at the basis of verb stems. The Semitic lexicon really starts with nouns of action, and all of its verbs are, strictly speaking, a sort of denominatives. Most of the so-called derivatives from Ayin Vau verbs it is impossible, as in the case before us, to derive from them directly, and yet such is the condition of current Semitic etymology and its terminology that it is necessary to correct the editors here, and to say that after all they are wrong in supposing that (il) is not

! אול derived from

The main etymological defect of the whole work, however, is that inherited from Dietrich, namely, the tendency to overdo the etymologizing. It is amazing what combinations are sometimes made in order to bring an immense stock of roots with similar letters under the same primary root, when the meanings alleged to be related are only brought together by a stretch of fancy. Almost every "Ayin Doubled" root gives occasion for an exercise of this kind, though occasionally one is more temperately treated, as 772, 77). It is hard, again to understand what the editors mean when, after citing proofs at y that the word is not the same as the Arabic wada'a, “ to place," they say: "nevertheless the Arabic wada'a can afford us the ground-meaning of y." But even if the comparison were admissible after they have proved that it is not, it would be hard to see how "placing" can have anything to do with knowing. They say, after Schultens, that it means "placing in the mind"; but in this case, not to mention other objections, the essential idea has been interpolated.

It will now be in place to note a few derivations of actual words. 7, "blood," is derived from 07. This means "to be like," and the problem is to bring these notions together. "To be like" is "to be level," i. e., of course, "level with the ground." A thing is made level with the ground by "pushing” or "pressing," from which we get the notion of "pressing together," and so "making close." The notion of being "dark" is connected with being "close and "thick," and what is "blood" but something "dark-red" and thick"? This is only a small part, however, of what the root is made to yield. Nobody can disprove all this, but some of the transitions of meaning are, to say the least, rather violent.

An error which, as far as we know, is found in all the dictionaries, is to be pointed out in the derivation of p, "eye-lid," found once and in the plural (Ps. 77, 5), as the editors have forgotten to note. This is connected with 0, "to guard," as though it meant "the guard" of the eye. There are two fatal objections to this. First, the form is that of the part. pass. so that it must have meant "the guarded thing," which is unsuitable. Second, the word comes out in Syriac with an original th sound: thimrâ, and has the same consonants in its Targumic form. The root is, in fact, unknown, and as the û is written defectively, it is not certain that the Massoretic form is correct. The current etymology of the quadriliteral an, to interpret," 'translate," is certainly wrong; and the way in which it arose is an interesting study. The simpler root Da in Hebr. Aram. and Arabic means "to stone," as a capital punishment. Hence it has been supposed that it originally meant "to throw," then "to throw over," and then "to translate." But it never means "to throw" simply, but "to throw stones," or "to stone." In Ezek. 23, 47; I K. 12, 18 the word for "stone" is a “cognate accusative,” and the meaning “jaculari,” attributed in the lexicon to the Chaldee, is a mistake. It also never means "to throw

over," much less "to translate." It has been attempted to deduce the famous longer word from an Indo-European root which would bring it remotely into connection with our talk, but here the many necessary historical links are left to the imagination. If the word is Semitic, and if a derivation is insisted on, it is best to connect it with Assyr. ragâmu, “ to cry out," "shout." Comparing this (Haupt, Sintfluth-Glossar in KAT2, p. 517) with the same root in Ethiopic and Arabic, where it means "to curse,' ," it is plain that the idea of "speaking" must have intervened; vgl. Hebr. 22, "to curse," with Assyr. qeba," to speak," A. S. andswarian, "to answer," with swerian, "to swear.", "the Flood," is still derived from, but the Eth. mabal," billow," "billows," from a root bay, seems to throw doubt upon this view. The root, already mentioned, can only be a secondary from the kindred primitive ; vgl. na, “stream," "flood." The origin of np, "pan," 2 Sam. 13, 9, is stated to be obscure, but Wellhausen has already acutely and satisfactorily explained it as = "doughplace," standing for mas'eret. The old derivation of , "face," from the similar-sounding root meaning "to turn," reverses the true order, for the latter is a denominative from the former, as the Arab. derived V. conj. of wagaha, "to turn towards," is also a denominative from wagh, “face." The problem is solved when we find that the Assyr. pânu, “face," is the strict plural form of pû (Arab. ƒû, =), "mouth," to which it bears the same logical relation as Lat. ora does to os.

These examples must suffice to suggest how much the etymological portion of the work requires to bring it up to the proper level of method and accuracy. A few words must still be said to show how indispensable it is for Semitic scholars to have a direct acquaintance with Assyrian or rather with the science of Assyriology in its widest sense. Many words, besides those already referred to, receive light, both as to their origin and meaning, from the Assyrian as well as from its local predecessor, the Akkadian. To distinguish between the last two sources and to control the material, generally needs a special training and preparation, and it is surprising how few, comparatively, have devoted themselves to Assyriological studies. The editors have, for this last edition, relied almost entirely upon Schrader's KAT-a work which, with all its excellences, labors under many grammatical and lexical defects; but this they have not used as fully and as intelligently as they might have done. The following are a few errors and deficiencies which we have observed. The now worldwide p," reed," is derived (without mention of any of its Semitic equivalents) from in its hypothetical sense of "standing upright." But, being the same word as the Assyr. qanû, it comes, as is now notorious, from the Akkad. gin the bending thing." Npp is derived from 5, "to cover"; but the word is the Akkad. guza, the equivalent of the Assyr. kuśś, in the bilingual syllabaries. "a mina," is an old pre-Semitic Babylonian weight, Akkad. and Assyr. mana (see, e. g., Delitzsch, Assyr. Lesestücke, 2 ed., 77, 36), and has only an accidental association of sound with the Semitic п, to reckon," "assign," from which the lexicon derives it.

44

=

1 This meaning is well established for the Arabic along with the more common sense of ❝stoning," see, e. g., the Koran, Sura III, 31: "Satan the accursed." This is usually, but wrongly, regarded as being equivalent to “Satan who is worthy of stoning."

The Assyrian must also be often called in to rectify or illustrate words that rarely occur or are doubtful in meaning and origin. For example, 27. “banner,” is the Assyr. diglu, evidently of a similar meaning, which is properly "something conspicuous," from Assyr. dagálu, "to see," "look at" (vgl. Delitzsch in Lotz, Tigl., p. 131 f.). The fact that Assyr. êlamú means "in front of" (e. g. Sennach. II, 77), calls for a new treatment of Hebr. oh and on. The Assyr. equivalent of, “thumb,” is ubânu and this settles the ProtoSemitic form of the root which the Arab. ibhâm had put in doubt. The correctness of the traditional form and rendering of the an. λey. np is placed beyond doubt by Assyr. abiktu, “overthrow," its exact equivalent. Assyr. siqpu,“ sharp point," shows that the Hebr. and Aram. pt, to impale," "hang aloft," are denominatives. What the Assyrian has contributed to the understanding of Hebrew proper names has become better known, and is more fully indicated in the lexicon.

We have noted a few omissions and mistakes from oversight in addition to those mentioned in the errata. At an, nr. 2, should be omitted from pas in the important syntactical and theological passage cited from Gen. 15, 6. The citation on p. 45 b, line 5 from bottom, should read Ri 9, 37. The plural of 10, "secret counsel," and of 23, "ornament" (Jer. 3, 19), are omitted, though they are both exceptional forms. The peculiar forms of the Hifil inf. construct found in 2 Sam. 14, II are also wanting, as well as the Semitic equivalents for the numeral "four," which are of great phonological importance. All of these oversights are transferred from the preceding edition, as is the omission already mentioned in connection with .

A word must be said in conclusion in praise of the thorough manner in which the long introductory article, "Von den Quellen," has been worked over for the present edition, where a great deal of new bibliographical material has been added. The attention that has been bestowed upon the geographical and archaeological departments of the lexicon is evident not only here, but throughout the work.

It has been necessary, in the interests of Hebrew and Semitic studies, to dwell upon the shortcomings rather than upon the merits of the work just reviewed. But, taken as a whole, the lexicon as it now stands is, in our opinion, by far the best Hebrew dictionary in existence, and it should either in form or in substance be speedily done into English. If the leading defects to which we have called attention were to be remedied and, in addition, due deference paid to all schools of exegesis and textual criticism which are intelligently and conscientiously seeking to arrive at the truth, the demands of Hebrew students would be fully met by the next edition.

J. F. MCCURDY.

Leitfaden für den Elementarcursus des Sanskrit mit Uebungsstücken und zwei Glossaren, von GEORG BUHLER. Wien, Verlag von Karl Konegen, 1883. Bühler's book may be said to be almost the first practical introduction into classical Sanskrit, published outside of India. It is written in the style of our introductory Latin and Greek books, presenting alternately grammatical lessons and practical exercises in which the student is from the very start brought face to face with the living language. When Bühler, a few years ago, came to the

University of Vienna,he commenced with a beginner's class of over fifty students, and so effective and attractive was the easy practical way of his instruction that over thirty of these were still pursuing Sanskrit under him two semesters later. This will be significant to any one who knows how beginners' classes in Sanskrit thin out after the first semester in German universities.

The very great value of the book lies in the graded practical exercises. These are not of the style of 'Ollendorff's method,' but consist either of extracts from the literature, or close imitations of passages actually to be found. Whatever one may say of the grammatical part of the book, this collection of good Sanskrit sentences, undoubtedly the result of a good deal of labor, can always be utilized in recasting the book upon some other plan; or they can be employed in practical exercises in prose composition, after the student has had some acquaintance with the language and can be made to judge the grammatical part of the book for himself.

Certainly the grammatical part will not commend itself to Western scholars generally. To the scholar who has been brought up in India, who believes that the grammatical rules and the forms of Hindu grammarians clear down to the Bhaṭṭikavya may be expected any day to receive practical illustration, or to turn up somewhere in the literature, the artificialities of the Hindu system seem an evil which can be tolerated. But he who believes only in phonetics, practically capable of illustration in the MSS. and in 'quotable' forms, will gladly and to his profit leave the Hindu grammar to the Hindus and to the special students of native grammar. He will prefer to take from the start the scientific and yet practical European view of Indian language, which certainly becomes unavoidable as soon as one leaves the domain of the classical language and turns to the Vedic Sanhitas, Brahmaņas or sūtras. Another general consideration militates against the introduction into the Hindu system which the book leads to, namely, the actually acquired position of Whitney's Sanskrit grammar. There can be no reasonable doubt that an overwhelming majority of European Sanskrit students, not to speak of the Americans, now actually use this book for daily reference to a large extent, to the exclusion of other grammars. The second edition of the book, with a list of all accessible verbal forms, may be expected at no remote future, and this appendix will enhance the value of Whitney's grammar-one may fairly say it without being accused of clannishness-out of reach of comparison with any other. Now if Bühler's exercises were combined with an introduction to Whitney, his book would be an unmodified blessing indeed. Those who teach Sanskrit according to Whitney's grammar know how difficult and unsatisfactory it is to make the necessary selections for beginners; Bühler with his practical knowledge of how to teach Sanskrit would have guided him better than almost any other scholar. To be sure, the fault can even now be removed, certainly for English-speaking students. It would not be too difficult a task to employ Bühler's valuable selections and his equally valuable method of grading the lessons, but to transfer these into the framework of Whitney's method. Should there ever appear an English version of the book-and there is some reason to hope that an American scholar will undertake the work-it is hoped that the wish expressed above will not have been uttered in vain, and that Bühler's well-known liberality may permit this free rendering of his valuable book.

MAURICE BLOOMFIELD.

ΑΡΙΣΤΟΤΕΛΗΣ ΠΕΡΙ ΨΥΧΗΣ. Aristotle's Psychology in Greek and English, with Introduction and Notes. By EDWIN WALLACE, M. A., Fellow and Tutor of Worcester College, Oxford. Cambridge University Press, 1882. Hegel brought modern philosophy to the consciousness of the organic oneness of its largest results with the corresponding results of earlier, and especially of Greek, inquiry; and one of the most characteristic features of the philosophic movement in Germany since Hegel's death has been the attempt to comprehend and teach philosophy in and through its history. Accordingly, and as a necessary incident of this attempt, the world—or at least the world of scholars-knows how assiduously and fruitfully German scholarship has been devoted, within the period mentioned, to the critical study of the texts of Greek philosophy, the preparation and publication of new editions and translations, and the elucidation of the same through note and commentary.

English scholars will need a long time to catch up with their German cousins in this field of good works. Meanwhile, all symptoms of a disposition to enter it will be doubly welcome to a growing band of philosophic students, who consciously stand in need of nothing so much as of all fair aids to the ready comprehension of Greek philosophy. It is to such students that Mr. Wallace's work is chiefly addressed, and to them, we say advisedly, it will be invaluable. And for their benefit we may add that main stress is laid by the author on the exhibition of the nature and extent of the contribution made in Aristotle's Epì xns to the theory of cognition.

While Mr. Wallace's aim has thus been (in his own words) "explanation, rather than textual criticism," he has not been unmindful of the requirements of the latter. In particular, the views of Torstrik respecting a double recension of the text (see Torstrik's edition of Arist, de Anima, Berlin, 1862) have been tested by him," in regard to several portions of the text," with results expressed by Mr. Wallace as follows: "Without denying the existence of repetition and disorder in much that Aristotle wrote, or rather left in notes, I have tried in several passages to maintain the general correctness of the ordinary text against Torstrik's objections and emendations.'"

G. S. M.