theologians to Philip of Hesse, December 10, 1539, De Wette-Seidemann, op. cit., pp. 238 ff. This is generally supposed, as stated here by Seidemann, to have been composed by Melanchthon. Rockwell proves that it was only copied by him, with a few alterations, from a draft sent by Philip of Hesse which had been written for him by his councilor, Justus Winter. It can be readily seen that this fact is of great importance in judging the attitude of the reformers.

B. Luther to the Elector John Frederick, in Seidemann, Lauterbach's Tagebuch auf das Jahr 1538, pp. 196 ff. This is set by Seidemann in April or June, 1540, by Kolde early in June, and by Rockwell on June 10. This epistle, the writer's justification for his attitude in the bigamy of Philip, has been often overlooked notwithstanding its great importance. One passage has excited some comment: "Hette ich aber gewust, dass der Lantgraff solche notturft [i. e., his need of women] nhulengsther wol gebüsset und bussen konte an andern, als ich nhu erst erfare, an der zu Esschweg, solte mich freilich kein engel zu solchem rath gebracht haben." On this Kawerau (Köstlin-Kawerau, op. cit., Vol. II, p. 477) remarks: "Dagegen ist dunkel, wie er das schreiben könnte . . . . denn dass Philip . . . . sich ‘der Hurerei, Unkeuschheit und Ehebruchs nicht habe erwehren können' stand in der Luther übergebenen Instruction Philipps deutlich geschrieben." Rockwell tries to defend Luther by saying that it is highly probable that Luther means that, had he known that Philip satisfied his desires with other mistresses, as with her of Eschweg, he would have advised him to marry one of them instead of an uncorrupted virgin. It is doubtful how far this explanation, if admitted, exculpates Luther. C. Luther to the landgrave, July 24, 1542, De Wette-Seidemann, op. cit., p. 275, l. 13. Rockwell supplies the word lacking, probably from the original in Cassel, as "poli."

D. Luther to the landgrave, April 10, 1542. The copy of this letter in De Wette-Seidemann, op. cit., p. 312, is very faulty; for the correct text one must turn to Lenz, Nachlese &c., p. 136.

E. Luther to Friedrich von der Grüne, Burkhardt, Dr. Martin Luthers Briefwechsel, p. 403. This is set at the end of 1541 by the editor on the ground that the Weimar archives show that this was the time when Grüne was employed at Wittenberg. Köstlin (Köstlin-Kawerau, op. cit., Vol. I, p. 680, n. to p. 491) thinks that the letter should be dated at the same time as the saying found in Förstemann-Bindseil, Luthers Tischreden, Vol. IV (1848), pp. 474 ff. The same saying is found in H. Wrampelmeyer, Tagebuch über Dr. Martin Luther geführt von Dr. Conrad Cordatus (Halle, 1885), No. 671, where it is datable in the first quarter of 1532. It is prob

able that the strong resemblance between the letter and saying do not prove that they were contemporaneous, but rather that the works of which Luther complained lasted a long time. The late date of the letter is made probable also by its great asperity of tone, more characteristic of the Reformer's closing years.

III. AN ENGLISH TRANSLATION OF LUTHER'S LETTERS

Anyone who knows Luther's letters in the original must agree with Coleridge that it is a little strange that for so long no one undertook to translate them into English. Such a work is now at last before us, The Letters of Martin Luther, selected and translated by Margaret A. Currie (London: Macmillan, 1908). The selection and translation are such as to make the appearance of the work a subject of regret rather than congratulation. It would be hard to make a worse choice of five hundred letters out of the nearly six thousand extant. Not only are many of the most important events in the Reformer's life, such as the Leipzig Debate, the Peasants' Revolt, and the bigamy of Philip of Hesse, totally passed over, but many letters are inserted which never should have been included, as for example that letter to Leo X of 1519 above mentioned (Currie, p. 43) and one long ago proved a forgery. This epistle, purporting to be from the Reformer to his sister Dorothy, is inserted not once but twice under different dates (pp. 200 and 376, dated variously 1527 and 1539), although its genuineness was suspected by Seidemann in 1856 and its falsity proved by Burkhardt in 1866 (Dr. Martin Luthers Briefwechsel, p. 338).

The errors in translation are too numerous to be pointed out and too bad to be corrected. Three specimens of the translator's scholarship may serve to give a faint idea of its quality.

Eoban Hesse appears throughout as Coban Hesse (pp. xiii, xiv, 243, 347, etc.).

On p. 142 we come across "the Cardinal of Eborack." The reference is to Wolsey, cardinal archbishop of York.

On p. 322, "He invited Bugenhagen and me to breakfast, because the day before I had forbidden the sacrament (Nachtmahl) in the bath." The true translation of these words, which appear in Latin in the source from which Miss Currie says she got the letter, though she thinks proper to insert a German word to explain her version, is: "He invited me and Bugenhagen to breakfast after I had declined his invitation to dinner after the bath the previous evening." Cf. Enders, op. cit., Vol. X, p. 267. KöstlinKawerau, op. cit., Vol. II, p. 371.

PARIS, FRANCE

PRESERVED SMITH

RECENT THEOLOGICAL LITERATURE

NEW LITERATURE ON THE SEPTUAGINT1

The appearance of a second Part of the larger Cambridge Septuagint2 is of interest for several reasons. It comes soon after the first Part, and it remarks that the preparation of the notes on Numbers and Deuteronomy will not furnish so many difficulties as in this portion where it was necessary to meet the difficulties connected with the Hexaplaric text of the later chapters of Exodus. There are two small modifications, one of them of great importance, of the scheme carried out in the first Part. The symbol b now denotes the readings of the closely related MSS 108 and 19 (HolmesParsons): Where they differ, 19 is denoted by b1, 118 by b. The New Testament evidence has been more fully presented. In cases where the first hand of B (B*) stands almost alone, and one or both of the correctors' hands (known as B, B") have introduced a reading which is supported by almost all the other authorities, the editors have usually given the latter the preference in their text. They have done so because they were thus able considerably to shorten and simplify their notes. They might have done so even more frequently, though it is no part of their task to construct a true text of the LXX. For it is incredible how frequently the merest blunders of the copyist of B have been adduced as "the LXX," merely because they are printed in the text of the smaller and now also in the larger Cambridge Septuagint. The margins of Kittel's Biblia Hebraica are full of such rubbish. A few examples from Exodus may be quoted:

No MS of the LXX except B writes para instead of onueîa at 4:28, but Ryssel-Kittel quote "rà pýμara."

No MS of the LXX except B has 435 years instead of 430 in 12:40, 41 but Ryssel-Kittel state: "+ (cf. Gen. 12:4; 21:5; 25:26; 47:9) Exod. 12:41, G+, ut 40.”

1 See American Journal of Theology, XIII (July, 1909), 446-50.

The Old Testament in Greek according to the Text of Codex Vaticanus. Supplemented from Other Uncial Manuscripts, with a Critical Apparatus Containing the Variants of the Chief Ancient Authorities for the Text of the Septuagint. Edited by ALAN ENGLAND BROOKE, B.D., AND NORMAN MCLEAN, M.A. Vol. I, “The Octateuch," Part II, Exodus and Leviticus. Cambridge: The University Press, 1909. viii+155 +405 pages. 12s. 6d. net.

No MS of the LXX except B has at 32:14 περιποιῆσαι τὸν λαὸν αὐτοῦ instead of περὶ τῆς κακίας ἧς εἶπεν ποιῆσαι τὸν λαὸν αὐτοῦ; but Ryssel-Kittel deem this blunder of B worthy to be quoted, and place before every reader of this Hebrew Bible the riddle to reconcile this Greek with the Hebrew. In two of these three cases the new editors have now changed the text of Swete (=B), but not in the first. As I said, they might have done so in many more cases, and I am glad to see that this principle is approved in the important notice of the first part published by Ernst Hautsch in the Göttinger gelehrte Anzeigen, 1909, pp. 563-80. The same reviewer compared the collations of seven Vatican MSS from photographs in the possession of the Septuagint undertaking of the Royal Society of Göttingen, and found them very accurate, while those of Holmes-Parsons needed many corrections and additions. For the present Part and notice, I compared the whole of Exodus with Thompson's photographic reproduction of the Codex Alexandrinus, with these results:

7:17: The reading peraßaoiλe instead of peraßaλeî is marked in the MS as a mistake; so likewise 15:21, ávaßárηp instead of åvaßáτηv. 12:29: A has άrò пρштóкоν (haplography).

20:6: A omits the v ἐφελκυστικόν in φυλάσσουσι, and in 28:12, εἰσί. 20:8, 12, 15: The numbering of the ten commandments on the margins of A is not given (see on this strange omission, The Expository Times). 23:4: A has here the spelling xx@pov.

29:38: After τὴν ἡμέραν A has ἑπτὰ ἡμέρας ; see Swete.

33:17: A* has had rō, i. e., Tòv instead of TOUTOV.

36:34: A has, like y, the reading rò Apa instead of Toû Awμaros; see Swete.

36:39: I do not think that A has èvretuπwμéva; I believe the second letter to be K.

In the subscription it would have been well to state that the MS joins the article and the preposition with the following nouns, writing odos τῶνυἱῶν τηλ ἐξαιγύπτου.

In the apparatus at 12:10 A is quoted on the wrong side, i. e., for καTαλ—instead of for the rival reading Karaλer. Surely, if all other collations, which I cannot compare, are of the same accuracy, we have every reason to be satisfied.

As to the quotations from ancient writers, E. Hautsch has shown that they are sometimes too short and therefore liable to mislead. But this is unavoidable, where space forbids a special investigation and discussion of quotations. As examples I may discuss one from the New Testament and one from Philo:

In Exod. 32:28, in the narrative on the golden calf, the received text gives 3,000 as the number of slain Israelites. As variations B-M quote: εἴκοσι τρεῖς χιλιάδες ἀνδρῶν, r Ε Lwa (vid.): χιλιάδες τρεῖς καὶ εἴκοσιν, Cyr.-ed.: χιλιάδες εἴκοσι τρεῖς, Cyr. cod.: χιλιάδες τρεῖς, Cyr. cod. Before vs.

6 they had quoted from the New Testament, I Cor. 10:7, but here they fail to quote the sequel, vs. 8, "and fell in one day three and twenty thousand"; of course because they see in it a reference to Num. 24:9, "twenty and four thousand." Likewise they omit a quotation from Philo, V, 186 (ed. of Cohn-Wendland), réσσapes #pòs ToiîS ELKOOL. But both passages, that of Philo and that of Paul, especially the latter with its 23,000, refer also to Exod., chap. 32, like the quotation from Cyril.

But I must cut short the discussion of detail. The praises given in the first notice to printers and publishers and to the care of revising must be repeated; some misprints I have corrected in the Berliner Philologische Wochenschrift. One desideratum is that the publishers might add to the next part a book-marker with the list of symbols, like those in Kittel's Biblia Hebraica or in my Greek Testaments; at present everyone is obliged to turn to the first part, if he has forgotten the meaning of one of the symbols.

And now some gleanings from the rich apparatus. The mountain of the law, which we are now accustomed to call Sinai, but which we ought to call Sină, is spelled Eva consistently through Exodus and Leviticus only in the MS r, and in Leviticus (not in Exodus) in g. Where did this spelling originate? And does the difference in g prove the work of two different copyists? Examples of the Iota adscriptum are found in A, Exod. 28:28; 36:31. Or does the spelling wua hint at a pronunciation wia (with diaeresis)? By the way, this diaeresis ought to appear in the word poíσkos, where the ending is long; compare 36:33, poíσκwv, which might be read as a diphthong, with polokos, vs. 34. In the latter form the diaeresis is not necessary, the accent marking the pronunciation. The Hebrew shekel is spelt σirdos instead of σíkos in x everywhere in Exodus (7 times), and in 39:7 this spelling is presupposed by the misrendering columnae of the Bohairic translation. The copyist of the Old Latin introduces "Pompejus" into the law for the day of expiation, Lev. 16:8, ad dimissionem pompeio. In the same law the Greek MSS might be classified by the use of xíuapos and τράγος.

Very important is the agreement of A, not only with y, but also with b and w; compare Exod. 38:8, 11, 12, 22; 39:1; 40:7; Lev. 7:2. For this agreement shows that the MS A or its ancestor stood under the influence of Lucian; a most welcome evidence for the theory of von Soden concerning the New Testament part of this MS.