he possessed. Further, Bobbio possessed a number of Greek MSS., and these occupied the eighth case in his library. This brilliant conjecture casts light on the provenance not merely of famous MSS. and palimpsests containing Virgil and writings of Cicero, but also on that of ecclesiastical works, e.g., the Bobbio version of the Gospels (k), which gives us the text of the old Latin translation used by Cyprian, and the Gothic version of Ulfilas, which comes from Bobbio. The last work would naturally be connected with Cassiodorus, who is described by Mommsen as vir Gotho-Romanus.

Dr Loew, the chief authority upon Beneventan MSS., has thrown light upon the provenance of the Monte Cassino MS. (Laur. lxviii. 2), which is our sole authority for Tacitus' Annals, xi.-xvi., and Histories.1 The other works of Tacitus were preserved in Germany, viz., Annals, i.-vi., at Corvey, and the Agricola, Germania, and Dialogus at Hersfeld. Loew points out that there were two German abbots at Monte Cassino in the middle of the eleventh century, the period to which the MS. belongs, and traces German influence in various MSS. written at Monte Cassino in their time. He therefore concludes that the parent of Laur. lxviii, 2 was brought from Germany. If so, it is to Germany that we owe the preservation of all the works of Tacitus.

It may be mentioned in this connection that E. Jacobs has drawn attention2 to a previously unpublished letter of Niccolo Niccoli, written about 1430, in which he gives instructions to an agent sent to Germany in search of MSS. This contains a description of the Hersfeld MS., furnished by the monk to whom Poggio frequently refers in his letters. Jacobs points out that in this the Dialogus is not ascribed to Tacitus, and concludes that it was anonymous in the Hersfeld MS. This view is sharply attacked by Gudeman, who thinks that the omission of the author's name was due to negligence. The number of folios in the three works is given by Niccolo, and Gudeman makes use of 1 International Congress of Historical Studies, 1913. 2 Woch. Kl. Phil. 1913, p. 701. s Ibid. p. 929.

this statement in support of his view that a page of the Dialogus has been lost.

P. Lehmann has published a monograph upon Sicardus, who was a notable discoverer of MSS. in the early part of the sixteenth century. One of the chief services which he rendered was to procure from Lorsch a MS., now lost, from which Cratander printed the second book of Cicero's Letters to Brutus, which are not found in any MS. now extant. This work is a welcome supplement to the same author's previous treatise upon Fr. Modius.

It is impossible in the limits of this article to include contributions made to the criticism of separate authors. Perhaps the most laborious piece of work is that undertaken by C. H. Beeson, who has made a list of all non-Spanish MSS. of Isidorus previous to the middle of the ninth century.2 The number of these is almost incredible, as will be seen from the index, in which over 400 MSS. appear. Beeson takes the various works singly and arranges the MSS. by countries.

3

T. Thulin's dissertation upon the Latin Agrimensores deserves special mention. The subject is one which had previously been treated by Lachmann and Mommsen. The chief MS., the Arcerianus, belongs to the sixth century, and came from Bobbio. Another MS., belonging to the tenth century, contains some leaves in rustic capitals copied from a sixth-century MS. Thulin's treatment of the evidence is masterly, and his work contains much of general interest to textual critics. I would finally mention a very suggestive work by E. Norden, in which he explains certain repetitions in Cicero's speech pro Caelio, by supposing that Cicero spoke from rough notes, and that the speech, which was taken down by shorthand writers, was not revised for publication in the usual way. He also explains curious repetitions in

Brut. 307 and 312, where the teaching of Molon is twice mentioned, and Cat. iii. 24, where the same words occur at the beginning of two contiguous sentences, by the theory that a correction of Cicero, who intended to strike out one of the 'doublets,' was disregarded by a copyist.

ALBERT C. CLARK.

XII

GRAMMAR, LEXICOGRAPHY, METRIC

THE only Latin grammar which has appeared within the period of time to which this article is limited is my own New Latin Grammar.1 The book is designed for use in schools, and I should hardly think it worthy of notice in this place, were it not that I have tried to bring it into touch with the well-established results of research in various

departments of Latin grammar. It may at any rate be regarded as containing in brief compass my confession of faith on many moot points. For example, I have given (in § 400 and § 405) the result of my inquiries on the use of the dative with certain verbs compounded with a preposition.

This same problem is discussed by Emory B. Lease 2 as a matter which must be faced by every student of Latin syntax. The value of the rule that verbs compounded with prepositions take the dative was called in question in 1911 by Prof. E. W. Fay,3 and the difficulties which it involves had been felt by other grammarians before him. The result arrived at by Lease is that 'regard must be had not so much to the group of letters found at the beginning of the verb, as to the meaning of the verb in its totality.' It follows, therefore, that if the rules of the grammar do not include the particular meaning that will account for the

1 Oxford: Clarendon Press, 266 pp. ; 2s. 6d. See Class. Rev. 1913, pp. 61 ff.; Woch. für Klass. Phil. 1913, pp. 603 ff.; Berl. Phil. Woch. 1913, p. 728 f.

2 Am. Journ. Phil. xxxiii. 1912, pp. 285-300.

3 Class. Quart. v., 1911, p. 104.

101

dative, the rules must be modified' (p. 299). The difficulty is that the meaning of the dative varies with the particular preposition used in the compound, and that to describe the dative simply as an indirect object or a dative of 'disadvantage' hardly does justice to the special shades of meaning which it expresses.

A fourth edition of Brugmann's Griechische Grammatik,1 containing a good deal of new matter, has been brought out under the care of Prof. H. A. Thumb of Strassburg. Prof. L. Cohn has added an appendix on Greek lexicography. The first edition of this important work (published in 1885) contained 125 pp.; the fourth edition has grown to 772 pp. The additions of Thumb strengthen the work on the historical, as distinct from the comparative side, and make it even more than it was in its earlier forms an indispensable guide to the student of Greek.

Professor Wilamowitz-Moellendorff's Sappho und Simonides - Untersuchungen über Griechische Lyriker 2 is essentially a work of literary criticism, but it contains a section entitled Die sprachliche Form der lesbischen Lyrik (pp. 79-101) which is concerned entirely with grammar and scansion. He discusses lesbische Schrift, Ionismen, orthography, inflexion, the digamma, and accent, as well as some metrical matters to which reference will be made below. "There is no more important problem for Greek grammar than the testing of the trustworthiness of our accentuation of words. We now possess so many remains of ancient books that we see clearly how late, how rare, and how incomplete is the indication of accents, and how far it departs from that which is customary among us at the present day. No one can any longer dispute the fact that our accentuation is a product of the time of Photius. But we are bound to go behind the Byzantine period, for we are able to do so. . . . Can it be maintained that we

1 In Iwan Müller's Handbuch der Klassichen Altertumswissenschaft; Beck, Munich, 1913; 16s. 6d.

2 Berlin: Weidmann, 1913. See also pp. 109, 160.