karyshkovsky, to Olbia, but J. B. Brashinsky expressed doubt, calling attention to the fact that many stones in the museum at Odessa had come from many parts of Greece.33 I am indebted to Eugene Vanderpool for a reference to some of these arrivals described in the Athenische Mittheilungen (10 [1885] 114) with the quotation of a news report of March 19,1827, taken from the local press at Odessa: "Un batiment arrive de l' Archipel avait decharge une partie de son lest pres du mole de la quarantaine. M. Sontag, capitaine du port d'Odessa, ayant remarque par hazard parmi les pierres qui composaient ce lest, quelques debris en marbre, crut de­ voir les examiner de plus pres; il se trouva que ces morceaux con­ stituaient deux pierres sepulcrales avec bas-reliefs et inscriptions parfaitement bien conservees." The captain of the vessel declared that the stones came from Delos, but the author of the news report thought Rhenaia more likely. There is also a report in the same article (op.cit. p.116) of other inscriptions in the museum at Odessa once in the collection of the Russian consul at Smyrna. The text at Odessa preserves one reading not attested by any of the other known pieces: in the bouleutic oath the words are K]at p.~ xp[ fjTat in the third line from the bottom. The connective Kat must now replace if (which was restored) in line 12 of the fragment from Siph­ nos34 and in line 7 of the fragment from Smyrna.3S The same change must also be made in the composite text in §12.36 The new fragment also adds certainty to two passages previously restored: (1) in §11 of the composite text l6vT~ oaa [K]~'\€Vov[aLv] replaces l6vT[a oaa K€'\€VOVatV]; and (2) in §12 OPKOV [T]6V TfjS' {3ovMjS' replaces 0pK[OV T6V TfjS'] {3ov'\fjS'. For the present at least the provenience of the new text must re- 33 The text of Karyshkovsky is published in SEG XXI IS, where the Russian references will also be found. 34 ATL II p.65. 35 A TL II p.66. 36 A TL II p.6S. BENJAMIN D. MERITT 129 main unknown. The only obvious epigraphical peculiarity (omicron with a dot in the center) does not localize the script, for such omicrons are known from various other parts of the Greek world, as well as from PontosP No Pontic city can have been a member of the Athenian Empire after the Peace of Kallias and before 425/4 when Athens threw off all restraint and openly assessed the cities of the Euxine. The heralds in 425/4 must have journeyed to Pontos, though nothing is said in the decree of assessment of that year about their doing so. If the coinage decree (DI4) is dated in 449/8 one does not expect mention of the Euxine in it as part of the route of the heralds. Were the decree later than 425/4, it is strange that the heralds were not instructed to include Pont os in their travels, since the Pontic district occupied so large a section of the actual roster of tribute-paying cities in that year (A9 IV, lines 126-173). But the decree of assessment itself does not mention the Euxine (lines 5-6) and one must fall back upon the instructions to the heralds (lines 40--41) to include these new names in the roster. Yet once included the names could hardly have been con­ sidered a mere appendage to the Thrakian district. The Hellespontine district would have been more appropriate, for surely the Euxine was visited by sea, not overland from Abdera and Ainos. The absence of any mention in the monetary decree of sending a herald through the Hellespont to Pont os is perhaps some indication that the date of D14 is not later than 425/4. If the decree was earlier than 425/4, i.e. 449/8, no mention of a Pontic district, of course, was to be expected. Mattingly would date the monetary decree later than the appear­ ance of Syme in the quota lists. One copy of the decree was set up on Syme, and Mattingly argues that Syme first began to pay tribute in 434/3 (List 21 VI, line 28).38 This is to confuse payment of tribute by Syme in her own name with membership in the Empire. The mone­ tary decree was to be set up in every city, and Syme had no doubt been a member of the Empire from the beginning.39 Announcement of its special status in the quota list of 434/3 cannot be used as an argu­ ment for the date of D14. The financial decrees of Kallias (Dl and D2) have been made the 37 See the note in The Collection of Ancient Greek Inscriptions in the British Museum IV (London 1916) pp.50-1. 38 See Ancient Society and Institutions; Studies presented to Victor Ehrenberg (Blackwell, Oxford 1966) 195-6. 39 Cf ATL I p.562, and III p.SO; also Mattingly, in Ancient Society and Institutions, p.216 0.15. 4-G•R •B•S• 130 THE SECOND ATHENIAN TRIBUTE ASSESSMENT PERIOD subject of recent study, with an attempt to date them down from the accepted year 434/3 to some time in the late 'twenties.40 The argu­ ments used involve so much hypothesis, or even misinterpretation, in my judgement, and so much neglect of good evidence, that it would be tedious to restate in full the case for 434/3. Suffice it to say, in brief, that the provisions ofDI refer, without any reasonable doubt, to the many local officials (whether called Tap,lat, or €7T£UTeXTat, or hU:P01TOtol, or IXPXOVT€~) who controlled the money of the various Other Gods before the creation of the new united board of Taf.Llat TWV IXMwv (lEwv. The new board was to be constituted on the model of the already existing board of Taf.Llat TWV Tfj~ 'A()'YJvala~ (D2) and share with them the use of the Opisthodomos as treasury. Pending the formation of the new board the various debts to the Other Gods were to be sought out (from wheresoever to be found) and liquidated from funds at the disposal of the hellenotamiai (Dl, lines 2-7; D2, lines 20-21). Temporarily, these repayments were to be deposited with the Treasurers of Athena, until the new Taf.Llat TWV IXMwv ()€Wv could take over and steward them in their own name. It is grotesque, for example, to claim that the €7T£UTCXTat of D1, lines 18-19, were the €7T£uTeXTat of the Athenian mint.41 The audited accounts of money borrowed by the Athenian State from the Sacred Treasuries (IG 12 324; M. N. Tod, A Selection of Greek Historical Inscriptions2 [Oxford 1946] 137-41, no. 64) show that the Taf.Llat TWV IXMwv ()€WV had opened their books for loans to the State in 433/2 and that they had loaned to the State in eleven years more than 821 talents.42 It is absurd to believe that the repayment of this, or of any part of this, was envisaged in D2, lines 21-23, or that the much greater debt to Athena was in any way repaid, even in part, by the 3000 talents of DI, lines 3-4. Athens was in no position in 421 to repay these debts, when the Archidamian War had exhausted not only the imperial revenues but also the accumu­ lated sacred treasure with which Thucydides tells us she began the war. And a better explanation can be offered for the 3000 talents in 434/3. This is of some epigraphical interest because of the dating thus early 40 H. B. Mattingly, ProcAfrCA 7 (1964) 35-55, argues a date for them in the seventh prytany of 422/1 B.C. See esp. his p.51 n.9O. Citations of the texts hereafter are either to the Corpus or to ATL II. n Mattingly, ProcAfrCA 7 (1964) 5l. U A new and complete text of this important document is to be published soon in CQ. See especially lines 102-124. BENJAMIN D. MERITT 131 of the datives in -cus of Dl. Such datives are known in the quota lists in 430/29, and, without laying emphasis on the Praxiergidai inscrip­ tion43 or the Phaselis decree,44 may be accepted as not unreasonable in 434/3. But other fallacious arguments have also been put forward to show that the Kallias decrees must be later: (1) the use of com­ pounds with avv- rather than xavv-, and (2) the use of imperatives in -laOw rather than -aaOw. Mattingly gives a table which purports to show that no text where avv- replaces xavv- throughout can have been inscribed much earlier than 420 B.C.45 The word "throughout" is Mattingly's, but it is hard to define. If a scribe had a predilection for writing avv- it is no more significant to find him using it five times, or twenty times for that matter, than to find him using it once. Mattingly notes that avv­ occurs in Dl "no less than five times, while the older Attic xavv- is not used once." The "five times" are inconsequential. The scribe used avv- because of his choice of this form, but the number of times he used it depended on the compound words he had to write, and means nothing for the date. Indeed, in a footnote46 Mattingly cites many earlier usages of avv-. It occurs in the datable records of the tribute lists exclusively in 454/3, 446/5, 445/4, and 442/1, yet there cannot be any question of dating these texts in the late 'twenties. The word "exclusively" here is my own. Each text offers only one example, but this is what the scribe chose and is just as good evidence for his predilection as if he had "no less than five" chances to display his choice. The other table exhibited shows the usage of the "old" imperative forms -aaOw, -aaOwv and the later forms -laOw, -laOwv.47 I do not know where Mattingly got the singular form in -aaOw; the only attested examples of the singular, at any date, are in -laOw, and one wonders whether he has taken into account the difference between singular and plural. Confusing the two, he has the Kallias decrees fall into the "period of transition" in the 420's. The text of D1 has avaa€/Lawaa8ov in lines 17-18 and D2 has [J1T£/LJ~Ma[OoJ in line 10. But again the criterion of date is illusory; it is equally appropriate to put 43 IG 12 80; cf David Lewis, BSA 49 (1954) 17-21. 44 IG 12 16. 46 Ancient Society and Institutions (supra n.38) 198. Mattingly seems to regard this as a clinching argument for the late date of the decrees of Kallias. 46 Mattingly, op.cit. (supra n.45) 217 n.n. 47 Mattingly, op.cit. (supra n.45) 199. 132 THE SECOND ATHENIAN TRIBUTE ASSESSMENT PERIOD both these texts into 434/3, or even earlier. Reference to singular imperatives in -ou8w should be eliminated,48 and one example of -l.u8w as early as 453/Z should be added to show the long span in years of this singular form.49 Mattingly has another table (p.ZOl) showing the "older" forms in -ou8wv in the monetary decree (014) and in the decrees for Chalkis (017) and Eretria (016). This is all quite in order for the accepted dates of these texts in 449/8 and 44615. He shows also the form in -l.u8w in 07 (the Kleinias decree), which is appropriate for 448/7. But he brings the "later" form of the plural in -l.u8wv into evidence from the Hestiaia decrees (IG 12 40 and 41). Both examples thus cited depend on erroneous restoration, for the Corpus text can in no sense be trusted.50 The stone in IG 12 41, line 6, shows clearly [- - ]8vvl.u8o (no final nu) where the Corpus reads [d]8vvl.u8w{ v) [even with an erroneous omega], and where Mattingly incorrectly restores a final nu, reading [EvJ8vVEU8o[v]' The text from IG 12 40, line 18, is also a bad restoration. There are, in fact, in the Hestiaia decrees no ex­ amples of imperatives in -l.u8wv. The imperatives in -Eu8w, which they do exhibit, are normal for the accepted date ca. 446/5, as well as for any other date in the fifth century. The bearing of all this on the coinage decree (014) is that the one preserved plural imperative (-oa8wv in the Siphnian text) inclines toward the earlier rather than the later date, but is in no sense probative. INSTITUTE FOR ADVANCED STUDY January, 1967 48 Naturally this caveat does not apply to a-stem yerbs like 8t8wJLt (cf, e.g., Dll, lines 37-38) or JLta86w. 49 DlO, lines 37-38. 50 Some attempt at restoring part of the text ofIG 12 40 was made in ATL III pp.301-2 n.4.