is ore, which may perhaps stand in a similar relation to the accusative suffix wo, which marks the direct object of the verb with a definite emphasis; or it is possibly only a variation of ware.

These simple forms of the pronoun were once no doubt sufficiently respectful for the purposes of polite intercourse, but they have long since ceased to be so, and are now used only by the vulgar, or by superiors in addressing their inferiors. Toward an equal or superior, etiquette requires a more humble designation of one's self, a more honorable appellation of another, so that ware employed as a pronoun of the first person exalts the speaker, and as a pronoun of the second person abases the person to whom it is applied. The change in the position of ware is only relative; it remains at its old level, which is now. from the changed requirements of the case, too high for the first, and too low for the second person. This simple explanation of the two-fold use of ware and of ore Hoffmann has missed (Gram., pp. 103,86). In poetry ware is still the indifferent pronoun of the first person, neither proud nor humble.

Of the numerous substi utes for these pronouns which have been introduced to express the various degrees of respect or disrespect which the rank of the speaker and of the person addressed may call for, some are peculiar to the written or the spoken language, others common to both. Chinese forms abound in the writings, and are not infrequent in the speech, of scholars; and it is said that the recent revolution, which has increased the political influence of the literary class, has also given a greatly extended currency to some of these forms. They are generally epithets and titles, similar in character to our "your Honor," while forms of native Japanese origin are more often of a demonstrative character, serving to point out in a more general and indirect and therefore more polite way than the simple ware, are, etc., the person intended. For example, the usual respectful form of address for the second person singular is anata, contracted from ano kata, that side,' and for the third person, ano o kata, which contains the same elements with the addition of the adjective o, 'great' or 'honorable,' and is thus equivalent to that honored side.' Sama, meaning originally form,' shape,' added to a proper name, corresponds nearly to our Mr., Mrs.,' and is also frequently joined to pronominal forms to make them more respectful. This is but one manifestation of a tendency which has modified in a remarkable manner the higher styles of Japanese, both spoken and written. The simple and natural forms of the verb are constantly replaced by periphrastic conjugations, which weaken the directness of the language, and render it to the popular apprehension more elegant. So also in the address on letters, the name of the person is followed by a phrase like the following: to those present,' to the attendants,' or if the letter is for a scholar, 'beneath the study table,' the writer not venturing to address even his friend except through some such form. The form of address to the Mikado is hei-ka, beneath the staircase' (leading to the throne), i. e. to the attendants who are there in waiting.

It is curious to trace the gradual descent of some of these forms, which by long use have become vulgar, to a less honorable position than they once occupied. Though still high-sounding, they have become disrespectful. Such are, of the second person, ki-sama, noble sir,' ki-kō, 'noble princ, nanji, according to Hoffmann for na-mochi, renowned,' omaye, 'honorable presence,' the exact Chinese equivalent of which, go-zhen, is still the Japanese form of address to one's prince. Something like this is seen in the German, where du, ihr, er, Sie have each in turn been the respectful form of the singular pronoun in address. The use of a plural as a more polite form of the singular (the speaker associating others with himself or with the pers in addressed, and thus avoiding the appearance of presumption). which is so widely spread in European languages, starting from the imperial'we' of the later Roman emperors, and then passing to the second person, 'you' fr 'thou,' finds also a parallel in the Japanese. Ware-ware, we,' is a very respectful form of the first person singular, and anata-gata, 'you,' of the second singular. In some cases the original plural signification is so far obscured that when an actual plural is required a doubly plural form is employed; thus, ware-ware nazo, or waga-hai nazo, 'we.' It is possible that ora and oira, which are pronouns of the first person singular used in talking to servants, are forms which contain the plural suffix ra, and are thus contractions of an original ore ra. The plural of pronouns is formed in the same way as the plural of nouns, either by reduplication, e. g. ware

ware, or by the addition of a collective word denoting side, row, and the like. These latter again differ in the degree of respect which they convey, the most respectful being gata, the same word which has already been united with the singular pronoun in anata, ano-o-kata.

A few

Aside from the differences in the use of pronominal forms which prevail among the different rauks and classes of the people, there are others of a local and dialectic character. Neither the forms nor the values which they bear are everywhere the same; e. g. omaye, given above, is not everywhere disrespectful. are peculiar to women. Thus waraba, properly young unmarried woman,' is, in books written in the conversational style, frequently put in the mouth of women for I;' waga-mi, my body,' i. e. 'I,' is mostly used by women, but in the province of Ichijen is common to both sexes. Watai and wachiki, both corruptions of watakushi, the usual polite form of the pronoun 'I,' are used by a lower class of women.

10. Remarks on the Relation of the Chinese and Mongolian Languages, by Rev. John T. Gulick, Missionary of the A. B. C. F. M. in Mongolia.

Mr. Gulick maintained that, although the Mongol belongs to a group of languages of which the grammatical structure is wholly different from that of the Chinese and of the cognate languages of Tibet and Siam, there are nevertheless plain evidences of an early connection between the languages.

There are a few roots which seem to be identical in the two groups; but, as these roots may have been introduced from Chinese into Mongolian, they give no satisfying evidence of original connection. The common element in the languages named is to be found in the fundamental conceptions which control the classification and nomenclature of certain groups of ideas. These, it was claimed, are less subject to change than either the verbal roots or the grammatical structure.

By way of example, Mr. Gulick called attention especially to two subjects: 1. The conception which determined the classification of the points of the compass was originally the same in Chinese, Tibetan, and Mongolian, although in modern Chinese its primitive form has been greatly modified and obscured. In these languages the south is always mentioned before the north, and, in ancient Chinese and Mongolian, the west before the east. We further find, in Mongolian, that the word for south means in front,' for north, behind,' for west. to the right,' for east, 'to the left.' In the other languages, some of the points are named in the same way. The character which in Chinese stands for north is composed of two men turned back to back, and originally meant to turn the back upon,' or behind.' In the compass, the index is placed upon the south point of the needle. In at least one passage of the ancient classics, Mr. Gulick had found the word right used for west.' In languages of other families, the points of the compass are sometimes named upon the same principle, but another point is usually assumed to be in front. Thus, in Hebrew, the words before, behind, right, and left are used, but the east is in front. With us, in giving the points of the compass in their order, the north is conceived to be faced. The Hawaiians front toward the west, and call north right,' and south 'left.'

2. In the classification of family relationships, the Chinese and Mongolian correspond in several of the points in regard to which they differ from European languages. Both make a distinction between older and younger brothers and sisters; both have separate names for paternal uncles and aunts, as distinguished from material.

11. On Rev. S. A. Rhea's Kurdish Grammar, by Prof. W. D. Whitney, of New Haven.

Mr. Rhea's Kurdish grammar and vocabulary were presented to the Society now two years ago (see the Proceedings for Oct. 1869; Journal, vol. IX., pp. lix., lx.), and have been recently taken up by the Committee of Publication with a view to their examination and preparation for the press, as they are expected to appear in the next number of the Journal. By way of calling attention to their approaching publication, and of giving some intimation of their value, Prof. Whitney had drawn up certain comparative statements, especially as to the verb in the dialect

here represented, in its relations to those in the dialects of the language heretofore made known.

The Kurdish presented by Mr. Rhea is that of the Hakari mountains, on the border between Persia and Turkey, especially in the southern part of the pashalik of Van, the late principality of Julamerk, and may be the vernacular (so Rev. Mr. Shedd estimates it) of 200,000 souls. It holds a middle place between the dialects known about Harput on the west, and those southward from Soujbulakh, including Suleimania and Senna, these last being so different from it as to be hardly intelligible to one who knows it. Mr. Rhea spent seven or eight years among the Kurds, in Gawar, and became quite familiar with their language, using it freely in conversation, although never preaching in it. The materials were collected and worked into shape before his return to Persia in 1860.

In this, as in the other dialects, there is but a single original and proper verbal tense, of which the inflection is as follows (taking as example the verb kirin 'make, do,' Persian kerden): Sing., 1. kem, 2. keï, 3. ket, Pl., 1. 2. 3. ken. It never appears, however without a prefix; with b', it forms a "generic present," az b'kem etc., Im ke;' and, with t', a (conunuous) "present," az t'kem etc., 'I am making.' Another prefix, de, makes of it a future, az de kem, I shall make.' Then a simple preterit is made by prefixing to a verbal, uninflected, the series of secondary or objective personal pronouns, min, ta, wi, ma, hava, wan: thus, min kir etc., ‘I made;' and this, with prefixed, becomes an imperfect, min t'kir etc.. 'I was making.' But two classes of verbs, those ending in the infinitive in an and in, add to the verbal the present of the substantive verb as auxiliary, and then use with it the primary or subjective pronouns, az, tu, au, am, hun, wân: thus, az chemiám, 'I bowed,' az kennım, I laughed,' az t'chemiám, 'I was bowing,' az t'kennım, ·I was laughing.' Another invariable verbal, with the secondary pronouns prefixed. forms a perfect: thus, min kiria etc., I have made:' and this, with bu, 'been,' added, mikes a pluperfect, min kiribû etc., · I had made;' and, with ba added and de prefixed, a conditional, min dê kireba etc., 'I should make;' while in all these tenses also, the verbs in in and an add the auxiliary and use the other pronouns : thus, az kennima etc., 'I have laughed,' az kennibúm etc., 'I had laughed,' az dê kennibám etc., 'I should laugh.' Finally, there are two persons of an imperative, in a and n thus, b'ka, 'make thou,' b'ken, 'make ye.'

A comparison of this scheme of verbal inflection with the very different ones reported by Garzoni for the dialect of Amadia, by Lerch for the Kurmanji and Zaza dialects, and by Chodzko for the dialect of Suleimania, affords an instructive measure of the peculiarity of the Hakari dialect.

12. On the Continuation of Westphal's Methodical Grammar of the Greek Language, by Prof. James Hadley, of New Haven.

After referring to his remarks, made at the meeting of October, 1870. on the first part of this work, Prof. Hadley observed that faults of haste and carelessness, similar to those which he then criticised. are too frequent in this second part, which treats the etymology of the verb. Thus, it is twice said (pp. 40,42) that Herodo tus uses ¿kelaro for ěkεTO, 3 pl. impf. of Keipat; while p. 111 gives the true statement, that the Ionic of Herodotus has exéaro, and that Keiaro is Homeric. On p. 54, it is suggested that the ending Tov of the second person dual arose from the primitive tas (Lat. tis) of the second person plural, by loss of final s, and subsequent addition of a euphonic nasal, just as uev of the first person plural arose from μɛs (primitive mas, Lat. mus) which appears in the Doric. But on p. 80, it is asserted, without any allusion to the view just given, that the v in Tov is a sign of plurality, added, with a necessary connecting vowel o, to the 7 (originally to) which marked the second person singular. Westphal shows, indeed, an evident desire to explain the v of ev in the same way, and to recognize two forms, mas and man, as existng together in the primitive Indo-European, the last of which is preserved only in Greek ev; but here the concurrence of all the si-ter languages with the Doric Greek is too strong for him, and he is forced to admit that more probably μev is only an euphonic variation of μeç.

Agun, on p. 54, it is affirmed that in the middle endings ofɛ, olov, olny, the σ was an euphonic addition, a mere strengthening of ; and that these endings were originally preceded by an a (whether long or short, is said to be uncertain), which

expressed plurality. The same view re-appears on p. 72 (where, however, the lost a is given as long, without expression of doubt); and again, so far as the σ is concerned, on p. 226. But in a table inserted opposite p. 80, the author withdraws this explanation, and puts forward another, which says nothing of a lost a, long or short, but recognizes o as the plural sign, and thus as an original element. In the same table he takes back also the assertion on the opposite page that v in Tov, TNV, is a plural sign, and returns to the explanation of it on p. 54, as a paragogic nasal. Opinions thoroughly considered and maturely formed would hardly show such frequent fluctuations.

Observing that Westphal recognizes five signs of plurality the vowel a, the liquid, the sibilant σ, the smooth mute 7, and the aspirate - Prof. Hadley remarked that, with such a multiplicity of factors, it is easy to etymologize, especially if one is not solicitous about improbabilities. That v in Tov, TV, should be wholly different from v in o0ov, σ0ŋy, a plural sign in one case (see p. 80), and a euphonic nasal in the other: that the vowel in Tov, TV, should be wholly different from the vowel in o0ov, om, a euphonic connective in the one case, a sign of the middle voice in the other; that the euphonic connective should be lengthened, without the least apparent reason, in T or rav-in all these things Westphal seems not to be conscious of any difficulty or improbability.

Some curious speculations of the author as to the genesis of the Greek verbinflection were noticed, without being discussed. Thus, that the σ of the future is an intensive affix (unconnected with the root eo or as, 'to be'), marking an energy of will directed toward the action; that the σ is also intensive in the first aorist, where it marked an energetic cutting short of that continuance which belongs to the natural conception of verb action; - that in the second aorist, the same intensity was expressed by a reduplication, which originally belonged to all second aorists; that the formation of a perfect was originally common to all verbs, but, from the use of an aorist instead, was to a great extent lost in the Homeric time, though it was partially revived at a later period; - that, in the pluperfect active (23 ἐπεποίθεις, earlier ἐπεποίθεας), ea is not for εσα (ef. γένεα, orig. γένεσα, Lat. genera), and thus derived from the past tense of eiuí, but that e in ea is a connecting vowel inserted (for what euphonic reason?) before the connecting vowel a of the perfect. Would he say that in the Latin pepend-era-s, the r is a connecting consonant used to connect the two connecting vowels e and a?

That the book contains ingenious views and valuable suggestions was fully acknowledged by Prof. Hadley; but he thought it a very unsafe guide for the student who is not yet able to subject its statements to independent investigation and criticism.

13. On the Collation of a new MS. of the Atharva-Veda Prâtiçâkhya, by Prof. W. D. Whitney, of New Haven.

Prof. Whitney recalled to the memory of the Society the circumstances under which his edition of this Prâtiçâkhya had been published in the Society's Journal (Vol. vii., 1862). It was founded on the sole known manuscript, of the Berlin library, an inaccurate and imperfect copy of the work and its commentary. Inquiry had first been made in India for further manuscript material, but wholly without success, and the interest and importance of the work seemed to justify its being brought out, in as good shape as the circumstances rendered possible. Recently, a new copy of the treatise itself, without the comment, had come to light in India: it was found at Broach, and was purchased for the Government, along with other valuable Atharvan manuscripts (a catalogue of them is given in the Monatsbericht of the Berlin Academy, the number for Feb. 1871), by Dr. Bühler of Bombay. On hearing of this, Prof. Whitney had lost no time in writing to Dr. Bühler to beg a copy of the manuscript; and this gentleman, with the prompt and open-handed kindness which distinguishes him, had immediately responded to the request, and sent a copy, carefully revised by himself. It was intended to publish in the next volume of the Journal the complete results of a collation of this new authority; without troubling the Society with the full detail of them, some of the more noticeable general items were reported.

The MS. fully supports the form of the name of the treatise, caturâdhyáyiká, adopted in the edition; its constant ending is átharvaṇe catur adhyayikáyám

adhyayah. It gives as first introductory rule atha 'ngirasaḥ. It contains, as part of the treatise itself, the intrusive passages introducing and winding up the third section of the third chapter (before rule 55 and after rule 74), and also that prefacing the fourth chapter, which were in the edition treated as belonging to the commentary. It fills up, of course, the two notable lacuna in the third chapter (after rules iii. 28 and 80), and also shows the loss of three or four rules in other parts, where such loss had been suspected, but not regarded as certain, by the editor, or where it had escaped his notice. None of these losses, however, nor the changes of reading suggested, affected the essentials of the treatise. or even touched points of especial importance: so that, Prof. Whitney said, if he had been able to foresee ten years ago that such a manuscript would turn up at this time, he should have been justified in not waiting for it-the gain by delay would have been less than the loss. If the comment had also now come to light, the case might be different. The new manuscript is a very good and correct one. Along with it was received a copy also of a little grammatical treatise entitled prátiçakhya-múla-sútra, and pertaining to the Atharva-Veda; it may perhaps be made the subject of a future communication. Both manuscripts appear to have been written A. D. 1660.

No further communications being offered, a vote was passed of thanks to the Faculty of the Theological Seminary, for their kindness in giving the use of their room for the purposes of the meeting, and the Society adjourned, to come together again in Boston, on the 22d of May, 1872.