Chi 
 
arcl^aeologfcal Slnistitute of ^mtvica 
 
 EXTRACT FROM THE JOURNAL OF ARCHAEOLOGY 
 VOL. Xrit, NO. 1 
 
 ATTIC BmLDEsra ACCOXmTS 
 
 BY 
 
 ■ A -KIT Tl T\T1kTmi«*/-\/-\T 
 
 CN384 .dm'" ""'"""'•>' '■"■™n' 
 *"iiiri{l™,'i!?/",? accounts. 
 
 olin 
 
 3 1924 029 831 272 
 
A.^'^ixi H- 
 
American .Scijool 
 of Classical StuJies 
 at atfjens 
 
 ATTIC BUILDING ACCOUNTS i 
 
 [Plates II-IV] 
 
 I. The Parthenon 
 
 The Parthenon was begun during the political contest be- 
 tween Pericles and Thucydides, the son of Melesias (Plutarch, 
 Pericles, 12). The decree which authorized the construction, 
 dating probably from soon after 450 B.C., has never been dis- 
 covered, but numerous fragments, seventeen of which may now 
 be accepted as correctly identified, have been from time to time 
 attributed to the stele containing the actual expense accounts. 
 It was Kohler who discovered the first clue ; to a series of seven 
 fragments assembled by Kirchhoff (Monatsb. Berl. Akad. 1861, 
 p. 860; I. Gr. I, 300-311), referring to a construction which was 
 carried on through a period of at least fourteen years, begin- 
 ning with 447/6,2 Kohler (^Ath. Mitt. 1879, pp. 33-35) united 
 four others (one of them mistakenly),^ of which two mention 
 marble for pediment sculptures. Kohler's identification is now 
 almost universally accepted.* To his ten authentic fragments 
 additions have been made by 
 
 Foucart — three new pieces — including I.Cr. I, 300-302 a, 
 
 1 1 am greatly indebted to Dr. B. Leonardos, formerly curator of the Epi- 
 graphical Museum at Athens, for the opportunity to study these accounts on 
 the original stones, and to Dr. A. D. KeramopouUos, the present curator, for 
 permission to publish the already known pieces and to add some new fragments 
 of the accounts of the Propylaea. 
 
 2 The date of these fragments had previously given no hint, because it was 
 then supposed that the I'arthenon was begun before 454/3, and to it were 
 assigned the fragments I. (?. I, 284-288 (Kirchhoff, Mem. d. 1st. 1865, pp. 129- 
 142 ; Michaelis, Parthenon, pp. 287-288). 
 
 3 I.G. I, 111 ; 297 a and 6, suppl. p. -37 ; 311 a, suppl. p. 74. The first of 
 these, I, 111, does not belong with the others ; the letters are too widely spaced. 
 
 * Bannier alone in recent times speaks of this Inscription as unidentified 
 
 {Bh. Mus. 1906, p. 220). 
 
 American Journal of Archaeology, Second Series, Journal of the co 
 
 Archaeological Institute of America, Vol. XVII (1918), No. 1. °" 
 
54 
 
 WILLIAM B. DINSMOOB 
 
 --^r^ 
 
 _L. 
 
 o 
 (J 
 
 Q Ul 
 
 p" ''" <«■ r 
 
 Q hJ < 
 
 o 
 
 IS L 
 
ATTIC BUILDING ACCOUNTS I 55 
 
 suppl. p. 147, and two pieces wrongly assigned, /. Gr. I, 312-313, 
 331 (B.C.R. 1889, 174-178), making thirteen in all.i His 
 results were followed by Michaelis (^Arx Athenarum,^ A.E. 10). 
 
 Bannier — four new pieces — adding I. G. I, 220-221 {Bh. 
 Mus. 1906, p. 223) and I. a. I, 327 (Ath. Mitt. 1902, p. 304), 
 and two pieces wronglj"^ assigned, /. Gr. I, 330, and I. Q-. II, 4323, 
 suppl. p. 293 {Rh. Mus. 1906, 221; 1908, 429, 434) .^ 
 
 Cavaignae — two new pieces — adding Fig. 20 and Fig. 26 
 above, of his Mudes sur I'histoire finanoiire d'Aihines au V 
 siicle (pp. 1, Ivi, Ix, Ixii, and pi. II.), to Foucart's eleven 
 authentic pieces, and one of Bannier's (Z Gr. I, 327), rightly 
 discarding Foucart's I. <?. I, 312-313 (I.e. p. Ixv) and 331, but 
 unfortunately omitting Bannier's I.Gr. I, 220-221; his total 
 was fourteen pieces. 
 
 Woodward — two new pieces (£.&A. 1909-10, pp. 187-198). 
 
 For convenience I number the seventeen authentic pieces 
 (Cavaignac's fourteen, Bannier's I. Gr. I, 220-221, and Wood- 
 ward's two new pieces) as follows: A = Z G'. I, 220-221; B = 
 297 a ; C = 297 S ; D + E = 300-302 ; F = 300-302 a; Q = 303- 
 305 ; H = 306-307; I = 308-309 ; J = 310 ; K = 311 ; L = 311 a; 
 M = 327 ; N = Cavaignae, Fig. 20 ; = Cavaignae, Fig. 26 
 above; P=B.S.A. 1909-10, p. 187; Q = ^.*S'.A 1909-10, p. 
 190. All of these seventeen fragments are now in the Epi- 
 graphical Museum at Athens, even including I.Cr. I, 327 (M), 
 which Bannier supposed to be missing (Ath. Mitt. 1902, p. 304). 
 Of the five fragments wrongly assigned to the stele, two, I. Gr. 
 I, 111 and 312-313, are in the Museum, while three, I. Gr. I, 
 
 330, 331, and II, 4323, are missing ; two, I. G. I, 312-313 and 
 
 331, belong to the Propylaea, while the others remain 
 unknown. 
 
 Five fragments of the Parthenon stele had been dated — 
 J in 444/3 B.C. and I in 444/3^43/2 B.C., on account of the 
 
 1 Both I. Cf. I, 312-313 and 331 belong to the accounts of the Propylaea, as 
 we shall see (Part III). 
 
 2 Neither I, 330 nor II, 4323 is now in the Epigraphical Museum at Athens ; 
 actual comparison is therefore impossible. But I, 330 would require drastic 
 emendation, and could not in any case be combined with I. G. I, 309, as Bannier 
 proposed, this position being occupied by Cavaignac's new fragment N. II, 
 4323, as represented in the Corpus, seems to preserve the width of the narrow 
 edge as 0.12 m., while in the Parthenon stele it was 0.198 m. 
 
56 WILLIAM B. DINSMOOB 
 
 names of the secretaries for the Hellenotamiae,i and D + E ^ in 
 434/3 B.C. and F in 433/2 B.C., on account of the names of the 
 archons — and formed an unquestionable sequence, before 
 Michaelis in 1901 attempted the first publication of the series 
 as a whole (^A.E. 10). His result was hardly successful; he 
 included I. Gr. I, 312-313 and 331, which really belong to the 
 Propylaea, and refused Christ's dating of fragment I. Bannier 
 suggested several combinations, as of B, C, and G (only B and G 
 really go together) and placing A in the fourteenth or fifteenth 
 year (actually the former). The first comprehensive recon- 
 struction of the series was attempted by Cavaignac in 1908 (I.e. 
 pp. 1-lxix, pi. II), on the basis of a study of the types of letter- 
 ing as shown by squeezes ; he decided that the stele was about 
 20 X 60 cm. in section and of great height, inscribed on the 
 front (six years), the back (six years), and the left edge (three 
 years). He failed, however, because of having used squeezes 
 rather than the original stones; in three cases (D, I, and 0) 
 fragments preserving the top or bottom of the stele are placed 
 in the middle, and the marked differences of weathering on the 
 back and front are completely disregarded ; it happens, there- 
 fore, that he mistakes the reverse for the obverse of the stele, 
 that he misdates (in some cases by as many as nine years) one 
 or more faces of nine of the fourteen fragments that he employs, 
 that he restores only one column instead of two on each broad 
 face of the stele, and that he does not recognize that the fourth 
 face of the stele was also inscribed. Cavaignac's publication 
 brought out two new combinations, the joining of C to 0, and 
 the association of I with N. The discovery of other combina- 
 tions, coupled with the recognition of Cavaignac's errors, caused 
 me to undertake, on the basis of the fifteen fragments known 
 before 1910, a new restoration which was already complete 
 when Mr. Woodward discovered the two new fragments which 
 gave rise to his publication (B S.A. 1909-10, pp. 187-198); 
 his own views as to the correction of Cavaignac's arrangement 
 he was so kind as to withdraw in favor of my restoration, of 
 which he publishes a description (I.e. pp. i88-190) and a 
 diagram (I.e. p. 198). 
 
 1 J. Christ, De publicis pop. Ath. rationibus, p. 34 ; cf . 7. &. I, 236, 237. 
 
 2 Joined by Kirclilioff in 1861, I.e. 
 
ATTIC BUILDING ACCOUNTS I 57 
 
 Of the seventeen authentic fragments, four (D, F, G, and N) 
 are opisthographic, and exhibit in common one face badly 
 weathered, with granulation of the marble near the letters and 
 with vertical grooves caused by the beating of heavy rains, while 
 the opposite face has the surface beautifully preserved, but with 
 the letters sometimes corroded where drops of rain in compara- 
 tive quiet trickled down and lodged in them. We shall find 
 reason for identifying the well-preserved face as the obverse, 
 the other as the reverse of the stele. Of the thirteen frag- 
 ments that are not opisthographic, five (A, E, I, J, and M) are 
 from the better preserved face or obverse, and eight (B, C, 
 H, K, L, 0, P, and Q) from the weatherbeaten reverse. 
 
 On three of the fragments, D and + P,^ the original top 
 surface, roughly dressed with the hammer and point, is pre- 
 served. D has the entire thickness of the stone, 0.198 m., so 
 that the other two cannot be from the opposite face at the same 
 corner ; these three pieces either were from two separate stelae 
 or formed the two upper corners of a single stele ; the latter 
 is certain, as appears from the uniformity of the marble grain 
 and of the corrosion of the surfaces. The words and letters 
 on these two upper corners of the badly weathered face cannot 
 possibly be combined in one column of inscription, so that we 
 must certainly restore two columns. ^ Furthermore, to these 
 three fragments from the upper corners may be joined live 
 others which preserve parts of one of the narrow edges and of 
 one or both of the broad faces. To D Kirchhoff had joined E, 
 to which in turn I fitted accurately the fractures of A and H. 
 To -f- P (joined by Woodward) Cavaignac had fitted C, and 
 to this again I fitted F which had formerly been placed just 
 below D + E.^ We now have two groups, each of four frag- 
 ments, from the upper right and left corners of the stele, and 
 preserving the upper portions of six columns, two on each 
 
 1 The junction between these was made by Woodward. 
 
 - This conclusion may be found repeated in Mr. Woodward's article (p. 189). 
 
 3 The arrangement of F below D + E was the primary cause of Cavaignao's 
 errors ; it brought writing characteristic of the obverse below that of the reverse, 
 wherefore Cavaignac assumed that the upper half of what is properly the re- 
 verse, being different in character from the lower half of the reverse which was 
 continued (!) into the upper half of the obverse, must have been inscribed earlier 
 and was therefore the obverse. 
 
58 WILLIAM B. DINSMOOB 
 
 broad face and one on each narrow edge. The narrow edges 
 contain the fourteenth and fifteenth years respectively, as ap- 
 pears from the archons' names, so that the accounts of thirteen 
 years must be distributed between the four other columns. 
 + P give the name of the first secretary of the senate for the 
 year concerned as Ilet^taSe?,! who appears also in the Propylaea 
 account of 437/6 B.C. {I.Cr. I, 314); this must, therefore, be 
 the eleventh year of the Parthenon series which began in 
 447/6 B.C. Since + P form the head of the right-hand 
 column of the weatherbeaten face, this column is obviously 
 the fourth of the series of six, and the weatherbeaten face is 
 the reverse of the stele ; for after only two more years both 
 broad faces had been filled and the accounts could be continued 
 only on the narrow edges. Column IV must have contained 
 the eleventh, twelfth, and thirteenth years; column III, also 
 on the reverse, began likewise with a new year (fragment D ; 
 cf. the discussion of the eighth year below), and must have 
 contained the eighth, ninth, and tenth years. On the obverse 
 appeared, therefore, the accounts of the first seven years. Frag- 
 ments I and M, both of the obverse, preserve the bottom of the 
 stele ; the latter has also a narrow edge of the stele at the right 
 and so forms the end of column II and of the seventh year's 
 account ; this conclusion is verified by the contents of M, ex- 
 penses and the surplus which concludes each annual account. 
 Fragment I cannot be combined with M and so forms the end 
 of column I ; it ends in the middle of a year's account, the fifth 
 year as appears from Christ's identification of the secretary for 
 the Hellenotamiae. Column II therefore contained only half 
 of the record of the fifth year and all of the sixth and seventh 
 years ; the seemingly disproportionate length of column I will 
 be accounted for when we observe that the second year was 
 apparently omitted. 
 
 Each annual account began with a prescript, followed by 
 two series of entries, receipts (Xefifiara) and expenses (avaXo- 
 fiara}. The first receipt was always that of the surplus from 
 the epistatae of the previous year ; the last expense was the 
 surrender of the new surplus to the epistatae of the next year ; 
 besides the surplus in silver drachmae each always included 
 1 Woodward, B.S.A. 1909-10, p. 188. 
 
ATTIC BUILDING ACCOUNTS I 59 
 
 "97^ gold staters of Lampsacus and Cyzicus which were never 
 spent. The heading for the expenses was simply the word 
 avaXofiara, occupying a line by itself. The heading for the 
 receipts was a phrase, (tovtok') Xe/ifiara to iviavro (tovto') rdSe, 
 incorporated with the end of the annual prescript ; the TovToa 
 and Tovro were sometimes omitted because of lack of space, as 
 in the third year and in the eleventh to fifteenth years inclusive 
 (as read below). Each annual prescript (except the first) 
 occupied the entire width of a column ; below the prescript the 
 column was divided, forming a double column for each annual 
 account, the left half recording only sums of money, the right 
 half the items. ^ Each entry begins with a fresh line (unlike 
 the Erechtheum accounts), regardless of where the previous 
 entry terminated, so that the columns have no proper right 
 edges. The letters, however, almost always observe the stoi- 
 chedon arrangement, spreading only when the amount of avail- 
 able space is excessive. The type employed in the reproduction 
 of these inscriptions resembles closely that on the original 
 stones, except that the E should have the three bars equal in 
 length,, and in M all four strokes should slope. 
 
 The annual prescripts are completely preserved only for the 
 fourteenth and fiteenth years (D and F); the fragments of the 
 eleventh year (0 + P) show that there the same formula ob- 
 tained, i.e., Tot? eTTierrdTecn hol<; iypaiMfiaTeve iirl re? 
 
 SeKare? /SoXe? hei 7r/30TO? eypa/ifiaTeve iirl dpypvTO'; 
 
 ^AOevaioia-(^iv^. But fragments D (reverse), F (obverse), H 
 (reverse), and N (obverse) all begin the prescript with E • , 
 EPI • • , or EPITE ■ • , and then after a gap appear the letters 
 • • SHEI (fragment I). The restoration might be (1) 'Eirl re? 
 
 dpxe<; hei (Kirchhoff, in I. &. I, 304), (2) 'Eirl re? e? 
 
 ap'xe<i hei, as in some tribute lists and in the Propylaea accounts 
 (used by Cavaignac for his ninth and tenth years), (3) 'EttI 
 
 Te? ap'x^e<; re? e? hit, as in some of the tribute lists, 
 
 (4) 'Etti t€? y8o\e9 het (proposed by Woodward, I.e. p. 193, 
 without analogy), or (5) 'EttI re? e? /3o\e? hei, on the 
 
 1 The same arrangement is characteristic also of the Propylaea stele. In one 
 earlier account {I.G. I, 286), the money column is at the right of the item 
 column ; in another (/. G. I, 289-296) , the sums are mingled with the items as 
 in the later accounts of the Erechtheum. 
 
60 WILLIAM B. DINSMOOR 
 
 analogy of the last years of the Partheuoa accounts (used by 
 Kirchhoff for I. Cr. I, 309, and by Cavaignac for his first eight 
 years of the Parthenon). The solution is given by fragments. 
 G and 1 ; in G we read EAP[a;[tAiaTet/e] and [w/oojTOS EAPA- 
 Ifi/j^dreve] in the first two lines, the latter referring to the first 
 secretary of the senate and requiring the word ^o\e, so that 
 Kirchhoff correctly used apxe with the other secretary ; in I we 
 read in the corresponding positions • ■ e]5 HEI and rPOT0<,^ 
 again requiring ap^e (formulae 1, 2, or 3) and /SoXe respec- 
 tively. The second phrase of the prescript might then be 
 restored, on the analogy of the Propylaea accounts, iirl re? 
 
 ^oXif hei irpoTO'i iypafi/idreve,^ were it not for the fact 
 
 that in the only two places at which the results can be tested, 
 fragments G and I, the words eVl re? /SoXe'i Mi would occupy 
 so much space as to leave none for the name of the first secre- 
 tary of the senate,^ which would be absurd ; I therefore use the 
 shorter form rei /SoXei on the analogy of I. Cr. II, 867. Then 
 the numeral must have appeared in the first phrase, which 
 would have been like formula (2) or (3); only (2) is of the 
 right length to fit the allowable projection of the prescript to 
 the left of the item column in fragments I -f- N, of the fifth 
 year,* and this may be retained as the invariable first phrase 
 
 1 Kirchhoff read here, by an impossible contraction of the lines, ['ETri res- 
 
 /SoXajS HEI TIMO0["s ]5 n POT [s ^7paMM«i;e], whereupon 
 
 Penndorf {De scribis, p. 132) restored TIMOO[«os 'Ara^Xiio-rio]?, identifying 
 him with the father of Conon, and this was retained by Ferguson (77ie Athenian 
 Secretaries, p. 14) with the date ca. 435/4 derived from Michaelis. Cavaignac 
 rightly expanded the lines, but rid himself of the secretary in the second line 
 by reading P P0T0[7^''"] ; yet a trace of the § of irporos is actually preserved. 
 
 2 This phrase ml.G. I, 814-315 is usually restored xal iwl res /SoXes k.t.X; 
 but there is no. space for the koI. 
 
 8 That is, if the lines are restored by means of the combinations G + H and ' 
 I + N, as discussed below. 
 
 * On N, the prescript of the fifth year begins two spaces to the left of the 
 money column of the year ^bove, the fourth (cf. Cavaignac, Fig. 20) ; on J, 
 the money column of the fourth year, as appears from the restoration of 
 [^^[s]§§H]EKTE, began sixteen spaces to the left of the item column of 
 the same year (cf . Cavaignac, Fig. 21) ; on I it appeai-s that the left edge of the 
 item column shifted one letter space between the fourth and fifth years (cf. 
 Cavaignac, Fig. 22). The prescript of the fifth year began then nineteen spaces- 
 to the left of the item column of the same year ; on I it appears that the item 
 Column begins under the E of HEI and EPI TE[s Triiarrcs dpx?]? H mate 
 up exactly the number of nineteen letters. 
 
ATTIC BUILDING ACCOUNTS I 61 
 
 of the first ten annual prescripts. The rest of each prescript 
 gave a list of the epistatae (omitted in the last five years) clos- 
 ing, instead of opening as in the Propylaea accounts, with the 
 word eTTicTTciTat. Finally came a phrase evidently beginning 
 with H01§ (fragment I) and ending with [eVp] AM MATEY [e] 
 (fragment G),^ as if the person named were secretary to the 
 epistatae, already named in the first line ; for this reason 
 Kirchhoff and Cavaignac restored the same name in the first 
 and fifth lines of the prescript on G ; hut in the fifth year 
 the two names in these positions were certainly different, 
 TIMOO'£o9] and A^ ■ • • .^ Analogous are the tribute lists 
 of 443/2 and 442/1 B.C. (J. a. I, 237, 238), when two secre- 
 taries were attached to each board of Hellenotamiae ; the 
 first iypafifidreve, the second ■x^crvveypafifudTeve, and the second 
 we note was permanent, at least for two years ; the Erech- 
 theum board of 408/7 had also a hviroypa/M/j^aTevi, Pyrgion of 
 Otryne (/. 6r. I, 324, a, e). Traces of the name of this third 
 secretary in the Parthenon accounts are preserved only in the 
 fifth year, A^ • ■ • , which Kirchhoff restored A h[TiKXe 9],^ caus- 
 ing Michaelis to date the fragment wrongly as oa. 435/4 B.C.* 
 In the eleventh year, however, fragment P gives the name of 
 the secretary for the epistatae as ANl, tempting Woodward 
 (I.e. p. 188)"^to restore 'ANT[t«:\e9 'Oe6'6i' ?] EAPAMMATE[i;6]; 
 but to place the demotic before the verb is unusual, the spacing 
 of ANl seems to indicate that there were twelve letters instead 
 of thirteen before the iypafi/idTeve, and the last of these appears 
 to have a vertical hasta (cf. Cavaignac, Fig. 26); I restore 
 therefore 'ANT[t«>.e9 xo-v] NEAP AM MATE [fe]. This longer 
 verb fills also the required space on G. There is no space for the 
 ■)(crvv- on the reverse of F which may be identified as a fragment 
 of the twelfth year (see below), and in the thirteenth year (Q 
 as identified below), where the end of the name is preserved,^ 
 the x'^vv- is certainly omitted. It would appear, therefore, 
 
 1 The unique last lines on the obverse of F, KAIHO • • • and P0I5I • •, 
 will be discussed later (third year). 
 
 2 It was to avoid the mention of three different secretaries in the fifth year 
 that these first two lines have always been wrongly restored. 
 
 3 Reiected by Cavaignac. * Accepted by Ferguson. 
 
 6 Woodward reads • • ]0§ EAP[aM]MATE["E] {I.e. p. 193), but it is a 
 vertical hasta, as for . . ] E§, that appears on the stone. 
 
62 WILLIAM B. DINSMOOB 
 
 that Anticles attained the full secretaryship in the twelfth 
 year. The annual prescripts may now be restored without 
 difficulty ; the formulae are given below in tabular form : 
 
 First Ten Yeaks Last Five Ybaks 
 
 Ett^ t€s €S apx^^ (To?$) iircaTdTetn 
 
 hec iypcLfi/jidTeve hots 'AvtikXcs (x(rt;i')e7pa/A/t(£Tei/e 
 
 TEi ^oXet i-rl res (• • ■ es Kal) SeKdret ^oXes 
 
 TrpoTos iypaiJifidTeve Aei Trporos iypa/iiidreve 
 
 " iirl &PXOVTOS 'A0evaiot(r(_iv) 
 
 (list of ca. 5 epistatae) 
 
 iTnardTai 
 
 Aots ('AcTiKXes) x^vv€ypa,p.pA,Tev€(^ ) 
 
 TOi/Tois Xi/iiMTO. TO inavTO (toiJto) rdSe \4p,pAiTa to iviavTO (roiro) rdSe 
 
 It is interesting to note that the change of the formulae coin- 
 cides with the dedication of the chryselephantine statue and 
 with the beginning of the Propylaea, and so perhaps marks an 
 entire change in the administration of the Parthenon. 
 
 The only fragment of the top of the obverse of the stele is D 
 (Z Gr. I, 300), which appears in the seventh year of Cavaignac's 
 scheme; here the first five lines contain larger letters than the 
 rest, seemingly part of the first annual prescript, though the 
 preserved letters are at the top of column II ; this indicates 
 that, as in the Chandler Inscription of the Erechtheum (Z. Cr. 
 I, 322) and, as we have good reason to believe, in the stele of 
 the Propylaea, the first annual prescript extended across the 
 entire width of the two double columns. The preserved letters 
 are spaced 0.02 m. on centres; if the total width of the stele 
 be taken as about 1.20 m., on the analogy of that of the Propy- 
 laea (cf. p. xxx), we should have space for no more than fifty- 
 eight letters in the longest line (1. 5) ; the first six lines may 
 then be restored by following the analogy of the first prescript 
 for the Propylaea, observing the stoichedon arrangement. 
 
 L. 1 — [Ilap^evov] OS is offered merely as a suggestion, the official 
 name of the temple at this period being uncertain. 
 
 L. 3 — [Avt] I KU E§, Bockh {Staatsh.^ ±1, p. 305) ; this is rather the 
 name of the first epistates ; there is here space for five epistatae, the 
 usual number in the Parthenon, as in the accounts /. G. I, 289-296. 
 
 It is interesting to note that in this inscription of the summer 
 of 446 B.C. the n and S appear regularly, perhaps for the first 
 
w 
 
 o 
 
 DU 
 
 > 
 
 o 
 > 
 
 W 
 W 'Z. Ui 
 
 z 
 s 
 
 D 
 
 o 
 o 
 
 
 OS 
 
 < 
 
 s 
 
 < 
 
 
 X 
 
 a. 
 
 r— 
 
 
 
 ;^ 
 
 d 
 
 
 -e- 
 
 
 
 o 
 
 
 <i 
 
 
 Q. 
 
 CO 
 
 
 a 
 
 
 CJ 
 
 - 
 
 vu 
 
 vu 
 
 
 a 
 
 X 
 
 p ' 
 
 l- 
 
 
 
 Q. 
 
 tu I 
 
 d 
 
 
 a 
 
 d 
 
 t- ' 
 
 a. 
 
 
 h 
 
 a. 
 
 d ' 
 
 a. 
 
 
 
 
 - 
 
 d. 
 
 d 
 
 
 t- 
 
 H 
 
 d, ' 
 
 Q. 
 
 
 b 
 
 - 
 
 d 
 
 f^ 
 
 
 - 
 
 t= 
 
 a. 
 
 uj 
 
 
 t= 
 
 w 
 
 >- 
 
 ^ 
 
 
 V 
 
 
 UJ 
 
 s 
 
 
 - 
 
 
 ty> 
 
 b 
 
 
 d 
 
 
 O 
 
 X 
 
 
 « 
 
 
 l- 
 
 
 
 o 
 
 
 
 
 1 1 
 
 ;^ 
 
 
 t= 
 
 
 c^ 
 
 -< 
 
 w 
 
 
 
 d 
 
 1- 
 
 p 
 
 
 
 t- 
 
 !< 
 
 w 
 
 
 
 o 
 
 (-H 
 
 l- 
 
 
 
 1- 
 
 - 
 
 d 
 
 
 
 p 
 
 a 
 
 i 
 
 
 (y 
 
 o 
 
 V 
 
 a. 
 
 
 - 
 
 l- 
 
 (/• 
 
 d 
 
 
 O 
 
 o 
 
 «4/ 
 
 Q. 
 
 ^ 
 
 -« 
 
 l- 
 
 1- 
 
 ?- 
 
 lif 
 
 
 p 
 
 d 
 
 w 
 
 -< 
 
 
 d 
 
 Q. 
 
 
 o 
 
 
 ^ 
 
 V 
 
 
 <52. 
 
 
 ;:4 
 
 - 
 
 
 - 
 
 
 w 
 
 '< 
 
 
 V 
 
 
 o 
 
 -< 
 
 
 h 
 
 
 t- 
 
 d 
 
 
 ^ 
 
 
 d 
 
 Ui 
 
 
 - 
 
 
 t- 
 
 {/« 
 
 |y^ 
 
 b 
 
 
 d 
 
 u 
 
 - 
 
 - 
 
 
 i 
 
 ;k 
 
 o 
 
 o 
 
 
 i 
 
 o 
 
 ^ 
 
 - 
 
 
 w 
 
 t- 
 
 cy» 
 
 d 
 
 
 X 
 
 s< 
 
 d 
 
 ;^ 
 
 
 (y 
 
 w 
 
 -. 
 
 w 
 
 
 - 
 
 (■ 
 
 b 
 
 ■^ 
 
 
 o 
 
 - 
 
 d 
 
 < 
 
 
 h 
 
 X 
 
 ?- 
 
 (/« 
 
 
 s 
 
 Q. 
 
 a. 
 
 o 
 
 
 o 
 
 d 
 1 — 1 
 
 1 1 
 
 1 1 L- 
 
 _l 1 1 
 
 1 1 
 
 
 
 1 
 
 
 d 
 
 
 i- 
 
 to 
 
 ;^ 
 
 ,, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 — 1 
 
 1 — 1 
 
 
 
 
 
 
 
 
 1 — 1 
 
 1 — t 
 
 
 
 
 
 
 
 
 1 — 1 
 
 
 
 
 
 u 
 
 
 
 l- 
 
 
 o 
 
 d 
 
 o 
 
 o 
 
 
 1^ 
 
 .J 
 
 
 1 — 1 
 
 
 
 
 
 
 
 1 — 1 
 
 
 
 
 p 
 
 W 1 
 
 
 
 
 1= 
 
 o 
 
 d 
 
 ^ 
 
 u 
 
 1 — t 
 
 o 
 
 
 
 i- 
 a 
 
 o 
 
 1 1 
 
 1 1 
 
 o o 
 
 
 
 1 — 1 
 
 1 — 1 
 
 UJ 
 
 
 
 
 
 
 
 
 t- 
 
 P 1 
 
 
 t 
 
 
 (A 
 
 I- 
 
 t- 
 
 w: 
 
 i:^ 
 
 t- 
 
 ^ 
 
 ;*. >^ 
 
 
 1 — 1 
 
 w 
 
 P 
 
 
 
 
 
 
 
 
 d 
 
 UJ 1 
 
 
 - 
 
 
 - 
 
 p 
 
 b 
 
 d 
 
 Ui 
 
 b 
 
 a 
 
 w <aii 
 
 
 
 
 o 
 
 UJ 
 
 
 t — 1 
 
 o» 
 
 
 
 
 
 d. 
 
 K 1 
 
 
 t 
 
 
 o 
 
 d 
 
 - 
 
 b 
 
 •< 
 
 ^ 
 
 b 
 
 « t/- 
 
 
 
 
 f^i 
 
 t- 
 
 
 (y 
 
 
 - 
 
 
 
 
 
 d. 
 
 d • 
 
 
 >k 
 
 
 h 
 
 - 
 
 t= 
 
 ■6- 
 
 -. 
 
 1= 
 
 -e- 
 
 ^ w 
 
 
 
 
 ;^ 
 
 d 
 
 
 .^ 
 
 
 o 
 
 
 
 
 
 a 
 
 d. . 
 
 
 
 
 - 
 
 1^ 
 
 UJ 
 
 a. 
 
 Osjl 
 
 u 
 
 a. 
 
 JsJ) t- 
 
 
 
 
 o 
 
 a. 
 
 
 o 
 
 
 rS 
 
 w 
 
 
 
 
 Q. 
 
 i 
 
 
 
 
 a 
 
 u 
 
 •;*. 
 
 a 
 
 P 
 
 ;^ 
 
 a 
 
 p d 
 
 
 
 
 - 
 
 a. 
 
 
 »< 
 
 
 ;:^ 
 
 p 
 
 
 
 
 ?- 
 
 a 
 
 
 
 
 Si 
 
 O 
 
 O 
 
 < 
 
 W 
 
 O 
 
 (y 
 
 
 ty 
 
 
 
 a. 
 a 
 
 a 
 
 
 
 
 o 
 
 ^ 
 t. 
 
 
 
 
 »*j 
 
 Q. 
 
 
 
 
 w 
 
 l- 
 
 o. 
 
 Ir 
 
 (y 
 
 V 
 
 w 
 
 >u O 
 
 ■^ 
 
 
 
 t- 
 
 ^- 
 
 
 sc 
 
 
 o 
 
 a 
 
 
 
 
 
 ?- 
 
 
 
 
 p 
 
 a 
 
 UJ 
 
 w 
 
 w # 
 
 h 
 
 Q. 
 
 =1. r< 
 
 O 
 1 1 
 
 UJ 
 
 1 1 
 
 o 
 
 UJ 
 
 
 .J UJ 
 
 1= 
 
 i. 
 
 1 — 1 
 
 a 
 
 
 
 
 w 
 
 
 
 
 UJ 
 
 t- 
 
 y- 
 
 Q. 
 
 Q. 
 
 o 
 
 V 
 
 w ;». 
 
 X 
 
 > 
 
 ^ 
 
 (y 
 
 
 CO p 
 
 o 
 
 i. 
 
 
 
 
 
 w 
 
 
 
 
 h 
 
 a 
 
 o 
 
 UJ 
 
 w 
 
 
 
 I 1 
 
 
 
 t — 1 
 
 1 1 
 
 c 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 a. 
 
 t- 
 
 ^ o 
 
 2 
 
 LU 
 
 LU 
 
 (f 
 
 O V 
 
 a. 
 
 d 
 
 
 
 
 
 o 
 
 
 
 
 a 
 
 a. 
 
 Q. 
 
 K 
 
 W 
 
 
 
 
 
 
 
 
 1 1 
 
 1 1 
 
 
 
 t- 
 
 1 — 1 
 
 
 
 t- 
 
 
 
 
 d. 
 
 a. 
 
 t: 
 
 d 
 
 a 
 
 N 
 
 d 
 
 a — 
 1 — 1 
 
 o 
 
 H 
 
 -A 
 
 X 
 
 U 
 
 1 — 1 
 
 w 
 
 Q- 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 ;^ 
 
 t- 
 
 - 2 
 
 >- 
 
 < 
 
 _\ 
 
 — 
 
 < 
 
 LU <, 
 
 _J 
 
 ^- 
 
 a 
 
 o 
 
 
 
 o 
 
 
 
 
 a. 
 
 w 
 
 ;^ 
 
 h 
 
 t- * 
 
 
 
 
 
 
 LU 
 
 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 (y 
 
 V < 
 
 LU 
 
 2 
 
 m 
 
 — 
 
 w 2 
 
 Q 
 
 
 
 t- 
 
 
 
 Q- 
 N 
 
 
 
 
 a 
 
 X 
 
 o 
 
 b 
 
 b 
 
 i- 
 
 o 
 1 I 
 
 O h- 
 
 — 
 
 2 
 
 X 
 
 2 
 
 LU 
 
 X 2 
 
 h 
 
 
 a. 
 
 d 
 
 
 
 
 
 
 Q. 
 
 if- 
 
 1- 
 
 o 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ?- 
 
 ^ 
 
 d 
 
 b 
 
 b 
 
 a 
 1 — 1 
 
 W 
 
 w < 
 
 2 
 
 < 
 
 < 
 
 O 
 
 -i 
 
 < < 
 
 <- 
 
 
 
 a. 
 
 
 
 
 
 
 
 UJ 
 
 o 
 
 Q. 
 
 p 
 
 p 
 
 — 
 
 >- 
 
 >- Q- 
 
 < 
 
 Q. 
 
 Q. 
 
 Q- 
 
 _1 
 
 Q- Q- 
 
 Q. 
 
 
 o 
 
 w 
 
 
 - 
 
 
 
 
 
 ;^ 
 
 (■ 
 
 "a 
 
 o. 
 
 a- * 
 
 :^ 
 
 Q- 
 
 Q- < 
 
 H 
 
 < 
 
 < 
 
 1- 
 
 o 
 
 < < 
 
 < 
 
 
 
 ;^ 
 
 
 w 
 
 
 
 
 
 p 
 
 p 
 
 1 — 1 
 
 X 
 
 1 1 
 
 X 
 
 1 — I 
 
 1_ 
 
 X 
 
 X C 
 
 LU 
 
 LU 
 
 C 
 
 W 
 
 X 
 
 H LU 
 
 H 
 
 1 — 1 
 
 X 
 
 O 
 1 1 
 
 
 '^ 
 
 
 
 
 
 b 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 J. 
 
 .1. 
 
 
 
 
 
 
 [/> 
 
 
 
 
 
 X 
 
 t- 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 u. 
 
 
 
 4 
 
 
 
 w 
 
 " 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 = 
 
 
 L. 
 
 
 
 
 
 
 X 
 
 u 
 
 
 
 
 
 b 
 
 
 
 
 
 
 
 -L. 
 
 
 
 c_ 
 
 
 < 
 
 
 
 
 
 
 Q. 
 
 X 
 
 
 
 
 
 w 
 
 
 
 
 
 
 
 < 
 
 
 
 <1 
 
 
 < 
 
 
 
 ;k 
 
 
 
 d 
 
 O 
 
 
 
 
 
 l- 
 
 
 
 1 — 1 
 UI 
 
 
 
 LU 
 
 < 
 
 
 
 <d 
 
 
 < 
 
 
 
 
 
 
 w 
 
 02. 
 
 
 
 
 
 d 
 
 
 
 1— 
 
 
 
 H 
 
 X 
 
 
 
 S. 
 
 
 < 
 
 
 
 a 
 
 1 1 
 
 
 ^_^ 
 
 w 
 
 " 
 
 
 
 
 
 1- 
 
 
 
 iC 
 
 
 
 ^ 
 
 X 
 
 
 
 X 
 
 
 X 
 
 
 
 
 
 
 l- 
 
 w 
 
 
 (y 
 
 
 
 b 
 
 
 
 
 
 
 LU 
 
 X 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 l- 
 
 
 O 
 
 
 
 ^ 
 
 
 
 LU 
 
 
 
 1 — 1 
 
 E- 
 
 
 
 E. 
 
 
 X 
 
 
 X 
 
 
 
 -1-3 
 
 Q- 
 
 v 
 
 
 .^ 
 
 
 
 t= 
 
 
 
 X 
 
 
 
 X 
 
 E. 
 
 
 
 X 
 
 
 X 
 
 
 1 — 1 
 
 
 
 P-l 
 
 
 
 
 
 
 
 
 1 
 
 
 w 
 
 
 
 
 
 
 
 X 
 
 
 t — 1 
 
 
 2 
 
 
 
 tC 
 
 1- 
 
 S 
 
 (/> 
 
 ;^ 
 
 
 w 
 
 w 
 
 
 
 
 
 
 w 
 
 ^— 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 i_ 
 
 1 
 
 
 
 ) 
 
 
 w 
 
 
 
 
 1 1 
 
 
 
 1 — 1 
 
 
 
 
 
 
 
 w 
 
 1 1 
 
 1 1 L 
 
 1l 
 
 — 1 
 
 - 
 
 1 1 
 
 1 — 1 
 
 , 
 
 I — 1 
 
 E. 
 
 1 
 
 1 — 1 
 
 w 
 
 2 
 
 
 
 2 
 
 
 
 
 2 
 
 
 
 H 
 
 < < 
 
 _i c 
 
 1- . 
 
 X V 
 
 V 
 
 o 
 
 X 
 
 W 
 
 
 1 1 
 
 1 1 
 
 <> 
 
 1 
 1 
 
 E! 
 
 1 1 
 
 E. 
 <»• 
 <• 
 
 2 
 
 
 
 2 
 2 
 2 
 
 
 • 
 
 
 2 
 EL 
 
 
 
 <• 
 
 C_ 
 
 < 1- 
 
 - « 
 
 t - 
 
 - 
 
 < 
 
 O 
 
 
 
 
 
 1 — 1 
 
 tXi 
 
 
 
 1 1 
 
 
 t 1 
 
 
 1 — 1 
 
 
 
 <j. 
 
 LU 
 
 I V 
 
 V V 
 
 w h 
 
 - 
 
 •^ 
 
 Q. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ xl 
 
 
 
 
 
 o 
 
 « 
 
 ^ 
 
 1 1 
 
 
 
 1 — 1 
 
 1 — 1 1 — 1 
 
 1 1 
 
 
 
 
 
 
 p 
 
 UJ 
 
 CO 
 
 
 
 
 
 
 
 
 
 l- 
 
 
 
 
 
 
 
 
 
 [ — 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 d 
 
 w 
 
 
 
 a 
 
 CO 
 
 o 
 
 o 
 
 
 
 
 
 
 w 
 
 a 
 
 
 
 
 
 
 
 
 
 p 
 
 b 
 
 ^ 
 
 • 
 
 1 — 1 
 
 a. 
 
 a 
 
 t- 
 
 ^ 
 
 o 
 
 
 
 
 
 l- 
 
 
 1 — 1 
 
 
 
 
 
 
 
 
 d 
 
 a. 
 
 0^ 
 
 W 1 
 P ' 
 <*< 1 
 
 
 
 a. 
 
 a 
 a. 
 
 i- 
 
 a 
 
 w 
 
 Ji 
 a 
 
 ;^ 
 
 
 
 
 
 d 
 
 a. 
 
 -Q- 
 
 o 
 
 p 
 
 
 
 
 
 
 
 
 ;>. 
 
 a 
 
 p 
 
 t- 1 
 
 
 
 ?- 
 
 p 
 
 
 « 
 
 
 
 
 
 a. 
 
 /T 
 
 l- 
 
 
 
 
 
 
 1 
 
 
 UJ 
 
 <! 
 
 tii 
 
 a w 
 
 
 
 
 
 "* 
 
 b 
 
 "* 
 
 
 
 
 
 d 
 
 ;^ 
 
 a 
 
 
 
 
 
 
 1 
 a 
 
 1 
 
 o 
 
 O 
 
 ut 
 
 a. 
 
 cy 
 
 a. p 
 a. u. 
 
 d ' t- 
 
 
 
 p 
 b 
 
 o 
 o 
 
 UJ 
 
 a 
 
 p 
 
 it: 
 
 
 
 
 
 a. 
 
 o 
 
 a. 
 
 d 
 
 a. 
 a. 
 a 
 a. 
 
 ° o 
 
 
 2 
 
 o 
 
 
 ' 
 
 1- 
 
 ' 
 
 t- 
 
 u 
 
 UJ 
 
 a. d 
 
 ?- a. 
 
 
 
 X 
 
 (A 
 
 p 
 a 
 
 a. 
 
 w 
 
 
 
 
 
 
 
 t- 
 
 ?- 
 
 !< -. 
 
 
 Z 
 
 UJ 
 
 1 1 
 
 i 
 
 a 
 
 t 
 
 UJ 
 
 h 
 
 \- 
 
 w a. 
 
 
 
 
 
 
 '*' 
 
 
 
 • 
 
 
 o 
 
 >4J 
 
 "* O 
 
 
 I 
 
 1 1 
 
 
 1 — 1 
 
 1 — 1 
 
 1 1 
 
 1 — I 
 
 1 — 1 
 
 ' 
 
 UJ 
 
 "* 
 
 1- 
 
 ^ 
 
 ki i^ 
 
 
 ;:k 
 
 
 ^ 
 
 V 
 
 CO ^ 
 
 O o 
 
 ^X 
 
 
 H 
 
 
 1 
 
 w 
 
 < 
 
 z 
 
 UJ 
 
 < 
 
 < 
 
 (A a 
 
 o a. 
 
 
 X 
 
 ^ o 
 ■" a. 
 
 h- _ 
 
 
 K O 
 
 a - 
 1 1 1 1 
 
 1 I 
 
 o 
 1 1 
 
 1 — i 
 
 UJ 
 
 1 1 
 
 O 
 
 1 1 
 
 
 < 
 
 a. 
 
 1 ' 
 
 ■ 
 
 
 z 
 
 < 
 
 w 
 
 w 
 
 o ^ 
 
 1 1 
 
 t — 1 
 
 a 
 1 1 
 
 1 
 
 H 2: 
 W < 
 
 < 
 
 CD 
 
 2 
 <: 
 
 < 
 
 LU 
 
 Z 
 
 LU _ 
 W LU 
 
 xh 
 
 << 
 
 '^ a. 
 
 ^^ 
 t — ( 
 
 
 UJ 
 
 I 
 f- 
 
 LU 
 
 o 
 
 H 
 
 
 - 
 
 c_ 
 
 — 
 
 LlI 
 
 O 
 
 O 
 
 O w 
 
 
 ^- 
 
 < 
 
 O 
 
 w 
 
 O H 
 
 < 
 
 h- 
 
 < 
 
 X 
 
 w 
 
 
 ' ' 
 
 1 
 
 LU 
 W 
 
 >■ 
 
 a. 
 
 w 
 
 >- 
 
 w 
 
 >- 
 
 2 O 
 
 - H 
 
 1- 
 
 w 
 
 < 
 
 w 
 
 < 
 
 < 
 
 O 
 w 
 
 w < 
 
 > Q. 
 
 LU 
 
 < 
 
 0. 
 
 W 
 
 1 — 1 
 
 < 
 
 1 — 1 
 
 >- 
 
 LU 
 
 03 
 a 
 
 > 
 
 1 1 
 
 , 
 
 O 
 
 LU 
 
 LU 
 
 Q- 
 
 Q. 
 
 H O 
 
 — 
 
 — 
 
 2 
 
 Q. 
 
 >■ 
 
 n < 
 
 < 
 
 <- 
 
 
 a. 
 
 1 — 1 
 X 
 
 1 — I 
 
 Z 
 
 D 
 
 8 
 
 o 
 
 1 1 
 
 1 
 
 1 — 1 
 1 — 1 
 
 C. 
 
 L 
 
 X 
 
 X 
 
 — CL 
 
 LU L 
 
 ■e- 
 
 O 
 
 X 
 
 5 
 u. 
 
 < 
 
 a. 
 
 V 
 
 
 1 — 1 
 1 1 
 
 1 — I 
 1 1 
 
 o 
 
 1 1 
 
 a 
 
 1= 
 
 1 — 1 
 
 
 
 
 
 
 
 
 
 X w 
 
 w 
 
 — 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 < 
 
 
 
 
 
 
 
 
 W O 
 
 o 
 
 •^ 
 
 cy> 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 1 — 1 1 — 1 
 
 i — 1 
 
 1 — 1 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 UI 1 
 
 
 K 
 
 - 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 
 
 X ' 
 
 
 a 
 
 O 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 
 
 
 
 
 a. . 
 
 
 h 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ffl 
 
 
 
 
 
 
 
 1 — 1 
 
 a 
 
 
 b 
 
 p 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 LU 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 (y w 
 
 
 -^ 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 H 
 
 u, ^ 
 
 
 t= 
 
 »- 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 UJ 
 
 
 
 
 LU 
 
 H 
 
 
 
 
 
 
 
 
 
 
 1 
 
 ' 
 
 ' 
 
 
 
 LU 
 
 a. - 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 \ 
 
 ' 
 
 
 
 X 
 
 \u «u 
 
 
 
 
 
 
 LU 
 
 
 
 
 
 
 
 
 
 '• -1. E. 
 
 < J- X 
 
 X 
 
 X 
 
 X 
 
 
 
 w 
 w 
 
 
 
 
 
 
 
 X ' 
 w • 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 1 1 1 I 
 
 1 1 
 
 
 
 
 
 w < 
 
 
 
 
 
 
 
 
 
 . C X 
 ' ' X 
 
 X 
 X 
 
 X 
 X 
 X 
 
 E. 
 
 X 
 X 
 X 
 
 X 
 
 
 <> 
 
 
 LU > 
 
 I- UJ 
 
 1 — 1 
 
 1 — 1 
 
 
 
 
 1 — 1 
 
 
 
 
 
 
 
 
 
 
 . . . 
 
 2 
 
 
 a * 
 
 C ' 
 
 
 
 
 3 
 
 0! <J» . 
 
 1 I 1 — 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 LJ • 
 
 
 , 
 
 
 
 4^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 i 
 
ATTIC BUILDING ACCOUNTS I 63 
 
 time in regular use ^ ; in the spring of the same year A/ and ^ 
 had been used in the tribute list (J.(r. I, 233). 
 
 The next prescript, on F (J. Gr. I, 302 a), was in 11. 34-40 of 
 column I, separated by a gap of twenty-eight lines from the 
 first annual prescript, as may be calculated by its distance 
 below the top of the stele.^ This prescript I assign to the third 
 year for the following reasons. We noted that in column I 
 were accounts reaching to the middle of the fifth year. That 
 of the first year occupies exactly as much space (averaging 
 thirty -three lines) as the accounts in columns III and IV, 
 though these contain only three years each; the fragments of 
 the fourth and fifth years seem to be no less voluminous ; one 
 of the first years was evidently therefore omitted, unless we 
 are to accept the improbable solution that two annual accounts 
 were very much abbreviated. In 1. 40 we read P0I5I, hitherto 
 uninterpreted, but which may be restored as \toI<; Tr/jore] P0I5 1 [v 
 67rf a-Tarecri] , as if the under-secretary wrote the following 
 accounts both for the epistatae of the current year (KAI H0[t9]), 
 and for those of the year before, i.e. for the second and third 
 years. The probability that the omitted year was the second 
 rather than the third is increased by the fact that in the second 
 year came the revolt and campaign in Euboea, which Cavaignac 
 too (Z.c. pp. 76, 82, 87) supposes to have had some effect on 
 the work. Michaelis had assigned this fragment to a supposed 
 sixteenth annual account (432/1 B.C.), when the work had, 
 however, ceased ; Cavaignac, supposing it to be the third rather 
 than the second prescript, accidentally obtained what I believe 
 to be the correct date, the third year. 
 
 L. 33 — [rpiVes ySoXes], Cavaignac. 
 
 L. 38— [2<^£t]TI0§, Cavaignac. 
 
 L. 40 — TovTo, omitted for lack of space. 
 
 Fragments J (I.Gr. I, 310) and N (Cavaignac, Fig. 20) + I 
 (^I.Gr. I, 308), dated by the names of the secretaries for the 
 
 1 They had appeared previously in the tribute list of 449/8 {I.G. I, 231), 
 and § in the prescript of 453/2 b.c. (I, 227); cf. Roberts, Ch-eek Epigraphy, I, 
 p. 102. 
 
 2 It will be remembered that F was fitted to C, the latter heing a piece of the 
 top of the stele. 
 
64 
 
 WILLIAM B. BINSMOOB 
 
 Hellenotamiae, give almost the complete account of the fourth 
 year, without the prescript. The numbering of the lines is 
 based merely on the average of thirty-three lines for each year. 
 
 L. 76 — [^4[il5SH]EKTE, restored from comparison with other 
 years, locates the left edge of the money column. 
 
 L. 77— [PMMM] \pa|F1HHHAAhh, made stoichedon with 1. 75, 
 and with the smallest amount possible ; any larger sum would require 
 T, which was not used as appears from the trace of the last M. 
 
 L. 80— [MMMM]XX|HH|aAAnillll, restored for reasons given 
 below. 
 
 L- 82 — X5Eh[o8tKov], restored by Cavaignac (I.e. p. Ixvii), identi- 
 fying them with the ypa^l fevtas who were active in this very year 
 (Philochorus, Frg. Hist. Gr. p. 398, 90). 
 
 Ij- 84 — "'"P[i£po7rotov], suggested by Meyer, who prefers how- 
 ever T[eix'"^<"o''3 (Forschungen z. alt. Gesch. II, pp. 100 n. 3, 104) ;. 
 cf. Andocides, III, 7. 
 
 L. 85 — [pMMM]1, the only possible restoration, since I read 
 the last figure as M, as does Cavaignac in his pi. II; others have 
 read it A, as does Cavaignac himself at his p. 87. 
 
 L. 95 — [tJOSEPI^I [aras], KirchhofE and Michaelis; Cavaignae 
 (I.e. p. 88) supposes an allusion to the makers of epistylia, but such, 
 specialization seems unwarranted. 
 
 L. 96 — rEYKINAK[al l\<£wa], Cavaignac (I.e. p. 88, pi. II). ' 
 
 The most important of the facts revealed by this account is- 
 the exact size of the receipt from the Hellenotamiae, obviously 
 a part of the tribute from the Delian confederacy. It was 
 evidently the use of this that caused the controversy between 
 Pericles and Thucydides, and most writers, therefore, as Meyer 
 (Forschungen, II, p. 99) and Furtwangler (^Masterpieces, p.424), 
 suppose it to have been a large amount. But it is noticeable 
 that though the Hellenotamiae appear in the accounts of the 
 fourth, fifth, and ninth years, and probably therefore contributed 
 during the first ten years for the erection of the Parthenon,^ 
 in the last five years they gave nothing. From 437/6 to 
 433/2 B.C., we find that they devoted their attention to the 
 Propylaea, and in the accounts of the Propylaea the sum they 
 gave is specified as ^-^ of the annual tribute, the airap'x^E which 
 
 1 The other seven (six) are so mutilated as to give no information. 
 
ATTIC BUILDING ACCOUNTS I 65 
 
 was the rightful shai-e of Athena (J. Gr. I, 260). ^ It is practically 
 certain that this was the very sum which had been diverted 
 from the Parthenon, and which had been expended for that 
 building during the first ten years. ^ Since Pericles therefore 
 "used only the airapxe in his constructions, and not the money 
 in the treasury of the Confederacy, it would appear that the 
 accusations by Thucydides were unfounded, a fact which would 
 perhaps account for the victory of Pericles. With the knowl- 
 edge that it is a sixtieth of the tribute, the exact amount is 
 easy to determine. Pedroli (in Beloch, Studi di St. Ant. 
 1891, pp. 101-207) estimates that in this Olympiad the total 
 annual tribute amounted to 414 talents 5170 drachmae, while 
 Cavaignac (I.e. pp. xliv, 91-92) gives 435 talents; the cnrapxe 
 would have been 41486|- or 43500 drachmae, respectively. 
 The nearest possible approach to this in our inscription is 
 [MM M MjXXfflHpA Annul, 426'i5f drachmae; all other possible 
 restorations vary so much from a sixtieth of any reasonable 
 tribute as to be certainly wrong ; ^ we may then be certain 
 that the total tribute in 444/3 B.C. amounted to 426 talents 
 4550 drachmae. 
 
 The remainder of column I is taken up by the prescript and 
 receipts of the fifth year, fragments N (Cavaignac, Fig. 20) + 
 I (J. Cr. I, 309), as dated by the name of the secretary for the 
 Hellenotamiae ; Michaelis had, however, assigned I to about the 
 twelfth year. 
 
 L. 101 — [/SoXejS, KirchhofE and Cavaignac. 
 
 LI. 101-102 — TIM0O[eos 'Ava<^Av(rTi]0§, Penndorf, de Scribis, 
 p. 132 (cf. p. 60, note 1). 
 
 L. 102— MATEY [e], Cavaignac. 
 
 L. 102 — rP0T0[7eV€s], Cavaignac. 
 
 L. 104 — AN[riKA.e?], KirchhofE (rejected by Cavaignac). 
 
 L. 109 — [tt] APATAN'[tov rov tes deo'], Cavaignac. 
 
 1 Purtwangler, to be sure, suggests that the sum given for the Propylaea was 
 not the d,irapxi but a second sixtieth {Masterpieces, p. 424 ; cf. Busolt, II, p. 568). 
 
 2 Meyer's conjecture that the Hellenotamiae gave money for the Parthenon, 
 for the buildings of the lower city, and for those at Eleusis (for all these except 
 the Parthenon we have no evidence) , at the same time that they gave the sixtieth 
 for the Propylaea, is disproved by my dating of fragments of the accounts for 
 the Parthenon. 
 
 SThe nearest are [MMMpa] or [pfj^XX], i.e. 37675| or 59675| dr. 
 
C6 WILLIAM B. DINSMOOR 
 
 L. 110 — EABA ■ - , a form of iy^aivo, Cavaignac {I.e. p. Ixvii). 
 
 L. 116 — X§E[vo8iKov], Cavaignac. 
 
 L. 116 — T E I X [ojrowv] , Meyer {I.e. pp. 100 n. 3, 103) ; cf . Andocides, 
 III, 7, and Plutarch, de glor. Ath. 8, Pericles, 13 ; Callicrates was 
 at the head of these epistatae as well as of those of the Parthenon. 
 
 The expenses of the same year appear at the top of column II, 
 just below the first annual prescript, on D + E (Z (r . I, 300) . This 
 and the following fragments of Column II are sc/ fragmentary 
 that they are printed here with little attempt at restoration. 
 
 avaA,o fi a t a 
 
 __..-_____ |MTI 
 
 ________ -ON 
 
 10 — — _ — — 
 
 * * * 
 
 — — — — — — — — -EPTE 
 
 ________ T05 
 
 20 — — — — — 
 
 — — — — — — — OTIME 
 
 — — — — — — Tt]\ETOYTO 
 
 * * # 
 
 L. 8 — [tp-yoXaySeVa] N T I or similar, Cavaignac {I.e. p. Hi). 
 
 L. 17— EPTE, Cavaignac. 
 
 L. 21 — [tovtov] TIME, Cavaignac. 
 
 Of the sixth year, we have only a few letters of the receipts, 
 some appearing on A (Z Gr. I, 220), which we have joined to E, 
 others on G (Z (?. I, 805) which, as will appear from the re- 
 verse (ninth year), falls in this position. 
 
 * * * 
 
 30 ________NTEN 
 
 — — — — — — — — ^EON 
 
 — — — — — — — — lEI 
 
 _________ I 
 
 ___ _ AP 
 
 35 r A[pa _ _ _ 
 
 _ _ _ 
 
 ANAUEo/iara] 
 
 * * * 
 
ATTIC BUILDING ACCOUNTS I 67 
 
 The fragment from the bottom of column II, M (J. Gr. I, 327) 
 gives us the conclusion of the seventh year (the sixth in Ca- 
 vaignac's scheme); the numbering of the lines is based on the 
 fact that this column ends seven lines higher above the bottom, 
 of the stele (0.195 rn.) than does column I (0.088 m.). 
 
 * * * 
 
 [XiCopyois T ojl K I ONON 
 
 -J OISXSYUINOIS 
 -J I TO 
 
 105 - - ]TOHY r o^r ouE I 
 
 ]l ISHYP0PA0I5 
 
 [irepieyevtTOTOEVtovToJTO YTO 
 [j(pi)(ro<7TaT£p£sAa;ii<)bcraK]E N I 
 [^p V IT o(TT are pcsJ^v ^ I K e. V o]l 
 (v a c a t) 
 
 L. 102 — a payment probably for channeling columns. 
 L. 103 — [cpyjOIS, Cavaignac; this was probably work on the- 
 ceiling and roof. 
 
 L. 106 — cf . the Propylaea stele, col. II, 1. 69. 
 
 The eighth year begins on the reverse of the stele, in column 
 
 III, fragment D (/. Cr. I, 302) ; the first line is three lines 
 lower, from the top of the stele, than is the first line of column 
 
 IV. Fragment E, which should have the continuation of this 
 account, is now badly worn and shows nothing. On the other 
 hand, fragment L (I. Cr. I, 311 a), obviously from the reverse, 
 cannot be assigned to the ninth, eleventh, or twelfth year be- 
 cause it contains payments for stonework which are duplicated 
 on fragments certainly to be assigned to these years ; these 
 formulae for stonework on L differ completely from those used 
 in the ninth to the twelfth years inclusive ; and the buying of 
 ivory and silver, and the payment to gold workers, on L, must 
 certainly antedate the selling of ivory, silver, and gold from 
 the ninth year onward. By elimination then the only suitable 
 year of the six on the reverse is the eighth ; here fragment L 
 does not come into actual contact with any other fragment, but 
 the length is such as to agree with the allowance of thirty-three 
 lines to this year, as determined by the location of H, of the 
 
68 WILLIAM B. DIN8M00B 
 
 ninth year. The last line is given by fragment G {I.G-. I, 
 303), which is discussed below. 
 
 L. 7 — TE[t0pa(Tios], the only possible demotic, if the line begins 
 with a demotic. 
 
 L. 21 — E[A£<^as ioviOi], Cavaignac; confirmed by the traces of U 
 and E; cf. I.G. I, 298, suppl. p. 146. 
 
 L. 24 — probably work on the ceiling and roof, continued from 
 the previous year. 
 
 L. 29 — perhaps for the painted decoration of mouldings, but 
 more probably, when taken into connection with 11. 21-22, 30-31, 
 for the pedestal of the Athena Parthenos, which contained decorar 
 tions in ivory and gilded silver {I.G. II, 662 B, 16; 664 6, 6; 660, 
 62; 661, 22; 676, 40; 701 ii, 60; 703, 10), and must have been 
 finished along with the statue itself in this same year (cf. the 
 account of the ninth year). 
 
 LI. 30-31 — this work in silver might, like the ivory in 11. 21-22 
 (cf. the accounts of the temple of Asclepius at Epidaurus, Baunack, 
 Aus Epidaurus, 11. 6't-66), appear to be part of the decoration of 
 the doorways {I.O. II, 675, 4; 678 A, i, 29; 701 ir, 47; 704; 708), 
 were it not that the doorways are not mentioned until the tenth year. 
 
 L. 33 — possibly {Jiv']^'^ ^[yo'O- 
 
 For the ninth year almost the complete record is preserved. 
 Fragment H we have fitted accurately to the bottom of E ; it 
 contains the beginning of a prescript, in 11. 38-41 according to 
 the distance below the top of the stele ; the line above this was 
 blank, wherefore the eighth year must have ended in 1. 36. 
 Bannier had correctly suggested (J?A. Mus. 1908, p. 429) that 
 B and G be associated on account of their contents ; he is fol- 
 lowed by Cavaignac, but Woodward remains unconvinced (I.e. 
 p. 195, n. 1); there is no actual point of contact. The date is 
 given by the fact that, although this pair of fragments is cer- 
 tainly from the reverse, the demotic of the secretary for the 
 Hellenotamiae in G (i". G-. I, 305) is [Pa/x,] NOSIO[s] or [Hay] NO- 
 §IO[s],^ which can belong to none of the years of the reverse 
 except the ninth ; the demotics of these secretaries for the five 
 other years are known, the eighth Rv/SdSei (I. 6r. I, 240), the 
 tenth 'Axapvevi (I, 242), the eleventh Kepafieik (I, 243), the 
 twelfth At^o-oi/ew (I, 244, 544), the thirteenth 'Axapvev<; (I, 
 
 1 Hitherto wrongly restored [Muppi] N05 IO[s]. ^ov which there is not space. 
 
tg 
 
 
 
 
 S 
 
 
 
 
 H 
 
 r-p 
 
 
 
 
 
 3 
 
 1 — 1 
 
 
 w 
 
 
 
 
 DU 
 
 ' 
 
 
 (< 
 
 
 
 
 ,_^ 
 
 ' 
 
 
 a 
 
 
 
 
 to 
 5^ 
 
 w w 
 
 3. 
 
 
 T 
 
 r^ 
 
 C- 
 
 S ? 
 
 3. 
 
 
 o 
 
 o 
 
 — 
 
 w w 
 
 8 
 
 
 t- 
 
 1 — 1 
 
 > 
 
 
 
 
 
 
 
 X 
 
 t- l- 
 
 a. 
 
 
 8 
 
 w o 
 
 J 
 
 8 B 
 
 ^- 
 
 
 l- 
 
 ■< a. 
 
 o 
 > 
 
 S- a. 
 
 w 
 
 
 b 
 
 8 w 
 
 
 s. s. 
 
 ^ 
 
 
 - 
 
 b « 
 
 
 a a 
 
 s 
 
 
 1= 
 
 -e- - 
 
 
 Q. Q. 
 
 b 
 
 1 — 1 
 
 w 
 
 4.»vP 
 
 
 ?- ^- 
 
 X 
 
 w 
 
 ;^ 
 
 a p 
 
 
 W VLl 
 
 (/• 
 
 «5 
 
 o 
 
 <1 ^ 
 
 
 
 (/» 
 
 tu 
 
 a 
 
 Q. 
 
 (/. CP 
 
 
 
 o 
 
 ^ 
 
 i- 
 
 vu 
 
 (U U 
 
 
 
 t- 
 
 v: 
 
 o 
 
 1- 
 
 Q. a. 
 
 
 
 o 
 
 -' 
 
 l~ 
 
 O 
 
 u w 
 
 
 
 
 a. 
 
 t- 
 
 p 
 
 Q. 
 
 I- l- 
 
 „ 
 
 
 
 =: 
 
 ;4 
 
 o 
 
 N 
 
 8 a 
 
 n^ 
 
 
 
 
 < 
 
 h 
 
 ;u 
 
 t- i- 
 
 2 
 
 
 
 
 tf 
 
 o 
 
 o 
 
 b b 
 
 S 
 
 
 
 
 - 
 
 t- 
 
 t- 
 
 o o 
 
 
 
 
 o 
 
 p 
 
 8 
 
 b b 
 
 o 
 
 ■^ 
 
 
 ■s; 
 
 a 
 
 Q. 
 
 p p 
 
 o 
 
 W 
 
 
 - 
 
 - 
 
 a 
 
 Q. Q. 
 
 
 -^^ 
 
 
 a 
 
 ^k 
 
 1 1 
 
 X X 
 
 1 1 1 — 1 
 
 w 
 
 (/• 
 
 
 (. 
 
 w 
 
 
 
 UJ 
 
 
 
 
 
 
 
 s 
 
 w 
 
 
 a 
 
 o 
 
 
 
 UJ 
 
 
 
 
 
 
 
 w 
 
 X 
 
 
 ^ 
 
 h 
 
 
 
 Q 
 
 
 
 
 
 
 
 2 
 
 Q. 
 
 u 
 
 b 
 
 a 
 
 
 
 O 
 
 
 
 
 
 
 
 o 
 
 B ^ 
 
 < 
 
 ^ 
 
 i- 
 
 
 
 W 
 
 Ut 
 
 3 
 
 t= 
 
 a 
 
 
 
 >-" 
 
 u CS 
 
 i 
 
 u 
 
 s. 
 
 
 
 2 
 
 O 
 
 UJ 
 
 <: 
 
 
 
 
 
 
 ( — 1 
 
 o - 
 
 
 w» 
 
 A 
 
 
 UJ 
 
 «D « 
 
 u 
 
 o 
 
 w 
 
 
 h- 
 
 o 
 
 ?- I 
 
 
 -J 
 
 -< 
 
 
 ^ 
 
 a; 
 
 
 
 
 
 
 
 < 
 
 O 
 
 
 b 
 
 w 
 
 
 UJ 
 
 [L. 
 
 o 
 
 (/> 
 
 
 8 
 
 - 
 
 
 X 
 
 < 
 "z. 
 
 K 
 
 w 
 
 
 Q. 
 
 o 
 
 I 
 
 w 
 
 h 
 
 
 <l) 
 
 (■ 
 
 ■ 
 
 w 
 
 1 1 
 
 D 
 O 
 
 -> 
 
 
 -> 
 
 p 
 
 1 
 
 <J.E. 
 
 ~> 
 
 1 1_ 
 
 L_ 
 
 _J 1 — 1 
 
 1 — 1 
 
 1 1 
 
 i_1j 
 
 2 
 
 _l "^ 
 
 . UJ 
 
 o 
 
 <it<y<3r 
 
 s 
 
 J. 
 
 1 -e 
 
 - X 
 
 : 1- 
 
 1- 
 
 <h 
 
 ^<f 
 
 o 
 
 a. 
 
 8 
 
 
 o I- 
 
 
 a. 
 
 ^,w VV w 
 o < — < 
 z — o _ 
 > o < 2 < 
 O -io- o O 
 H _iO 1- < 
 UJ < _iO < 
 
 2 z > O O 
 — O w 
 
 H I- X _i_i 
 
 b 
 8 
 
 ?- 
 
 u 
 
 a u, 
 
 I 
 
 •" I o 
 
 " 1 111 
 Q. < — 
 Www 
 
 < < — 
 - -^o 
 ^ ^x 
 
 _iQ- O 
 O O w 
 
 O O > 
 
 a. 
 
 I — I 
 
 8 
 O. 
 
 < ■ 
 
 Q. 
 
 < 
 
 < 
 
 I- 
 
 < 
 
 < 
 
 8 "J 
 
 b 4; 
 
 8 >- 
 
 <! ^ 
 
 (/. w 
 
 u UJ 
 
 < I 
 
 2 . 
 
 O r- 
 OC 
 w L 
 
 x.x. 
 
 p 'u' 
 
 u p 
 (- w 
 
 a t> 
 8. a 
 
 4. S. 
 
 a =*- 
 I 1 1 1 1 — 
 
 a. < H 
 
 < a. w 
 
 UJ < < 
 w UJ — 
 O w w 
 _iO ^ 
 
 - h- ^ 
 
 I — 1 1 — 
 
 ■©• o 
 
 b t= 
 
 p 
 
 I I "^ 
 
 u> X 
 
 w o 
 
 <t UJ 
 
 X H 
 
 W < UJ 
 
 UJ S < 
 
 < 2 < 
 
 - < T" 
 
 I — 1 1 — I 
 a. o 
 
 b ?-. 
 
 8 
 
 l- 
 
 b 
 I I 
 
 C_ 
 UJ 
 
 o 
 
 UJ 
 
 I — I 
 
 o 
 
 a. 
 
 B ^ 
 
 ■e- — 
 
 2 H 
 
 < > 
 
 w w 
 
 X — 
 
 z w 
 o o 
 
 — Z o 
 
 2 s. i- 
 
 I — I 
 
 a a :■ 
 
 t- PL, w 
 
 
 < 
 
 Q. Q. O 
 
 b o 
 X i- 
 
 < 
 
 tA X o 
 
 w a :i_ 
 I I 
 
 I I I — 1 1 1 
 
 - O CL >- 
 
 C Q- < VV 
 UJ C- LlI X 
 
 * p 
 
 8 
 
 a. 
 
 8. 
 a a 
 Q. a- 
 8 ^- 
 
 0: 
 
 UJ 
 
 UJ 
 
 X 
 
 w 
 
 jW 
 
 E. 
 <1" 
 
 8 
 
 a. 
 
 t= 
 
 o 
 
 -= 8. a. 
 
 I — I 
 a o 
 
 a. h 
 b P 
 
 o o 
 
 I I 
 
 b UJ 
 
 p 2 
 
 a. — 
 t I 
 
 \ H 
 
 8 
 
 b 
 w 
 
 t- 
 
 iC 
 
 UJ 
 
 ^? 
 
 N 1- 
 
 UJ .^ 
 
 ^ B 
 UJ ;^ 
 X u, 
 
 UJ 8 
 
 a-Q. 
 ^UJ 
 
 z -e- 
 
 ,o>- 
 
 _iUJ 
 
 >- a. 
 w < 
 X c_ 
 
 z 
 o 
 
 Q. < 
 
 w < 
 < 
 
 < 
 
 UJ < 
 
 I- H 
 
 z w 
 
 UJ UJ 
 
 L. Z 
 
 z w o 
 
 O - I- 
 
 Z H O z 
 
 O < S O 
 
 H 2 O O 
 
 < O I- - 
 
 2 0^^ 
 
 UJ W "^ !>• 
 
 I — I 
 
 Z - - o 
 
 8 - 
 a 8 
 
 B ^ 
 
 I I 
 
 ^< 
 
 OH 
 
 — w 
 
 — w 
 
 <C o 
 
 •^ >- 
 
 VV 8 
 
 — X 
 
 O x: 
 
 ■^ P 
 
 O Si 
 1= 8 
 O l- 
 CO -> 
 Ob- 
 
 ^ 
 
 
 
 
 o 
 
 b 
 
 
 
 
 I- rr 
 
 ^ 
 
 
 
 
 s o 
 
 r-, a. 
 
 
 
 
 » a^ 
 
 a »■ 
 
 
 
 
 h „ o 
 
 - o 
 
 
 
 
 O s< s^ 
 
 a. - 
 
 
 
 
 *- 8 " 
 
 ^ UJ ». 
 
 1 
 
 
 
 == b * 
 " 1 1 
 
 w t- kf 
 
 
 
 
 ^ b - V 
 
 ' 
 
 
 
 8 -9-- 
 
 
 
 
 1 1 
 
 1 II — II — 1 r 
 
 w< 
 
 • 
 
 
 -2 H 
 
 1 — 1 
 
 _i< z 
 
 1 a 
 
 , 
 
 
 — < >- 
 
 UJ 0. UJ 
 
 1 — 1 
 • Ql 
 
 1 — 1 
 O 
 
 
 UJ _ii£: 
 
 H UJ \ 
 
 z<^ 
 
 , UJ 
 
 1 1 
 
 X 
 
 
 O w w 
 
 1- UJ UJ 
 
 ^1 a 
 
 W X 
 
 < 
 
 
 E T 
 T E P 
 T E P 
 
 RTHENOl 
 
 1 1 1 1 
 
 T< w 
 
 ^ 
 
 z 
 
 w w a 
 
 o 2 UJ 
 
 1 1 
 
 w 
 
 o 
 
 B tA b 
 
 -OH 
 
 UJ 
 
 — 
 
 Z < < < 
 
 Ol, 
 
 i^ t- 1- ,, 
 
 - a o 
 
 * < a. — 
 
 >- 
 
 z 
 
 ^. - i- 
 
 ' 
 
 
 
 ■^ UJ 
 
 O SC 8 
 
 r >- < 
 
 UJ 
 
 UJ 
 
 < w w I 
 
 ^-^ a. 
 
 :> o^ 
 
 :^ 
 
 2 
 
 UJ o O 
 
 '^V? o 
 
 a o ^ 
 
 O w w 
 
 w 
 
 < 
 
 <o ■as a 
 
 1 — 1 
 
 b 
 
 * 1 — ^ 1 — 1 
 
 1 — 1 
 
 1 — 1 
 
 -= P H 
 
 - - ^, 
 
 - o • 
 
 ' 
 
 8 
 
 « a. a. D 
 
 X X a 
 
 a. (- ■ 
 
 , 
 
 :>£ 
 
 b X X 8 
 
 1 — 1 1 11 1 
 
 1 — 1 1 1 1 — 1 
 
 1 1 
 
 1 — 1 
 
 1 — ii II — 1 O 
 
 < 
 
 o 
 z 
 
 5 
 5 
 
 CQ 
 
 1 — 1 
 UJ 
 
 1- 
 
 it: 
 
 UJ 
 
 X 
 
 w 
 
 w 
 
 1 — 1 1 — 1 
 •cJ-E. 
 
 EK" 
 
 s 
 
 < 
 
 (3 
 
 OQ 
 
ATTIC BUILDING ACCOUNTS I, 69 
 
 ^16). 1 The connection of H and G to form the prescript has 
 already been discussed, and confirms the location of these frag- 
 ments in the ninth year.^ The ninth annual account was com- 
 posed therefore of H (1.0-. I, 306), G (I, 305), and B (I, 
 
 297 a). 
 
 L. 38 — [Em TEs apx^s A"]> Kirchhoff. 
 
 L. 38 — [Aioj/w(rt]<t>IL,0§, rests only on the amount of space to be 
 filled; ['Apto-rdj^lUOS restored here by Cavaigaac. 
 
 L. 39 — rPO[Tos] hitherto restored, but B exists as the fourth 
 letter. 
 
 L. 40 — EAP[ajn/xaTeue], the usual restoration, is impossible after 
 "the same word at the end of the preceding line ; it must be some 
 proper name. 
 
 L. 41 — X5Y[Tr£Taov], Cavaignac; this does not seem the proper 
 place for a demotic. 
 
 L. 42 — [ — (^i\o9 eyp], Kirchhoff; ['Apio-Toc^tXos eyp], Cavaignac. 
 
 L. 48 — [Mvppt]NOSIO[s], Kirchhoff, as discussed above. 
 
 L. 49 — [dp-yupto irpoidevTo]^, restored because the number of letter 
 spaces fits, and because the weight is omitted, as it should be in the 
 <3ase of unwrought silver. 
 
 L. 50 — [irpa&€VT']OHP , Cavaignac; perhaps [o-Tafl/*] ON P might 
 be better. 
 
 L. 61 ■ — / [^pva-io] ; Cavaignac saw here a trace rather of E. 
 
 L. 53 — XSYL,ON[7rpa^eVTo]l, Bannier; this was evidently surplus 
 -wood left from building the ceiling and roof in the two preceding 
 years. 
 
 L. 54 — EY<l>EP[o], probably the father of Ilpcms (Aristophanes, 
 Acharn. 843; I.G. I, 225 Ic, suppl. p. 174; Ath. Mitt. 1894, p. 163). 
 
 L. 55 — SAYPONC [s], who contributed also to the Propylaea in 
 437/6 B.C. 
 
 L. 59 — n ENTE[\€^ev koI k.t.X.], Kirchhoff; the xot I omit because 
 ■of lack of space. 
 
 L. 61 — [AoSoTTot] 1 S, Woodward (I.e. p. 191, n. 3), who, like Liddell 
 
 1 For the two last see the discussions of the Propylaea accounts of the corre- 
 sponding years. 
 
 2 Bannier {Rh. Mus. 1908, p. 429 n. 2) had proposed that G and B be placed 
 below C, which conflicts, however, with my junction of C and F. Cavaignac 
 (I.e. p. lix) rightly dated G and B in the ninth year by elimination, but rather 
 ■accidentally, since he supposed that the secretary's demotic in the twelfth year 
 
 was 4k TO (/. Gr. I, 244 ; really Alxcoi/eis, cf. the Propylaea stele, col. 
 
 I, 1. 63), and that the thirteenth year was on a narrow edge. Woodward would 
 ■assign B to either the tenth or the thirteenth year. 
 
70 WILLIAM B. DIJSrSMOOR 
 
 and Scott, cites Xenophon, Cyr. VI, 2, 36 and Aeschines, 57, 27 ; 
 a better analogy is the speech of Pericles (Plutarch, Pericles,. 12), 
 mentioning the employment of oSon-oioi on this very building ; cf. 
 I.O. II, 834 c, 28, also Francotte, L'industrie dans la Or&ce ancienne 
 (II, p. 86), and Ferguson (Trans. Am. Phil. Assoc. 1904, p. 17). 
 
 L. 62 — [eiri TO. kvkXo], Woodward (I.e. 191-192) ; previous restorar 
 tions were iwl t6s ovos (I.G. I, 297 6) or em Ta hvrro^vyta (I, 297 a). 
 
 L. 63 — [tos heKovras ^]ENTE^ L[6£v] in previous readings. 
 
 L. 65 — restored by Woodward (I.e. p. 192). 
 
 The selling of gold and silver, to which ivory was added at 
 least as early as 437/6 B.C., began in this ninth year (439/8) 
 and continued until the close of the work. These material's 
 were sold in such large quantities (cf. the fourteenth year) that 
 they cannot have come merely from the pedestal of the chrysele- 
 phantine statue; the epistatae of the Athena Parthenos must 
 have turned over to the treasury of the Parthenon the surplus 
 material remaining after the completion of the statue. We 
 may be certain, therefore, of the correctness of the generally 
 received view, which makes the date given by the chronogra- 
 phers (Eusebius Armen. 440/39 B.C., Hieronymus 439/8, Philo- 
 chorus 438/7) that of the completion of the statue, and not of 
 its inception, as Pareti would prefer (^Hom. Mitt. 1909, p. 271 f .). 
 The evidence seems to show merely that the actual work at 
 least was finished in 439, though the dedication may have been 
 postponed to the Panathenaic festival of 438 B.c.i In this same 
 
 1 In the dilapidated condition of the Geneva Papyrus (Nicole, Le proces de 
 Phidias), little reliance can be placed on Nicole's and Pareti's rival readings 
 of what are supposed to be archons' names, MoMwx'Sou = ['A]i't[i]oxWou, and 
 [Eifiu] (/ii)^i/ous = ['A]i^e[i)]Sous. Prom these Nicole attempts to show that the 
 statue was not quite finished even in 438/7. Pareti is even more revolutionary 
 and would place all the work on the statue between 439 and 434, but his evi- 
 dence (I.e. p. 278) is really adverse to his theory ; the sale of wood in the ninth 
 year is hardly the surplus left from the frame of the statue, which would have 
 been a petty amount, but rather that left from the ceiling and roof constructed 
 in the seventh and eighth years ; the sale of surplus gold and ivory does not 
 begin in 434/3, for gold was first sold in the ninth year (439/8), ivory perhaps 
 in the tenth and certainly in the eleventh (437/6), when according to Pareti's 
 view the epistatae should really have bought gold and ivory. The actual expense 
 accounts of the statue (I.G. I, 298 and suppl. p. 146, I, 299, 299 a, and 556), 
 which might be expected to settle the matter, unfortunately cannot be dated ; 
 they are probably of the years between 443/2 and 439/8 b.o. They include 
 receipts from the treasurers of Athena, who in 443/2 ceased to contribute to the 
 Parthenon. 
 
b 
 
 W < O 
 
 o 
 
 H CL 1- 
 
 — UJ LU 
 
 W h- — 
 
 UJ w < 
 
 < 
 I- a. 
 
 & UJ 
 
 a H 
 
 < 
 
 UJ 
 
 >- 
 
 "'an 
 
 ui 
 w 
 
 O 
 
 C 
 
 i_ 
 o 
 
 UJ 
 
 O 
 
 u O 
 
 b ^ 
 
 •6- — 
 
 < >- 
 _\^ 
 
 w w 
 
 UI UJ 
 
 I I 
 
 !^ Ul 
 
 u < 
 
 8 - 
 S w 
 
 b b 
 
 Q. Q. 
 
 X X 
 
 Ul 
 
 H 
 
 UJ 
 
 X 
 
 w 
 
 w 
 I — I 1—1 
 <• E. 
 
 3. 
 
 3. 
 
 a. 
 
 I- o- 
 I — I 
 
 < I 
 
 8 
 
 4. 
 S. 
 
 ^ * 
 
 < 
 QL 
 
 < 
 
 w \ — 
 
 
 -« 03. 
 
 b w 
 
 I- 8 
 
 8 X 
 
 b «3 
 
 J= a 
 
 a * 
 
 <! 
 
 Ul up 
 a 
 
 o 
 
 a. 
 a 
 
 w 
 o 
 
 b 
 
 o 
 t: 
 
 o * 
 
 a 
 
 a. 
 
 a # 
 ?- 
 
 a 
 
 o 
 
 z 
 
 o 
 
 2 
 
 Ul 
 X 
 H 
 
 <: 
 
 bJ 
 
 H 
 
 b. 
 O 
 
 Z 
 
 D 
 O 
 
 o 
 o 
 < 
 
 o 
 
 2 
 
 5 
 5 
 
 CO 
 
72 WILLIAM B. DINSMOOE 
 
 year appears on B, and because if K were placed in the 
 thirteenth year at the end of column IV it would appear, to 
 avoid colliding with fragment I, too Jiigh above the bottom 
 of the stele. 
 
 L. 95 — these doors, evidently of the naos, Cavaignae supposes to 
 have been those of a workshop or temporary gates of the Acropolis. 
 
 L. 96 — HIEP[as oiKt'as], Michaelis, on the analogy of the wrongly 
 identified fragment I.G. I, 312-313. 
 
 L. 97 — [eTri]§KEYE?, Bannier. 
 
 L. 98 — the salaries of the epistatae and their secretary, cf. 
 col. V, 1. 38. 
 
 The record of the eleventh year, at the top of column IV, 
 can be almost completely restored from C + + P (/.(?. I, 
 297 b, Cavaignae, Fig. 26, B.S.A. 1909-1910, p. 187), as joined 
 by Woodward and Cavaignae, and identified by the name of 
 the first secretary of the senate, Peithiades. 
 
 L. 1— ANI[«\?s 'OS6ev?'\ EAPAM, Woodward {I.e. p. 188). 
 
 L. 3 — Ififidrevev ctti Ewflu/xe'v] OS, Woodward. 
 
 L. 9 — restored according to the number of missing letters, this is 
 the first recorded sale of ivory. 
 
 L. 13 — a sale of the wooden wheels used for transporting marble; 
 the last two words read by Cavaignae. 
 
 L. 15 — [Ka]TTITEPA, Cavaignae; but the final is plainly to be 
 seen in his Fig. 26 ; this is probably the surplus of what was used 
 for the trappings and decorations of the Panathenaic frieze. 
 
 The beginning of the twelfth year appears on F (/. Gr. I, 
 300 a), which joins C The fragment Q (B.S.A. 1909-1910, 
 p. 190) was correctly dated by Woodward (I.e. pp. 194-196) 
 as of the twelfth year with the beginning of the thirteenth, 
 though on the basis of a hypothesis which does not exclude 
 the eighth and ninth years or the ninth and tenth ; I therefore 
 note that my location of G and B confines Q to the twelfth 
 and thirteenth years. The numbering of the lines on F is 
 made certain by the junction to C ; the fragment Q is located 
 by allowing to the twelfth year the average thirty-three lines. 
 
 L. 34 — the unusually long space for the name of the first secre- 
 tary of the senate is confirmed by a similar long gap in the Propylaea 
 record of this year. 
 
> 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 — 1 
 
 
 
 1 — 1 
 
 
 
 
 w 
 
 
 
 ' 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 
 1-H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 < 
 
 
 
 
 
 
 
 
 
 w 
 
 1: 
 
 w 
 
 
 
 
 O 
 
 < 
 < 
 
 z 
 
 Ul 
 
 
 
 
 "* 
 
 
 
 
 
 
 
 1 — 1 
 
 1 — r 
 
 
 
 1 
 
 5^ 
 
 
 1 — 1 
 
 8 
 «!. 
 
 ?- 
 
 1 1 
 
 LlI 
 
 
 
 
 
 
 
 
 o 
 
 < 
 
 < 
 
 a. 
 
 UJ 
 
 
 Z 
 
 o 
 < 
 
 >■ 
 
 
 
 
 
 
 UJ 
 
 -i 
 
 Ul 
 
 < 
 1- 
 
 Ul 
 
 UJ 
 
 
 
 
 W 
 
 o 
 O 
 
 
 O 
 
 1- 
 
 
 
 eo 
 
 o 
 
 so 
 
 1 — 1 
 
 W 
 Ul 
 
 141 
 
 w 
 1 1 
 
 >- 
 
 Ul 
 
 H 
 
 < 
 1- 
 o 
 
 1- 
 
 o 
 
 1- 
 
 >- 
 
 
 
 
 
 I — r 
 
 b 
 w 
 
 <* 
 
 d 
 
 9k 
 
 
 1 — 1 
 
 o 
 
 
 
 
 
 w 
 1 — 1 
 
 Ul w 
 
 w p 
 
 a 
 o 
 
 w 
 o 
 
 
 > 
 X 
 
 
 
 W 
 O 
 
 
 
 
 
 o 
 
 
 
 >- 
 
 < 
 
 
 
 c 
 
 Q. 
 
 
 
 
 c 
 
 UJ 
 
 < 
 
 w 
 
 UJ 
 
 o 
 
 1- 
 
 < 
 
 z 
 
 Ul 
 
 
 < 
 
 > 
 
 w 
 
 2 
 
 
 >- 
 
 o 
 
 O 
 
 z 
 
 o 
 
 1 — 1 
 
 < 
 
 a. 
 
 > 
 
 1 — 1 
 
 < 
 
 1 — 1 
 
 
 1 — 1 
 
 
 1 
 
 a 
 
 d 
 
 
 b 
 
 
 1 — 1 
 
 o 
 
 t- 
 
 1 — 1 
 
 
 - 1- 
 a. a 
 ^ a. 
 
 p 
 a 
 
 b 
 
 
 :5 
 
 
 < 
 
 H 
 O 
 a. 
 
 z 
 
 
 W 
 UJ 
 
 
 z 
 
 UJ 
 
 o 
 
 z 
 
 
 O 
 
 1- 
 
 Ul 
 
 w 
 
 Ul 
 
 z 
 
 w 
 O 
 
 z 
 o 
 
 
 UJ 
 
 H 
 
 
 
 
 < 
 
 z 
 
 z 
 
 UJ 
 
 < 
 
 1- 
 w 
 
 
 1 — 1 
 
 a. 
 
 1 — 1 
 
 UJ 
 
 L 
 O 
 
 1 — I 
 1 1 
 
 
 1 
 
 1 — 1 
 
 e 
 
 1 — 1 
 w 
 
 a 
 
 1 — 1 
 
 p 
 
 a 
 
 a. 
 
 o 
 
 Hi 
 
 o 
 
 o 
 
 o 
 
 
 
 ;^ 
 
 d 
 1 1 
 
 1 — 1 
 
 w 
 o 
 
 < 
 
 9k 
 1 1 
 
 
 p 
 o 
 t- 
 o 
 1 — 1 
 
 o 
 ft. 
 
 w 
 
 o 
 
 9k 
 
 «• a. 
 u a 
 t- a. 
 - ?- 
 
 9k 
 w 
 O 
 
 a. 
 o 
 
 
 
 
 < 
 
 c 
 
 W 
 
 
 H 
 
 
 < 
 
 UJ 
 
 
 2 
 
 o 
 
 z 
 
 X 
 
 
 1- 
 
 
 
 
 UJ 
 
 < 
 H 
 w 
 
 UJ 
 
 z 
 
 < 
 
 Ul 
 
 1- 
 
 1 — 1 
 
 I- 
 
 
 p 
 
 b 
 
 )< 
 
 w 
 
 u 
 
 t- 
 
 w 
 
 X 
 
 ;k 
 
 
 
 w 
 
 O 
 
 H 
 
 a. 
 
 H 
 
 
 1 
 
 a 
 
 1 — I 
 
 t= vu 
 
 l- 
 
 i- 
 
 
 
 
 < 
 UJ 
 
 w 
 
 UJ 
 
 < 
 
 O 
 
 
 UJ 
 
 z 
 
 
 w 
 
 ■e- 
 
 2 
 
 H 
 
 w 
 
 o 
 
 1- 
 
 1- 
 1 — 1 
 
 00 
 
 1 — 1 
 
 o 
 
 1 — 1 
 
 o 
 
 1 — 1 
 
 a. 
 
 
 H 
 1 — 1 
 
 a 
 
 
 
 
 O 
 
 UJ 
 
 _V 
 
 z 
 
 < 
 V 
 O 
 
 1 — 1 
 
 w 
 o 
 
 H 
 
 
 Ul 
 
 _i 
 
 w 
 < 
 
 < 
 
 L 
 
 d 
 
 
 d 
 
 •0- 
 
 a. 
 d 
 
 9 
 
 a 
 a. 
 a. 
 
 o 
 o 
 
 a 
 
 a 
 a. 
 
 o 
 
 a 
 
 b 
 •6- 
 
 a. 
 
 
 
 
 w 
 
 
 
 -i 
 _i 
 
 w 
 
 Ul 
 
 1 — 1 
 
 UJ 
 
 h- 
 w 
 
 < 
 
 Ul 
 
 < 
 
 z 
 
 
 < 
 
 Z 
 
 b 
 
 ■«• 
 
 1 — 1 
 
 H 
 
 >- 
 
 w O 
 
 U o 
 
 a 
 
 a 
 a. 
 
 d 
 ft 
 
 z 
 o 
 z 
 
 UJ 
 
 
 
 1 1 
 
 < 
 
 < 
 
 
 Ul 
 
 
 < 
 
 > 
 
 ~~ 
 
 
 1 
 
 
 9k 
 
 
 
 
 
 
 UJ 
 
 c 
 
 ■) 
 
 ^~ 
 
 UJ 
 
 a. 
 
 
 ;k 
 
 
 
 <j 
 
 Ui 
 
 a 
 
 Q. 
 
 a. 
 
 o 
 
 a 
 
 p 
 
 
 
 l- 
 
 o 
 
 
 < 
 
 UJ 
 
 
 UJ 
 
 _\ 
 
 i^ 
 
 H a. 
 
 a. 
 
 iti 
 
 
 
 p 
 
 O 
 
 z 
 
 
 s 
 
 
 _i 
 
 ^ 
 
 
 
 
 
 o 
 
 
 
 
 
 
 »- 
 
 1 — 1 
 
 1 — 1 
 
 -J 
 
 1 — 1 
 
 1 1 
 
 1- 
 
 UJ 
 
 
 b* 
 
 
 o 
 
 t/» 
 
 w 
 
 Q. 
 
 1= 
 
 w 
 
 i- 
 
 < 
 
 ^ 
 
 
 1 
 
 ^ 
 
 1 — I 
 
 t- 
 
 Q. 
 
 w 
 
 
 O 
 
 W 
 
 w 
 
 1 1 
 
 < 1 
 
 w 
 
 (/• 
 
 a; 
 
 
 
 b 
 
 UJ 
 
 O 
 
 
 2 
 o 
 
 W 
 
 UJ 
 
 w 
 
 UI 
 
 u. 
 
 ^ 
 
 tj 
 
 
 
 
 1- 
 
 
 
 
 
 
 z 
 
 UJ 
 
 1 — 1 
 
 z 
 
 
 1 — 1 
 
 z 
 
 < 
 1- 
 
 V 
 
 
 
 l- 
 
 u 
 
 V 
 
 
 
 
 '< 
 
 ^ 
 
 
 
 
 * 
 
 ' 
 
 
 (3 
 
 ;k 
 
 UJ 
 
 < 
 
 
 
 H 
 
 Ul 
 
 UJ 
 
 _— 1 1 
 
 ^ lH 
 
 << * 
 
 - * 
 
 < 
 a. 
 
 
 > 
 z 
 
 3 
 
 t/» 
 
 IIP 
 1 1 
 
 < 
 
 
 z 
 
 X 
 
 Q. 
 
 a. 
 
 Q. 
 
 
 
 
 w 
 
 
 
 
 
 
 C- 
 
 o 
 O 
 
 ( — 1 
 
 < 
 
 UJ 
 
 c_ 
 
 
 o 
 
 
 o 
 
 Q. 
 
 Q. 
 
 (/< 
 
 
 
 ^ 
 
 o 
 
 a. 
 
 a. 
 
 
 w» 
 
 « 
 
 B 
 
 1- 
 
 o 
 
 
 1 1 
 
 o 
 
 1 — 1 
 
 a. 
 
 c_ 
 
 2 ^ 
 
 t/» 
 
 o 
 
 X 
 
 
 
 \jj 
 
 w 
 J 
 
 1 
 
 z 
 
 UJ 
 
 UJ 
 
 o 
 
 a. 
 
 < 
 w 
 
 Ul 
 
 1- 
 < 
 
 u. 
 
 H 
 
 o 
 
 
 
 
 a 
 
 a. 
 
 '7' 
 
 o 
 
 
 '7 
 
 1 — 1 
 
 o 
 
 w 
 
 1 — 1 
 
 o 
 
 w 
 
 1 — 1 
 
 1 — ) 
 
 w 
 
 I — 1 
 
 a 
 
 u 
 
 1 — 1 
 
 
 a 
 
 
 I- 
 
 a 
 
 a 
 
 3« 
 
 
 
 b 
 o 
 
 
 V 
 
 a 
 
 d 
 
 # 
 
 o 
 1 — 1 
 
 1 1 
 
 < 
 < 
 
 1 1 
 
 \ 
 o 
 
 1 — 1 
 < 
 
 W 
 
 UJ 
 
 w 
 
 C 
 
 i_ 
 o 
 
 
 1 — 1 
 
 _l 
 
 H 
 < 
 
 < V 
 
 < 1 
 
 o 
 
 o 
 
 H 
 
 u. 
 o 
 
 D 
 
 o 
 o 
 o 
 
 < 
 
 o 
 
 z 
 
 
 8 
 
 1 — 1 
 
 X 
 
 w 
 
 UJ 
 
 o 
 
 X 
 
 Q. 
 
 1 — 1 
 
 Li 
 1 ■ 
 
 c 
 
 h 
 
 z 
 
 
 I — 1 
 
 b 
 
 o 
 
 b 
 
 o 
 
 
 o 
 
 a. 
 
 
 a 
 
 o 
 
 9 
 O 
 
 
 d 
 
 o 
 a 
 
 d 
 a. 
 
 o 
 
 a. 
 
 o 
 
 a 
 
 o 
 
 o 
 
 d 
 
 o 
 
 
 
 o 
 
 d 
 
 a. 
 
 a. 
 
 
 b 
 
 o 
 
 b 
 
 o 
 
 (- 
 & 
 
 
 
 a 
 
 
 b 
 
 o 
 
 i- 
 b 
 o 
 
 
 o 
 
 1 — 1 
 
 o 
 
 < 
 
 o 
 
 L 
 
 >- 
 
 < 
 
 5^ 
 
 < 
 
 w 
 
 1 I 
 
 UJ 
 
 < 
 
 W 
 
 1 
 \ 
 
 UJ 
 W 
 
 
 a * 
 
 9^ 
 
 o # 
 
 d 
 
 s 
 
 
 < 
 
 _\ 
 
 < 
 
 - 
 
 l- 
 
 b 
 
 b 
 
 ■6- 
 
 ^ 
 
 
 Q. 
 
 < 
 
 w 
 
 
 a. 
 
 a. 
 
 «i 
 
 o 
 
 b 
 
 o 
 
 » 
 
 a 
 
 o 
 
 1 
 
 '< 
 
 a 
 
 - 
 
 b 
 
 b 
 
 i: 
 
 
 
 u 
 
 
 b 
 
 b 
 
 
 o 
 
 I — 1 
 
 d 
 
 1 — I 
 
 o 
 
 <- 
 r— 1 
 
 ( — 1 
 
 1 — 1 
 
 a. 
 
 1 — 1 
 
 I — 1 
 o 
 
 1 — 1 
 
 o 
 
 7 
 
 
 ^ 
 
 a. 
 ■< 
 a * 
 
 OS 
 
 u 
 
 
 w 
 
 o 
 m 
 
 w 
 
 > 
 
 Q. 
 
 o 
 
 ? 
 
 p 
 
 w 
 
 a 
 
 
 w 
 
 s 
 
 3. 
 
 
 ■ 
 
 w 
 
 b 
 
 ^ 
 
 w 
 
 60 
 
 X 
 
 ^ 
 
 ■«> 
 
 ' 
 
 a 
 
 t- 
 
 a. 
 
 p 
 
 p 
 
 u* 
 
 
 
 ■^ 
 
 a. 
 
 p 
 
 p 
 
 * 
 
 ■^ 
 
 <l> 
 
 OO 
 
 l- 
 
 o 
 
 d 
 
 d 
 
 Q. 
 
 b 
 
 p 
 
 b 
 
 P 
 
 !C 1 
 
 
 
 
 o 
 
 c 
 
 u 
 
 Q. 
 
 Q. 
 
 Q. 
 
 -< 
 
 t- 
 
 
 l- 
 
 b 
 
 ** 
 
 
 > a 
 
 ;^ 
 
 ^ 
 
 " 
 
 t- 
 
 o 
 
 d 
 
 •* 
 
 ^ 
 
 1 
 
 ^- 
 
 d 
 
 V 
 
 Q. 
 
 a. 
 
 - 
 
 
 
 <! 
 
 w 
 
 a. 
 
 a. 
 
 
 - 
 
 ■^ 
 
 o 
 
 t= 
 
 
 ?- 
 
 a 
 
 w 
 
 a. 
 
 a. 
 
 t- '" 
 
 o 
 
 ?» 
 
 5 
 
 o 
 
 55 
 
 
 -) 
 
 w 
 
 
 h 
 
 I — 1 
 
 o 
 
 N 
 
 t: 
 
 X 
 
 X 
 
 w 
 
 b 
 
 
 b 
 
 ;>< 
 
 (■ 
 
 
 , 
 
 o 
 
 4 
 
 -< 
 
 Ul 
 
 HS 
 
 (■ 
 
 -< 
 
 ■< 
 
 , 
 
 a 
 
 i« 
 
 b 
 
 X 
 
 X 
 
 o 
 
 
 
 w 
 
 1= 
 
 X 
 
 X 
 
 
 X 
 
 ■< 
 
 rfS 
 
 w 
 
 '< 
 
 d 
 
 s< 
 
 i 
 
 >< 
 
 >< 
 
 9k ^ 
 
 t- 
 
 
 J 
 
 8 
 
 til 
 
 
 1 — 1 
 
 1 — 1 
 
 (LI 
 
 1 — 1 
 
 1 — 1 
 
 1 I 
 
 l-_i 
 
 1 1 
 
 1 1 
 
 lZj 
 
 
 1 1 
 
 1 1 
 
 1 1 
 
 1 1 
 
 1 1 
 
 1 — I 
 
 I — 1 
 
 1 — 1 
 
 1 1 
 
 1 1 
 
 1 1 
 
 1 1 
 
 1 i 
 
 1 — 1 
 
 lIj 
 
 1 1 
 
 1 — 1 
 
 I — 1 
 
 1 1 
 
 << 
 
 
 
 o 
 
 1 1 
 
 1 1 
 
 1 1 
 
 
 L_J 
 
 1 1 
 
 U— 1 
 
 1 — 1 
 
 1 — 1 
 
 1 1 
 
 1 1 
 
 U— 1 
 
 1 — 1 
 
 ' ' 
 
 < 
 
 y 
 
 ;^ 
 
 d 
 1 — 1 
 
 5 
 
 CQ 
 
 >-" 
 O 
 
 o 
 
 
 b 
 w 
 
 
 a. 
 
 
 
 
 
 1 — r 
 UJ 
 
 
 
 
 
 
 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 — I 
 
 Ul 
 
 H 
 
 b 
 
 w 
 
 9k 
 o 
 
 
 
 1 — 1 
 Ul 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 — I 
 UJ 
 
 Ul u 
 
 o o 
 
 o 
 X. 
 
 a. 
 
 
 
 o 
 
 u 
 
 < 
 
 
 
 
 CO 
 
 P4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ;^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Ul 
 
 t- 
 d 
 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 
 
 d 
 
 
 
 g 
 
 
 t- 
 
 ;>. 
 
 b 
 
 o 
 
 
 
 
 Ul 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 l- 
 
 w 
 
 a 
 
 
 
 UJ 
 
 
 
 
 
 
 
 
 
 
 
 UJ 
 
 b 
 
 u 
 
 o 
 
 
 
 < 
 
 
 b 
 
 w 
 
 w 
 
 (- 
 
 
 
 
 X 
 
 
 
 
 
 
 
 
 tl 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 w 
 
 b 
 
 'i 
 
 o 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 u 
 
 l- 
 
 eo 
 
 
 
 u. 
 o 
 
 
 •O 
 
 -<S 
 
 ;^ 
 
 d 
 
 
 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 w 
 
 -• 
 
 o 
 
 X 
 
 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 w 
 
 l- 
 
 d 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 — 1 
 
 1 — 1 
 
 t= 
 
 OSL 
 
 a 
 
 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 w 
 
 d 
 
 ■* 
 
 X 
 
 
 
 a 
 
 o 
 
 
 b 
 
 Ut 
 
 w 
 
 (- 
 
 
 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 <> 
 
 E. 
 
 
 
 
 
 \ — 1 
 
 1 — 1 
 
 
 
 
 
 
 
 
 
 
 1 — 1 
 
 r — 1 
 
 
 
 
 
 
 2 
 
 
 w 
 
 w 
 
 s 
 
 d 
 
 
 
 Of 
 
 1 — 1 
 E. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 <»• 
 
 <u 
 
 w 
 
 U> 
 
 a. 
 
 
 <, E. 
 
 
 
 
 
 
 
 
 
 
 <*« 
 
 E. 
 
 i- 
 
 w 
 
 o 
 
 
 
 o 
 z 
 
 
 Wi 
 
 h 
 
 w 
 
 i 
 
 
 
 <• 
 
 <!• 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 B; <> 
 
 t/> 
 
 m; 
 
 "* 
 
 
 <• 
 
 <• 
 
 
 
 
 
 
 
 
 
 
 <!• 
 
 <• 
 
 b «3 
 
 ^ 
 
 
 
 
 - 
 
 - 
 
 l- 
 
 a. 
 
 
 
 12: 4- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 1 1 
 
 ■* 
 
 b- 
 
 b 
 
 
 §2 
 
 
 
 
 
 
 
 
 
 
 1 1 1 1 
 
 t= 
 
 cj 
 
 
 
 
 < 
 
 
 o 
 
 t= 
 
 d 
 
 V 
 
 
 
 1 1 
 
 L_l 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 4 
 
 p 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 
 (- 
 
 w 
 
 a. 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t- 
 
 V 
 
 < 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 w 
 
 s< 
 
 < 
 
 
 
 u 
 
 s 
 
 < 
 
 
 1 — I 
 
 t 1 
 
 1 1 
 
 l2j 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 1 — 1 
 
 1 — 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 — 1 1 — 1 
 
 1 1 
 
 1 — 1 
 
 
 
 
 
 
 
 lO 
 
 
 
 
 
 O 
 
 
 
 
 
 <A 
 
 
 
 
 9 
 
 
 
 
 
 » 
 
 
 
 
 
 s 
 
 
 
 
 
 !S 
 
 
 
 
 
 
 
 
 
 S 
 
 
 
 
 
 a 
 
 te 
 
 S 
 
 ffi 
 
 S 
 
 
 Q. O 
 
 O 
 
 o* 
 
ATTIC BUILDING ACCOUNTS I 73 
 
 L. 35— [evtauTS to] YTO TA[8£], Poucart and Gavaignac. 
 
 L. 56— [\i0ay]OA[«is], Woodward. 
 
 L. 58 — [t]OL, UIO0[v], Woodward implies that only one stone 
 was transported in this year, and that the plural T05 in the next 
 line is therefore a mistake ; it is better to suppose that tos Xtflos was 
 intended as in other years, and that the mistake lay in cutting U 
 twice. 
 
 L. 62 — omitted by Woodward. 
 
 L. 63 — [totjto TaSe], Woodward. 
 
 The first line of the thirteenth year appears on the same 
 fragment Q, as was recognized by Woodward. Here too must 
 come the reverse of Cavaignae's smaller fragment N, the ob- 
 verse of which belongs in the fourth and fifth years ; it con- 
 tains only two letters, OS, perfectly preserved as when the 
 chisel left them, so that it is difficult to comprehend Cavaignae's 
 inability to find them a second time (Z.e. p. Ix, n. 2), and his 
 representation of them as one above the other. These two 
 letters appear at the level of 1. 96 of column I, and so 0.43 m. 
 above the bottom of the stele ; the last certain line above this 
 is 1. 39, opposite 1. 35 of the obverse ; the i-estoration of the 
 obverse requires sixty-one lines in the gap, equivalent to sixty- 
 seven of the more closely spaced lines of the reverse, so that 
 the two letters OS would have appeared in 1. 106. 
 
 L. 67 — the choice of formulae allowed here by Woodward (I.e. 
 p. 193) is impossible ; we must restore that of the five last years. 
 
 L. 67 — •[• • -JOS, Woodward, but a vertical hasta, as of E, ap- 
 pears ou the stone. 
 
 L. 106 — the length of the line is evidence for the restoration 
 proposed. 
 
 On the analogy of the eleventh and twelfth years, we should 
 restore four more lines after this entry, making 110 lines in all, 
 with the account ending 0.37 m. above the bottom of the stele. 
 
 Evidently the work was now nearly finished, and therefore, 
 though both sides of the stele had been filled, it was not 
 deemed necessary to set up a second stele. The edges, hitherto 
 probably left rough as they still remain in the Propylaea stele, 
 were worked smooth and formed additional space for the 
 accounts. That of the fourteenth year, inscribed on the right 
 
74 WILLIAM B. DiySMOOB 
 
 edge (with reference to the obverse), is now complete, begin- 
 ning on fragments D + E ( J. Gr. I, 301), and continued on A (/. Gr. 
 I, 221) and H (I, 307), which I fitted side by side to the bottom 
 of E. 
 
 10 
 
 15 
 
 20 
 
 25 
 
 , 1 
 
 § 
 
 E 
 
 P 
 
 1 <^ 1 
 
 A T E 5 1 H 1 
 
 § 
 
 ANT 
 
 1 
 
 K 
 
 U 
 
 E 5 E 
 
 APAMMATE 
 
 Y[^] 
 
 _ P 1 
 
 T 
 
 E 
 
 § 
 
 T ET 
 
 A PT E § K A 1 
 
 A E 
 
 K A T 
 
 E 
 
 § 
 
 B 
 
 UE 
 
 5 H E 1 M E T A 
 
 
 A E N 
 
 E 
 
 § 
 
 P 
 
 POT 
 
 OSEA PA MM 
 
 
 ATE 
 
 Y 
 
 E 
 
 E- 
 
 P 1 K 
 
 PATE TO SA 
 
 P X 
 
 N T 
 
 
 
 t 
 
 A 
 
 O EN 
 
 A 1 1 § 1 N 
 
 
 V E M 
 
 M 
 
 A 
 
 T 
 
 A TO 
 
 E N 1 A Y T 
 
 
 TOY 
 
 T 
 
 
 
 T 
 
 A A E 
 
 
 
 X H H 
 
 H 
 
 
 P 
 
 E P 1 
 
 A ENO ME NO 
 
 M 
 
 H p A 
 
 A 
 
 
 M 
 E 
 
 E N E 
 N 1 A 
 
 KTOP POTE 
 Y T 
 
 P C 
 
 n A A 
 
 
 
 ) 
 
 [A 
 
 a /x </> 
 
 OSTATE P E 
 0-] A K E NO 1 
 
 t 
 
 -^ 4 p§ § 
 
 
 Cx 
 
 p V (T 
 
 oJ<- T A T E P E 
 
 t 
 
 H E KTE 
 
 
 K 
 
 [v t ' 
 
 K e V o] ' 
 
 
 
 
 
 P 
 
 A P A 
 
 [t a ;m tj N [7i o 
 
 6 T a] 
 
 M M p: 
 
 
 
 T 
 
 E § O 
 
 E 0[e r] A M 1 E 
 
 Y[o .] 
 
 
 
 
 H 
 
 1 t 
 
 KPATE5EA 
 
 P[a] 
 
 
 
 
 M 
 
 MAT 
 
 EYEUAMPT 
 
 P [. V .J 
 
 X H H 
 
 H 
 
 
 X 
 
 P Y § 
 
 1 OP P A O E[v 
 
 T s] 
 
 D AA hh 
 
 
 § 
 
 TAG 
 
 M0N[3 AAA [A 
 
 P H h hj 
 
 
 
 
 T 
 
 1 M E 
 
 T Y T 
 
 
 X H H 
 
 H 
 
 
 T 
 
 t E * 
 
 A N T t[n p a]0[€ v] 
 
 n 1 1 1 1 
 
 
 
 T 
 
 § 5 
 
 A:T 1 
 
 T A O M [v] T T 
 
 M E T [o w r o] 
 
 (?) 
 
 
 f' 
 
 I 
 
 N 
 
 A U C 
 
 1 M A T A 
 
 
 -- H 
 
 -- -L 
 
 H 
 h 1 
 
 
 
 
 N E M 
 
 A T -" [v] 
 
 
 30 - - MI§O0/^[aTOv] 
 
 X, [7i]YP0P[yot sHevreXej 
 
ATTIC BUILDING ACCOUNTS I 75 
 
 H H Z [A] [a- I K a I \ I 6 OS u. V ari^] 
 Prll L[o-teirira KVKXaJ 
 
 35 M p3 X H H ^A" A[aA/xaTO 7ro]IO IS 
 
 HpAAA ENACtc-rto../. 0^005 
 A I- h 
 
 [X] [H H H H K A T A IV [e V I o] I S (?) 
 
 [tt € p]l E[y e v] E T 
 
 40 [t o e V I tt V t] '^ T V [t o] 
 
 D? ^ ^] [XP iiTOirru Tip e s] 
 
 LA a fji <I> (t a k € V o ij 
 
 C^ '^ P § §] [XP vo-ocrra T € p e s] 
 
 [H E K T E] [K V ^ t K £ 1/ u i] 
 
 L. 17 — the treasurers of Athena, having devoted their attention 
 exclusively to the statue of Athena between 443/2 and 439/8 b.c, 
 and to the Propylaea in 437/6 to 436/4, now return to the Parthenon. 
 
 L. 22 — pAAA (?), Kirchhoff; I complete the weight to fit the 
 ratio 14 : 1, as described below. 
 
 L. 25 — the weight of the ivory was more than 2 talents 60 dr., 
 but less than 4 talents 5960 dr. ; the value, in proportion to silver 
 or gold, was at most less than half of the quotations for the best 
 ivory in recent times, though this is offset by the modern decrease 
 in the actual value of gold and silver. 
 
 LI. 35-36 — the formula was suggested by Woodward {I.e. p. 192, 
 n. 4) in connection with I. O. I, 307, and is now confirmed by the 
 junction to I.G. I, 221, which adds the interesting fact that in 
 434/3, the sixth year of work on the pediments, the sculptors were 
 paid 16,392 dr. 
 
 L. 38 — no word indicating the class of people who received this 
 payment, such as the use of the dative case would require, can fit 
 this space ; the only possible word is KaraaeviOK, which should prop- 
 erly be in the genitive case ; this is the usual place, just after the 
 payment to the sculptors, for the entry Karap-evlov. These are the 
 monthly salaries of the architect, the epistatae, and their secretary ; 
 the amount, 1800 dr., is correct for six persons paid 30 dr. a month, 
 i.e.. the architect, the secretary, and four epistatae (whose number 
 varied, however, from three to five). 
 
 L. 39 — the amount of the surplus is unknown; the receipts 
 amounted to 29,147 dr. 4 ob., of which at least 20,620 dr. 3 ob. were 
 expended. 
 
76 WILLIAM B. BINSMOOn 
 
 Herodotus (III, 95) says the ratio of values of equal weights 
 of gold and silver was as 13:1; but from the slight traces on 
 our inscription Kirchhoff conjectured that the ratio rose to 
 rather more than 14 : 1 (Mbnatsb. Berl. Akad. 1861, 860 f.). On 
 a fragment of the accounts for the chryselephantine statue (Z Gr. 
 I, 298 suppl. p. 146) one may read, according to the usual in- 
 terpretation, that 6 talents, 1528 (or -9) or 1618 (or -9). 
 drachmae, and 5 obols weight of gold cost 87 talents 4652 dr. 
 of silver. It seems better to read the lllll with the marginal 
 entry, as 87 talents 4652 dr. 5 ob., a fourteenth of which would 
 be 6 talents 1618 dr. -i- ob. (PTXIR[H]APhH-[C]), exactly 
 fitting the indications on the stone and giving the ratio as 
 exactly 14 : 1. 
 
 By the end of 433/2 B.C. the work had been completely 
 finished.^ As appears from the junction of F (J. G-. I, 301 a) 
 to C, the account of this fifteenth year was inscribed, not on 
 the narrow edge below the fourteenth year, as has been sup- 
 posed, but on the opposite narrow edge, so that all four faces 
 of the stele were used. The position of the account is remark- 
 able; instead of beginning at the top of the stele, as on the 
 three other faces, it begins 0.57 m. below the top. 
 
 [tJOISEP I5TATE § I H 
 [O^ANTI KL.E5EA PAM 
 MATEYE:Eni TE5rEM[,r] 
 
 ~E5KAIAEKATE5B0l 
 5 E5HEIKPITIAAE5n PO 
 TOSEA PAMMATEYE 
 EP IA*§EYA0§APX0 h 
 [r]0§AGENAIOI5:L.EMN 
 [..rjATOENIAYTOTAAE 
 10 [TrejPIAENOMENO 
 
 [/x,jLiey]-KTOnPOT 
 [ep oev I a u to] 
 
 The dimensions of the stele can be calculated with a fair 
 degree of accuracy. The thickness, as already noted, was uni- 
 
 1 Michaelis attempted to assign to 432/1 e.g., the fragment of the third year 
 /. G. I, 302 a. 
 
ATTIC BUILDING ACCOUNTS I 77 
 
 formly 0.198 m., except at the very bottom, where it was 
 increased to 0.21 m. by a werkzoll (as on the Propylaea stele) 
 covering the reverse to a height of 0.20 m.^ As for the width, 
 we may estimate from J that the money column was at least 
 0.21 m. wide, the item column at, least 0.31 m., both together 
 amounting to a width of more than 0.52 m. The length of 
 the prescript of the eleventh year (0 + P), i-e. the combined 
 widths of the money and item columns, was 0.58 m.^ The- 
 total width of the two double columns would be somewhat 
 more than twice this amount, or about 1.20 m., a result similar 
 to that which we shall obtain for the Propylaea stele. The 
 height can be estimated only by the spacing of the lines. On 
 the obverse, the first twenty-nine lines occupy a space of 0.51 m.,. 
 after which the lines are spaced regularly 0.0165 m. In the 
 first column we require one hundred and sixteen lines, measur- 
 ing 0.51 m. (11. l-29)-f-1.43 m. (11. 30-116) = 1.94 m.; the 
 last line is 0.088 m. above the bottom of the stele, which would 
 have been therefore about 2.04 m. high. Each broad face con- 
 tained two double columns, each narrow edge one. In column 
 I, including the first annual prescript which extended entirely 
 across the stele, were one hundred and sixteen lines, ending 
 0.088 m. above the bottom ; in column II, one hundred and 
 nine lines ending 0.195 m. above the bottom; on the reverse,, 
 the werkzoll covered about 0.20 m. at the bottom, and the 
 inscriptions ended about 0.16 m. higher; on the right edge, 
 the fourteenth year's account began at the top and covered 
 forty-four lines ; and on the left edge the fifteenth year's 
 account began 0.57 m. below the top, and was of unknown 
 length. 
 
 In conclusion, the historical facts gleaned from this inscrip- 
 tion may be summarized, with the assistance of other chrono- 
 logical sources, somewhat as follows : 
 
 Year I, 447/6 B.C.; Timarchides archon, of Aphidna 
 
 undersecretary of the epistatae, Diodorus of Paeonidae secre- 
 
 1 This appears on Mi but is badly footworn, having evidently formed the- 
 surface of a mediaeval pavement. 
 
 2 The letters are spaced 1^ cm. on centres ; the first letter of P is 0.36 m. 
 from the right edge of the stele, and sixteen more spacings are needed to com- 
 plete the restoration. 
 
78 WILLIAM B. DINSMOOR 
 
 tary of the Hellenotamiae. The Parthenon begun with Ictinus 
 and Callicrates as architects. 
 
 Yearll, 446/5 B.C.; Callimachus archon. Work temporarily 
 suspended on account of the campaign in Euboea. 
 
 Year III, 445/4 B.C. ; Lysimachides archon, Eu secre- 
 tary of the Hellenotamiae. Work on the Parthenon resumed. 
 
 Year IV, 444/3 B.C. ; Praxiteles archon, Strombichus of 
 ChoUeidae secretary of the Hellenotamiae. The earliest extant 
 notice of the contribution of the airapxe (yi^^ri. 42,675 dr. 5 ob.); 
 Pericles falsely accused of the misuse of the Delian funds, by 
 Thucydides ; the charge dismissed and Thucydides ostracized. 
 The new fleet of triremes finished and the surplus money 
 (90,000 dr.) turned over to the Parthenon. 
 
 Wood is purchased, probably for scaffolding. 
 
 Year V, 443/2 B.C. ; Lysanias archon, Timotheus annual 
 secretary of the epistatae, Anticles appointed permanent under- 
 secretary ; Sophiades of Eleusis secretary of the Hellenotamiae. 
 The middle long wall finished by Callicrates, and the surplus 
 funds devoted to the Parthenon. The treasurers of Athena 
 cease to contribute to the Parthenon, apparently because their 
 money was needed to begin the chryselephantine statue. 
 
 Year VI, 442/1 B.C. ; Diphilus archon, Anticles undersecre- 
 tary of the epistatae, Chalcideus of Melite secretary of the 
 Hellenotamiae. 
 
 Year VII, 441/0 B.C. ; Timocles archon, Anticles under- 
 secretary of the epistatae. The columns channeled, and work 
 begun on the ceilings and roof. 
 
 Year VIII, 440/39 B.C. ; Morychides archon, Anticles under- 
 secretary of the epistatae, Sosistratus of Hybadae secretary of 
 the Hellenotamiae. Work on the ceilings and roof completed ; 
 ivory bought, and gold and silver workers paid, for decorating 
 the pedestal of the chryselephantine statue ; marble brought to 
 the ateliers, probably to be in readiness for beginning the pedi- 
 ment sculptures. 
 
 Year IX, 439/8 B.C. ; Glaucinus archon, Dionysiphilus (?) 
 of Probalinthus secretary of the epistatae, Anticles under- 
 secretary, of Rhamnus (or Hagnus) secretary of the 
 
 Hellenotamiae. Contributions from private individuals, 
 hitherto given for the chryselephantine statue, now given to 
 
ATTIC BUILDING ACCOUNTS I 79 
 
 the Parthenon. The statue completed by Phidias; its sur- 
 plus material turned over to the Parthenon by the epistatae 
 of the statue, and the gold and silver begin to be sold, perhaps 
 to assist in paying for the carving of the pediment sculptures 
 (which were probably likewise the work of Phidias and of his 
 assistants). The Parthenon virtually completed, and surplus 
 wood from the scaffoldings and roof is sold. The pediment 
 sculptures begun and no other work done in this year. 
 
 Year X, 438/7 B.C. ; Theodorus archon, Anticles under- 
 secretary, of the epistatae, of Acharnae secretary of the 
 
 Hellenotamiae. The doors of the naos set in place, completing 
 the work on the temple ; the statue of Athena Parthenos dedi- 
 cated at the Panathenaic festival. 
 
 Year XI, 437/6 B.C. ; Euthymenes archon, Peithiades first 
 secretary of the senate ; Anticles acting secretary of the epis- 
 tatae, cus of the Ceramicus secretary of the Hellenotamiae. 
 
 The Propylaea begun, whereupon the Hellenotamiae cease to 
 contribute to the Parthenon, and the treasurers of Athena, 
 instead of resuming their contributions for the construction of 
 the Parthenon (now that the statue has been completed), like- 
 wise divert their funds to the Propylaea. Surplus ivory 
 and tin sold, and work henceforth confined to the pediment 
 sculptures. 
 
 Year XII, 436/5 B.C. ; Lysimachus archon, Anticles ap- 
 pointed permanent secretary of the epistatae, Philetaerus (or 
 Philemonides) of Aexone secretary of the Hellenotamiae. 
 
 Year XIII, 435/4 b.c. ; Antiochides archon, as first 
 
 secretary of the senate, Anticles secretary of the epistatae, 
 Thoinilus (of Acharnae) of the Hellenotamiae. 
 
 Year XIV, 434/3 B.C.; Crates archon, Metagenes first sec- 
 retary of the senate, Anticles secretary of the epistatae. Crates 
 (of Lamptrae) of the treasurers of Athena, Protonicus (of the 
 Ceramicus) of the Hellenotamiae. As the Propylaea approach 
 completion, the treasurers of Athena give part of their funds 
 to the Parthenon sculptures. Of the total receipts 29,147 dr. 
 4 ob. in this year, 16,392 dr. given as wages to the sculptors, 
 and 1800 dr. as salaries to the epistatae and their secretary. 
 
 Year XV, 433/2 B.C.; Apseudes archon, Critiades first secre- 
 tary of the senate, Anticles secretary of the epistatae, Euthias 
 
80 WILLIAM B. DINSMOOB 
 
 (of Anaphlystus) of the treasurers of Athena. The Parthenon: 
 sculptures completed and the accounts closed ; Phidias accused 
 of embezzlement and impiety, imprisoned, and later executed ; 
 the outbreak of the Peloponnesian War at the same time 
 terminates work on the Propylaea and the Erechtheum. 
 
 William B. Dinsmooe. 
 
 Amekican School of Classical 
 Stddies, Athens. 
 
 COEREOTION 
 
 In the last line of the note on page 542 of the last volume of this 
 Journal (XVI, 1912, p. 542), for 30.95 m. read 26.40 m. 
 
80 WILLIAM B. DINSMOOB 
 
 (of Anaphlystus) of the treasurers of Athena. The Parthenon: 
 sculptures completed and the accounts closed ; Phidias accused 
 of embezzlement and impiety, imprisoned, and later executed ; 
 the outbreak of the Peloponnesian War at the same time 
 terminates work on the Propylaea and the Erechtheum. 
 
 William B. Dinsmoor. 
 
 American School of Classical 
 Stddies, Athens. 
 
 CORRECTIONS TO PP. 53-80. 
 
 A new examination of the accounts of the Parthenon necessitates 
 one change in the order of the fragments; the letters on the broad 
 face of H are too large for the reverse, bfit are identical with those on 
 the obverse; I was misled by the present mutilation of the fragment. 
 The following corrections are chiefly the results of this change. 
 
 Transfer the beginnings of 11. 38 41 of Plate III to 11. 68-71. of Plate 
 II, reading tEtdQxei; in 1. 68, EAP[an(4,dteue] and XSYf^tsxaov] in 11. 70-71. 
 Figure 1 (p. 54): transpose H, and read «VII» instead of «VIII» (col.ll). 
 P. 57, 11. 10-12: read .six (A, E, H, I, J, and M) and seven (B, C, K» 
 
 1. 9 from bottom: read «four' instead of «five» 
 
 1. 6 from bottom: omit «and H» 
 
 1. 3 from bottom : read «groups, one of the three and one of four* 
 P. 59, 11. 9-10 from bottom; read «H (obveroe)* 
 P. 60, n. 3: omit .G+H and« 
 P. 61, last line: read «N» instead of «/v» 
 P. 64, 1. 2: insert «To these must be added H (L G. I, 306); it contains 
 
 the beginnings of a prescript, with letters characteristic of the 
 
 obverse, and belonging to the left edge of the obverse; other 
 
 fragments preserve the prescripts of years III and V, leaving 
 
 IV for H.» 
 
 1. 4: insert .L. 69 — ppo[iocI etc.» as on p. 69, II. 9 10. 
 P. b7, last line: omit «as determined — — ninth year.» 
 P. 68: omit II. 13 16 from bottom. 
 P. 69, 11. 1-3: omit ^The connection ninth year.» 
 
 I. 4: omit -H (L O. I, 306). 
 
 II. 9-15: omit comment on 11. 39-41. 
 
 P. 71, n. 1: read -It actually lasted seven years, however. » 
 
 P. 74, 1. 3: omit «and H (1, 307)« and «side by side» 
 
 P. 75, 11. 3 8 (35-40): end the lines with minuscule letters, 1. 38 as [eviov] 
 
 11. 1617 from bottom: omit «the formula fact that» 
 
 1. 9-12 from bottom: omit "no word — — genitive case* 
 P. 76, 1. 21: insert «The location of the obverse of H (col. I, 11. 68-71) is 
 
 such as to bring the piece of narrow edge in 11. 35-40 of year XV.» 
 
 1. 2 from bottom: insert, as 11. 35-40, the ends of 11. 35-40 printed 
 
 on p. 76, 1. 38 reading [hi):toQYo]lS 
 P. 77, 1. 7 from bottom: read «and was likewise forty four lines long.» 
 P. 78, 1. 5: insert « — of Halae secretary of the epistatae,» 
 
 1. 7: insert « — of Probalinthus secretary of the epistatae,» 
 
 1. 4 from bottom: omit «of Probalinthus». 
 

 arcl^aeological 9!n0t(tttte of America 
 
 EXTRACT FROM THE JOURNAL OF ARCSAE.OILOOY 
 
 roL. xrii, NO. 2 
 
 AMERICAN SCHOOL OF CLASSICAL STUDIES AT ATHENS 
 
 ATTIC BUILDING ACCOimTS 
 
 11. THE ERECHTHEUM 
 
 BY 
 
 WILLIAM BELL DINSMOOR 
 
American Srijool 
 of Classical StuUics 
 at atijcns 
 
 ATTIC BUILDING ACCOUNTS 
 
 II. The Eeechtheum^ 
 
 Op the records of the first period of the construction of the 
 Erechtheum, including, as we may imagine, a decree of about 
 435 B.C. authorizing the erection of the building, and a stele 
 with the expense accounts down to the outbreak of the Pelo- 
 ponnesian War in 432, not a fragment has been identified. Of 
 the later series, however, beginning with (A) a fragment of 
 the decree of Epigenes in one of the last prytanies of 410/9 
 (/. G-. I, 60, and suppl. p. 18), which provides for a commission 
 to examine the state of the unfinished building, we have a fairly 
 complete array. To the stele containing the report of this 
 commission, and dating from the first prytany of 409/8 B.C., I 
 assign five fragments. 
 
 B. The upper portion of a marble stele (J. 6r. I, 322), found 
 on the Acropolis by Chandler in 1765, and because of its con- 
 tents immediately identified as referring to the Erechtheum, 
 was taken to England for the Society of Dilettanti and by them 
 presented to the British Museum. Early writers all supposed 
 that the stele was complete; it remained for Boeckh {C.I.Cr. 
 1 60) to note from the context that more lines were needed, and 
 that therefore the lower end of the stone must be missing. 
 
 C. A fragment of the left edge of a stele (J. Gr. I, 322 S), 
 found on the Acropolis in 1838 and now in the Museum at 
 Athens, was identified by Pittakis ('E^. 'Apx- 1839, no. 215), 
 Rangabe (Ant. hell. I, p. 86), and Stephani QAnnali, 1843, 
 p. 286), as forming part of the missing continuation of the 
 first column of the stele. 
 
 1 1 repeat my acknowledgments to Dr. KeramopouUos, of the Epigraphical 
 
 Museum ; and to Dr. Caskey, whose studies of the Erechtheum inscriptions have 
 
 made me more intimately acquainted with the fragments and their contents, I 
 
 am under many ohligations. 
 
 American Journal of Archaeology, Second Series. Journal of the 949 
 
 Archaeological Institute of America, Vol, XVII (1918), No. 2. ^^ 
 
ATTIC BUILDING ACCOUNTS II 243 
 
 D. A fragment of the right edge ofj a stele (J. Gf-. I, 322, 
 suppl. p. 152), bought by the Greek Archaeological Society 
 and said to have been found in Athens (before 1889), was 
 suggested by Kirchhoff as part of the second column of our 
 Erechtheum stele, but this suggestion has not always been 
 received with favor by recent writers. The contents, style of 
 writing and size of letters, and the distance of the left side of 
 the column (as restored) from the preserved right edge of the 
 stone, are exactly the same as in the Chandler stone. The fact 
 that B, C, and D all belong to one and the same stele seems to 
 me beyond doubt. But there is one strong objection. The 
 Chandler stone (B) is 0.09 m. thick and roughly tooled on the 
 back, without tapering; C is a mere sliver 0.05 m. thick, and 
 so does not enter into the question ; but D, even though broken 
 at the back, is at least 0.135 m. thick, far thicker than the 
 Chandler stone. Either therefore B and D do not belong 
 together, or the back of B is modern. 
 
 Now what do we learn from Chandler's own account ? In his 
 Travels in Greece (1776, pp. 57-58 = 1821, pp. 71-72), we read 
 as follows : " Another marble, which has been engraved at the 
 expense of the Society of Dilettanti,^ was discovered at a house 
 not far from the temple of Minerva Polias, placed, with the 
 inscribed face exposed, in the stairs. The owner, who was 
 branded for some unfair dealing with the appellative -Jefiit' 
 or ' the Jew,' prefixed to his name, seeing me bestow so much 
 labour in taking a copy, became fearful of parting with the 
 original under its value. When the bargain was at length 
 concluded, we obtained the connivance of the disdar, his brother, 
 under an injunction of privacy, as otherwise the removal of the 
 stone might endanger his head, it being the property of the 
 grand signior. Mustapha delivered a ring, which he commonly 
 wore, to be shewn to a black female slave, who was left in the 
 house alone, as a token ; and our Swiss, with assistants and 
 two horses, one reputed to be the strongest in Athens, arrived 
 at the hour appointed, and brought down the two marbles,^ for 
 which he was sent, unobserved ; the Turks being at their devo- 
 
 1 The engraving is published by Chandler, Inscr. gr. 1774, II, I ; Wilkins, in 
 Walpole, Memoirs, p. 580 ; Kose, Inscr. gr. pi. XXII. 
 
 2 One was a fragment of a treasure list. 
 
244 WILLIAM BELL DINSMOOB 
 
 tions in the mosque, except the guard at the gate, who was in 
 the secret. The large slab was afterwards rendered more portable 
 by a mason." ^ When we remember that another slab of the 
 Erechtheum accounts (J. Gr. I, 321) has on its back an inscrip- 
 tion (^A.J.A. 1906, pi. Ill) which was not discovered until forty 
 years after the inscription on the front had been published, we 
 may well hesitate to imagine what may have been on the back of 
 the Chandler stone, buried under the crust of lime mortar which 
 probably covered it at the time of the extraction from the stair 
 of the house of Jefiit Mustapha. Yet it is to the problem of 
 what was lost by the mason's endeavor to render the stone 
 " more portable " that we must now turn. 
 
 E. A fragment of the right edge of a stele (/. G. I, 282), 
 found on the Acropolis in 1839 and now in the Museum at 
 Athens, was first recognized bj"- Rangabe (^Ant. hell. I, no. 88) 
 and Jahn (^Arx Athenarum\ 1880, no. 16), as part of the speci- 
 fications for work to be done on a wooden ceiling of the 
 Erechtheum. The stone is broken away 0.025 m. below the 
 last line, leaving, however, a vacant space which is enough to 
 show that this was the bottom of a column of text. For various 
 reasons, now unnecessary to relate, it appeared to me probable 
 that this was to be associated with the three other pieces ; the 
 matter was finally settled by the discovery that the fracture at 
 the back exactly fits the broken ba^ck of C- The two pieces 
 thus placed together give the original thickness of the stele as 
 0.139 m., sufficient to include D and show that on the back of 
 the report were inscribed the specifications for new work, most 
 of which were chipped off by the mason and are now hopelessly 
 lost. 
 
 F. A fragment of a lower corner of an opisthographic stele 
 (unpublished), now in the Museum at Athens, was there 
 recognized as forming part of the inscription to which E 
 belonged, i.e. the specifications. ^ The other face has, however, 
 a few letters from the last two lines of an inscription exactly 
 like that on the Chandler stone — 
 
 N I A K / 
 
 OSTPinOAE (vacat) 
 
 1 The italics are mine. 
 
 2 It is now marked " Eis I, 282," and will be published by Mr. L. D. Caskey. 
 
ATTIC BUILDING ACCOUNTS II 
 
 245 
 
 and the thickness 0.139 m. is the same as that reached by 
 fitting C and E together. The broken surface at the bottom 
 of D, moreover, exactly fits the top of F, and enables us to 
 terminate the Chandler inscription thus : 
 
 [II 'EXei, n-t]NIAKO,iJ;; KOSHEKPOAE 
 
 [irXaT]0§TPinOAE 
 
 Our five pieces of the stele now form three groups, B (itself 
 broken into two pieces), C + E, and D + F, which cannot be 
 joined together. We may, how 
 
 '"■•'' * «^'?»V'-"''*^^'^^^*"*. 
 
 '; .*'''.**•*' 
 
 ever, form an idea of the gen- 
 eral contents of the stele and 
 of the gaps that separate the 
 fragments (Fig. 1). , 
 
 The report proper on the 
 obverse, below the prescript 
 which extended across the en- 
 tire width (11. 1-7), falls into 
 two general divisions, an in- 
 ventory of the building and an 
 inventory of what lay on the 
 ground about the building. 
 The first division is subdivided 
 into two sections, containing a 
 list of the blocks missing from 
 the building (col. I, 11. 8-43),i 
 and a list of those already 
 placed but unfinished (11. 44- 
 92). The second division is 
 subdivided into three sections, containing lists of stones 
 then on the ground (;)^a/u.at) ; first are those completely 
 worked (vrai'TeXo? i)(^crep<yaafieva), second those partly worked 
 (he/jLiepya), and third those which were still unworked. The 
 list of stones completely worked begins in B, col. I, 1. 93, and 
 extends to C, 1. 2 ; we cannot be certain as to how much is 
 missing in the gap, but I should estimate that the loss was 
 
 1 For this portion of the text see A.J.A. 1908, pp. 184-185. 
 
 Qbvers'^ 
 
 Heverse 
 
 FiGDRE 1. — The Chandler Stele. 
 
246 WILLIAM BELL DINSMOOR 
 
 about 30 lines,^ so that we may tentatively number the 
 14 lines of fragment C as col. I, 11. 131-144. The list of 
 he/j,iepya would then begin with col. I, 1. 133, and continue 
 through at least 93 lines of column II of fragment B (11. 8-100) ; 
 how much is lost at the bottom of column I is uncertain, but 
 can be estimated as at least 9 lines,^ giving the total length of 
 column I as 153 lines at a minimum. In the list, part of which 
 is preserved on fragments D + F, the descriptions are so sum- 
 mary, merely a series of dimensions, that we are evidently con- 
 cerned with rough unfinished blocks. The end of this column 
 II is 0.256 m. above the bottom of the stele; column I may 
 well have been longer, for on the reverse the lowest letters of 
 the first column are 0.152 m. above the bottom of the stele. If 
 we suppose that the end of column I of the obverse (1. 153) 
 was likewise about 0.152 m. above the bottom, then column II 
 would have been 9 lines shorter,^ including 144 lines in all. 
 The last 27 lines (11. 118-144) of column II appear on frag- 
 ments D -f F, so that the gap between B and D would have con- 
 tained 17 lines (11. 101-117), with the conclusion of the list of 
 hefiiepya and the title and beginning of the list of unworked 
 blocks, an allowance that is by no means impossible. The total 
 
 1 The estimate includes the following items : 
 
 101-103 two or three epikranitides (of. B I, 16-18 ; C, 4-11 ; D, 3-5). 
 104-106 one capital of metopon (or in list of ke/depya ; cf. B I, 29-32). 
 107-109 perhaps one antimoros of epikranitis or epistyle (cf. B I, 22-28 ; D, 
 
 10-12) . 
 110-112 five or less epistylia (cf. B I, 33-35). 
 113-115 some eight-foot frieze blocks (cf. B II, 8-10; LG. I, 821, 1-2 as 
 
 revised below, and 29-30). 
 116-117 some six-foot frieze blocks (cf. B II, 11-24 ; LG. I, 321, 3-5, 21-22). 
 118-119 at least one two-foot frieze block (cf . /. G. I, 321, 5-6). 
 120-121 several four-foot flank geisa (cf. B II, 25-48). 
 122-124 seven four-foot fagade geisa (cf. B II, 49-52 ; /. G. I, 3212, 20-22). 
 125-126 one seven-foot fagade geison (cf. LG. I, 321^, 23-24, where the length 
 
 has dropped out) . 
 127-130 some raking geisa (cf. B II, 80-86) . 
 
 2 The estimate includes the following : 
 
 145-147 perhaps one antimoros of epistyle or epikranitis (cf. 11. 107-109 above). 
 
 148-150 one or more epistylia (of. 11. 110-112). 
 
 151-153 some eight-foot frieze blocks (continued in col. II, 8-10). 
 
 3 The lowest lines on the obverse are spaced about ll|mm. ; 0.256 — 0.152 = 
 0.104 = 9 X 0.0115 m. 
 
ATTIC BUILDING ACCOUNTS II 247 
 
 height of the stele was probably about 1.835 m., reconstituted as 
 follows : 1.085 m. (the Chandler stone) + 0.190 m. (the missing 
 17 lines spaced about 11;| m.) + 0.560 m. (fragments D + F). 
 
 On the reverse the spacing of the lines is 14 mm., so that in 
 the first column (= column III), ending 0.152 m. above the 
 bottom of the stele, there were evidently 120 lines, of which 
 11. 114-120 appear on fragment E. The end of column IV is 
 preserved on E, as appears from the vacant space below ; its posi- 
 tion can be determined only by the adjustment of the obverse. 
 The fractures of E and C fit together so that the bottom of the 
 last line (1. 14) of E is 0.035 m. lower than the bottom of the 
 last line (1. 14) of C= col. 1, 1. 144, which was, perhaps, 0.256 m. 
 above the bottom of the stele ; the last line of E would be there- 
 fore 0.256 — 0.035 = 0.221 m. above the bottom, and corresponds 
 to 1. 115 of column III.^ Of this reverse, 78 lines of each column 
 were probably preserved on the back of the Chandler stone until 
 they were removed in the effort to render it more portable. 
 
 Next in the series of Erechtheum inscriptions come the 
 expense accounts of this same year 409/8, of which four large 
 fragments have already been published : 
 
 G=I.G. 1, 321 ; Arx Athenarum\ A.E. 24 ; Ath. Mitt. 1911, pp. 320- 
 
 321. 
 H = /.(?. I, 321^ suppl. p. 148; A.E. 25 (cf. Ath. Mitt. 1911, p. 322). 
 I = I.G. I, 321^-^ suppl. p. 150; A.E. 26 (cf. Ath. Mitt. 1911, p. 334). 
 J =I.G. I, 321, suppl. p. 75 ; A.E. 27. 
 
 The relative order of the four fragments, as worked out by 
 Michaelis (Ath. Mitt. 1889, pp. 357-361) and reestablished by 
 Caskey (Ath. Mitt. 1911, pp. 318-320), calls for no further 
 discussion ; it is now only a question of the intervals that 
 separated them.^ The number of stelae, and their height and 
 
 1 0.221 - 0.152 = 0.069 = 5 x 0.014 m. 
 
 2 As Caskey remarks, the attempts to assign /. G. I, 321*, suppl. p. 151 (by 
 Kolbe, Alh. Mitt. 1901, p. 229) and I.G. I, 326 (by Frickenhaus, A. J. A. 1906, 
 14), i.0 the series of 409/8 are mistaken ; I shall soon state my reasons for dating 
 them in the last prytany of 408/7 (fragments R and T). Also wrong are Ban- 
 nier's suggestions (Bh. Mus. 1906, p. 226) tov I.G. I, 331 a + b, suppl. p. 39, and 
 323 ; the latter really belongs likewise to the tenth prytany of 408/7 (frag- 
 ment S), as does 331 c, suppl. p. 39, proposed by Michaelis (A.E. 27 a) and 
 Bannier as part of our inscription of 409/8 (fragment U). Bannier includes 
 also I.G. I, 325, which I assign to 407/6 (fragment X). 
 
248 WILLIAM BELL DINSMOOB 
 
 width, are alike uncertain. Frickenhaus assumed that on each 
 stele were three columns, Caskey that there were four, Bannier 
 that there were five. Both Frickenhaus and Caskey find that 
 the number of stelae was four, but Bannier concludes that there 
 was only one. Nobody has yet succeeded in bringing into rela- 
 tion with the rest the newly discovered inscription on the back 
 of one of the stelae, published by Washburn {A.J.A. 1906, pp. 
 2-3, pi. III). We are certain only that the stelae were 0.154- 
 0.157 m. thick, polished on both sides and set up like a parapet 
 with double anathyroses at the joints, so that both back and 
 front were intended to be visible. 
 
 The only attempt to join together pieces of this inscription 
 by actual contact, that of Professor Heberdey as reported by 
 Washburn and confirmed by Frickenhaus, was actually a 
 failure, as Caskey was able to prove (^Ath. Mitt. 1911, pp. 319- 
 320). But, though the actual junction is impossible, there is 
 no reason for placing fragments H and I, as Caskey does, in 
 separate stelae; as Washburn says Q.c. p. 2, n. 2) "from 
 the line of direction of the top of the fragments, from their 
 thickness, and from their nature at the back, there can be no 
 doubt, to one who examines the actual stones, that they belong 
 together." Washburn's conclusion in this regard is unques- 
 tionably correct. The principal fracture which appears on both 
 fragments is not that at the upper right edge of each (^Ath: Mitt. 
 1911, p. 320), but that at the upper left edge. Therefore, 
 while accepting Caskey's association of G and H because of 
 their contents and the anathyrosis at the left edge, I should 
 place in the same stele fragment I, as in Figure 2. It then 
 appears that this stele was at least five columns in width,^ but 
 probably no more. 
 
 G is assigned to the first stele of the series by Frickenhaus, 
 while Caskey prefers to place it in the second, because of the 
 presence of the anathyrosis and the fact that it indicates a stage 
 
 1 Bannier likewise restores the stele with five columns, but for very difierent 
 reasons (Bh. Mus. 1906, p. 226). He would join the last preserved line of G to 
 the first line of the first preserved column of I, placing H above G ; the third 
 column of I would be continued, according to Bannier, by the first column of J. 
 The actual fragments, however, will not allow the junction of G and I ; the frac- 
 tures are such that to bring the first preserved column of I in the same column 
 with G, the lines which Bannier combines must come 21 lines apart. 
 
ATTIC BUILDING ACCOUNTS II 
 
 249 
 
250 WILLIAM BELL DIN8M00B 
 
 of the work somewhat later than that of the report ; and it 
 carries with it, as we have seen, the two fragments H and I. 
 The date of G and H is evidently no earlier than the third 
 prytany of 409/8 ; in the first prytany three frieze blocks were 
 laid (^I.G. I, 322, col. I, 11. 42-43), and in the second prytany 
 could have been accomplished all the work which preceded 
 that recorded in G.-' All this work could not have occupied 
 an entire stele of five columns, preceding G ; on the other 
 hand, actual experiment shows that the whole could be written 
 in 95 lines of text, about 20 devoted to the first prytany and 
 75 to the second, by no means enough for even a single column 
 of the stele if it were of ordinary height, at least 1.80 m. 
 I assume, therefore, that G + H, of the third prytany, are to be 
 placed below the 95 lines devoted to the first two prytanies in 
 column I. To the question of the anathyrosis point we shall 
 return. Allowing for the prescript of the third prytany, it 
 seems fair to number the 43 lines of G as column 1, 11. 101-143. 
 H follows almost directly after G, as Mr. Caskey notes, with a 
 small gap ; the direction of the principal fracture indicates a 
 gap of 15 lines, so that we may number the 29 lines of H as 
 column I, 11. 158-186. The relation of the oblique fracture on 
 fragment I to that on H (the principal fracture of the stele) is 
 such that 1. 1 of the latter (=col. I, 1. 158) comes opposite 1. 
 43 of fragment I ; according to the system of enumeration here 
 adopted, fragment I contains 11. 134-146 of column II, 11. 116- 
 160 of column III, and 11. 133-156 of column IV. Column I 
 contained the accounts of the first and second and part of those 
 of the third prytany, including the stonework and part of the 
 woodwork ; column II, 11. 134-146 are evidently the end of the 
 stonework which formed the beginning of a new prytany, 
 the fourth, of which the account is closed in col. Ill, 1. 117 ; 
 the heading in the following line would then be that of the 
 fifth prytany, the accounts of which filled columns III (lower 
 
 1 This work included the laying and iTepyaala of four plinths and the 
 maschaliaia, the completion of some epikranitides, the laying and iTrepjoffla 
 of six epikranitides and the metopon capital, the completion of epistylia on the 
 ground and of one in place, the laying of five epistylia and the iT£pja<rla of these 
 and three others, the laying of the three central frieze blocks on the east and of 
 their antithemata, the iirepyaala of these and of the three frieze blocks previously 
 laid at the southeast corner, and the laying of the west half of the north frieze. 
 
ATTIC BUILDING ACCOUNTS II 251 
 
 portion) and IV. The five columns of the stele evidently 
 contained the accounts of about 5J prytanies. The three pre- 
 served fragments, G, H, and I, all came from the lower portion 
 of the stele. The total width of the five-columned stele 
 would come to 1.048 m.; the height must have been about 
 2.10 m. 
 
 This stele was opisthographic; the only fragment that re- 
 tains any of the back, G, has the same writing, width of column, 
 and general contents, that are characteristic of the other frag- 
 ments of the accounts. The reverse of G must have been in 
 the fifth column of the reverse of the stele. 
 
 After these three fragments, as Michaelis ascertained, belongs 
 J, containing the record of the laying of roof tiles and other 
 matters of completion ; the right edge is polished and intended 
 to be visible, as Caskey notes (I.e. p. 319), forming the close of 
 the series.^ It is evidently the record of the last, or tenth 
 prytany. The back of this piece is blank, so that it cannot be 
 fitted into the opisthographic stele composed of G, H, and I. 
 Now the explanation of the anathyrosis at the left edge of G -f- H, 
 where we need no earlier columns of accounts, becomes clear. 
 As in all previous building accounts, e.g. those of the Parthe- 
 non, the Propylaea, and probably also the original work on the 
 Erechtheum, it was intended that the accounts should be in- 
 scribed on the obverse and reverse of a single slab ; with the 
 gradual lengthening of the prytany accounts, however, both 
 obverse and reverse had been occupied while yet two prytanies, 
 perhaps, remained to be inscribed. There was no alternative 
 but to set up a second stele, to the left of the first with an 
 anathyrosis joint between, and to continue on its reverse the 
 accounts of the final prytanies, while the obverse remained 
 blank. The reverse of G was probably concerned with the 
 stonework forming the beginning of the ninth prytany at the 
 bottom of column X, while J contains parts of columns XIII- 
 XV belonging to the tenth prytany. 
 
 A few notes on the published portions of Mr. Caskey's read- 
 ings of the accounts of 409/8 {Ath. Mitt. 1911, pp. 317-343) 
 may conveniently be inserted here. 
 
 1 Friokenhaus uses this right edge as if it were a joint surface, abutting 
 against his fourth stele. 
 
252 WILLIAM BELL DINSMOOR 
 
 G, 11. 1-2 Q.e. p. 320) might perhaps be read : 
 
 [yoTO Totxoi'' jn?] K0§ [oKTOTToSas, ^]Y$§OS 
 
 [SiTToSas, eivn] P: TEN TETP[a,ro]A[i'a] N HEK- 
 
 This entry need not refer to iirepyaaia, and therefore to the 
 epistylia, merely because the rate is given by tetrapodies (I.e. 
 p. 323); it is more probable that we are concerned with the 
 frieze. Rates are given for all kinds of work done on the 
 south frieze (the beginning of the account — no rates are given 
 for the east frieze, and only half of those for the north) ; since 
 the frieze blocks themselves vary in length, a common rate 
 could be given only in tetrapodies; and it happens that the rate 
 for laying the Eleusinian stone of the frieze was actually 5 dr. 
 per tetrapody (11. 5, 6, 28, 31), just as here, I think, we have 
 the statement of that fact. On the other hand, the universal 
 rate for iTrepjaaia on the frieze, even when the wide Pentelic 
 antithemata were included, was only 3J dr. per tetrapody (11. 
 18, 28, 43).^ The missing letters at the beginning of 1. 2, re- 
 stored by Mr. Caskey as the dual \_Bvocv ttoSoii'], can find no 
 parallel among the other dimensions given in these inscriptions 
 except where the inclusion of a fraction did not allow the 
 simple adjective St7ro(u)? (B, col. I, 11. 34, 38). 
 H, 11. 8-9 (I.e. p. 322) I should read as 
 
 Ul- 
 
 lOw ipiOfjibs ■. ArllM] ANTlO[e]MATA 
 [P£VT£A«Kai pi: AiJAINAI/ /.[A]PII 
 
 With Xt'l \_0oi, 'EXeuo-twaKot] we should have only one space for 
 the number of frieze blocks, of which fourteen were set on the 
 
 1 A seeming discrepancy of price in 1. 28 is easily explained. The text (I.e. 
 p. 321) reads P II (as on the stone), the rate then being 3Jdr. per tetrapody; 
 the sum should have been, more accurately, 5 dr. IJ obols for the six feet. 
 Mr. Caskey later refers to this item as 7 dr. (I.e. pp. 327, 331), as if he had 
 read it P HH, and thereupon draws the conclusion that iirep-yaala on the east 
 frieze was more costly than elsewhere, 4 dr. 4 obols per tetrapody. The distinc- 
 tion between irXdros (1. 2.5) and hi(p(7os (11. 9, 33) does not seem so certain as to 
 give a plausible reason for a difference in price; the text gives only the final s of 
 one hv(t>(70i and nothing of the other, while I seemed to see [.] ' [. . .]S and 
 [• • • ■]*-'[•] respectively, as if we should read irXiros for the three sides 
 of the east portico, and so for all five cases in which antithemata are 
 mentioned. 
 
ATTIC BUILDING ACCOUNTS 11 253 
 
 south, east, and north sides. The number of Pentelic antithe- 
 mata is probably not 1 1 1 1 (pp. 322, 326) but P I (cf . G, 11. 7-12, 
 24—26, 32-35); for as Mr. Caskey notes elsewhere (pp. 326, 
 327, 331) the antithemata on the east fagade seem to have been 
 of Pentelic marble, though this fact is not expressly stated in 
 11. 24—26. Of the number of Aeginetan antithemata there 
 seem to be more traces than Mr. Caskey shows, and the number 
 seems to have been larger than the four outer walls of the 
 temple would have allowed ; perhaps some of these blocks 
 ■were on the cross wall, since 21 lines (15 in the gap between 
 G and H, 6 on H) are more than we need for the west wall 
 alone. 
 
 1, col. I, 1. 23, might be read: 
 
 [P : — ; /jbiKO'S lleirTawov, 7rXaTo[§ TPI P ON 
 
 As Mr. Hill and Mr. Caskey point out, the southwest geison 
 was unique with the longer face, given as 1^ feet, toward the 
 south flank (I.e. pp. 335, 337); the normal angle geisa are 
 given as 6 feet long, their greater lengths towards the fagade 
 (I.e. pp. 335, 336). 1 On the analogy of the southwest geison 
 Mr. Caskey likewise revolves that at the northwest corner 
 
 1 The foot used in the Erechtheum and the Propylaea seems to have been 
 0.32725 m. long, the tetrapody, in the blocks furnished by the contractors, being 
 then 1.309 m. But the blocks were trimmed for their places to an average length 
 of 1.2985 m., so that the surveyors measuring the blocks in situ should, to be 
 accurate, have used a unit of 0.3246 m. The total length of the south geisa was 
 70J of these feet, as if it were composed of Z\ feet of the east angle geison, six- 
 teen regular four-foot geisa, and 3^ feet of the vrest angle geison. The returns 
 of the normal angle geisa were Z\ feet, not Z\ feet as the inscriptions generally 
 give them (though 1. 27 correctly gives 3} feet) ; the variation is due to the fact 
 that the dimension needed to be merely approximate in the inscriptions as a 
 means of identification. The regular geisa average exactly four feet of 0.3246 
 m. in length, the approximate designation by the surveyors here coinciding with 
 a stock size, the tetrapody furnished by the contractors. The unique block 7J 
 feet long was more probably 7J feet long, its end coinciding with the normal 
 location of a joint and not quite reaching to the pilaster of the metopon, so that 
 it would have required a thick frieze below it, with the sofiit exposed in the 
 niche, as along the west side of the southwest wing of the Propylaea. If the 
 longer face of this angle geison had been toward the west, it would have needed 
 to be at least 9 feet long, since here a thick frieze below it would have been 
 impossible. And if, with the longer face toward the south, it had been intended 
 to disregard the jointing system and obtain a bearin;; on the pilaster of the met- 
 opon, this bearing would surely have been made greater than the almost useless 
 amount of 5 cm. obtained even with a length of 7J feet. 
 
254 WILLIAM BELL DINSMOOR 
 
 (which was, however, of normal size) so as to bring its longer 
 dimension toward the flank of the building. But it seems diffi- 
 cult to imagine that, when at the southwest corner the abnor- 
 mal conditions (the niche and the Kekropion) were not allowed 
 to affect the jointing system of the geisa, these same conditions 
 should have entirely disarranged the northwest corner, where 
 there was no reason for any change.^ The sole reason for re- 
 volving the northwest geison seems to be, to give the special 
 intermediate geison (1. 23) a length in even feet without a 
 fraction, that it may be fitted to the lacuna in the inscription 
 (I.e. p. 338). If we follow the approximate measurements of 
 the inscription less literallj'', we may subtract from the total 
 length of the west geison, actually 86^ instead of 36 feet, the 
 lengths of the northwest (6^ instead of 6 feet) and southwest 
 (approximately 3| feet) angle geisa, leaving an intermediate 
 space of approximately 26f feet. The length of the special 
 block must have been 2|^ or 6| feet, or rather to fit the lacuna 
 in the inscription, rphrov or hetrrdTrov; the latter must be pre- 
 ferred in order to obtain a bearing above the southernmost 
 column (cf. I.e. p. 338). As the short returns of the other 
 angle geisa, given as 3|^ feet, must be reduced to 3^ feet, so for 
 the southwest geison we are probably to interpret 3|^ feet as 3| 
 feet. On the west, as on the south, the jointing system was 
 not disturbed by the abnormal conditions at the southwest 
 corner; the difficulties were overcome by inserting a joint in 
 one case, by suppressing a joint in the other. The total number 
 of four-foot geisa on the west must have been five, not six; but, 
 as Mr. Caskey points out (I.e. p. 338), the angle geisa were not 
 laid in this prytany,^ and the same may have been true of 
 others, thus reducing the number at the beginning of the line 
 to even less than H. 
 
 I, col. II, 11. 8-42, show that the jointing of the blocks of 
 the west tympanum was unlike that in the east pediment. The 
 variations in height and length are explained by Mr. Caskey as 
 the simple result of an attempt to break joints with the pecu- 
 liarly arranged west geisa, though the variation in thickness 
 
 1 With the length of 6 (or rather 6^) feet toward the north, the next geison on 
 the north must have been 5^"^ instead of 4 feet long. 
 
 2 But at least one angle geison must have been laid in a preceding prytany, so 
 that the restoration iKiroii<rapTi (11. 28, 31), denoting virork done on the ground, 
 does not appear to be as suitable as iirepya<rii4mi or a similar word denoting worls; 
 done in place for the reception of the end of the raking geison. 
 
ATTIC BUILDING ACCOUNTS II 255 
 
 was intended to decrease the load on the lintel of the small 
 west door and on the thin epistyle at the southwest corner (I.e. 
 pp. 338—341). Now that we have found no peculiarities in the 
 geisou jointing that could have affected the tympanum, I feel 
 that another explanation of these variations is required. It is 
 apparent that the builders of the Erechtheum employed every 
 device known to them to decrease the tremendous strain on the 
 lintel spanning the tomb of Cecrops. A possible device would 
 have been the cantilever system employed in the Propy- 
 laea (^A.J.A. 1910, pp. 146 ff.); the south KepKiS(.aio<; could 
 have been so designed as to balance itself and its superposed 
 load exactly above the southernmost column, relieving the 
 south anta of much of the weight. Then the KepiciSialo<; would 
 have been seven feet in length; to allow for this increased 
 length, the next block and the Kopv<f)aio<; would have been re- 
 duced, the latter, as we learn from the inscription, to four feet. 
 The north half of the tympanum seems to have remained undis- 
 turbed, so that the Kopv^aco'i would have been centred one half 
 foot north of the axis of the pediment. 
 
 To the accounts of the next year, 408/7, have been assigned 
 seven pieces, I.Gr. I, 324 a-e (which I number K to 0), and 
 Ath. Mitt. 1901, pp. 223, 224 (P and Q); the date was obtained 
 by KirchhofE {Abh. Berl. Akad. 1864, p. 52,i by Ferguson {The 
 Athenian Secretaries, p. 27) from the order of prytanizing 
 tribes, and by Kolbe (Ath. Mitt. 1901, p. 225) from the new 
 fragment P with the archon's name E[u]«:[Te/tiovo9]. I include 
 also six other pieces, I.a. I, 321 ^(R),^ 323 (S), 326 (T), 331 c 
 (U), 331 g (V), and I. a. II, 4331 (W).^ 
 
 These accounts were inscribed on slabs only about 10 cm. 
 thick, evidently intended as a revetment; the backs are all 
 roughly picked except in the case of T and U.* The thinness 
 accounts for the small size of the slabs. The complete height 
 given by M is 0.950 m. The width is in no case preserved, 
 for no piece is wider than two columns ; but the broken left 
 
 1 This was rejected by Michaelis {Ath. Mitt. 1889, p. 356 ; A.E. 28) in favor of 
 the second half of 409/8. 
 
 2 As assigned by Michaelis, A.E. 28 /. 
 
 3 W assigned to this account by Bannier, Berl. Phil. W. 1911, p. 854. 
 
 * It was evidently these two fragments that led Washbarn to say {A.J. A. 
 1906 p. 3) that these slabs are smooth behind. 
 
256 WILLIAM BELL DINSMOOB 
 
 edge of M shows that the width was originally more than two 
 columns, and the dowel cuttings in the top and bottom of M 
 lie at one side of the intercolumnar space, not centred on it as 
 we should expect if there were only two columns,^ while the 
 mediaeval splitting along the middle of the second column of 
 K implies that here was the axis of a three-columned slab ; the 
 original width would have been about 3 x 0.224=0.672 m. 
 Each slab had thus a quarter of the area of a normal stele ; 
 could it have been that four were first set up together, with 
 the total dimensions about 1.344 x 1.900 m.? 
 
 The relation of the number of columns to the number of 
 slabs has not been considered in previous restorations. Kirch- 
 hoff indeed suggested that there was originally only one slab, 
 later cut into smaller pieces (Z G-. I, p. 172). Robert arranged 
 the accounts in thirteen or fourteen columns, the first of which 
 contained the four fragments of 409/8, according to the theory 
 of Michaelis (Eermes, 1890, pp. 439-442). Kolbe returned to 
 Kirchhoff's theory, separating the two years and associating L 
 and M ; he restored nine columns for the accounts of 408/7 
 (Ath. Mitt. 1901, pp. 231, 234). 
 
 The fact that the bottom of the first preserved column of M 
 cannot be combined with the top of the second has long been 
 recognized as evidence that the inscription was more than one 
 tier of slabs in height. To assume with Kolbe (I.e. p. 232), 
 however, that there were three tiers merely because M has dowel 
 cuttings both on top and bottom, and so might have been 
 dowelled to a tier below and a tier above, seems impossible.^ 
 In the eighth prytany the total expenditure was 1239 dr. 1 ob., 
 of which we have preseryed 907 dr. at the bottom of one 
 column of M and 126 dr. 1 ob. at the top of the next ; above or 
 below this slab, therefore, were listed the remaining 206 dr. 
 which, as they must have been expended on very detailed wood- 
 work, could conceivably have filled the height of one more tier, 
 but certainly not of two more. In close proximity with M 
 was K, containing the beginning of the account of the seventh 
 
 1 There were probably two dowels as in the decrees for the Temple of Athena 
 Nike,'E0. 'Apx- 1897, pi. 11. 
 
 2 His other reasons I disregard because they are based on the false assumption 
 that L, N, and P are to be associated with K and M. 
 
ATTIC BUILDING ACCOUNTS II 
 
 25T 
 
 WW JIf7FJMg Bi » 3 MTgafl a «agftTn i lgLiK.T2 flf Q3L^J"Lf 
 
 
 O 
 
 
258 WILLIAM BELL BINSMOOR 
 
 prytany ; it has always been assumed that the two are to be 
 placed side by side in the same tier. These two cannot, how- 
 ever, have formed the left (K) and right (M) edges of one 
 and the same slab, which would then have been four columns 
 in width ; for the top of K is roughly tooled ^ and without a 
 dowel cutting, while the top of M is smooth and has a dowel 
 cutting. Nor can we place them side by side in separate slabs, 
 each of three columns (if not more); for while the eighth 
 prytany occupied exactly one column, the seventh prytany 
 would have occupied almost four columns at least, an impos- 
 sible proportion. Therefore we must place K vertically above 
 or below M, making, if there were originally three columns in 
 each, the second column of the one coincide with the first pre- 
 served column of the other. Since the first preserved column 
 of M is not continued by the second column of K, and since the 
 top of K has no dowel cutting for the bottom of M and is too 
 rough to form a bed, it is certain that K is to be placed above 
 M. Then the rough top of K is paralleled by the unpolished 
 tooth chiseling on the top of L (the top of P is worn away), 
 and is without dowels, forming the top of the upper tier.^ 
 Such a disposition would be confirmed if the first column of K 
 were continued on the missing first column of M, or if the 
 second column of K were continued on the first preserved column 
 of M. The first test is now impossible. The last remaining 
 items of the second column of K (11. 22-40) are concerned with 
 the removal and resetting of the sculptors' scaffolding. In the 
 next missing 47 lines ^ should have followed payments for the 
 frieze sculptures placed by means of this scaffolding ; and it is 
 noteworthy that 11. 1-22 of the same column on the lower slab 
 M are concerned with such frieze sculptures ; nine figures or 
 groups costing 867 dr. occupy the 22 lines, and to make up 
 the missing 2448 dr. (the total for sculpture on this prytany 
 is 3315 dr.) we should require about twenty-five figures or 
 groups occupying about 61 lines, fitting the allowance of 47 
 
 ' It is original, though Kolbe thinks otherwise (I.e. p. 233, "oben ein Tail 
 abgeschnitten ist "). 
 
 " There is a slight difference, owing to the employment of a different work- 
 man, in that the tops of L and P are level (cf. the Parthenon and Propylaea 
 stelae), while that of K slopes down toward the back. 
 
 8 Supposing that there were 87 lines in the upper as in the lower slab. 
 
ATTIC BUILDING ACCOUNTS II 259 
 
 lines if the proportion of the cost devoted to groups was some- 
 what larger. 
 
 With this combination of K and M, there is no possibility of 
 bringing into relation with them the fragments correctly united 
 by Kolbe, L, N, and P. His reasons for combining all five 
 pieces are : (1) the fact that Kirchhoff had already done the 
 same (in the case of K, L, and M), and (2) a desire to bring 
 the title (composed of P and L) in the centre of the series of 
 slabs. I may observe that it is most improbable, judging from 
 experience with other stelae, that the entire number of slabs 
 should have been so accurately foreseen at the beginning of 
 the year as to allow of any such axial position. Kolbe attains, 
 moreover, what is by no means an axial position for the title 
 only by the most violent compression of the first six prytany 
 accounts, and by placing K in an impossible relation to M. 
 Robert (^Hermes, 1890, p. 439) had long before set aside the 
 assumption that Kirchhoff's reasoning must be correct ; but 
 he, too, was- bound by the impression that L with the title 
 must have an axial position ; he had attempted to demonstrate, 
 however, what Kolbe afterwards rejected, the freedom of L from 
 K and M.^ Kolbe's certain arrangement of L, N, and P — 
 
 P LI L II 
 N I N II 
 
 would require four columns in each slab if P and L are to be 
 united in one ; but if there were only three columns in each 
 slab, we must suppose that a joint lay between P and L. Not 
 only is it impossible to combine them with K and M, but they 
 occupy so much space that we may with certainty place them 
 before K and M, in the first six prytanies. Since each of the 
 prytany accounts on K and M occupies one column or slightly 
 more (when it is a question of sculpture), we may conclude 
 that for the first six prytanies we should require two slabs in 
 each tier, giving six columns besides the 58 lines of the sixth 
 prytany on K. To these first two vertical pairs of slabs belong 
 the three fragments L, N, and P ; P and L, with their roughly 
 
 1 Pallat and Frickenhaus follow Kolbe in the association of K + M and 
 L + N + P {A.J.A. 1912, p. 188, n. 1). 
 
260 WILLIAM BELL DINSMOOB 
 
 tooled tops, give the beginnings of columns III and I V-V ; N 
 has the ends of columns II-III. The bottom of column II and 
 the top of column III are concerned with stonework, the begin- 
 ning of a prytany account, evidently the third ; ^ the bottom of 
 column III and the top of column IV are concerned with sculp- 
 ture, the end of a prytany account, evidently again of the third 
 prytany, which would have been particularly long because it 
 included work on sculpture (as in the seventh and tenth 
 prytanies).^ At the top of column V appears woodwork, near 
 the end of the fourth prytany. ^ 
 
 Two other fragments have been accepted without question, 
 and Q. The former is always placed at the very end of the 
 inscription, in the tenth prytany, because the right edge is 
 preserved. The vertical edges of K and M, on the other hand, 
 are said by Kolbe (I.e. p. 2-32) to be treated as joint surfaces 
 with anathyroses ; these are not anathyroses, however, but 
 decorative marginal draftings, and they are no more treated as 
 joint surfaces than is the edge of 0;* that they -were used as 
 joint surfaces, abutting against other slabs, is nevertheless 
 certain, and the same is true of the right edge of 0, for what 
 we must identify as the final slabs of the series had writ- 
 ing of an utterly different character. The edge of has the 
 same rough toothed chiseling that appeared on the top of the 
 original pair of slabs (as shown by L). As appears from the 
 position of the title, the first two pairs of slabs (fragments L, N, 
 and P) were set up together and may have had an anathyrosis 
 joint between them ; the place for is the once exposed right 
 edge of the second pair of slabs, where in the upper part of 
 column VI it would form the conclusion of the fifth prytany 
 
 1 Kolbe assumes that because the scaffolding of the columns in the north 
 porch was taken down in the sixth prytany, the fluting of the eastern columns 
 (on N and P) was begun only after that date. But Mr. Hill has shown (A.J.A.^) 
 1910, pp. 292, 294) that the scaffolding of the columns of the north porch was 
 transferred, not to the east portico, but to the west cella. 
 
 2 Kolbe assumes that because a great deal of sculpture is missing from the 
 account of the seventh prytany, N and L must be fragments of that account. 
 
 8 It is unnecessary to suppose that, merely because the bent beam was set in 
 the sixth piytany, L must be later ; here it is a case of working out on the ground 
 the woodwork connected with the beam, before it was set in place. 
 
 * For a proper anathyrosis in a vertical joint surface, compare the left edges 
 of G + H. 
 
ATTIC BUILDING ACCOUNTS II 261 
 
 account, with the items referred to in the account of the sixth 
 prytany (column VII, K 52-53). The reason for the empty 
 lower part of the fragment would seem to be that the account 
 ended so near the bottom of this upper slab that the title for 
 the sixth prytany was placed on the lower slab, as was the 
 case also in the tenth prytany. The small fragment Q seems 
 to refer to similar work of gilding ceiling rosettes and eyes of 
 columns ; it cannot be combined with O and could hardly be 
 earlier ; its place would be therefore in the ninth prytany, 
 immediately following M, which closes in the midst of similar 
 items. 
 
 The method followed in setting up the slabs of 408/7 is now 
 certain. To approximate the size of a normal stele, two vertical 
 pairs of slabs were set up side by side, the upper tier doweled 
 to the lower, and the lower to a plinth ; the exposed vertical 
 edges and the top of the upper tier were roughly tooth chiseled 
 but not polished; the backs of the slabs were roughly picked, 
 except near the edges where a narrow projecting margin (a 
 rough anathyrosis) was left to make a close joint against the 
 background. Symmetrically disposed about the joint marking 
 the axis of the combined pairs of slabs was the title [eVi] 
 E[u]K[Te/toi'o? ap)^ov]TOt.} occupying exactly the width of the 
 two central columns of the original series of six. When these 
 six columns, each evidently of 163 lines,* had been filled, an- 
 other workman set up a third pair of slabs, with columnsVII- 
 IX, each of 174 lines ; the slabs were similar to the original 
 four, except that the edges were picked, with a decorative 
 smooth marginal drafting, and the top of the upper slab was 
 left rough. Column IX was entirely filled when only 64 lines 
 of the ninth prytany account had been inscribed ; there was no 
 alternative but to set up a fourth pair of slabs, with columns 
 X-XII. The small sliver Q, from the uppermost part of 
 column X, gives no hint as to any technical peculiarities of this 
 
 1 Kolbe believes it necessary to precede these three words by ['ASerawc 
 dpdXoffavl, to be symmetrical about the words [dpx] ITE KT[oi' | ' Apx^\l,OXOl^ 
 I 'A7pu]KEGE[''] ; tliis formula does not appear on other building inscriptions 
 and cannot be symmetrically related to any disposition of the slabs. 
 
 2 The upper tier is supposed to have been of the same height as the lower, 
 which contained 87 lines ; the space allotted to the title is equivalent to the 
 height of 11 lines. 
 
262 WILLIAM BELL DINSMOOB 
 
 fourth pair of slabs ; but the fragments of the lower slab of the 
 pair, now to be discussed, show that they were smooth behind, 
 and so again the work of a different stonecutter. 
 
 The ninth prytany account, to be of the customary length, 
 must have filled at least the upper half of column X ; this 
 would leave most, if not all, of the part of column X on the 
 lower slab, for the beginning of the tenth prytany account. 
 The small fragment R, with the smooth top and dowel hole 
 characteristic of the lower tier of slabs of 408/7, the same size 
 and style of letters, the same formulae and lengths of lines, as 
 in the earlier prytanies of 408/7, exactly fits here ; ^ the dowel 
 cutting is at the left of the intercolumnar space, so that R is 
 from the upper left-hand corner of the slab, the opposite of M 
 where the right-hand dowel is at the right of the intercolumnar 
 space. R contains therefore in column X the beginning of the 
 tenth prytany account, with a margin of 0.025 m. to separate it 
 from what came above ; in column XI is the continuation of 
 the same account (work in sculpture) with the same margin 
 continued, doubtless a mistake of the stonecutter. The fact 
 that the letters are irregularly arranged, i.e. not gtoiehedon as 
 in the rest of the account of 408/7, seems to have been a pecu- 
 liarity of the tenth prytany.^ Two small fragments, S and U,^ 
 clearly belong together ; the writing is larger than that on any 
 fragments that I have yet mentioned, but the formulae and 
 names of workmen, and the thickness of the slab (U= 0.092 
 m. with smooth back), are the same as in the accounts of 408/7, 
 with the non-stoiohedon arrangement characteristic of the tenth 
 
 1 Kirohhoff {I.G. I, suppl. p. 151) suspected that this was perhaps of 408/7 
 rather than of 409/8, with which it was classed ; this was confirmed by Ferguson 
 {The Athenian Secretaries, p. 26), who restored ['E/)exSe(]A05, fitting tbe 
 tribal order of 408/7, and it was so accepted by Michaelis {A.E. 28/). Kolbe 
 {he. p. 229) and Frickenhaus {A.J.A. 1906, p. 14) return it to 409/8, which 
 Caskey rightly says {Ath. Mitt. 1911, p. 318) is impossible. 
 
 2 The two other fragments hitherto assigned to the tenth prytany, N and 0, 
 are stoichedon ; this attribution of N, by Fabrioius and Michaelis, was proved 
 impossible by Kolbe ; has been discussed above. I have assigned N to the 
 third prytany, to the fifth. 
 
 3 S was attributed to the series by Stephani (Annali, 1843, p. 327), but 
 doubted by Kirchhoff ; Michaelis includes it as a sort of appendix (A.E. 29) ; 
 Bannier (flft. Mus. 1906, p. 226) assigned it to 409/8. U was attributed to some 
 Erechtheum account by Kirchhoff (I.Gf. I, suppl. p. 151) ; Michaelis {A.E. 27 a) 
 and Bannier (I.e.) include it in the account of 409/8. 
 
ATTIC BUILDING ACCOUNTS II 263 
 
 prytany. U is concerned with woodwork, S with sculpture ; if 
 the writing on this slab increased in size toward the bottom, S 
 would come below R in column XI, and U, earlier in the account 
 because woodwork always precedes sculpture, would come at 
 the same level in column X. The supposition that the writing 
 enlarged toward the bottom of the slab is confirmed by T,i 
 which has letters larger even than those of S and U, and pre- 
 serves the original smooth bottom of the slab ; ^ it is smooth on 
 the back, and of the same thickness as U. The items refer to 
 woodwork, so that it must be placed in column X, below U ; 
 the last line is not continued by the first line of R in column 
 XI, so that, as in the previous pairs of slabs, we may be certain 
 that column XI began in the upper tier. Fragment W has 
 even larger letters inscribed somewhat carelessly and not stoi- 
 ehedon ; ^ it continues S with entries of sculpture,* and so may be 
 placed in column XI. Finally the fragment V,^ with extremely 
 widely spaced and careless letters, is certainly to be combined 
 with W, though not in actual contact ; it has the narrow inter- 
 columnar space characteristic of the Erechtheum accounts, and 
 to the left of this appears one letter (E) of the more closely 
 spaced column X; it concludes the tenth prytany account, i.e. 
 the entire account of 408/7, near the bottom of column XI, 
 with the total payment for sculpture and the balancing of re- 
 ceipts and expenses. 
 
 The accounts of 408/7 closed while the Erechtheum was still 
 unfinished, and while column XII was unoccupied. It is natural 
 to suppose that what little remained to be done in 407/6 was 
 recorded in column XII. A fragment, I.Gi. I, 325 (X),^ has 
 the same width of column (0.22 m.) and the same size of 
 
 1 Schone {Serines, 1870, p. 54) refused to assign T to the Erechtheum ; 
 Frickenhaus {A. J. A. 1906, p. 14) attributed it to the account of 409/8, whence 
 Caskey {Ath. Mitt. 1911, p. 318) rightly ejected it. 
 
 2 Kirchhoff wrongly says that it is broken on all sides. 
 
 3 Assigned to the account of 408/7, but without closer identification, by 
 Bannier {Berl. Fhil. W. 1911, pp. 853-854). 
 
 i Cf. 1. 5 [t^]N AYN[ar/ca], and 1. 8 [riv 7]YNA[iKa] as in column VIII, 
 
 M 21. 
 
 s Assigned to the Erechtheum, hut without closer identification, by Bannier 
 (iJA. Mms. 1906, p. 226). 
 
 <• Piitakis ('E0. 'A-px- no. 418) had assigned it to the Erechtheum accounts, 
 and Bannier (Bh. Mus. 1906, p. 226) placed it in the accounts of 409/8. 
 
264 WILLIAM BELL DINSMOOB 
 
 letters as in the accounts of 408/7 (the extravagant spacing 
 of the tenth prytany being repressed), and betrays the hand of 
 the same " Cockney " scribe who has seized the new fashion of 
 omitting the aspirate H only to apply it where it was never 
 used. The form of the account, giving the total of the wages 
 paid day by day, is very different from the accounts of 408/7. 
 This common authorship and the change of formulae cause me 
 to date X as of 407/6; it comes from the lower corner (with a 
 smooth right edge) i of a slab somewhat thicker than those 
 previously noted (about 0.105 m.), and so not from column 
 XII, but from the lower of a fifth pair of slabs (column XV). 
 The last line ends 0.167 m. above the bottom of the slab, while 
 the account is not yet finished, so that it must have been con- 
 tinued on a sixth pair of slabs. 
 
 Probably the temple would have been finished in 407/6 but 
 for an untimely accident, the fire of 406 in the 7ra\ato9 veto? 
 (Xenophon, Hell. I, 6, 1). Whether the iraXaiK veck was the, 
 Erechtheum or the old Hekatompedon matters little in this 
 connection ; the fire, if not actually in the Erechtheum, cer- 
 tainly injured it at least. For we possess a fragment of an 
 inscription, /. G: II, 829 (Z), which seems to record repairs 
 after a fire,^ with formulae exactly like those of the Erechtheum 
 accounts.^ The thickness of the slab, moreover, is 0.098 m., 
 the back roughly picked with a smoother margin near the edge, 
 exactly as on the original four slabs of 408/7, The text of 
 Z has the same form of daily entries that appears on X. The 
 reduction in the size of the writing would be explained by 
 reasons of economy; otherwise the repairs occasioned by the fire 
 would have necessitated a great increase in the number of the 
 
 1 Kirchhofi and Bannier say that the left edge is preserved. 
 
 2 The restoration KeKalv/iha] in 1. 4 has always been accepted, except by 
 Judeich (TbpojrrapAie, p. 244, n. 6), who suggests as alternatives KeKa[\vti,ijL4va] 
 or some form connected with koimSs. Perhaps the most conservative restoration 
 would be a verb derived from Kaivis, since Kaim(^6)s itself occurs in line 14 ; but 
 even these renovations would probably have been caused by a fire, so that we 
 may return to the original interpretation of the inscription. 
 
 ' Those who identify the TraXaiJs i'et6s with the Erechtheum naturally assign 
 this fragment to the Erechtheum accounts ; and even of those who follow Dorp- 
 feld's view as to the actual location of the fire, Michaelis (Jb. Arch. t. 1902, 
 p. 3, and A.S. 30) and Judeioh (Topographie, p. 244, n. 6) agree that the in- 
 scription refers to repairs in the Erechtheum. 
 
ATTIC BUILDING ACCOUNTS II 265 
 
 slabs. The generally accepted date of Z, [i-n-l A]t[o(^a//To] 
 apxov[TO<;^,^ is due to the first publication by Kohler (^ffermes, 
 1867, p. 21), when it was supposed that the lines began one 
 letter farther from the left edge than is actually the case, so that 
 line 1 would have begun with [A]f[o^cti/To]; republishing it in 
 tlie Corpus (I. Gr. II, 829), Kohler showed traces of the actual left 
 column of letters but did not draw the obvious inference from 
 the new arrangement. We should now be forced to read [ett] 
 in one line and [} A]i[o(^ai/To] in the next, a most improbable 
 manner of beginning a new annual account. Better is [e7r]l 
 [- - - - -] ap'x^ov[To<!'\ beginning the new line ; the solution 
 proposed by Cooley and Dorpfeld {A.J.A. 1899, p. 352, n. 3), 
 [cTTJl [KaXXto] a/jp^oi'[To?] (406/5), would fit exactly; the only 
 other possible name between 412/1 and 377/6 is 'AXeft'o 
 (405/4). 2 I select 'A\ef to rather than KaWio for the following 
 reason. There is in the Epigraphical Museum at Athens an- 
 other fragment, I. Gr. II, 845 (Y), which joins accurately the 
 top of Z, and evidently therefore dates from the year previous 
 to that of Z. Together they seem to have formed the left edge 
 of a sixth pair of slabs. ^ We may then be certain that the 
 date of Z is 405/4, not 395/4; Y and Z together betray too 
 much uncertainty in the use of Attic and Ionic letters to date 
 from a later decade. Y, of the previous archonship, is therefore 
 of 406/5, the year of the fire, when the unexpected continua- 
 tion of the work led to the adoption of smaller letters. X, of 
 407/6, forms the connecting link, with the lettering of 408/7 
 and the formulae of 406/4.* 
 
 William Bell Dinsmooe. 
 
 Athens, 1912. 
 
 1 Doubted only by Dorpfeld (Ath. Mitt. 1887, p. 47), Cooley (A.J. A. 1899, p. 
 352, n. 3), and Judeich (I.e.). 
 
 2 This inscription has i instead of X5- 
 
 3 In the combined fragments, 11. 1-22 = 1-22 of Y ; 1. 23 is obliterated ; 11. 24- 
 44 = 1-21 of Z. 
 
 4 The date of the Carpathian inscription (A.E. 31), given by Poucart as 394 
 or 393 B.C., rests primarily on the supposed date 395/4given to the fragment Z 
 (B.O.S. 1888, p. 158); if the inscription is to be connected with the Ereohtheum 
 at all, it may be ten years older. Whether the date of the fire mentioned by 
 Demosthenes (XXIV, 136, p. 743) can likewise be pushed back to 406, as Dorpfeld 
 would prefer (Ath. Mitt, 1887, p. 44), depends largely on the identity of the 
 TraXaiJs veus. 
 
American .Sctjool i'l .- , 
 
 of Classical Stutii?si^\>i^- ■ '---i','''/ 
 at mm ' ^(xjy 11 iSii ) 
 
 ATTIC BUILDING ACCOUNTS 
 
 III. The Pkopylaba 
 
 Of all the buildings on the Acropolis, the Propylaea alone 
 are exactly dated by literary accounts. " Philochorus (in his 
 fourth book) and others record that the Athenians began to 
 build them in the archonship of Euthymenes (437/6 B.C.), 
 with Mnesicles as architect ; and Heliodorus in his first book 
 about the Acropolis at Athens says, among other things, the 
 following : ' In five years they were entirely , finished ; two 
 thousand and twelve talents were expended; and five were the 
 gates they made, through which they entered the Acropolis.' " ^ 
 It was without difficulty, therefore, that a fragment of marble 
 found near the east portico of the Propylaea in 1830, inscribed 
 with (n/3o)rYLAIOEPAAS[ia9] and [^7r'Eu]OYMEN0§APX0[z^- 
 To^^ on one side and ['Etti re? t6t] APTESAPXE5, with Metage- 
 nes as first secretary of the senate (cf. I.G. I, 301, of 434/3) 
 and Crates of Lamptrae as secretary of the treasurers of Athena 
 (cf. I. G-. I, 141, 142) on the other side, was recognized by 
 Rangabe as part of . the building accounts of the Propylaea 
 (Ant. hell. I, 1842, pp. 88-91, No. 89). This fragment (I.G. 
 I, 314 J-315 b) is necessarily the starting point for any consid- 
 eration of the building inscription. 
 
 The next advance was made by Kirchhoff, who united the 
 Rangabe fragment to another (I. G. I, 314 a-315 a) which had 
 been likewise published by Rangabe (I.e. pp. 233, 394, No. 128) 
 but had not been identified (Jh. Phil. Pad. 1861, pp. 47 ff.). 
 Subsequently Kirchhoff suggested other candidates for the stele 
 of the Propylaea, such as I. G. I, 312-313, 315 a-c (suppl. 
 p. 38), and 554 (cf. p. 222). Unfortunately these last were 
 not generally accepted ; Kohler had already proposed to assign 
 
 1 Harpocration, s.v. Upoiri'Saia Tavra ; Photius and Suidas, s.v. RpoTiXaia 
 Toura ; Plutarch, Pericles, 13. 
 
 371 
 
372 WILLIAM BELL DINSMOOB 
 
 I. Gr. I, 312-313 to that series which he later identified as be- 
 longing to the Parthenon (J. Gr. I, p. 161), and in this he was 
 followed by Foucart (B. C.H. 1889, p. 176, n.) and Michaelis 
 (^A.E. 10). 1 Michaelis tentatively admitted I. Gr. I, 315 a-o and 
 554 {A.JE. 5), but Bannier (Ath. Mitt. 1902, p. 303, n., Eh. 
 Mus. 1908, p. 429) and Cavaignac {Etudes, p. Ixxi) rejected 
 I. G-. I, 815 a-c, as well as 312-313, and admitted only I. Gr. I, 
 554 of Kirchhoff's three suggestions. Thte net results of Kirch- 
 hoff's work seemed to be, therefore, the addition of I. Gr. I, 
 314 a-315 a, and 554, to the Rangabe fragment. 
 
 Eduard Meyer (Forschungen, II, pp. 99, 101, n.) rather casu- 
 ally speaks of another fragment as belonging to the Propylaea, 
 I. Gr. I, 316. This was confirmed by Bannier, who identified 
 one more fragment, I. Gr. I, 331 d, suppl. p. 77, and found that, 
 according to the lettering at least, it must have joined I. Gr. I, 
 316 accurately {Ath. Mitt. 1902, pp. 302-303) ; I.a.I, 316 itself 
 is lost. 
 
 Cavaignac's list of fragments of the inscriptions of the 
 Propylaea is the same as Bannier's, viz. I. Gr. I, 314 5-315 b, 
 314 a-315 a, 316, 331 ci, and 554 {Etudes, p. Ixxi). 
 
 This was the state of the inscription when I worked on it in 
 1909-10 in connection with the study of the Propylaea. At 
 that tihie I enlarged the number of fragments from five, the 
 total proposed by Kirchhoff, Bannier, or Cavaignac, to seven- 
 teen. Delays in the presentation of this material allowed me 
 to supplement it by similar investigations of the inscriptions of 
 the Parthenon {A.J.A. XVII, 1913, pp. 53-80) and of the 
 Erechtheum {ihid. pp. 242-266), and also to benefit by the 
 subsequent studies of Woodward {B.S.A. 1909-10, pp. 198- 
 205) and Bannier {Berl. Phil. W. 1911, p. 853; 1913, pp. 
 317-319). 
 
 Woodward's studies, proceeding in two directions, led him 
 to enlarge the number of known fragments from five to ten ; 
 he included the two rejected Kirchhoff fragments and three 
 others hitherto unpublished. Measuring the two opisthographic 
 fragments united by Kirchhoff (J. Gr. I, 314, 315) and the two 
 opisthographic fragments proposed by Kirchhoff but rejected by 
 subsequent authorities {I.Gr. I, 312-313, 315 a-c), he observed 
 1 Cf. A.J.A. XVn, 1913, p. 56 and n. 1. 
 
ATTIC BUILDING ACCOUNTS III 373 
 
 that they all taper in thickness to such an extent as to prove 
 that they all belong to a single stele. Kirchhoff, indeed, had 
 already suggested that to include these fragments the stele 
 must have tapered from bottom to top (J. Gr. I, suppl. p. 38), 
 and this solution overcomes the objection of Bannier and Ca- 
 vaignac that they differ greatly in thickness. Woodward's 
 conclusions as to the exact positions of the four opisthographic 
 fragments, as indicated by their thicknesses, I am, however, 
 unable to accept ; and in the following discussion I have re- 
 tained my own restoration of the stele, which was completed 
 several months before he measured the fragments in question. 
 On the other hand, Woodward identified with certainty one 
 new fragment (0 in the following list), which he correctly 
 joined with two others which I liad placed side by side in the 
 Epigraphical Museum; since his transcript (^B S.A. 1909-10, 
 p. 199) does not represent the three fragments separately, I 
 give them here : 
 
 I Y I '0 I E 
 
 HVrOPA APb 
 K. A I L I O 
 
 K AT AAA \ P X t 
 
 M I S O S B 
 
 AN""' Y E:E 
 
 E 
 
 E 
 
 
 ■) AA 
 
 / 
 
 
 
 N 
 
 :)AA 
 
 § 
 
 
 L 
 
 - 
 
 
 
 
 — 
 
 
 
 
 / 
 
 frg. M frg. N frg. 
 
 Besides all these, I include five fragments already published 
 but never yet assigned to the inscription of the Propylaea. 
 Bannier (Bh. Mus. 1906, p. 223) had suggested that 1. &. I, 
 116 i, suppl. p. 25, and Kirchhoff that LG. I, 317 a, suppl. 
 p. 88, and 5551, suppl. p. 128, were parts of some building 
 
374 WILLIAM BELL DINSMOOR 
 
 accounts, but they were not more definitely identified. Foucart 
 {B.C.R. 1889, p. 176, n.) and Michaelis {A.E. 10) assigned 
 /. G-. I, 331 to the accounts of the Parthenon. I. Gr. I, 555 k, 
 suppl. p. 56, was supposed to be a part of the tribute lists. 
 
 Finally, I must mention four fragments which have not yet 
 been published. The most important is numbered P in the 
 following list ; I found it in the fourth room of the Epigraphi- 
 cal Museum. Q, R, and S were scattered in the first room of 
 the museum ; my attention was called to Q by Woodward ; 
 attached to S I found a manuscript note to the effect that it 
 joined the missing fragment I. G. I, 331. The four unpub- 
 lished fragments are represented below : 
 
 r 1 § I /\ I 
 
 lEPA EYSBU 
 
 I S T I C 'T I S L 
 
 E 1 T / 3 D A 
 
 => N § 
 
 \ L E frg. R 
 
 T A 
 N 
 
 frg. P I 
 
 L I C 
 
 L I C 
 
 \ I L I G 
 
 5O0MAT L\r 
 
 frg- Q frg. S 
 
 For convenience I number the nineteen above-mentioned 
 fragments as follows: k = I.G-. I, 116 1; B = 312-313; C = 
 314a-315a; = 314 6-315 6; E = 315a-e; F = 316 ; G = 
 317 a; H = 331 ; I = 331c?; J = 554 ; K = 555/fc; L = 565Z; 
 M, N, 0=B.S.A. 1909-10, p. 199, and as shown above; 
 P, Q, R, S as shown above. Seventeen of the nineteen frag- 
 ments are now in the Epigraphical Museum at Athens ; two, 
 /. Q-. I, 316 and 331, are missing. 
 
ATTIC BUILDING ACCOUNTS III 375 
 
 There is one fundamental piece of external evidence which 
 affects the arrangement, i.e. the testimony of Heliodorus and 
 Plutarch that the building was erected in five years. This has 
 always been accepted, and in view of what we shall learn from 
 fragment E, I may here state that the information is exact. 
 With the aid of this, and because the two fragments (C and D) 
 from the top of the stele show on one side the prescript of 
 the year 437/6 (ErE[i/]OYMEN0§APX0[i'To?], the first year of 
 the work, according to Philochorus), and on the other side the 
 prescript of the fourth year ([eVt re? rer] APTE5 APXES), Kirch- 
 hoff correctly deduced the fact that the accounts of the first, 
 second, and third years were inscribed on the obverse of the 
 stele, and those of the fourth and fifth years on the reverse. 
 
 The first attempt to arrange other fragments of the stele was 
 by Cavaignac (^Etudes, pp. Ixx-lxxi). He observed that I.Cr. 
 I, 564 (J) has writing characteristic of the obverse of the frag- 
 ments C + D, and he assigned it to the third year, supposing 
 that the fragment I. Gr. I, 244, of the tribute lists of the Delian 
 Confederacy, would exclude it from the second year (436/5). 
 That this conclusion is erroneous, even if we can suppose that 
 J. Gr. I, 244 is rightly dated in the Corpus, will appear below. 
 Again, he decided that the two fragments F + I joined by Ban- 
 nier have writing characteristic of the reverse of the stone, and 
 that they must date therefore from the fifth year ; both the 
 writing and the preservation of the surface show clearly, how- 
 ever, that these fragments belong to the obverse. 
 
 Here, too, I may briefly describe Woodward's restoration. 
 On the four large opisthographic pieces (B, C + D, and E) he 
 distinguishes the obverse from the reverse correctly, but unfor- 
 tunately supposed that the three fragments first published by 
 him (M -)- N -I- 0) had the shape of the letters and the finish of 
 the surface characteristic of the reverse (I.e. pp. 201, 202), and 
 accepted without question Cavaignac's similar error in the case 
 of F -f- I. With J the actual date, the second year, is obtained 
 accidentally, by placing it in the upper part of the obverse, while 
 the blank space at the end of the stone shows that it belongs at 
 the bottom. The two opisthographic pieces dated by Rangabe 
 and Kirchhoff (C -f- D) remain correct in Woodward's restora- 
 tion, but the two other opisthographic pieces are assigned to 
 
376 WILLIAM BELL DINSMOOR 
 
 the very dates at which they could not have been inscribed.* 
 Furthermore, he accepts the arrangement foUovred in the case 
 of the Parthenon (cf. A.J.A. 1913, p. 54), a pair of columns 
 side by side, only v^here I. Cr. I, 315 a-h (E) proves that there 
 was more than one column ; elsewhere he has a single column, 
 subdividing into two columns at the bottom of the obverse, and 
 returning into one column on the reverse. This unusual dis- 
 position is represented in tabular form, B.S.A. 1909-10, p. 205. 
 The four opisthographic fragments, B, C, D, and E, form the 
 skeleton about which all the smaller pieces must be placed. 
 As given in the Oorpus, B is either 10 cm. (Pittakis) or 14 cm. 
 thick (Kohler), C and D are 11 cm. thick (Pittakis), and E 
 varies from 17 to 18 cm. (Kohler). According to my meas- 
 urements, C and D, at the top of the stele, are 0.111 m. thick ; 
 at the bottom of the lowest line of letters on C the thickness 
 has increased to 0.138 m., with a taper of 0.027 m. in 0.400 m. 
 or of 0.0675 m. per metre. In fragment E, the thickness is 
 0.159 m. at the top of the topmost letters, and 0.186 m. at the 
 bottom of the lowest letters, a taper of 0.025 m. in 0.370 m. or 
 of 0.0675 m. per metre. In fragment B I find a variation be- 
 tween 0.136 m. and 0.155 m., a taper of 0.019 m. in 0.277 m. 
 or of 0.0685 m. per metre. The uniformity of these four frag- 
 ments is emphasized by the high rate of the taper, 68 mm. per 
 metre. ^ To compare building inscriptions only, the decrees for 
 the Temple of Athena Nike were inscribed on a slab which 
 tapered, to be sure,^ but only 0.008 m. (from 0.093 m. to 
 0.101 m.) in a distance of 0.295 m., a rate of 27 mm. per metre. 
 On the other hand, the building inscription of the Parthenon 
 was cut on a slab of uniform thickness, 0.198 m. throughout its 
 
 1 I.Cr. I, 313 is assigned to year II instead of year I, as is obtained below ; 
 I.Q. I, 312 is given to year IV instead of year V ; I.Q. I, 315 a-h is assigned to 
 years IT and III instead of I and III ; and I.G. I, 315 c is assigned to year V in- 
 stead of year IV. 
 
 2 Woodward gives for C, 0.11 m. at tlie top and 0.13 m. at a point 0.35 m. 
 lower ; for E, 0.17 m. at the top and 0.186 m. at 0.30 m. lower ; for B, 0.14 m. 
 at the top and 0.155 m. at 0.30 m. lower ; this is a taper of 50-67 mm. per 
 metre. These inaccuracies of measurement gave him false positions for B and 
 E. 
 
 3 It has not the uniform thickness 0.09 m. given by Kabbadias, 'E0. 'Apx- 
 1897, p. 176. 
 
ATTIC BUILDING ACCOUNTS III 377 
 
 height.* The opisthographic building inscriptions of the 
 Erechtheum are cut on slabs of uniform thickness, 0.139 m. and 
 0.155 m. throughout. 
 
 In fragments C + D the face with the prescript of the first 
 year (J. Gr. I, 314) is in most part well preserved, with the 
 original polish of the surface. The reverse (J. <J. I, 315) is on 
 the contrary so badly weathered that it is extremely difficult to 
 read, just as is the case with the Parthenon inscription {A.J.A. 
 1913, p. 57). The same characteristics distinguish the two 
 faces of fragment B, of which I. Gr. I, 313 must therefore be 
 the obverse, and I. Gr. I, 312 the reverse. Likewise it appears 
 that in fragment E the face with I. Gr. I, 315 a-5 must be the 
 obverse, and I. Gr. I, 315 o the reverse. 
 
 As regards the relative positions of the fragments in the en- 
 tire height of the stele, the taper gives us absolute evidence. 
 C + D combined would contain portions of 11. 1-13 of the 
 obverse and of 11. 1-18 of the reverse. The increase in 
 thickness of the stele is practically 2 mm. for each line of 
 inscription (on the obverse). We may calculate from this 
 that fragment B includes 11. 16-28 of the obverse and 11. 19- 
 32 of the reverse. And E contains parts of 11. 31-47 of the 
 obverse and 11. 41-46 of the reverse.^ 
 
 Laterally these four fragments may be located with as much 
 certainty. The key is given by the obverse of fragments C + D. 
 The first preserved letters of line 2 are A I, the ending of ['ETrto-- 
 rarjAI, and this has always, from the time of Kirchhoff, been 
 restored as the first word of the inscription. That it was not the 
 first word is shown by the entries below the prescript. In the 
 combined fragments C + D we find below the prescript traces 
 of six lines in smaller letters ; on C, at the left, are sums of 
 money, and on D, at the right, are words. That these two col- 
 umns, one of money and the other of items, belong together is 
 shown by the fact that where the heading ['AvaXo]MATA occurs 
 
 1 Strange to say, it is in this stele that Foucart and Michaelis attempted to 
 insert our tapering fragment B . 
 
 ■' Woodward (I.e. p. 201) by means of his measurements of the taper calculates 
 that fragment B was separated from C by a gap of at least 0.175 m., so that B 
 would contain 11. 25-37 of the obverse, and E, of which the lowest line would 
 be, according to Woodward, 1.435 m. below the top of the stele, would contain 
 11, 49-65 of the obverse. The gaps are really much smaller. 
 
378 WILLIAM BELL DINSMOOR 
 
 in the item column, we find the money column empty.' Below 
 this heading must therefore be expenses; and there in fact we 
 read . . . MATON, which should be, on the analogy of the Par- 
 thenon accounts, [we] MATON. . Above the heading ['Ai'aXo]- 
 AAATA are two items which must be receipts. These two items 
 are supposed to represent all the revenues of Athena for that year, 
 which were so great that when they had been diverted into the 
 treasury of the Propylaea, it became unnecessary to appeal to 
 the state for. funds, as was done in other years.^ Let us see of 
 what these revenues consisted. The first item is a sum of 132 
 
 drachmae received as ... . ASHIEPASMIS (an item which 
 
 appears on both sides of fragment B as OIKIA^HIEPA . . . ); the 
 second is a sum of 6 drachmae 1^ obols received as ... . KON 
 TIME (which appears on fragment B as PINAKO. II . .). The total 
 receipts for the first year would thus have been 138 drachmae 
 IJ obols,^ out of which 27,200 drachmae,* and much more now 
 lost, were expended! These two very small sums cannot be 
 from the regular sources of large supplies, such as the state 
 treasury, the board of treasurers of Athena or of the Delian 
 Confederacy, or other boards. Moreover, it must be noticed 
 that Tt/ie, indicating the money derived from the sale of objects, 
 appears only among the last items of each year in the Par- 
 thenon accounts ; house rental also, when it is mentioned else- 
 where in the Propylaea accounts and in those of the Eleusinian 
 epistatae (/. (r. I, 288 a, suppl. p. 145), appears among the last 
 items. It seems extremely probable therefore that these two 
 
 1 This combination of two columns, money on the left and items on the right, 
 is found also in the Parthenon inscription (cf . A.J.A. 1913, p. 59). 
 
 2 Cavaignac, Etudes, p. 101. 
 
 3 It is true that Kirohhoff {I.Q. I, p. 162) and Cavaignac {Etudes, p. 102) 
 suppose that these are only the ends of sums, other figures at their left having been 
 lost because of the breaking of the stone. The same is said of the three items of 
 expense, which have certainly lost some lesser figures at the right. But in the 
 second of the expenditures at least enough of the original surface is preserved to 
 show that at the left of the extant MpXfFI . . . there were no more figures, and 
 this determines the left edge of the money column. This is confirmed by the 
 size of the first item of receipts, which could not have had another H or any 
 higher figures before the extant H A AA h V , without making it too large for oUias 
 hiepas fjd(r9o<ns, as we shall observe later. 
 
 * Kirchhoff and Cavaignac read the total sum remaining as even greater, 72,200 
 dr. 
 
ATTIC BUILDING ACCOUNTS III 379 
 
 small sums are the last items in a series of receipts, most of 
 which have been previously mentioned. Such a previous men- 
 tion could occur only if there were another double column of the 
 inscription to be read before C + D, which would therefore be at 
 the top of a second double column. 
 
 That there were two such double columns is certain. Actual 
 traces of both appear on the obverse of fragment E (I. Gr. I, 
 315 a— J, suppl. p. 88). If we were to attempt to restore only 
 one column, the top of B, as located by the tajDer, would collide 
 with the bottom of C ; we must therefore place B at the left or 
 right of C, but not below it; this again results in two columns. 
 That the same arrangement must be extended to the whole stele 
 will become apparent from a comparison of the lengths of the 
 lines of the inscription. 
 
 Fragment I, shown to be of the obverse by its well preserved 
 surface, very unlike that on the reverse faces of the four opis- 
 thographic fragments, and by the larger size and spacing of the 
 letters,^ belongs to the right edge of the stele (the edge is pre- 
 served) and so to the second double column. As combined with 
 F, the obvious restoration of line 7 (Bannier, Ath. Mitt. 1902, p. 
 303) has 26 letters, spaced 0.0133 m., so that the 25 spacings 
 (0.332 m.) with the two half letters at the ends (0.010 m.) and 
 a blank surplus of 0.026 m. at the right, give 0.368 m. as the 
 distance of the beginning of the item column from the right edge 
 of the stele.^ On the obverse of C-l-D, the left edge of the 
 money column, as obtained above, is 0.225 m. from the left edge 
 of the item column. The entire width of the second double 
 column was then 0.225 -|- 0.368 = 0.593 m. If there were 
 two such double columns, the width of the stele would have 
 been approximately 1.19 m. More than two double columns 
 would give a very impracticable width for a thin free-stand- 
 ing slab. 
 
 We now place the obverse of C 4- D at the top of the stele, 
 
 1 Disregarding these facts, Cavaignac and "Woodward assign it to the reverse. 
 
 2 A similar result for the width of the item column may be obtained on the 
 reverse of C -|- D, where we restore in lines 9-10 (an item repeated word for 
 word in the Parthenon account, Col. V, 11. 17-20, of the same year) 30 letters, 
 spaced 0.0126 m., so that the 29 spacings (0.365 m.) with the two half letters at 
 the ends CO.OIO m.) would give the width of the item column as about 0.375 m. 
 
380 
 
 WILLIAM BELL DINSMOOR 
 
 
 " 
 
 T1 
 
 
 
 5 
 
 
 -* 
 
 S 
 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ,H 
 
 CO 
 
 
 < 
 
 
 5 
 
 ■*f< 
 
 rrt 
 
 
 1-9 
 
 
 n 
 
 
 
 
 
 ^ 
 
 N 
 
 
 
 
 
 
 
 Ph 
 
 £8 
 
 S 
 
 H 
 
 T! 
 
 
 
 
 
 
 
 
 El 
 
 
 
 T3 
 
 
 
 
 
 O 
 
 iH^ 
 
 +3 
 
 m 
 
 
 
 H 
 
 
 d 
 
 r, 
 
 a. 
 
 
 o 
 
 rt 
 
 1^ 
 
 c 
 o 
 o 
 
 t-^ 
 
 T3 
 
 -») 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 R 
 
 ;^ 
 
 % 
 
 
 
 
 p 
 
 
 O 
 
 W 
 
 
 S 
 
 
 
 
 S 
 
 "t 
 
 «-i 
 
 H 
 
 & 
 
 
 ^ 
 
 13 
 
 +^ o 
 
 .a s 
 %'i 
 
 
 
 
 
 
 
 
 S S 
 
 n 
 
 
 
 ^ -K 
 
 ^ 
 
ATTIC BUILDING ACCOUNTS III 381 
 
 above the second double column, the beginning of the money 
 column 0.594 m. from the right edge of the stone. As a re- 
 sult, the axes of the fii-st preserved letters in 11. 4 and 5, in 
 both cases I, are 0.038 m. to the right of the left edge of the 
 money column. The axes of and M, the last preserved let- 
 ters in 11. 3 and 4, are 0.895 m. farther to the right, and there- 
 fore 0.594 -_ (0.038 -f- 0.395) = 0.161 m. from the right edge 
 of the stone. In this space we could restore at most, in 11. 2 
 and 8, six more letters with the average spacing of 24 mm. 
 But it is noticeable that when we add four more letters'in line 
 3, we arrive at the end of the word APXO[z^to?] ; and siinulta- 
 neously 11. 2 and 6 arrive at the ends of words, EPAA§[ta9] 
 in line 2 being spread to occupy nine spaces instead of eight, 
 and AM4>ITPO[7re0ei'] in line 6 being squeezed into eleven and 
 a half spaces so as to finish evenly at the right. Here then we 
 may terminate the prescript, leaving the remaining 0.058 m. in 
 blank.i 
 
 On the obverse of C 4- D, the prescript of the year 487/6 has 
 11. 2-5 perfectly stoichedon, an arrangement which is varied in 
 11. 6-7 of fragment C, though it is regained on fragment D 
 with the purpose of ending line 6 evenly with those above. 
 This variation was caused by the insertion or omission of one 
 letter in line 6. There, the name EP . XA . . . . occupies 9|- 
 letter spaces, so that it contained either 9 or 10 letters ; and 
 judging by the manner in which the letters diverge more and 
 more fi'om the axes of those above, it is evident that the pur- 
 pose was to insert an extra letter ; the name was therefore 
 ' Eirixa-pi'VO'; or '^irc^apiBe;. Line 7 follows the arrangement of 
 line 6, except that the last two letters, having plenty of room, 
 are spaced farther apart. 'ETTLxapi. . . ? ' Kfj-i^npoiredev is the 
 fifth and last epistates ; next must follow the regular formula,^ 
 which closed the prescript and announced that receipts were to 
 be entered next, tovtok Xe/J^fiara to iviavro tovto rdSe, and the 
 last four letters of this are preserved in line 7. Between the 
 beginning of EP . XA . . . . and the beginning of TAAE, which 
 
 1 Four letter spaces at 24 mm. occupy 0.096 m., and the last half letter 0.007 
 m., total, 0.103 m. ; and 0.161 - 0.103 = 0.058 in. 
 
 2 In this restoration I follow Kirchhoff ; cf. the Parthenon accounts, A. J. A. 
 1013, p. 59. 
 
382 WILLIAM BELL DINSMOOR 
 
 comes exactly below it, we must expect 50 letters, 22 for the 
 name and deme of the fifth epistates, and 28 for the closing 
 formula (omitting the rdSe). But since these include one 
 extra letter, the ordinary stoiehedon arrangement must have 
 had 49 letters. In these the average spacing is 24 mm., so 
 that with 48 spacings (1.152 m.), and adding the two half let- 
 ters at the ends (0.013 m.), we obtain the total .width of the 
 prescript as about 1.165 m. To this must be added the sur- 
 plus of 0.058 m. at the right edge of the stele, giving a mini- 
 mum of 1.223 m. for the total width of the stone ; we may 
 assume, therefore, that the width was about 1.23 m. 
 
 We may be certain that the first annual prescript extended 
 over the two double columns, as on the stele with the report of 
 the Erechtheum Commission in 409/8 (Z. Gr. I, 322) ; such an 
 arrangement accounts for the appearance of the first annual 
 prescript on fragments C -f- D of the second double column. 
 Another important result is the discovery of the number of 
 letters that must be restored, at the left of those preserved in 
 the prescript on C + D. In 11.4 and 5, in which the first pre- 
 served letters are both I, we read 22 letters to the right edge ; 
 the total number of letters in a regular line being 49, we must 
 restore 27 at the left of the I. In line 2, where the first pre- 
 served letter A is two spaces to the right of the I in 11. 4 and 5, 
 we must restore therefore 29 letters at the left, instead of the 
 7 given by the restoration [eTria-Tar] A I. These new considera- 
 tions lead to the following restoration : 
 
 [®eot : Anemia : Tv^e] 
 
 2 \^a p \ L T € K T O V M. V £ CTIK A.€Si«at ETT I 
 
 o-Ta t] A IP [pojrYUAIOEPA A §[tu.s 
 
 3[7luts as£ ■ypa/iju.aTeve 
 
 £]YS EP E[t,]OYAA EN05APX0[v to s 
 i \^A. f V a i o L n- I V t ir I T c s/8oA.es/t e i 
 n £] I O [t] A A E [s] -I P T 5 E A P A A/ [ft a T e 
 
 5 [vertTTto'Ta t a i 
 
 - : T] I M A E N [e s I k] A P I E Y § : A [- 
 
 6 [- . 
 
 ]§ EP[i]XA[pi S£s]AM<l>ITP'"[7r£e£ V 
 
 7[tOVTOIsX CJLI/X.U T u. T EVtaVTO T o 
 
 V t] T A A E 
 
ATTIC BUILDING ACCOUNTS III 383 
 
 L. 1. — No trace of a letter is now visible ; the invocation is supplied on 
 the analogy of that which heads the reverse of the stele; the latter could 
 hardly have occurred without the former (cf. Cavaignac, Etudes, p. Ixx). 
 
 L. 2. — Before the word [cTrto-TaT] Al, hitherto always restored as the first 
 word of the inscription, were 22 letter spaces. The only possible specifica- 
 tion of time, Em tcs irpdres apx^i, falls short by 5 letters. The gap must 
 be filled by words which shall have a relation to the whole, equivalent to 
 that of iiri<TTa.Tai. Now the epistatae and their secretary are all provided 
 for in the following lines ; but on the analogy of the very similar prescript 
 to the report of the Erechtheum Commission (/.(?. I, 322) the architect's 
 name also should appear. And his name may have a very prominent place, 
 as in the Erechtheum accounts of 408/7 (Ath. Milt. 1901, 22i). It so hap- 
 pens that 22 letter spaces will be filled exactly by the words 'ApxireKTov 
 Mveo-«A.es Kai. Mnesicles is mentioned by only two ancient authors, Philo- 
 chorus (in Harpocration s.v. Upo-n-vXaia Tavra) and Plutarch (Pericles, 13). 
 Philochorus mentions him in the same breath with the fact that the 
 Propylaea were begun in the archonship of Euthymenes, as if he had derived 
 the entire sentence from the prescript of this inscription. The passage in 
 Plutarch is supposed to have been derived from Ephorus, who may have 
 copied the statement from this official record (cf. Fowler, Harvard Studies, 
 1901, p. 215). 
 
 L. 2. — The building is here called the npoTrvXaioj/, perhaps in reference 
 to the fact that only the central hall was then under consideration. Three 
 years later, when the wings too must have been well under way, we find it 
 named, in the plural, to. lipoirvKaui (Col. Ill, 1. 45). And so it is in later 
 inscriptions, beginning with I.G. I, 32 B^, whether we assign this to the 
 year 435/4 (Kirchhoff, Abh. Berl. Akad. 1864, p. 8 f.; 1876, p. 21 f.) or to 
 419/8 (Cavaignac, Rev. Phil. 1900, p. 135 ; ktudes, pp. xxii, 138), and in 
 literature, beginning with Thucydides (II, 13). 
 
 L. 3. — The name of the secretary of the epistatae here occurred before 
 the mention of the archon, as Kirchhoff recognized, and so must have been 
 introduced by the word hoh, referring back to the board consisting of the 
 architect and the epistatae ; elsewhere the introductory word is hii, referring 
 to dpxis. There are 30 letter spaces before the archonship is given; sub- 
 tracting the necessary Aots and iypa/i,fidT€ve, 15 letter spaces remain to be 
 divided between the secretary's name and his demotic. Both were there- 
 fore comparatively short. The demotic ends with . . £]Y§, and might have 
 6 (four examples in the list of Attic demes), 7 (five examples), or, more 
 probably, 8 letters (twenty-one examples, in the pre-Euclidean spelling). 
 The name would have had 9, 8, or 7 letters, according to the length given 
 the demotic ; in any case, the sixth letter of the name seems to have been 
 A (Col. I, 1. 59). I therefore restore [ as e]YS. 
 
 L. 4.— I'AOevaloKTivl, restored in agreement with the prescript for the 
 last five years of the Parthenon (in the thirteenth and fifteenth years 
 
 1 As restored by Wilhelm, Sitzungsb. Wien. Akad. 1901, p. 13 ; cf. Cavaignac, 
 Etudes, p. xxi and pi. I. 
 
384 WILLIAM BELL DINSMOOR 
 
 of which, however, it was shortened to 'A^eratois because of lack of 
 
 L. 4. — [xai €7ri Tes ^oXis Aei], Kirchhoff ; the koI seems unnecessary, and 
 does not seem to have occurred in the Parthenon accounts, nor is there space 
 for it iu the present line if we accept ['A^evatottrii'] and [He] I0[t] AAE[s], 
 both of which seem necessary. 
 
 L. 4. — The name of the first secretary for 437/6 reads . . lO . AAE ., 
 restored by Kirchhoff as HcpotaSts, accepted with reservations by Ferguson 
 (The Athenian Secretaries, p. 14) ; Michaelis read . l6 . X8£[s] {A.E. 5), and 
 
 Baniiier 8£[s] {Rh. Mus. 1908, p. 430). The same name is preserved 
 
 on fragment of the Parthenon account for this year (Cavaignao, jStudes, 
 p. Ixii), again with the beginning of the name broken away. The restora- 
 tion of 1. 4 l_' AOcvaLouTw im res /ioXei hit . .] leaves only two letters for 
 the beginning of the name, which must therefore be [He] IO[t]AA E[s]. 
 This has been confirmed by Woodward's discovery of the sma)l fragment 
 of the Parthenon inscription (B.S.A. 1909-10, p. 188; cf. A.J.A. 1913, pp. 
 58, 72). 
 
 LI. 5-6. — [e-irto-TaTai] should probably be repeated here, to fit the usual 
 formula and to reduce the space for the first name to less than 24 letters. 
 There were five epistatae, as is shown by the amount of space allotted to 
 their names; there were five epistatae likewise in 434/3, but only four in 
 436/5 and 435/4. A similar irregularity appears in the first ten years of 
 the Parthenon accounts, where there were probably five epistatae in the 
 first year, six or seven in the third, three in the fifth and eighth, and four 
 in the ninth year.^ 
 
 The receipts for each year are classified and follow each other 
 almost in a fixed order. They may be divided into five depart- 
 ments : (1) the surplus from the epistatae of the previous year ; 
 , (2) grants from various boards, the treasurers of Athena, the 
 Hellenotamiae (both from the tribute of the Delian Con- 
 federacy and from the army chest), and the treasurers of 
 Hephaestus ; (3) money received from the sale of useless mate- 
 rial, tiles, wood, pinakes, hides of animals, and coloring matter ; 
 (4) regular income from 'house rental; and (5) contributions 
 from private individuals. 
 
 Fragment B, which cannot be placed in the second item 
 column of the obverse without colliding with the bottom of 
 C -f- D, apart from the fact that the receipts which it contains 
 cannot continue the expenses begun on C -(- D, must be located 
 
 1 Five seems to have been the usual number, as in the early account, /.(?. I, 
 293-296. But only three appear in the account for the statues of Athena and 
 Hephaestus {I.G. I, 318) and in the report on the Ereohtheum (/.(?. I, 322). 
 
ATTIC BUILDING ACCOUNTS III 385 
 
 in the first item column of the obverse, 11. 16-28. The small 
 unpublished fragment P makes an actual contact joint with the 
 obverse of B, so that no doubt of its attribution is possible ; it 
 contains portions of 11. 23-31 of Col. I, and shows that the 
 heading 'AvoKofiara for the first year was in 1. 29. Amono- 
 the items in the missing 11. 8-15 we may restore the usual for- 
 mulae for receipts from the treasurers of Athena and from the 
 Hellenotamiae. The remaining items are as follows : 
 
 [X <r] Y L [v 
 X § YMM[t ^ TO V 
 
 KE PAMC[t oairo T o V 
 
 0NKAOA[i p c 8 € vrov 
 
 2oONTONE[- 
 r 1 A K /A [- 
 
 X < YLO '^C"' pi 6 € V TOVTtylie] 
 
 r I N A K [vj r I [,. e] 
 
 I KIAS I IE PA[s fti-T^oo-. s] 
 
 25 rAPAnL[€ji§Ti C[-- 
 
 § K Y T N [tt] E P IT / \_€ ^a.Tov T I A £] 
 r A PA§[av]DONO §[ 
 
 [ej O E N ~ [ij \ U E 0[v T 9 
 
 LI. 16-17.— [xo-JYLOW and X§YMM(iktov) both suggested by 
 Kirohhoff ; the formula for the receipt from the Hellenotamiae, XSYMM" 
 [a)(i,Ko <fi6po], cannot be restored here because the traces in line 16 will not 
 fit. 
 
 LI. 18-21. — Evidently these go together to form a single item, in con- 
 nection with the sale of material from the old Propylon and adjacent build- 
 ings, now demolished to make way for the new structm-e. With 11. 18-19, 
 compare the sales of old roof tiles from demolished buildings in Deles (/.C. 
 XI, 144, A 21 ; I.O. XI, 287, A 22-23 ; B.C.ff. 1905, p. 489, a 7). 
 
 L. 22 A sale of old wooden beams from demolished buildings; simi- 
 larly the woodwork of a propylon at Delos was sold before making repairs 
 (I.G. XI, 161, A .38; cf. B.C.H. 1882, p. 20, 1. 157). 
 
 L. 2-3. — The sale of pinakes was continued into the next year (Column 
 II, line 9). Could they have been revetment slabs from the old Propylon, 
 or the marble metopes of the Hekatompedon ? 
 
 L. 24. — This item will be discussed in connection with Col. II, line 8 ; 
 the word //.ia-Boaii is correctly restored by Bannier (Rh. Mus. 1906, p. 221, 
 n. 5; cf. also I.G. I, 283, 15-19; 288 a, suppl. p. 145; I.G. II, 814, a, 
 
386 WILLIAM BELL DINSMOOE 
 
 A 26/30, and b, 24/25), but is rejected by Cavaignac {Etudes, p. xxx, n. 2) 
 and Woodward {B.S.A. 1909/10, p. 202). 
 
 L. 25. — A private contribution; Pleistias was probably the same as the 
 ambassador to Perdiccas of Macedonia in 426 B.C. (I.G. I, 40) and the sec- 
 retary of the senate of 435/4 {I.G. I, 37, 273). 
 
 L. 26. — An entry of unusual character ; it is the year immediately after 
 the Panathenaic festival of 438, and this is almost certainly a reference to 
 the SepfiariKov (cf. Hesychius, s.v. ctkCtos ' irav Sip/m), the money accruing 
 from the sale of the hides of the cattle slain in that festival (Harpocration, 
 s.v. Aep/mnKov) ^ ; the rather rare word TrcpiT/nt/taTa is usually found in con- 
 nection with Sip/uiTa (cf. Hesychius, s.v. KdAAea; Moeris, s.v. HcTTVKia), so 
 that the association with ctkvtov is perfect.^ 
 
 LI. 27-28. — Contributions from two private individuals, Sauron, who 
 contributed also to the Parthenon in 439/8 (A.J.A. 1913, p. 69), and 
 Timoleon. ■ 
 
 The expenses of the first year, as has been mentioned, begin 
 in line 29. The large fragment E, which has on the obverse 
 the straggling ends of the lines of an item column at the left 
 and the beginning of a money column at the right, must be 
 placed midway between the two columns, and in 11. 31-47 ; 
 the items at the left must, therefore, belong to the expenses of 
 the first year. The three fragments, M -|- N -|- 0, joined together 
 by Woodward, obviously contain expenses, the conclusion of an 
 account, and the prescript of a new year. The preservation of 
 the surface is totally unlike the granulated and badly weathered 
 reverses of B, C, D, and E. Woodward unfortunately thought 
 that they resembled the reverse of E (B.S.A. 1909-1910, 
 p. 201) ; placing them on the reverse, he could assign them 
 only to the end of the fourth year and the beginning of the 
 fifth. 3 On the obverse we must choose between the first and 
 second years or the second and third. As we shall find, the 
 beginning of an annual prescript, evidently the third, appears 
 on E, in the second column 11. 3,5-39 ; the fragments M -f N -f- 
 
 1 On the Sep/iaTtK6v, of. LG. II, 163 and 741. 
 
 2 It is used also in connection with thin silver plates in the Parthenon inven- 
 tories (/.(?. II, 695, 8-9 ; 698, II, 29-30 ; J. U.S. 1909, pp. 182, 184, U. 12-13). 
 
 " Woodward believes that this dating is confirmed by restoring, in the annual 
 prescript on I -|- J, as first secretary of the senate, KpiridSes Teiffpdtrios, who held 
 that office in 433/2 ; this would approximately fit a stoichedon arrangement. 
 But not only is the number of letters in these lines completely unknown, so that 
 such a restoration cannot be checked ; in no case in these inscriptions is the 
 demotic of the first secretary given. 
 
ATTIC BUILDING ACCOUNTS III 387 
 
 cannot be fitted to this with any sort of stoiohedon arrange- 
 ment. On the other hand, it is possible to make a perfectly 
 stoiohedon arrangement if vv6 consider that the prescript belongs 
 to the second year. What is more important at present is that 
 the phrase hoSoTrocol<; ical XiOov avaridea-i eTrl to, kvkXu, which 
 appeared so frequently in the accounts of the Parthenon (Col. 
 Ill, 61-62 ; Col. IV, 22-23, 58-59 ; Col. V, 32-34), and was 
 indeed suggested by Woodward in the case of E (I.e. p. 205), 
 proves that E and M + N + all belong together. When thus 
 fitted together, it appears that the accounts of 437/6 closed 
 in line 51, the expenses having occupied 23 lines. 
 
 [AyaXo/x, a] TA 
 30 [ove/iu. r o]N 
 
 [ -]"[-- -]^o§ 
 
 [- -- --] I § 1 E I A§ 
 
 35 l\ I 6 o T oynotsnei/TeAejOEN 
 
 ] E N 
 
 8] ) t 
 
 t] § I E I A S 
 
 40 
 
 1^ 
 
 45 H Y r I [y o . . 
 
 HYPO P A[o tsT £VTeAe0ev/lo8o]rOIO[is] 
 KAIL I 0[o sav aTt6'€0-ie7riTa]<Y KLA 
 
 KATAM E[./ 10 v] 
 MI5O0A/\A[tov] 
 50 A N E ' ~ N [- - 
 
 ~APb[S]DAA[£VT otsvcoiseirin-T.^ re n-t] 
 
 L, 45. — [^v] P [pyots] , Woodward ; another possibility is H Y P 1 [.oXlv'] , 
 as iu the Parthenon inscription (Col. II, line 105), and perhaps also in the 
 Propylaea accounts, Col. II, next to last line. 
 
 L. 47.— KAI LIO [oTo/iois], Woodward. 
 
 Ll.48-49.— KATAA/\E[vta]lMI§O0M[aTa], Woodward, in the nomina- 
 
388 WILLIAM BELL BINSMOOR 
 
 tive ; he questions the case of Kara/xtviov {B.S.A. 1909-10, p. 199, n. 2) 
 where it appears in the Parthenon inscriptions, but it should obviously be 
 in the genitive case (cf. A.J.A. 1913, p. 75," and Berl. Phil. W. 1913, 
 p. 317 2). 
 
 L. 50.— AN E[Xo(ra]M [«»'], suggested by Bannier {Berl. Phil.' W. 1913, 
 p. 318). 
 
 L. 51. — Only the first word is restored by Woodward; Bannier now 
 reads the line as I have given it, on the analogy of I.G. I, 289-296 (Berl. 
 Phil. W. 1913, p. 318). 
 
 I have observed that the annual prescript on fragments 
 N + can be restored to fit the requirements of the second 
 year; slight traces of the beginning of the receipts appear on 
 the same stones. 
 
 [eriTcsSevTepasjA PX E §[fee6 
 
 [- £ir I T eJSB OLu[sA€t 
 
 55 \j7r p T o i; € y p a jj, (J, a T £jYE:E[Tr i cr T a t a. i 
 
 [ - - - -] E [ 
 
 [-- -- TO VT]0[isA.c/AjU.aTaTO etc. 
 
 [Trap a] [^p, ir p o t € p o v etc. 
 
 [- -] /[s£ypa/A/iaT£ve 
 
 I 
 L. 54. — Woodward restores KptnaSes Tetepao-ios {B.S.A. 1909-10, 
 p. 203), to fit his supposed date of the fragment, and an assumed number 
 of 43 letters to the line ; but the nuulber of letters in these lines is unknown, 
 and the demotic of the secretary should not be given. 
 
 Another fragment, J, also from the obverse, as is shown by 
 its preservation and lettering (Cavaignac here agrees), comes 
 from the bottom of a column, as is shown by the fact that 
 4J era. of blank space still remain below the last line.^ Be- 
 cause this fragment shows that the column ended in the course 
 of a year's receipts. Column II, which must have ended the 
 account of the third year (since the fourth year began afresh 
 
 1 The use of KaTa/ierlois, in the dative case, in one portion of the Parthenon 
 account, is not a stone-cutter's mistake, but mine ; that we should read 
 Karaixevlov here also, 1 shall show in a later number of this Journal. 
 
 2 Bannier is here mistaken in supposing that Woodward intends to read the 
 two lines as one entry; the divisional mark is given, B.S.A. 1909-1910, p. 199. 
 
 8 This fact was not observed by Woodward, who places it almost at the top 
 of the stele, above B. He thus accidentally arrives at the correct date, tne sec- 
 ond year. Cavaignac assigned it to the third year, for reasons given below. 
 
ATTIC BUILDING ACCOUNTS III 389 
 
 at the top of the reverse, Column III), is out of the question. 
 We must place it at the bottom of Column I, where its receipts 
 will continue those on fragment 0. The receipts, from the 
 treasurers of Athena and from the Hellenotamiae, are those 
 which generally follow directly after the entry of the surplus 
 (as in the Parthenon account. Col. I, 76-81 ; Col. Ill, 47-48 ; 
 Col. V, 16-20 ; and elsewhere in the Propylaea, Col. Ill, 11. 
 9-13) ; but in some cases other receipts are entered first (as in 
 the Parthenon, Col. I, 109-114; and elsewhere in the Propylaea, 
 Col. II). Rather than make J follow directly, we may inter- 
 pose some such entry as Trapa heWevoTafiiov cnro a-Tparia'; (cf. 
 Coh II; Col. Ill, 16). J should be placed in 11. 61-65 of 
 Column I, or perhaps a little lower. In any case, it was contin- 
 ued directly by the receipts in 11. 8-9 of Column II. 
 
 [irapa] 0[ftirp o t e p o v eTricTTaTovhoi's'J 
 
 [- ]/[sey/oa/x,/u.aTeve eus] 
 
 60 [- 
 
 [jTupa T a /J. I o V A]OIT[a T£s8co(.TafjL,i.£v 
 
 O 1/ 1 
 
 ['iots eJAPAM[/iiaTfii£ 
 
 [irapa h e XA.eJ^ OTAM[t ov/lot s- 
 [eypa /t /iut]E Y EA I / \^cr o v € v STo;(crii/x,p. 
 
 a )^ t K o] 
 \^<j) o p o p. vaa 7rJ0T0T/[Xu,VTo] 
 
 n8[otviJA SHIE P A §AA\ *=[i9o(7is] 
 
 9[,r.va]K ONTI M E 
 
 Sums of money are preserved at the left of the last two items : 
 II, 8 lAAAI-h 
 
 9 rhic 
 
 L. 59. — [ ] \ [s], in which the five spaces at the beginning are given 
 
 by the Trapa of the line above, is the name of the secretary for the epistatae 
 of the first year, as was explained in the comment on line 3. 
 
 LI. 61-65. — These lines are repeated according to Kirehhoff's restoration, 
 I.G. I, p. 222, except that hots is moved to the beginning of line 62, in 
 line 6-3 I read [s , and in line 64 / — x- 
 
 L. 64. — The only demotics beginning with A I being AlyiXiem, Ai^oAiSes, 
 and Alx(Tov€ik (all of nine letters), the additional trace / makes the resto- 
 
390 WILLIAM BELL BINSMOOB 
 
 ration Alxiroveik certain. Cavaignac (Etudes, pp. Ixx-Lxxi) attributes J 
 to the third year because, he says, the demotic of the secretary of Helleno- 
 tamiae (beginning with Al) is not the same as that for the second year, 
 436/5, as given by I. G. 1, 244. This is not a valid reason, however, because the 
 demotic of the secretai-y of that year, <)> I LEiratpcK 0€o8] E KTO[s] as restored 
 by Dittenberger (5^Z7.i No. 17), or <t>l LEC/uovtSes Eip] EKTO[s] (Kirchhoff, 
 Abh. Berl. Akad. 1870, p. 108) is lost; Cavaignac must have attempted to 
 restore EKTO as a demotic. On the contrary, the trace on J fits the 
 nine spaces for the demotic in line 4 of /. G. I, 244, so that if the latter is 
 correctly dated in the Corpus, we should read the name of the secretary in 
 
 II. 63-64 of the Propylaea inscription as ^iXiraipos or ^iXeixoviSei Aixa-ovcvs 
 (cf. A.J.A. 1913, pp. 68, 69 n. 2, 79). 
 
 Col. II, line 8. — [xop]A§HIEPA5MI*[(9os], Cavaignac (Bf«rfes,p.lxx); 
 to this Woodward rightly objects (B.S.A. 1909-10, p. 202, n. 1-2). For 
 fi.ta-6o(ns, compare Col. I, line 24. I deferred until this moment the consid- 
 eration of the size of the annual rental. This is probably the rental of one 
 of the houses from which Athena derived part of her income,^ the amount 
 being transferred in this case by her treasurers to the account of the epis- 
 tatae. The Delian treasury received only 297 drachmae for the yearly rental 
 of several houses (7. G. II, 814 a, A 30 ; b, 25), so that a larger sum than the 
 132 drachmae, obtained by completing the H of which one hasta is visible 
 in line 8 of our inscription, would be surprising for a single house, besides 
 extending beyond the left edge of the money column. '^ The rate was proba- 
 bly about 8 per cent ; ^ the valuation of the house was probably about 1 650 
 dr., which, if not in connection with a large estate, would imply a house 
 of the better class.* This rental of 132 dr. probably appeared in each 
 annual account of the Propylaea. 
 
 L. 9. — Cavaignac supposed that the space was not great enough for the 
 four missing letters of [Trtva] KON unless the I was "en surcharge." 
 
 The expenses of the second year begin in line 10 of Column 
 II; and on fragment E, just above the five letters which form 
 the beginnings of the five lines of the third annual prescript, we 
 observe in line 33 the traces of / ( = A/\) of the surplus. The 
 fracture of L (J. &. I, 555, 1) fits the bottom of the obverse of 
 
 1 Compare the lists of ok/oi iepal and of olK'/niara belonging to the Delian 
 Apollo (I.G. II, 814, 817), 
 
 2 Kirchhoff and all later commentators prefer to make it larger. 
 
 8 A workshop, dwelling house, and another building in the Peiraeus, repre- 
 senting a total value of 700 dr. , were rented for 54 dr. yearly, or 7f per cent (I. G. 
 
 III, 1058 ; Frankel, Hermes, 1888, pp. 814-318) ; a house in Melite worth 8000 
 dr., together with a house in Eleusis worth 500 dr., brought in as rental 300 dr. or 
 8f per cent (Isaeus, Uepl toO ' Ayvlov Kkiipov, 42). 
 
 * Compare the prices, for instance. In the inscription of Tenos: Dareste, 
 Haussoullier, and Eeinach, Inscr. juridiques, I, pp. 64-87. 
 
ATTIC BUILDING ACCOUNTS III 291 
 
 C, and completes the sums of money in 11. 11-13. Fragment 
 S likewise belongs to the obverse, according to the preservation 
 of the surface; with it must be considered the missing H, which 
 joined it accurately. The items are clearly expenses; the spac- 
 ing of the lines is slightly too great for the similar items of 
 stonework on E, Column I; we must therefore place them in 
 Column II, among the expenses of the second or third year. 
 A more definite attribution seems to be impossible; at present 
 I include them in the second year's account. The tiny frag- 
 ment Q seems, according to its surfaces, to belong to the obverse 
 of the stele. The items which it contains, fjua-dofjidTov, did not 
 occur at the beginning of the expenses of the first, second,^ or 
 fifth year, but did occur at the end of the expenses of the 
 first and third years; probably the end is the natural place 
 for it.^ With the end of the first year's account occupied by 
 M + N + 0, and that of the third year apparently by A, we 
 should place Q near the end of the account of the second year. 
 
 10 [a V u, X o] M A T A 
 
 X XX XpHlpAhhl [ov £]MA"f N 
 MpaXpHAAPI-IC 
 r Xp HHHHArH 
 [ ] X p H H H H 
 
 15 [ - ] VVV 
 
 # * * 
 
 I 
 L I C 0[ 
 
 L I C T [yu. O t s - 
 
 LI0OTOMO l[s 
 LI '^ OTOMO l[s- 
 
 [- 
 
 [\i 6l]A AO A[i as- 
 
 # # * 
 
 [k u T u,] \ I [v I O v] 
 
 [^i ij > O A^ A T[o v] 
 
 # * * 
 
 33/ [M irapeSo;iiei/ etc. 
 
 1 At the beginning of the expenses of the second year, Col. II, line 11, could 
 be'read [/iio-eoMa] "'[oi'] only by an impossible crowding of the letters ; Cavaignac 
 seems to be mistaken in saying that the second N\ of luaSoixirov is still 
 distinct. 
 
 2 The Parthenon accounts, however,' give this item twice at the beginning 
 and twice at the end. 
 
392 WILLIAM BELL DINSMOOR 
 
 L. 12. — |>to-6o]M [a] ~ [ov], Cavaignac Q.c. p. Ixx). 
 
 L. 13. — The first figure is probably p3 rather than P, as it is usually read. 
 
 L. 33. — The amount is repeated in line 40, from which this is restored. 
 
 The accounts of the third year begiri in line 35 of Column II; 
 here are letters which can only be the beginnings of the five 
 lines of the prescript. Restoring 11. 35-36 according to tlie 
 usual formulae, it appears that the O at the beginning of line 
 
 36 was almost certainly the first letter of the demotic of the 
 secretary of the epistatae ; and from Col. Ill, line 8, we learn 
 that the demotic of the secretary for the third year was ®opir 
 Kio<i, a slight addition to our evidence. Fragment R, containing 
 a few letters of an annual prescript, must likewise be assigned 
 to the third year, the only vacancy on the obverse. Its letters, 
 moreover, are so arranged that they fit the restoration of lines 
 
 37 and 39. 
 
 E^TT tresTpiresa p ^ e <s h e lETTixXcseypa/x 
 
 /x a T e V t] 
 0[u ptKioseiriTe s /3 o \ € s/tct — 
 
 TT (0 O T O S 
 
 E /[pa/x/AaT€t)e:e7r tJSTAl[at — 
 
 A I [- - - J E Y § B L [ 
 
 A ^[a- tos:T o]' TO IS L[£/i/*aTaTO£Vtau 
 
 TOrnVT O TaS £ 
 
 LI. 35-36. — 'E7ri(cA.£s ©optKios is named in Col. HI, line 8. 
 
 L. 38. — Here appear the names and demotics of the second and third 
 epistatae (the first appeared in line 37), and the name of the fourth (his 
 demotic is given in the following line). The name of the third epistates is 
 perhaps B L [ettopos] . 
 
 L. 39. — The demotic of the fourth and last epistates, beginning with AN, 
 can be only ' AvayvpdcrLos, Ava(j>Xv(TTios, or 'AvaKat£vs ; either of the two 
 former, 11 letters in length, would fit tlie space exactly. 
 
 The receipts begin in line 40 ; traces are preserved on both 
 fragments E and R. The two fragments F + I, united by Ban- 
 nier and assigned to our stele because of the formulae and the 
 rough surface of the right edge (like that on the top of C + D), 
 belong to the obverse, as is shown by the surface and the size 
 and spacing of the letters.^ They come, therefore, from the 
 
 1 Cavaignac and Woodward assign them to the reverse, and so to the fifth 
 year, for the same reason which identifies them as of the obverse, the style of 
 the writing. 
 
ATTIC BUILDING ACCOUNTS III 
 
 393 
 
 right edge of Column II, and contain receipts, so that they 
 cannot be assigned to the second year, which closes its receipts 
 in line 9. The third year is the only remaining possibility ; 
 they would then be placed below and to the right of E and R. 
 
 b 1 
 
 b ' 
 
 4. 
 
 W w 
 
 < o 
 
 d < 
 
 W 
 
 o 
 
 w > 
 
 < O uj 
 
 UJ L < 
 
 - § < <: 
 
 5 W Z Q. 
 
 O llI O < 
 
 - I 1^ 
 
 W ^ O 
 
 -- <£ Lu 
 
 I — I I — I 
 ■ d I 
 
 -si 
 d 
 
 w ~ 
 
 O O 
 
 LU O 
 
 ^ w 
 
 _i — 
 
 -e- o 
 
 o X 
 
 c O 
 
 >■ , 
 
 w 
 
 X I 
 
 O o ^ 
 H H a. 
 
 w 2: < 
 > < 1^ 
 
 u 
 
 \ 
 
 z < 
 
 Q- 1- 
 
 < C 
 
 I — I 
 
 I — I 
 
 < a 
 
 
 d = 
 
 o 
 d-d 
 
 «- ;>< Q. 
 >- d a 
 
 In restoring fragments F + 1, I have usually followed Bannier's readings 
 (Ath. Mitt. 1902, p. 303). 
 
394 WILLIAM BELL DINSMOOR 
 
 [A€k]A§TE§MEA, Bannier (I.e.). 
 ['Av8/)OKA.es'Ay]ASIPP0<t>U[u£u's], Rangab6 (Ant. hell. I, p. 164). 
 
 The remaining portion of Column II was occupied by the 
 expenses of the third year. Of these no fragment can be 
 placed with absolute certainty, but their conclusion is probably 
 to be read on fragment A (J. G. I, 116 i), which is certainly 
 from the end of a column, with 13 cm. of blank space now 
 remaining below the last line. The surface is not so perfectly 
 preserved as on most of the obverse, and there is a possibility 
 that the fragment belongs to the reverse. I give it here, how- 
 ever, as of the obverse. 
 
 [- ]rOLINANEYI[- 
 
 [7rap£8oyU,]_NTO I §N EO l[s£7rtOrTaT£(Tl] 
 
 Fragment G, with numbers only, must be assigned because 
 of its surface to the obverse, and because of the amount of 
 space at the left of the sums, to Column II. Here it can be 
 referred only to an expense account, for the spacing of the en- 
 tries does not fit that of the receipts of the third year ; it re- 
 minds one of the two-line expense items which come near the 
 end of a year's account, as, for instance, in M. The grain of 
 the top of the stone would seem to associate it with the mica 
 veins at the bottom of C, so that it probably contains expenses 
 of the third year, rather than of the second. Another frag- 
 ment with numbers only, K, seems to record expenses rather 
 than receipts, since even obols are mentioned. But whether it 
 comes from the first, second, or third years can hardly be 
 determined. It seems to me unnecessary to reprint these frag- 
 ments, and I can add nothing to Kirchhoff's comment in the 
 Corpus. 
 
 On turning to the reverse, we find that C -|- D, which formed 
 the right-hand upper corner of the obverse, here belong to the 
 left-hand upper corner. The text of these two combined 
 fragments begins with the prescript of the fourth year, writ- 
 ten in smaller letters which do not extend across the entire 
 width of the stele, as at the top of the obverse, but only across 
 Column III. 
 
ATTIC BUILDING ACCOUNTS III 
 
 395 
 
 
 Ul 
 
 lu -^^ 
 
 i. 
 
 1 1 
 
 
 b 
 
 
 -1 
 
 4 '" 
 
 
 < 
 
 p h- 
 
 ci 
 
 <- 
 
 
 o 
 1 — 1 
 
 
 S. «5 
 
 
 O 
 
 UJ 
 
 Q. 
 
 w 
 
 
 5 
 
 
 ■u a 
 1 1 
 
 
 — 
 
 5 
 
 ?- 
 
 O 
 
 
 — 
 
 
 -1 f 
 
 
 < 
 
 I- — 
 
 w 
 
 \ 
 
 
 1 
 
 
 W o 
 
 
 __ 
 
 d LU 
 
 (/I 
 
 1 — 1 
 
 
 1 — 1 
 
 
 ~ t- 
 
 1 — 1 
 X 
 
 UJ 
 
 X 
 
 
 O 
 
 p 
 
 
 1 1 
 
 -J 
 
 UJ 
 
 O s 
 
 p 
 
 w 
 
 =^UJ 
 
 o 
 
 b 
 1 1 
 
 — 
 
 \ 
 
 ^ 
 
 >- — 
 
 h- 
 
 I 1 
 
 UJ 
 
 a -i 
 
 Q. 
 
 — 
 
 r 
 
 1 1 
 
 1 1 
 
 1 — 1 
 
 O z 
 
 <: 
 
 X 
 
 Q- O 
 
 C 
 
 Q. 
 
 1 — 1 
 
 <_> 
 
 o 
 
 H UJ 
 
 — 
 
 0. 
 
 -• m 
 
 1 — 1 
 
 < 
 
 s 
 
 ^ 
 
 
 w O 
 
 <: 
 
 <: 
 
 1 — 1 
 
 '" w 
 
 w 
 
 — 
 
 w 
 
 w 
 
 t- 
 
 1 — 1 
 
 z 
 
 w 
 
 w UJ 
 
 5. 
 
 < 
 
 h 
 
 >- 
 
 a 
 
 ui a 
 
 LU 
 
 UJ 
 
 ^ 1- 
 
 tj 
 
 1- 
 
 - 
 
 h- 
 
 Q. 
 
 — t- 
 
 o 
 
 H 
 
 s» — 
 
 ^- 
 
 < 
 
 ■< 
 
 ^ 
 
 i- 
 
 < a 
 
 1 — i 
 < 
 
 CL 
 
 c 
 
 
 1- 
 
 
 — 
 
 
 Q. 
 
 - 
 
 < 
 
 UJ 
 
 
 W 
 
 
 < 
 
 
 O 
 
 o 
 
 1 — 1 
 
 1 — 1 
 
 1 
 
 
 1 — 1 
 
 
 w 
 
 
 ® 
 
 V 
 
 w 
 
 
 
 1= 
 
 
 1 — 1 
 
 
 
 © 
 1 — 1 
 
 ^ 
 1 — 1 
 
 
 
 p 
 
 a 
 .5 
 
 
 1 1 
 
 
 1 1 
 
 L. 1. ^This invocation, restored also by Bannier (Berl. Phil. W. 1911, 
 p. 853), is similar to those which headed the accounts of the statue of 
 Athena Parthenos (^I.G. I, 298) or the Parthenon inventory for 422/1 (/.(?. 
 1, 170; Wilhelm, Hermes, 1901, p. 449). 
 
 L. 4 — A P I [o-T . . . ] OSM [eXtreus] , Kirchhoff ; a trace of -- as the eighth 
 letter indicates that the name must be 'Apio-TvAAos. 
 
 L.5. — ['^]EM( )E[. ., Kirchhoff; I restore TIM[6o-r/3aTos] 
 
 KE[ ] to fit the space. 
 
 The receipts of this year, so far as they are preserved oa 
 fragments C + D, may be repeated almost without variation 
 from Kirchhoff's restoration. 
 
396 
 
 WILLIAM BELL DINSMOOB 
 
 u> 
 
 
 & 
 
 t/» 
 
 t/l 
 
 >> 
 
 
 ' 
 
 
 w 
 
 1 1 
 
 C/i 
 
 1 
 
 - 
 
 
 o 
 
 P 
 
 o 
 
 s 
 
 
 ■ 
 
 
 ^ 
 
 Jj 
 
 (-» 
 
 o 
 
 
 s 
 
 w 
 
 « 
 
 i 
 
 
 1 — 1 
 
 1 — 1 
 1 1 
 
 
 o 
 
 z 
 
 rfS 
 
 
 lU 
 
 Q. 
 
 1 — 1 
 
 1 1 
 
 
 1- 
 
 o 
 
 1- 
 
 UJ 
 
 UJ 
 
 ^ 
 
 1 — 1 
 
 - 
 
 ^ 
 
 o 
 
 ^ 
 
 
 r — 1 
 
 Q. 
 
 w 
 
 _i 
 
 O 
 
 o 
 
 o 
 
 a. 
 
 1 — r 
 
 C 
 
 >- 
 
 
 lIj 
 
 1 — ) 
 1 1 
 
 < 
 
 >- 
 
 w 
 
 1 1 
 
 lIj 
 
 ^ 
 
 ^ 
 
 O 
 
 w 
 
 
 _i 
 
 ~~ 
 
 1- 
 
 0. 
 
 o 
 
 
 is: 
 
 h- 
 
 < 
 
 Q_ 
 
 X 
 
 
 o 
 
 UJ 
 
 < 
 
 < 
 
 < 
 
 — 
 
 UJ 
 
 _i 
 
 
 O 
 
 O 
 
 c 
 
 h- 
 
 CL 
 
 < 
 
 — 
 
 H 
 
 Q. 
 
 O 
 
 UJ 
 
 w 
 
 1- 
 
 H 
 
 <r 
 
 z 
 
 \- 
 
 o 
 
 1- 
 
 W 
 
 O 
 
 UJ 
 
 1 — 1 
 
 p 
 
 — 
 
 w 
 
 z 
 
 o 
 
 UJ 
 
 w 
 
 c_ 
 
 < 
 
 — 
 
 
 
 < — ) 
 
 
 
 
 
 
 
 c 
 
 1 
 
 
 o 
 
 
 
 UJ 
 
 O 
 
 >- 
 
 < 
 
 ^ 
 
 I^ 
 
 o 
 
 
 
 r 
 
 
 
 
 1 — 1 
 
 i — 1 
 
 
 
 
 
 
 d 
 
 T? 
 
 hi 
 
 w 
 
 1 
 
 .a; 
 
 V 
 
 -1 
 
 — 
 
 § 
 
 C- 
 
 X 
 
 u 
 
 r"~l 
 
 1 — 1 
 
 f— -) 
 
 
 
 1 — 1 
 
 
 
 r—i 
 
 
 
 
 P 
 
 w 
 
 ? 
 
 ;^ 
 
 s. 
 
 a 
 
 1 — 
 
 O 
 
 d 
 
 V 
 
 1 
 
 ;;k 
 
 
 
 
 
 
 
 1 — 1 
 
 1 — 1 
 
 
 
 
 
 w 
 
 l- 
 
 4. 
 
 o 
 
 d 
 
 t- 
 
 b 
 
 
 ^ 
 
 
 
 
 
 
 
 
 1 — 1 
 
 1 I 
 
 
 
 
 
 1 
 
 
 t- 
 
 d 
 
 s. 
 
 1 — 1 
 
 - 
 
 O 
 
 .-', 
 
 . 
 
 o 
 
 
 
 Q. 
 
 d 
 
 t- 
 
 1 — 1 
 
 ^ 
 
 UJ 
 
 H 
 
 < 
 
 1 
 1 1 
 
 s. 
 
 
 
 V 
 
 i. 
 
 iJli 
 
 Q- 
 
 < 
 
 is: 
 
 O 
 
 ■e- 
 
 c 
 
 
 
 
 l- 
 
 a. 
 1 — 1 
 
 o 
 
 < 
 
 H 
 
 UJ 
 
 c 
 
 UJ 
 
 H 
 
 d 
 
 
 
 o 
 1 1 
 
 < 
 
 X 
 
 UJ 
 
 O 
 
 > 
 
 < 
 
 X 
 
 1 — 1 
 1 1 
 
 i- 
 o 
 
 
 
 Q. 
 
 CL 
 
 z 
 
 w 
 
 Z 
 
 UJ 
 
 < 
 
 z 
 
 c 
 
 (* 
 
 
 
 c_ 
 
 < 
 
 O 
 
 UJ 
 
 UJ 
 
 1- 
 
 > 
 
 o 
 
 I — 1 
 
 w 
 
 
 
 
 UJ 
 
 
 
 1- 
 
 _s 
 
 
 ^ 
 
 
 
 
 ^ 
 
 
 
 ^ 
 
 
 
 
 
 < 
 
 
 1 — 1 
 
 
 
 
 
 
 W 
 
 § 
 
 < 
 
 -i 
 
 
 o 
 
 i. 
 
 
 ■< 
 
 
 
 O 
 
 
 
 
 
 1 — 1 
 
 
 
 
 
 
 
 1— 
 
 LlJ 
 
 < 
 
 Q. 
 
 UJ 
 
 Q. 
 
 s 
 
 
 w 
 
 
 
 
 _i 
 
 f- 
 
 is: 
 
 — 
 
 s. 
 
 O 
 
 t- 
 
 
 >«S 
 
 
 
 < 
 
 ^ 
 
 < 
 
 w 
 
 < 
 
 a 
 
 -©• 
 
 s 
 
 
 d 
 
 
 
 Q. 
 
 — 
 
 Q. 
 
 — 
 
 a. 
 
 a. 
 
 o 
 
 Q. 
 
 
 o 
 
 
 
 < 
 
 c_ 
 
 < 
 
 o 
 
 < 
 
 ^ 
 
 :« 
 
 S 
 
 
 d 
 
 
 
 C 
 
 UJ 
 
 C_ 
 
 X 
 
 T 
 
 u 
 
 •a 
 
 t= 
 
 
 t= 
 
 
 
 At the left of the entry in line 7 is the sura -] HHHAPhl-l-l-. 
 
 LI. 11-12. — rP0T6A[ei/es]KE[<^i(ri£iJs], Kirchhoff; but Cavaignao 
 {jStvdes, p. Ixxi) pointed out that instead of the supposed A a vertical hasta 
 was visible; I verified this, and found that the letter was clearly N. The 
 only possible Attic name of the proper length is P POTON[tKos]. Again, 
 in the demotic, there are too few spaces for K E[(^i(ri£]YS ; we must read 
 
ATTIC BUILBTNQ ACCOUNTS III 397 
 
 K E[po/*e] Y^- It so happens that the only lipuyroviKiK known in Attica is 
 €K Kepa/AcW, on sepulchral inscriptions of about 400 B.C. {I.G. II, 1235, 
 1238) ; but he is not necessarily the same as our secretary. 
 
 L.16. — dor AOr'O^TPATIASP, Kirchhoff; the last letter, H , is cer- 
 tainly to be read Tl, and I can only suggest that the somewhat inappropriate 
 word Tifi.(. was written here by mistake ; it does not appear in the similar 
 entry in Column II. 
 
 A few traces of the expenses of the fourth year appear on 
 the back of fragment E (/. Gr. I, 315 c) ; they show that the 
 account ended in line 46. 
 
 45 
 
 nejNTEL E$l 
 
 KaTa/u,£vtov] 
 
 -- -- n]POPYLAIA 
 
 TrapeSofjLevT otsveotseiri n-]TATE S I 
 
 L. 41. — Xt^oTo/iois nejNTELESI, Woodward (I.e. p. 203); the space 
 is, however, too great for this. 
 
 LI. 45-46. -. — Woodward believes that this is a final statement of the 
 financial position at the close of the work, and so would assign it to the 
 fifth year. 
 
 It might be assumed that the prescript of the fifth year and 
 the receipts are continued in Column III, but for the fact that 
 the receipts appear also in 11. 19-26 of Column IV, on the back 
 of fragment B (J. Gr. I, 312). From the obverse we learned that 
 one year's accounts normally occupied two-thirds of a column ; 
 the receipts at the top of Column IV could not be those of an 
 additional sixth year (throwing aside the testimony of Helio- 
 dorus), unless we suppose that the fifth year was squeezed into 
 the lowest third of Column III, which seems impossible.* On 
 the other hand, agreeing that the receipts on the reverse of B 
 (Column IV) are those of the fifth year, they would have been 
 overwhelmingly numerous if they had begun in Column III. 
 It seems more probable that at the end of the year 433/2, the 
 
 ^ The accounts of the fourth year occupy about as much of Column III (11. 
 1-46) as those of the first, year cover in Column I (11. 1-51), i.e. about two- 
 thirds of a column. 
 
398 WILLIAM BELL DINSMOOK 
 
 secretary to the epistatae, finding that the work was now 
 ended because of the transfer of the energies of the state to 
 financing the operations which began in the winter of 433/2 
 against Potidaea, and that economy of space on the stele was no 
 longer necessary, began anew with Column IV for the accounts 
 of the fifth year. There was a similar extravagance in spacing 
 the account of the last year of the Parthenon, the same year 
 433/2 (^A.J.A. 1913, p. 76). After the receipts of the fifth 
 year, in line 27 of Column IV appears the heading ' AvaXofJ^ara, 
 and parts of these expenses are preserved in 11. 28-32. My 
 reading does not differ from that of the Corpus (/. 6r. I, 312), 
 and I shall not therefore repeat the text. The expenses of the 
 fifth year cannot have continued after line 40 ; for from line 
 41 to the bottom of the stele, a rough werkzoll (visible on E) 
 occupies the space which would have been filled by the accounts 
 of the completion of the building, if the work had not been in- 
 terrupted at that point by the outbreak of the Peloponnesian 
 War. The werlezoll here is as characteristic of the sudden ces- 
 sation of work as are the werkzoll and bosses on the pavements 
 and walls of the Propylaea themselves. 
 
 William B. D'insmooe. 
 American School op Classical Studies, Athens.