I give here a passage from the tablet of legal precedents, which will serve to explain the reason for the giving back of the slaves and the refunding of the money by the seller :

"A man who has sold slaves for money, and has kept a claim upon them; the sum received by the seller, the money, according to the contract, in its full amount, he shall give back. If children have been born, he shall pay for them at the rate of one-half (or one and a-half) of a shekel of silver."

The above text and extract will show that the translation proposed by me for the tablet containing the unknown characters (Proceedings for April, 1883, pp. 103-107, and June, 1883, pp. 152154), with the corrections afterwards made by me (Proceedings for February, 1884, pp. 102-106), is substantially correct, the slave being really sold on the condition that, if the seller thought fit to buy her back again, he could do so upon refunding the money. That published in the Proceedings for February, 1884, is also, as there pointed out, a transaction of the same class, and well illustrates the law of slave-trading in Babylonia in ancient times. All three documents explain, and are in turn explained by, the extract from the tablet of legal precedents given above.

The tablet S.+, 431 differs from almost all other trade-documents in beginning, instead of ending, with the date. containing a declaration by a witness that the really been paid, and in the expression instead of the more usual

It is also peculiar in money refunded had Eina manza-zu,

Mu-kin-nu for the

witnesses. We have here again (1. 3), as in the two other tablets already translated (Proceedings for June, 1883, and February, 1884), the verbal form from the root p, here given II { Q→, umarga', with the usual change from to which occurs so often in the Babylonian dialect. The meaning is, "to forfeit," or "give up." The word latani, in line 7, I conjecture to mean "wife," when speaking of a slave. In line 9 occurs the word ašibu, connected, most likely, with the Akkadian išib, "incantation." This root seems to have become quite naturalized in Babylonian. line, is for îdê-su, "I know it." This word is preceded by anaku, "I," because the first and third person singular masc. of idû, "to know," are the same. Innaš seems to be the third person singular masc. niphal of našu, "to raise," "take," "bring," Heb. N, (in The loss of the last radical () is rather common in words of this class—indeed, the Assyrian was as lightly pronounced as the Hebrew. The name in the last line , Niri-abiknu, is there stated to be that of

.(נשה Babylonian

Ides, in the 12th

an Elamite, and it is not therefore unlikely that the other seemingly non-Semitic names,, Razam, Razam-ubba, l> ~ —, Asbu-tatika, ✦ =

~, Asbu-metana',

Kardara', KW A», Ḥattiza', and ▼ <> { < X, Artarušu, are Elamite also. The first three characters of the second line are doubtful, but they seem to be intended for rab humadi, "chief of the humadi." If this be the case, the words are written in a most unusual way, the usual prefix for an officer, (see lines 9, II, 15, and 18) being left out, and substituted for E (see line 11). The position of the word also is unusual. As, however, most of the people engaged in the contract were certainly foreigners, it is likely that not only the expression, but the unusual

In W.A.I. III, plate 59, line 40 (gloss), occurs the interesting form ME, it-ta-na-as-i, for it-ta-na-si, where the use of the closed syllable before the i is intended to indicate the breathing which represented the T.

form of the drawing-up of the whole, arises from the fact that it was either written or dictated by one of these strangers. Humadi, if it be the title of Razam-ubba, is probably, therefore, a foreign word.*

In the extract from the table of legal precedents the following words are interesting:

The form ameluttu (=amēlūtu) is an abstract from amēlu "man" (fem. ameltu, "woman," occurring also in this tablet). It is also expressed on the contract-tablets (see Proceedings for April, 1883, p. 107) with the ideograph and phonetic complements (E. ame-lut-tum, or, amelu-tum). Abkatu comes probably from the common Babylonian word âbāku, “to take," "to acquire." The forms nadinanu, "giver," and maḥiranu, "receiver," are also very noteworthy. It may here be remarked, that the Babylonian verb nadānu, from which come the forms iddinu and inamdin, has probably nothing to do with the Hebrew, as there was in Babylonian, much used at Sepharvaim, the verb natānu, having the same meaning. The third pers. sing. aor. Kal of natānu is ittan.

The following Communication has been received from Professor Sayce:

THE KYPRIOTE GRAFFITI OF ABYDOS.

Since the publication of my Paper on the Kypriote inscriptions I copied at Abydos, in the last number of the Proceedings, I have received notes upon them from Dr. Deecke and Mr. Pierides, which, as coming from such experienced decipherers of this class of inscriptions, are particularly valuable. Dr. Deecke, indeed, has made a discovery which is of high importance for the future interpretation of the Kypriote texts: he has pointed out that among the characters I

* It may be, however, that the four initial wedges, which are rather faint and uncertain, are not to be read, in which case we should have the determinative prefix instead of. It is possible also that the two following characters are intended for ("to ride "), and if so, we should get the group , which may be transcribed amēl šadadi, “man of a chariot,"

These are evidently archaic forms, used in a legal sense. Upon the same model is formed the common word pakiranu, "claimant," from the same root as pakaru, "claim." The same form occurs in line 16 of the tablet S. +, 431 above, in the expression, Bit-kaṣiranu.

copied there are some which denote closed syllables, i.e., syllables beginning and ending with a consonant.

Like M. Six, Dr. Deecke and Mr. Pierides have noticed that the character I inadvertently read ma in No. III is really ku, the name being Κυπραγόρω.

In IV they propose to read Móyans instead of Móĝavıs, and Dr. Deecke compares the termination with that in -ávos (-avías) as in Phanis by the side of Phanios and Phanias.

In VI Dr. Deecke would read Maλakλeidw, considering it to be a name compounded with μάλα.

In VII he would read Μέλερμος, comparing Μελήσερμος and Meλéaypos. He interprets the name Phaukrês as meaning "Light in need” (comp. φαύ-βιος, Φώ-κριτος).

In IX he proposes Νοστ-αμαυσα(ς), suggesting the names ̓Αμευσίας and 'Apevoravías in explanation of the second part of the compound. See his note on XL.

In X he conjectures ëøv 'Hdaλieú(s). My third copy, however, marks all the characters as certain except the second se.

In XI he conjectures Ελλο-νίκιος and Ελλ-ονίνης, His conjecture is confirmed by my copy, which has Yni, though the engraver has

turned it into the usual form of te.

On XII he remarks that anions actually occurs by the side of ἀπειθής.

In XV he suggests Ιφι-ολάτωρ for Ιφι-ολέτωρ.

In XXIII he points out that the name is 'Akeσтo-μé(μ)øns, formed like αμεμφής.

In XXV he notes that the reading is no doubt Vages, since yéoua originally began with digamma.

In XXVI he compares the Kypriote name 'Api-Baos, and adds that perhaps Orekles is for 'Ορεσ-κλῆς.

In XXVII he regards the names as compounded with vévvotos, the superlative of νέννος, i.ε., νάννος, the second name being Νενιστοβάλην.

In XXXIV he reads 'O(μ)po-kλéfns, and in XLII Thyrsiya(s). His important discovery is in XL, where he shows that must be ros, the words being Πέτρος and θυρα ωρός. I may add that the first letter of the first line is engraver has made it the ordinary po.

in my copy, though the For he proposes the

value of nos, and identifies the character with | in No. IX, the names being 'Απυγνώμονος or ̓Απυμνάμονος and Nostamausas. In this has the value of gno or mna. He also regards hin No. XX and No. XXV as new characters.

case
No. XVII,

Mr. Pierides has suggested a few conjectural readings, through a misconception of the notes I have attached to my copies. It is only where I mark a character as doubtful that I consider it to be so; where I mention that it is clear or distinct, this only means that it was unusually large and deeply cut. In all other cases the character seemed clear to me after at least three examinations on different days and in different lights; and though every individual epigraphist is necessarily fallible, the clearness of the Egyptian atmosphere, and the satisfactory state of preservation in which the graffiti are, render the chances of error unusually small.

Thus in No. I Mr. Pierides proposes to add the characters te-ke, so as to get the words 'avéŋkɛ. There was, however, no trace of any letter after the final ne, and avéŋke, “he dedicated," would be singularly inappropriate in a graffito.

In No. XIV he suggests that the second character may be read la, giving deλauíoews, "of Selamisis." It is, however, quite unlike Σελαμίσεως, any known form of la.

In No. XVIII he corrects my reading into eeidoriya.

In No. XIX he conjectures Protagoras in the genitive as the father's name.

In XX he suggests τόδε Ερέθιμος.

In XXIII he conjectures 'AKEσTÓ-Kupos for the first name.

On XXIV he says: "The first two characters in the first name may be a-u, giving 'AUTOкpérηS. I cannot help thinking that the second name is Bioaos; the third is, I believe, rightly interpreted; and the last was probably "Ayns."

In XXVI he ingeniously suggests 'OFλns, "the cattle-owner," which seems certainly right.

In XXVII he would read Nηptoтóriμos and (like M. Six and Dr. Deecke) Baλǹv.

In XXVIII he suggests ó vavπayos "the ship-builder," and it is quite possible that the two lower bars of the last character but one in the graffito may have been added to it by some later visitor to Abydos.