I would call especial attention here to Henry's acute Antinomies Linguistiques belonging to the Bibliothèque de la Faculté des Lettres de Paris (1896). EDWIN W. FAY.

Julius Firmicus Maternus, der Heide und der Christ, von CLIFFORD H. MOORE. Inaugural-Dissertation. München, 1897. (Leipzig, Th. Stauffer.) Pp. 54.

Attention is called to the above-named monograph because of the importance of the result arrived at by the author for the history of later Roman literature, and also because it seems appropriate that the Journal should keep its readers informed of the work of American scholars which may appear in a form not generally accessible to American readers. The dissertation of Mr. Moore has for its main thesis the identification of the author of the Christian treatise de Errore Profanarum Religionum, with the author of the fourth-century astrological work in eight books entitled Mathesis, both of which are attributed to one Julius Firmicus Maternus. Doubt as to the identity of these two homonymes is old; for the subscription of the Mathesis reveals the addition of the word iunior, and in our own time Bursian and Halm, who have edited the de Errore Profanarum Religionum, have supported this doubt with arguments that have found general acceptance.

Bursian's reasons for attributing the works to different authors were chiefly chronological. The Christian treatise, he held, must be placed in the year 347 A. D., while the Mathesis could not be dated earlier than 354; and the assumption that the Christian author of the former had in the meantime gone back to Paganism seemed inadmissible. But the date 354 was based upon a false assumption—namely, that the anonymus in a genitura, Math. II 32, was identical with Lollianus, to whom the work is dedicated. But Mommsen had already shown that the anonymus here alluded to was rather one Albinus, consul in 335, and thus the chronological difficulties in the way of identification were at once removed. To have applied Mommsen's observation to the question of identity is the service of Mr. Moore, but of far greater importance is his study of the relationship existing between the language and the thought of the two works. Here the author has pointed out such resemblances as raise the a priori probability of identity to essential certainty. One or two examples of this kind will suffice for the purposes of illustration. The adverb artuatim is found only in the Math., the de Errore, and in a letter of Ps.-Jerome. Furthermore, the verb artuare is found only in Math. VI 31, so that it may be inferred with some probability that the adverb is a new formation of the author. A

...

further proof of identity is afforded by an allusion in the intro-
duction of the Christian work to views which are found advanced
in the Mathesis: quod in fabricatione hominis artifex effecit, ut
antea diximus
specialiter retexenda sunt. With this cf.
Math. prooem. 3, 7 and 8. The relationship is clear and striking,
and it is remarkable that it could have escaped the attention of
earlier scholars. After demonstrating that the author of the two
works was one and the same person, the author devotes the
remainder of his dissertation to an instructive survey of the
sources of Firmicus Maternus in his two works.

In thus giving a résumé of the contents of this valuable treatise I have shown, I hope, how penetrating and thoroughgoing a work of investigation it is. The author's main result is an important contribution to the history of later Roman literature, and in arriving at it he has, in addition, made most valuable independent contributions to the history of the Latin language. G. L. HENDRICKSON.

Asadi's neupersisches Wörterbuch 'Lughat-i Furs,' nach der einzigen vaticanischen Handschrift, herausgegeben von PAUL HORN. Berlin, 1897. (Abhdgn. d. k. Ges. d. Wiss. zu Göttingen, Phil.-Hist. Kl. N. F. Band I, 8.) Pp. 1-37+ 1-133.

An important contribution towards advancing Persian lexicography has been made in the publication of an edition of the early work above cited. Asadi's lexicon is the oldest native Persian dictionary; it belongs to the eleventh century of our era, and it has been preserved in a unique manuscript found in the Vatican. The name of Asadi, moreover, is of interest because he was a nephew of Firdausi, the renowned author of the Shah Namah. He is known to have had two predecessors in Persian lexicography; but as their works are lost, Asadi's dictionary heads the list in the history of native word-treasuries. His position in Persian lexicography, in the eleventh century, may be compared, perhaps, with that of Milton's nephew, Edward Phillips, whose 'New World of Words' (1656) occupies a somewhat similar place in England in the seventeenth century.

Dr. Horn has given an elaborate introduction to this editio princeps of Asadī; and he prints the full text in the original Persian. The system which Asadi adopted in arranging words is not the arrangement according to initial letters, but according to the final syllable of the vocable. Such an arrangement, of course, had a practical value for the Persian rhymesters, although the dictionary does not seem to have enjoyed very extensive use. In the present edition, the alphabetic index to the folios enables one readily to find the words that occur.

Like modern lexicographers, Asadi has endeavored to illustrate the meaning and usage of words by drawing examples from Persian literary works that were standard in his time. Seventysix authors are thus found to be cited. The quotations are usually introduced by guft 'says X'; for reference by page, column and number was then unknown. Fragments from missing works have happily come to light in consequence of Asadi's citations. The most interesting, perhaps, (p. 19) are four verses from Rudhaki's lost version of Kalila and Dimna' which will interest Semitic and Sanskrit scholars. Starting with this identification as a hint, Dr. Horn has succeeded in finding several other verses which have come from the same poem.

The old lexicographer had an ear also for the different dialects of his day, and his linguistic distinctions have a value. The dictionary likewise gives a number of new forms or of less common words which are noteworthy, e. g. from Daqiqi xōz 'helmet' beside xōd; or similarly murvā 'good omen, augury,' cf. Pahlavi murvāk. There are a number of others to which the editor has drawn attention in his partial summary on pp. 32-4.

The Introduction is of interest, and the publication is welcome, and it was a worthy undertaking for the Göttingen Academy. We shall look forward, at no distant date, it is hoped, for the appearance of a new Persian-German dictionary which Dr. Horn himself promises to prepare and which he is admirably qualified to make a valuable production. Best wishes for his success! A. V. WILLIAMS JACKSON.

COLUMBIA UNIVERSITY, NEW YORK, Dec. 4, 1897.

REPORTS.

Revue de PHILOLOGIE, Vol. XXI.

No. I.

1. Pp. 1-7. On two Greek papyri in the British Museum, by F. G. Kenyon. One contains a small fragment, which K. considers as belonging to a Aakedaiμoriwv toλireia, possibly by Aristotle. The other throws light on the impressment of work animals in Egypt under the Romans.

2. P. 7. O. Keller adds optio to his examples of differentiation of homonyms.

3. Pp. 8-10. B. Haussoullier discusses the first of the two papyri published by Kenyon in the first article.

4. P. 10. R. Pichon points out two peculiar uses of servire in Seneca (De Tran. An. 9, De Vit. Beat. 17).

5. Pp. 11-25. The theatres of Rome in the times of Plautus and Terence, by Philippe Fabia. An interesting discussion of the question when seats were provided in the theatres. He rejects Ritschl's conclusions, and places the date much earlier. The article contains many important details.

6. Pp. 26-8. Critical remarks on fifteen passages of Hdt. I 27-91 (history of Croesus), by Éd. Tournier.

7. Pp. 29-37. Critical notes on thirteen passages of Tacitus, by Léopold Constans.

8. P. 37. In Quint. Curt. III 1, 11, J. Keelhoff transfers quia to the next clause.

9. Pp. 38-49. The demes and tribes, the arpiai and the phratries, of Miletus, by B. Haussoullier. This article is based on twenty-one inscriptions (including the five discussed in a former article, Rev. de Phil., 1896, pp. 95 ff.). Several interesting

facts are established.

10. Pp. 50-57. A new MS of the Letters of Seneca, by É. Chatelain. The author shows that the Vossianus F 70 of Leyden, and fol. 63-86 of the Canonicianus 279 of Oxford, are supplementary of the same MS. He describes the MS and shows its relation to others, and illustrates its great value by emending, with its aid, fourteen sample passages.

11. Pp. 58-66. Notes on Thucydides, by E. Chambry. Notes, chiefly explanatory, on twenty-six passages. These notes merit close attention.

12. Pp. 67 f. On mis, tis honoris gratia (causa) in Plautus, by L. Havet. Restores the metre in several passages by changing mei into mis, tui into tis.

13. P. 69. Éd. Tournier raises the question whether Toû μý can not and should not be changed to rò un (or something else) in every case where μý has no neg. force, but is induced by the neg. idea of the leading verb.

14. P. 70. In Phaedrus, IV 9, 2, L. Duvau proposes "Repente effugium quaerit alterius malo."

15. Pp. 71-9. Book Notices. 1) B. H. announces the forthcoming Bacchylides of Kenyon. 2) Contents of Part I of Michael's Recueil d'inscriptions grecques, with very favorable comment. 3) Epistola critica ad amicos J. van Leeuwen et M. B. Mendes da Costa continens Adnotationes ad Odysseam scripsit J. J. Hartman (Leyden, 1896). Noticed by P. Čouvreur, who thinks that the author often sees difficulties where there are none. He asks what the author means by 'spurious' verses in Homer, and makes some sensible remarks on the subject. 4) Die homerische Batrachomachia des Karers Pigres, nebst Scholien und Paraphrase, hrsgg. und erläutert von Arthur Ludwig (Leipsic, 1896). P. C. considers this "une œuvre définitive," but thinks thirty years a long time to devote to the subject. 5) Albert Martin finds Wedd's Orestes of Euripides (Cambridge, 1895) too conservative, and commends Herwerden's critical edition of EvpiTidov 'Eλévn (Leyden, 1895), though he seems to consider it also rather conservative. 6) Albert Martin briefly describes Mitchell Carroll, Aristotle's Poetics, ch. XXV, in the light of the Homeric Scholia (Baltimore, 1895) and W. H. Kirk, Demosthenic Style in the Private Orations (Baltimore, 1895). He considers these dissertations as excellent testimonials of the efficiency of the Johns Hopkins University. 7) B. H. highly praises G. Glotz, Lectures historiques. Histoire grecque (Paris, 1897). He objects only to a picture of the Propylaea, still containing "la vilaine tour." 8) Navarre (Octave), Dionysos, Étude sur l'organisation matérielle du théatre athénien (Paris, 1895). Adrien Krebs gives an outline of this work, finding it most excellent, but not entirely free from slight errors. He thinks the author has convincingly established the existence of the λογεῖον.

No. 2.

I. Pp. 81-90. On the expression of the aorist in Latin, by A. Meillet. Supported by analogies in other languages, the author undertakes to show that in Latin a prefix may be used solely to impart aoristic sense, as is often the case with com-, or