tions, whether on stone or copper, of the time of the Ganga king Śrî-Purusha (who reigned during the greater part of the eighth century), are engraved.

And here it may be permitted to leave this interesting subject. From the tenth century onwards the information. already published provides ample materials for the history of the language and literature. Down to that time the present article will I hope show that a mine of unexplored wealth awaits the researches of scholars.

ART. VI.-Was the Book of Wisdom written in Hebrew? By D. S. MARGOLIOUTH, M.A., M.R.A.S., Laudian Professor of Arabic at Oxford.

§ 1.

THAT the Proverbs of Ben-Sira were written in some kind of Hebrew has never been seriously questioned, and the number of sources in which clues to the original have been preserved would be sufficient to silence any doubts that might be raised. It would be natural to suppose that the Book of Wisdom, which bears so close a relation to those Proverbs, which enlarges on so much that Ben-Sira suggests, and endeavours to be deep where he is shallow, appealed to the same public, and was composed in the same language. But although this theory would receive a primâ facie plausibility from the Hebraisms with which the pseudo-Solomon's style abounds, his affectation of Greek eloquence, noticed by very early critics, his allusions to Greek customs, and his reminiscences of Greek authors, have seemed to put it out of court; and the best editors of this century only mention this theory to reject it. In the last century, however, it was supported by some eminent names; and early in this found an advocate in Bretschneider, the author of three dissertations on the Book of Wisdom, who had an adherent in Engelbreth. Bretschneider's arguments in support of a Hebrew original for 'Wisdom' were by no means convincing, for he produced no case in which a Hebrew word appeared to be mistranslated, but only cases where he fancied the supposed original was corrupt and required emendation; this made his restorations suffer from a compound improbability, whereas had he first obtained a solid basis for his hypothesis, the method would have been unobjectionable, as it seems in one or two cases to have been successful. Whether any one

since Engelbreth (whose contributions are of less consequence than Bretschneider's) has endeavoured to prove the proposition suggested in the title of the article, it is difficult in the vast mass of theological literature to discover; it is however unlikely that quite the same method will have been employed as that which the present writer will endeavour to follow, so that some, at least, of his observations are probably fresh. The question is of interest for the history of Jewish thought, for the history of the Hebrew language, and, of course, for the interpretation of the book itself.

If all the subsidiary sources of the text of Ben-Sira were wanting, and we had but one Greek copy, we could still discover with certainty what its original language was, by the observation that in a number of passages verses, which are obscure in the Greek, become clear when translated into Hebrew. A well-known passage is xxv. 14, 'There is no head above the head of a serpent,' words which are only intelligible when we remember that head' in Hebrew also means 'venom.' No less clear a passage is xviii. 17, 'the fool reproaches thanklessly,'. which gives a meaning when for 'reproaches' we substitute 'does a favour,' the alternative sense of the Hebrew word which signifies to reproach.' Of course one such passage would not prove very much; but a very few would create a probability; and each fresh enigma which was solved by this key would render that probability infinitely greater.

The earliest versions of the Scriptures, the Peschitto Syriac and the Vetus Latina, were made, the first of them to a great extent from the Hebrew, the second probably from a very independent recension of the Greek. If the same key which explained puzzles in the Greek also explained differences in any number between the Greek and these ancient versions, the probability that it was a true key would be greatly confirmed. For a single case might well be attributed to accident, but a number of cases could not be due to that cause.1

1 The hypothesis that the Syriac version was made from the Chaldee was started by J. M. Faber, Prolusiones de Libro Sapientiae Ornoldi, 1776.' This

6

It is very noticeable in Ben-Sira that many Greek manuscripts, and all the secondary versions, contain revised renderings, a comparison of which with the renderings given by the bulk of MSS. tells us with certainty what the original words must have been. If any of the varietas lectionis of the Book of Wisdom can be explained by the suggested key, the probability that it is a true one will become greater still.

And, lastly, if, when the key has been tried on all these locks, it has been found to open any considerable number of them, we shall be entitled to use it as a genuine instrument; to abandon the language of metaphor, where the text of the Book of Wisdom offers difficulties, we shall be justified in endeavouring before anything else to reconstruct the original; and if we cannot always find a satisfactory answer immediately, we shall nevertheless know where to look for it.

I will endeavour in the present paper to examine these different sources in the order in which I have enumerated them, and have hopes that the reader may find reasons for thinking it possible that the pseudo-Solomon wrote in NewHebrew, the language of Ben-Sira, and was a Jew, not of Alexandria, but of Palestine.

§ 2. Internal Evidence of the Greek Text.

I. xiv. 10. καὶ γὰρ τὸ πραχθὲν σὺν τῷ δράσαντι κολαστ θήσεται·

διὰ τοῦτο καὶ ἐν εἰδώλοις ἐθνῶν ἐπισκοπὴ ἔσται:

For

For that which is done shall be punished with the doer. this reason there shall be a visitation upon the idols of the Gentiles. The former of these clauses is unsatisfactory; it ought to mean 'that which is made shall be punished with the maker,' viz. the idol with its maker; this however the Greek would not allow. There is a passage in the Midrash Tanchuma

pamphlet I have not seen; it is discussed by Eichhorn, Einleitung in die Apocryph-Schriften,' p. 199. It may be presumed that Eichhorn quotes the best of Faber's arguments; he says that he quotes nearly all of them. These are all liable to the same objection as Bretschneider's (v. supra). I find only one of my own observations anticipated in that list, and this one which I have not given in § 3.

which suggests to us what the original idea here was (on Gen. xlvii. 12; ed. Warsaw, 1879, i. p. 56b). DI DU

העובד כן נפרעין מן הנעבד דכתיב ובכל אלהי מצרים אעשה שפטים וכן אתה מוצא בדניאל כיון, שפתר חלומו של . אבל דניאל לא קבל למה שכשם שנפרעין נבוכדנצר ... אבל דניאל לא מעובדי כו'ם כך נפרעין ממנה.

'As the worshipper is punished, so is the thing which he worships. For it is written "And on all the gods of Egypt I will wreak vengeance." And so you find in Daniel, that when Daniel interpreted Nebuchadnezzar's dream, he ordered sacrifice to be made to him. But Daniel would not accept it. Why not? Because just as the worshippers of idols are punished, so are the idols.' (The same passage is to be found in the Midrash Rabba.) It is possible from this that in the passage of Wisdom and have been wrongly translated doer and done, when they should have been rendered worshipper and worshipped. Of course the first is the Aramaic, the second the Hebrew sense of the words. T'worshipped,' occurs frequently in the Mishna of Aboda Zara. made is used in the Chaldee of Daniel, of the image which Nebuchadnezzar made.

This observation will further elucidate the preceding verses; 'That wood is blessed by which righteousness comes about; but that which is made by the hand is accursed, both it and its maker, because he was making it, and it being corruptible, was named God. τὸ χειροποίητον δὲ ἐπικατάρατον, αὐτὸ καὶ ὁ ποιήσας αὐτὸ, ὅτι ὁ μὲν ἠργάζετο, τὸ δὲ φθαρτὸν θεὸς ὠνομάσθη.

The wood which is blessed is the ark, or any ship; which is made by the hand quite as much as the idol; nay, more so; for some of the idols were not wrought at all, a fact with which the writer is familiar. The idol cannot therefore be differentiated from the ship by the fact that it is made by

1 In Buber's edition for 72 the words 'y are substituted. This makes the coincidence less striking, but does not seriously affect the argument. It will be assumed throughout that these Midrashim contain old materials, at whatever period they may have been drawn up.