A Comparative Grammar of the Indo-Germanic Languages: Volume IV/3 Morphology, Part III [Transl. from German, Reprint 2021 ed.] 9783112410622, 9783112410615

Citation preview

A

COMPARATIVE GRAMMAR OF THE

INDO-GERMANIC LANGUAGES.

Da muss sich manches Rätsel lösen Doch manches Rätsel knüpft sich auch. Goethe,

Faust.

A

COMPARATIVE GRAMMAR OF

THE

INDO-GERMANIC LANGUAGES. A CONCISE OF T H E

EXPOSITION HISTORY

OF SANSKRIT, OLD IRANIAN (AVESTIC AND OLD PERSIAN), OLD ARMENIAN, •GREEK, LATIN. UMBRO-SAMNITIC, OLD IRISH, GOTHIC, OLD HIAN GERMAN, LITHUANIAN AND OLD CHURCH SLAVONIC BY

KARL BRUGMANN. PROFESSOR OF COMPARATIVE PHILOLOGY IN THE UNIVERSITY OF LEIPZIG.

VOLUME IV. M O R P H O L O G Y ,

P A R T

III:

VERBS : FORMATION OF T H E STEM, A N D INFLEXION OR CONJUGATION.

TRANSLATED FROM THE GERMAN BV

R. SEYMOUR CONWAY, M. A. AND "LATE FELLOW OF GONVILLE A N D CMOS COLLEGE, CAMBRIDGE, AUTHOR OF 'VRRNER'S ¿ A W IN I T A L Y ' , PROFESSOR OF £,A?IN IN THE UNIVERSITY COLLEGE, CARDIFF.

W. H. D. ROUSE, M. A. SOME TIME FELLOW OF CHRIST'S COLLEGE^ CAMBRIDGE, ASSISTANT MASTER A T CHBLT B N H A M COLLEGE.

STRASSBURG. K A R L J. T R t f B N E R . 1895.

Printed by O. 0 1 1 a , Darmstadt.

TO

JOHN PEILE, DOCTOR OP LETTERS, MASTER OF CHRIST'S COLLEGE, THE FOUNDER OF THE STUDY OF COMPARATIVE PHILOLOGY IN CAMBRIDGE

THIS TRANSLATION IS DEDICATED

IN TOKEN OF THEIR GRATITUDE AND AFFECTION

BY

HIS OLD PUPILS.

AUTHOR'S PREFACE. When in 1889 I brought out the part of my work containing the account of Noun Morphology, I had in my mind, and partly on paper, a simpler plan for the remainder of the work than that which the reader has before him. I meant it to include the presentation of the forms of declension and conjugation, and little more. In view of the confusion and uncertainty that reigned in this department, where many questions of origin and history seemed utterly unsettled, I then thought it best to restrict the work to these limits; and I only hoped that perhaps after the lapse of years, if a second edition should be called for, the further developement of a science which had already made such rapid progress would have put me in a position to give a practical and useful history of Noun and Yerb Inflexion. But in the course of my work I was gradually converted from this pessimism; the difficulties no longer seemed insurmountable; and I at length decided to attempt a more complete account, not merely exhibiting the results of the different developements, but even now as far as possible tracing their course. Thus the work grew; and thus it comes about that the size of the latter part is so greatly out of proportion to that of the former. In giving up my first plan, I was influenced not a little by my belief, that from a comprehensive work such as this,

vm

Author's Preface.

a work in which it is sought to present the facts and problems of language in connected form, more might reasonably be expected than what I had at first proposed. A student might fairly ask that the many questions which await an answer should not be simply avoided, but that some honest attempt should be made to advance a step towards their answering. It must surely be useful that he should not only read that which can be called certain, Dot only be taught well establisht facts, but that he should at the same time find the various problems and puzzles, with which the study of Indo-Germanic inflexion abounds, at least briefly mentioned and conveniently arranged. So will the scholar guard best against the mistake which not the best scholar is wholly free from, the danger that in trying to bring order and light into his palace of knowledge, he may leave some dark riddle unattempted, and only move it from one corner to another. If amidst these shifting theories I have often taken a decided stand, and declared myself for one or other of them, adding therewithal other and many new views and explanations, I am yet far from believing that I have placed beyond all doubt the view which I have preferred. In these matters to indicate a path for future research or simply to establish a prima facie case is far harder than most people think; and many a theory which seemed to be fixt on the firmest foundation and to offer no point to attack, has been broken down in the end. I can only hope that the mistakes which these volumes must inevitably contain, may help to supply the means for their own correction. A few of my readers perhaps may wonder why certain new and some very recent theories upon Ablaut, proethnic Accent, formation of Roots and Suffixes, and other such matters, have in these last volumes been either altogether disregarded or only just glanced at. A good deal of the most recent work I would indeed have included in this last volume

Author's Preface.

IX

bat that it had to be finisht in 1891.]) In other cases I saw before me hypotheses, which attractive as they are, and fruitful as they may prove to be, at the time of their publication were too slightly worked out by their authors, and had been too little tested to allow of my making them the basis of my own account. In this volume I have practically not touched the newest theories of Ablaut; I confess that I approach the glib and symmetrical systems of Ablaut Series (cp. Bartholomae in Bezzenberger's Beiträge, XVII 105) with very little confidence, and I must refer to what is said on this matter in Yol. I § 309. Even a question of Yerb Morphology so important as the form of proethnic Roots (whether they were monosyllabic or not) I have left on one side; I believe neither the one thing nor the other, but only that in the present state of our knowledge we can know nothing about it. If in spite of this I have used hyphens freely, I would remind the reader that the hyphen means a real point of composition in such words as dioo-xovpoi,

xov-ro,

ava-ßa%ho,

e-tpspov, but in

ay-o-ftsv or .

hinav

with Aw'nw,

B u t for convenience'

sake verbal nouns will be only cited now and then while we deal with the various groups of the verb infinite, and the whole of them will be afterwards collected and examined in detail. Verbal C o m p o u n d s , such as would be formed by joining a real verb-stem

to a form of the finite verb, and would

answer to O.Pers. hama-pitd

— Gr. ¿fio-naT(op among noun

compounds, never existed as a distinct category, either in the later stages of the parent language, or since. ceptions

The only ex-

are reduplicated verb forms such as Skr.

dar-darti

'he bursts, breaks to pieces', so far as these can rightly be called compounds (§ 464).

W e must however not forget that

the reduplication of uninflected "roots" was certainly older, on

8

The Verb: General Remarks.

§462.

the whole, than the fusion of roots with personal pronouns into verbal forms. Whether the tense sign -s- in Gr. dsix-au e-deix-aa etc. was the verb subst. es-; whether -dh- in Skr. sd-dhati Gr. to9-« (\^ed-) e-oxs'9ov O.C.S1. i-dq, etc. is the verb dhS- 'u9-evou ; and whether these originally acted as auxiliaries, are questions which must be left alone. Even if this be the truth about them, they must have sunk to the level of inflexions long before the end of the proethnic period, and they could no longer be the type for compounds consisting of verb + verb. Nor do we find in the periods for which there is direct evidence either noun stems compounded with genuine verb forms, or genuine verbal stems compounded with nouns. We cannot class under the second head words like Gr. ¿(t/s-xaxog (from ap/iu), ¿Tti/atpe-xaxog (from eui/aipm), 2r>]a-ayoQijg (from Z-orrjoa), H.G. wetz-stein 'whetstone' melk-fass 'milkpail' (from wetze, melke). These are due merely to a perversion or interpretative corruption, and the imitation of older compounds which had a noun stem for the first member. They are not real compounds of a verbal stem with a noun. See I I § 30 pp. 51 ff., § 41 pp. 74 f., § 47 p. 86.1) Forms of the finite verb are clearly seen in composition only in the following classes of words; and here too one of the two parts has usually sunk to a kind of suffix or prefix. 1. A Yerb form is compounded with Adverbial words; as GT. an-sifii Lat. ab-eo, Lat. ne-scio, pr. Idg. *e-dfh,om, = Gr. ¿-¿Qaxov (I regard the augment as a temporal particle); Pruss. quoiti-lai 'he may wish, he might wish', pr. Idg. *bhiret-u Skr. bhdrat-u, — and also -i in *bMret-i *bMres-i (beside *bhiret *bhSres) was probably a demonstrative particle. 1) There is a new essay by W. Christ, AbhSngigkeitskomposita des örieohisohen, Berichte der k. bayer. Akad., 1890 pp. 187 ff. I oannot agree with the theory for which Christ takes up the oudgels again (pp. 184 ff.), that 'Aye-laoe arose from the imperative phrase aye iao'r, and that got oame from aQxe-&e'a>gof by phonetic change of t to t.

§462.

9

The Verb: General Remark*.

2. A Verb form is compounded with a Case, being itself the final member of the compound. The case, at the time of combination, was either a living case, or some kind of infinitive. Examples: Skr. dattismi (periphrastic future) for d&td asmi

'dator sum', Lat. possum

pessun-do,

Lat. are-ham

for potis

dre-facio

p. 177), Fr. aimerai

(O.Fr. amerai)

1" pi. suktum-bime,

etc.

sum,

vSnun-do

O.C.S1. nes6-achu for amdre

venum-do (III § 275

habeo, Lith. opt.

The line between these two classes is not absolutely fixed, as we see from such words as Skr. astam-eti 'goes down' (of the sun), which has for its first part asta-m 'to one's house, home'. R e m a r k . There is no olass of Verbal Compounds answering to Class I of Noun Compounds (pfio-miziog), whioh might have been a model for later formations. It seems therefore best not to giye a speoial chapter to Yerb Compounds, as was done to Nouns, but to describe the •ariouB kinds in the place they most naturally come, along with the Tenses, the Personal Endings, or as it may be. One point, however, should be touohed on here. It is a matter of wider interest, and well illustrates an important prinoiple of word-formation. "When a sentence becomes a single word, it may be treated like a simple word, and it may be infleoted or otherwise modified just as the simple word can. Lith. d'&'-k 'give* et-k 'go' (both imperative), consisting of the 2 nd sing. *du (cp. Lat. ce-do) and *ei + the particle -k, are the foundation for a 2 nd pi. dlikite, eikite eto.; and again eVcaz 'oome here', which itself is ei-k + szi 'here', produced eiksz-te. Pruss. quoitl-lai 'he might wish' ([lat = O.C.S1. li 'vel, aut'_) suggested the 2 nd sing, quoifilai-si and 2 nd pi. quoitilai-ti. Slav, daj-mi and Mod. Gr. 'give me* produoed the 2 nd pi. daj-mi-te and JO'S-UOV-TS. Of the same kind were Gr. 3 rd pi. aycrio-v ¿yCTU-aar (2 nd sing. iX9trm-i Hesyoh). Lat. 2 nd pi. agito-te, since Idg. *age-i5d was a compound of the imperative *age with the adv. (abl.) *tdd 'thence, then'. Also, according to Osthoff and Bugge, Gr. SeSmxauev SeSiaxare; they explain SeSioxe as *StSw = Skr. dade(u) + the particle *e = Lat. ce in cedo. Analogous forms of nouns or pronouns are: Gr. ovro-g AVRIJ aoo. TOVTO-V Boeot. OVTO-V etc., formed from the nom. *OV-TO *AV-ro, whose second part was a partiole (cf. O.C.S1. ku-to 'who'), see II § 4 p. 9, III § 415 p. 337; Mid. Ir. instr. pi. donafi-b built up on O.Ir. donaib-(h)i, see footnote on page S57; O.H.G. gen. deses dat. desemo following de-se 'this* with the partiole se, see III § 414 pp. 335 f.; Skr. dat. asmd-bhyam Lesb. nom. ¿nne-( following aco. *ns-me (Lesb. S/i/ie) with the partiole sme, see § 436 pp. 367 f., § 443 p. 3'.'9; Pol. gen. sing. sztukamiqsy following nom. sztukamiqsa 'a piece of meat', where miqsa is the gen. of mi^so.

10

Reduplicated Verb-Forms.

§462.

REDUPLICATED VERB-FORMS. Reduplication, the repetition of a word or other element of speech with the same grammatical force, to express that an action or state is repeated, or to intensify it, is certainly older than the modes of forming cases or parts of the which we actually see in use.

finite

verb

It had at first no special con-

nexion with either verbs or nouns, but was used with both ; verb types such as Skr. ddr-dar-ti

'it bursts, breaks up'

'we have ventured' ci-kit-e

'he knows', and noun

Skr. dar-dar-a-s

'broken' da-dhps-d-s

'knowing' may have other.

da-dhfs-imd types

like

'venturesome, bold'

ci-kit

been formed quite independently of each

Compare II § 6 pp. 12 if., § 51 ff. pp. 94 IF. !)

Root reduplication in verbs came to be very and this very early in the history

of the parent

important, language,

because it was turned to account in the formation of tenses. 1) Works on Reduplication in general have been cited in the footnote to vol. II page 12. On Yerbal Reduplication see the following. A. W i l l i a m s , On Verb-Reduplioation as a Means of Expressing Completed Action, Transactions of the Amer. Phil. Assoc., 1875 pp. 54 ff. P a u l i , Das praeteritum reduplicatum der idg. Sprachen und der deutsche Ablaut, Kuhn's Zeitsohr. xn 50 ff. O s t h o f f , Zur Geschiohte des Perfects, pp. 264 ff. and passim. — E r n a u l t , Du parfait en greo et en latin, pp. I f f . — E b e l , Reduplicierte Aoriste im Griech., Kuhn's Zeitschr. Ii 46 ff. — V o n d e r P f o r d t e n , Zur Gesch. des griech. Perf., pp. 42 ff. — D e e c k e , De reduplicato linguae Latinae praeterito, Lips. 1869. — S t o k e s , Reduplication im altir. Verbum, Kuhn-Schleioher's Beitr. Ii 396 ff. W i n d i s c h , Das reduplicierte Perfectum im Ir., Kuhn's Zeitschr. xxni 201 ff. — A. M o l l e r , Die reduplicierenden Verba im Deutschen als abgeleitete Verba, eine etymologisohe Untersuchung, Potsdam 1866. S o h e r e r , Die reduplicierten Praeterita, Zeitschr. f. österr. Gymn. xxiv 295 ff., and Zeitschr. f. deutsch. Altert, xix 154 ff., 390 ff. S i e v e r s , Die reduplicierten Präterita, Paul-Braune's Beitr. I 504 ff. P o k o r n y , Über die redupl. Praet. der germ. Sprachen und ihre Umwandlung in ablautende, Landskron 1874. H o l t h a u s e n , Die reduplicierenden Verba im Germ., Kuhn's Zeitschr. xxvn 618 ff. H o f f o r y , Die reduplicierten Praeterita im Altnord., ibid. 593 ff. — See also the works cited under the head of Tense-stems.

§465.

Reduplicated Verb-forms.

11

It was used to distinguish various kinds of action, and also the time at which the action took place. For these purposes Reduplication was very widely used, not only in the parent language but in most of its offshoots far on into the historical period. The wide use of verbal reduplication gave great impulse to similar reduplication in nouns. Noun - reduplication was at first a principle not very active or far-reaching; but thus reinforced it lasted much longer than it would have done, and in later times, under the influence of reduplicated forms in the verb infinite, reduplicated noun-forms appeared again where the proethnic reduplicated type had lost all its reproductive power. Examples are Gr. xexpay-/uo-g (Eur.) xexpay-^« (Aristoph.) cry, outcry', xfxgaa-ry-g 'one who cries aloud' (Aristoph.) beside xt-xgay-co'e xs-xpay-a, Titnoifr-rjOt-g 'trust' (Josephus, Philo) beside ns-noid--m ter-) 'pass over', var-v^t-ati 3 r d pi. of vart- (\/^uert-) 'vertere', mar-mfj-yd-tS 3 r d sing, of marjr d ( y f i n e r § - ) 'wipe', jata-ghan-ti 3 sing, of han(\fghen-) 'strike, kill', nam-nami-ti 3 r d sing, of namnem-) 'bend', dan-dai-Ona-s partic. of dqSda'flk-) 'bite', jafi-jabh-y&-te r d 3 sing, of jambhr ( v ^ gembh-) 'snap', ve-ve-ti 3 r d sing, of vi- 'appetere', ve-m-ya-te 3 r d sing, of v%- 'flutter', nS-nik-te 3 r d

12

§465.

Reduplioated Verb-Forms.

sing, of nijneig-) 'wash', nd-nu-mas 1st pi. of nu- 'cry out', cö-$kü-yä-te 3rd sing, of sku- 'cover, ¿ö-iuc-ana-s partic. of Sue- 'light, brighten', a-nö-nud-ya-nta 3rd pi. of nud- 'push'. st A v es tic car'-ker"-mahl 1 pi. of kar- 'think of' (Skr. imper. ca-kf-dhi). dae-döis-t 3rd sing, of dis- (\fde$-) 'show' (Skr. 3rd sing, de-di^-fe). zao-zao-ml lBt sing, of zu- call' (Skr. joham-mi).

Greek

'I move restlessly' for *-ipvp-i.ro: cp. 'moves convulsively, throbs, palpitates', fiogmurmurüre, -f.ivQ(o 'I roar, murmur' for *-/.ivp-i.w: cp. Lat. 0.H.G. murmurön (§ 595). yap-yaip-co 'I swarm' for *-yaQ-im neut. pi. 'swarm"), yay-yaivaiv ' TO fitzd ysXmrog VQOO(ydpyaga nai^ftv Hesych. (Schmidt conjectures yayyavtvuv): cp. O.C.S1. gq-gnajq 'I murmur, growl' (adj. gqgnivu). nafi-cpuivw 'I shine clearly' for *-),

¡¡axijna:

rjaxtfriai

(anxito),

¿tp-txofitjr : aqriyfiai. But if so we should expect similar forms in verbs which began with one consonant, and such forms as *e-ßqxa instead of ßt-ßtjxa

(see § 475).»)

We cannot suppose that the form i-qgaya stands for *ft-ggo>ya, and that it gave the type for i-ßli«rt/xa etc., because the dialect of Gortyn a has t-yQafjuai, and this dialect kept initial J- before an «-sound. The Cretan perfects ¿n-rjareXxe ij-yoaTrm mentioned in § 472 Rem., page 17 above, with r'-, only give a fresh problem to solve.

THE §

477.

The

syllable, Idg. *e- =

AUGMENT.2)

Augment Skr. a-

(avfyotc),

as it is called,

Armen, e- Gr.

is

a

which prefixed

to verbal forms serves to mark past time. 1) eXoyxa instead of IrXoyya in two late sepulchral inscriptions is probably not a mistake in the graving, but a misformation, due to contamination of the perfect HHoy/« or eil/j/n at a time when this kind of perfect had become unfamiliar. (Thumb, Mitteil, des deutsch, arch. Inst, in Athen, xvi 176). 2) R. G a r n e t t , On the Origin and Import of the Augment in Sanskrit and Greek, Proceedings of the Philol. Society I (1844) p. 265 ff. F r . M ü l l e r , Einiges über das Augment, Kuhn-Schleicher's Beitr., in 250 ff. J . D a v i e s , On the Temporal Augment in Sanskrit and Greek, Hertford 1865. F a u s t , Zur idg. Augmentbildung, Strassb. 1877. A. H. S a y c e , The Origin of the Augment, Transactions of the Philol. Sooiety, 1885—1887, pp. 652 ff. B r 6 a l , De l'augment, M6m. de la Soc. de lingu. vi 333 ff. J . A v e r y , The Unaugmented Verb-Forms of the Rig- and AtharvaVedas, Proceedings of the Amer. Orient. Soo., May 1884, pp. XI f., und Journal of the Amer. Orient. Soc. xi 326 ff. E b e 1, Die scheinbaren Unregelmässigkeiten des griech. Augments, Kuhn's Zeitschr. iv 161 ff. L a R o c h e , Das Augment des griech. Yerbums, Linz 1882. P d h l m a n n , Quomodo poetae epici augmento temporali usi sint, Tilsit 1858. G r a s h o f , Zur Kritik des homer. Textes in Bezug auf die Abwerfung des Augments, Düsseldorf 1852. K. K o c h , De ['augmento apud Homerum omisso, Brunswick 1868. S k e r l o , Über den Gebrauch (die Bedeutung) des Augments bei Homer, Graudenz 1874. M o l h e m , De augmenti apud Homerum Herodotumque usu, Lund 1876. B i n k e , De augmento verbi Herodotei, Braunsberg 1835. H. L h a r d y , Quaestionum de dialecto Herodoti caput primum: De augmento, Berl. 1844.

24

§477.

The Augment.

R e m a r k 2. It is an obvious suggestion that in tßiatiTijxa etc. we have the augment in place of the reduplication, since in verbs with a vowel initial the augmented preterite and the perfect oame to have the s a m e b e g i n n i n g : e . g . tjgtxhtor:

¡¡(e&iauai

{tge9t^a>),

¡¡axijna:

rjaxtfriai

(anxito),

¿tp-txofitjr : aqriyfiai. But if so we should expect similar forms in verbs which began with one consonant, and such forms as *e-ßqxa instead of ßt-ßtjxa

(see § 475).»)

We cannot suppose that the form i-qgaya stands for *ft-ggo>ya, and that it gave the type for i-ßli«rt/xa etc., because the dialect of Gortyn a has t-yQafjuai, and this dialect kept initial J- before an «-sound. The Cretan perfects ¿n-rjareXxe ij-yoaTrm mentioned in § 472 Rem., page 17 above, with r'-, only give a fresh problem to solve.

THE §

477.

The

syllable, Idg. *e- =

AUGMENT.2)

Augment Skr. a-

(avfyotc),

as it is called,

Armen, e- Gr.

is

a

which prefixed

to verbal forms serves to mark past time. 1) eXoyxa instead of IrXoyya in two late sepulchral inscriptions is probably not a mistake in the graving, but a misformation, due to contamination of the perfect HHoy/« or eil/j/n at a time when this kind of perfect had become unfamiliar. (Thumb, Mitteil, des deutsch, arch. Inst, in Athen, xvi 176). 2) R. G a r n e t t , On the Origin and Import of the Augment in Sanskrit and Greek, Proceedings of the Philol. Society I (1844) p. 265 ff. F r . M ü l l e r , Einiges über das Augment, Kuhn-Schleicher's Beitr., in 250 ff. J . D a v i e s , On the Temporal Augment in Sanskrit and Greek, Hertford 1865. F a u s t , Zur idg. Augmentbildung, Strassb. 1877. A. H. S a y c e , The Origin of the Augment, Transactions of the Philol. Sooiety, 1885—1887, pp. 652 ff. B r 6 a l , De l'augment, M6m. de la Soc. de lingu. vi 333 ff. J . A v e r y , The Unaugmented Verb-Forms of the Rig- and AtharvaVedas, Proceedings of the Amer. Orient. Soo., May 1884, pp. XI f., und Journal of the Amer. Orient. Soc. xi 326 ff. E b e 1, Die scheinbaren Unregelmässigkeiten des griech. Augments, Kuhn's Zeitschr. iv 161 ff. L a R o c h e , Das Augment des griech. Yerbums, Linz 1882. P d h l m a n n , Quomodo poetae epici augmento temporali usi sint, Tilsit 1858. G r a s h o f , Zur Kritik des homer. Textes in Bezug auf die Abwerfung des Augments, Düsseldorf 1852. K. K o c h , De ['augmento apud Homerum omisso, Brunswick 1868. S k e r l o , Über den Gebrauch (die Bedeutung) des Augments bei Homer, Graudenz 1874. M o l h e m , De augmenti apud Homerum Herodotumque usu, Lund 1876. B i n k e , De augmento verbi Herodotei, Braunsberg 1835. H. L h a r d y , Quaestionum de dialecto Herodoti caput primum: De augmento, Berl. 1844.

§§ 477,478.

The Augment.

25

It was originally an independent word, an adverb, followed by the verb, which then became enclitic; e. g. liqet 'he left* (Armen. e-lill Gr. e-Xms), and it may be compared with the Irish particle ro ( = Gr. ngn) which is used before verbs of the historic tenses. But in all the languages which have kept the Augment, it has become an inflexional prefix (cp. II § 4 page 6). A trace of its original adverbial character remains in the accentuation of Greek forms like 7iag-t-a%ov ("I offered'), which involves the same principle as that of rrag-sv-SsQ ('put in between') and of Skr. sam-d-cinute ("he heaps together, collects'). As to the origin of this adverb *e, and of *e, which as we shall see later was used in the same way in the parent language, only uncertain conjectures are possible. R e m a r k . Older explanations are cited by Curtius, Verb I s 109 ff. Sayce's new suggestions do not commend themselves to me (see page 24 footnote). It would be best to regard *e as a locative of the pron. stem 0-, with temporal meaning (see III § 409 p. 329); compare He (Lith. tl O.C.81. te) from *to- and the like (III § 424 p. 349). The relation of *e : has plenty of parallels, as *te : *te, *ne : *ne (IH p. 349 footnote, §415 Rem. p. 337). Compare also Per Persson, Studia etymologica, p. 78.

If the verb had other prefixes besides the Augment, this stood immediately in front of the verb. But sometimes a verb compounded with a preposition became to all intents and purposes a simple form, and then the augment came right in front. Skr. a-pidaya-t„ 'pressed' for *pi-zd- ("sit upon'), Gr. 1-nis^ov for m(a)sd-, see § 795. Skr. epic a-sambhramat 'he trembled'. Gr. Att. h-xud-svSov beside xa&->jvdov 'I slept'. When the structure of verbs was thus forgotten, there could even be a double augment: Skr. epic apraiMt beside prcli$it = pra-aisit 'he drove out', Gr. rjv-sixinijv 'I endured*. The same thing occurs in reduplicated forms; see § 850. § 478. The augment with verbs b e g i n n i n g in a C o n s o n a n t . Examples: Pr. Idg. *l bherorn 'I bore': Skr. a-bharam E l o p p e , Dissert, de augmento Herodoteo, cp. I. II., Schleusingen 1848. Sor o f , De augmento in trimetris tragicis abiecto, praemissa de crasi, elisione, aphaeresi quaestione, Breslau 1851.

26 Avest.

The Augment.

§478. rd

O.Pers. a-baram, Gr. t-ytgov. 3 sing. Skr. Armen. e-d Gr. e-ri&si 1 st pi. s-9-efttv, y/^dheplace'. 3 rd sing. Skr. a-bodhat a-buddha a-bubudhat Gr. s-nm>9STO t-NV&TXO snsnvaro, Y/^bheudh'awake, notice'. 3rd sing. Skr. a-dista a-diksat Gr. s-Stixvv \fdeik'show, point'. 3 rd sing. Skr. a-gan Armen. e-kn, y/^qetn- 'go, come'. 3rd sing. Armen. e-tes Gr. s-depxsro, derk- 'see'. All that is left of the augment outside of these three groups are a few obscure Germanic forms: Goth, iddja 'he went' = Skr. d-yat (I § 142 p. 127), A.S. 3 rd pi. eodun = Goth, iddjsdun, cp. §§ 587, 592, 886 Rem. But these are not free from doubt, because^ we find in Sanskrit epics the unaugmented form iyCL-t as well as a-yd-t (with iy- instead of y- like iy-e, § 493). So iddja too may represent the un-

d-da-dh&t

a-berem

a-dhcLt

augmented Idg. *iie-t.

In G r e e k , i- was often obscured by being contracted with the following vowel, after a or / which once began the root had dropped (cp. I § 165 p. 146, § 564 p. 421, § 603 pp. 455 f.); e. g. slno/Mjv for *s-(n)tnof.iav from sno/nai 'sequor', slpnov for *s-(o)fgnov from epnco serpo', slSov 'I saw' for *f-(f)i6ov (Horn, std'ov, Lesb. svid'ov), HgyaConitjv for *s-(f)Egya^Ofii]v (an inscr. of Hermione has sfepydtfazo) from 'eQydCof.au 'I work'. On tl&tlov 'I was accustomed', orig. *t-oft&iZov, nXxov 'I dragged', orig. H-ofslxov, compare I § 563. 7 p. 420, and the Author Gr. Gr. 2 § 13 p. 33. The aspirate of elno/Lojv elkxov slpnov, like that of tr/xa rjy.a pi. slfisv (for t-(d)rj- ¿-(it)«-, from "qui 'I send forth* for *ai-a>jfu) is doubtless due to the transference of the internal h (*e~hsnof*8v etc.) to the beginning; so it was in legos for *lhsgo-s (Skr. isird-s) and other words, see Kretschmer, Kuhn's Zeitschr. xxxi 421. In Greek, again, the augmented preterites of verbs which have lost their initial consonant are often treated like those of verbs that never had any (§ 480). This is commonest in later times. An example is Att. wxijaa instead of *i-(Jz)ot-xi]au, from (f)otxfM 'I live, dwell', following such forms as wStjaa (offico 'I swell': Armen. aitnum 'I swell' O.H.G. eii 'sore, abscess').

§479.

27

The Augment. § 479.

In Yedic, some verbs beginning with v, y, or r

have Or for their augment; as a-vpnak round' (see Delbrück, Ai. Verb.,

79).

augment with verbs beginning in 'wets' (see § 481).

from vptdkti

'turns

a - is also found as as ätinat

from

unätti

That this a- was original, at least in some

of these forms, is made probable by the use of rf- instead of Inot infrequently before Greek verbs with initial u. Horn, ijeidrj takwv

'I

'he knew'

was

captured'

for

; A t t .

for *q-(h)opaov

Examples:

SWQCOV

*rj-(h)a\aw

'I saw* (I § 611

p. 4 6 2 ) ; the aspirate in the last two is to be explained like that of sino/urjv eyxu (§ 478).

Words like ijpyu^o^itjv (beside

tlgyaQoi-iriv) are

It is true

less certain.

this form may

be

derived from *q-(f)sQy-; but so may it be an ad-formate of verbs which began with e- in the pre-Greek period (cp. toxyoa § 478, above). On the relation of *e- and *e- see § 477 with the Kemark, page 25. R e m a r k . On Greek forms with ij- compare G. Meyer, Gr. Gr, 8 pp. 421 ff.; the Author, Gr. Gr. 2 p. 150, and the works there oited. Another yiew, which I think not probable, is that certain verbs with initial f have a prothetio I-, and that from these were made preterites with the temporal augment (yetfij being to huidutvoi what ¡¡aevSov is to fgev&w X^reifdh-, cp. I § 626 pp. 470); afterwards, according to this view, other verbs with initial f but without prothetic t- took y- for augment. On the other hand, I agree with those who refuse to see 'the Idg. augment *e- in frßovXö^v 'I wished', y-Svvanyy 'I could', and %-neXXov 'I intended, I was to* do so and so. These forms occur in Attic from 300 B. c. onwards, and tjueMov as early as Hesiod (Thesg. 478, 888, 898). It is a fair conjecture that these were modelled on r r »tiov 'I wished', from »(Ila). In this Terb tJ- was no augment but a preposition, another ablautfrom of w- in w-rpeiiw etc., and identical with Skr. d 'to towards*. The iof l-9fita is a third ablaut-form of the same prefix, to be compared with a in Avest. a-säs- etc,; but the use of ¡¡- with S-e'ioi was confined to its preterite in prehistorio times, beoause ¿ttttio : fötXor seemed to be related as were igifa : {¡gifyr. Cp. Osthoff, Perf. 129, 604; Bartholomae, Ar. Forsch, n 169; above, vol. I l l § 246 p. 145. 1) Some scholars would write eve0rj in the Aeolio fashion, for which there is no authority at all. Attie ¡¡SU cannot be contracted from HISbut only from j ] H S - .

28

The Augment.

§480.

§ 480. In verbs with i n i t i a l S o n a n t the augment has everywhere ceased to be a separate syllable. It was contracted with the root-initial in the original language (cp. I § 114 p. 107). Examples. Pr. Idg. *esqi for *e esqi or *e estp, cp. pres. *es-mi = Skr. asmi etc.: Skr. dsam Avest. 3rd sing, as O.Pers. aham i. e. aham, Gr. Horn, rja Att. t} 3 rd sing. Dor. yg;') cp. O.C.S1. -jachu for *esom in imperfects like nesiachu (§§ 493, 510, 903). Pr.Idg.: from *ei-mi 'I go': Skr. rd rd dyatn 3 sing, alt Avest. 3 sing. Hip O.Pers. ay am i. e. ayam, Gr. fja instead of *>ju for (§ 502); compare Lith. ejau 'I went' from the stem *ei-a- (§ 586). Gr. jjgtCov from ipitui 'I strive'. O.C.S1. s-aorist jasu — *St-so-m, y/^ed- 'eat'. It is extremely probable that the same augment is seen in Lithuanian present forms of the substantive verb beginning with 8-, as pi. esame esate dual esava esata beside esame etc. and hsme (esme) etc. Like O.C.S1. -(j)achu -(j)ase etc. (see above), these were originally imperfect. But after all the other preterites of present stems with thematic vowel had fallen into disuse, this imperfect of es- was quite isolated; step by step it gave way to buvau, while at the same time the forms which ended like those of the present system came to be used as equivalent to them; and later the participle esfys was formed and used side by side with esqs, and in some dialects esii esi beside esu esi. Perhaps Lat. es 'thou art' (also es) is also an augmented form, and represents Idg. *es-s.2) Remark. Osthoff (Perf., 184 if.) assumes that Lat. es est estis from edd, and Lith. $dmi idu etc. O.C.S1. jami (emi) are forms of the augmented imperfect used as present. I think that their e- may very well have this origin. But another supposition is quite as good, nay 1) We are oertainly tempted to follow Bopp, Lagarde, and Bugge, and add Arm. ei 'eram' 3rd sing, er; but Idg. g seems always to become Arm. i. Compare Hubsohmann, Kuhn's Zeitschr. xxin 12. 2) So too the augment has crept into the present and future in Modern Greek, as na? c/Siena, 9a aus iSmata (Hatzidakis, Kuhn's Zeitschr. xxx 375); and so the augment of Armen. e-kn 'he came* and e-d 'he placed* has found its way into allied forms, as fut. ekic and edic (Hubsohmann, Arm. Stud, l 28; Bugge, Kuhn's Zeitschr. xxxn 38).

§§ 480,481.

The Augment.

2»

better — that they represent the ablaut-grade Idg. *ed-. Then *edmi: *edmi (Skr. ddmi) as Skr. imper. mid. sdk-sva : sdk-sva (v/"1 segh-), and as Skr. sdhati : sdhate (Gr. r^erat), dh&vati : dhavate (Gr. 9ia>), Or. ft^Sopm : peSo/jm, Lith. b&gu: Gr. ((¡¿fiouat, and so forth (§ 471 p. 16); and, if so, the ed- of *edmi must be identified with that of the s-aoriat Lat. essem O.C.S1. jasil (em), and probably with that of the perfect forms Lat. edi, Lith. M%s O.C.Sl. jadu (edu); see Inc. cit. above. One view only I must distinctly oppose; the view of those who regard this root as not belonging to the recognised e-series, but to an ¿-series, and who regard Skr. ddmi as not original, while the *edmi of Latin and Balto-Slavonio is. Compare § 494. How came this e-grade {mid- etc.) to exist in «-roots? It is impossible to say. It is found, in the proethnic stage and later, not only in the present and aorist, but in the perfect too (§ 848. 3), and also in nouns (e. g. Gr. /ur,Sos fii,aru>Q Armen. mid); and we have no right to limit this e to any single tense. It is certainly remarkable that the perfeot forms with e never had the reduplication (Skr. sah-vds- etc.). But there are other original unreduplioated perfeets, as for example *iioide 'knows') gee loc. cit.).

What was originally the quality of the resultant vowel, when an augment was contracted with the initial a- or o- of a root, it is hard to say. The Aryan languages of course always have O-, whether the root began in e, o, or a; as Skr. djat from djati 'he drives' Gr. aytt, dpasyat from apasydti 'is active' Lat. operatwr. In Armenian, verbs beginning with a- have apparently no augment in the preterite, as ac 'he led', ar "he took' auc 'he anointed' anc 'napijlfrs'; we also find a re-formation with augment i-anc (with later e instead of e). Greek forms like ayov Ion. ftyov (from ayto 'I lead'), w£of (from opu 'I smell"), codtjOu (from oldtio 'I swell') are suspicious, because their long vowel might come from analogy, once such forms as *esti (son) : *est (ys) had produced a belief that the lengthening of an initial vowel marked the past tense. Beyond all doubt this is the cause of the long vowel in such words as ixtTtvoa (from ixsreva) 'I beseech') and v/usvaiovv (v/nsvaiw 'I sing the bridal song'); compare what is said is § 643 on ¿Q-vv-fu : op-vv-fis». § 481. In the plural and dual of the pret. of 'I am' and *ei-mi 'I go', if Idg. *e- is allowed to be augment, we should expect forms like Skr. *a-sma Gr. *(f/ii(v Lesb. *e/nfiev and Skr. *ema *a-yan, Gr.

*es-mi

their *a-san, *el(*fr,

30

The Augment.

§§ 481,482.

op. pres. l 8 t pi. Skr. s-mas, i-rnds and the unaugmented imperf. Skr. Yed. s-an Avest. h-en and Avest. i-ta Gr. '¿-TIJV. One such form is Avest. dhma Gathic ehma — pr. Ar. *a-sma. Otherwise we find only asma dsan aima dyan, Avest. 3rd dual aitem, O.Pers. 3rd pi. aha aya i. e. doubtless aha Clya, Gr. rjfxev ¿¡ore, O.C.S1. 21"1 pi. -(j)as-te. If we suppose that the augment here was e (cp. § 477 and Rem. p. 25, § 479 pp. 26 f.), the sing, and dual-pl. agree in their initial syllable right back as far as the parent speech. However, it is possible, and probably better, to assume that the long vowel came from the singular, the initial of dsam fja as compared with dsmi d/ui being classed in the popular imagination with that of the preterites djam dnam ijpiov etc., which had a long initial vowel in all persons. R e m a r k 1. ^ in fair ¿¡re must be a re-formation (cp. I § 611 p. 461). But there is no need to bring in the influence of the sing. since *e + i-ent may oonoeivably have been the 3 r d plural (cp. § 1020.

1. a).

R e m a r k 2. Osthoffs view (Perf., 151 f.) that Skr. dsta Gr. r,are came from *i esti, and that *esti was the weak-grade form of \/es- with secondary or bye-acoent, is no longer tenable. See Bartholome, Bezz. Beitr. xvii 105.

In Sanskrit, verbs beginning with i-, or f - have Ou-, Hi-, and Or- in their augmented preterite. dun at from unatti 'wets' (wd-). duhat from tihati removes, pushes', aichat from ichdti 'wishes'. diiata from iSate 'owns'. drchat from ¡rchdti 'reaches, gets'. The augment here was probably e; see § 479 above, on d-vptak etc. Other attempts to explain these are given by Schleicher Comp.4 p. 738 (cp. J . Schmidt, Yocalismus, i 44) and Bartholomae, Ar. Porsch. ii 74 f. § 482. In Herodotus are a series of apparently unaugmented forms, of which aXxst {ahsw 'I ask*), tv/exo (sv/oftou 'I pray), avj-ero (av'gco 'I increase") are examples. These may be quite regular, and come from older forms with initial CLjr, eyr, Ciu-, as laid down in vol. I § 611 p. 461. The vowels in the first syllable of such forms as Att. TjTOvv (alreta), ipi^d/xtjv (fti/Ofiai), rjvS/ov {uv% oadunt? Oppeln 1836. G. M e k l e r , Griech. verba oontraota mit langem Themavooal, in Beiträge zur Bildung des grieoh. Verbums (Dorpat 1887) pp. I f f . H. von d e r P f o r d t e n , Zur Gesohichte der grieoh. Denominativa, Leipz. 1886. L. S ü t t e r l i n , Zur Geschichte der verba denominativa im Altgriech. I, Strassb. 1891. L o b e c k , De mutatione terminationum coniugationis circumflexae, Königsb. 1845. G. C u r t i u s , Zur Geschichte der grieoh. zusammengezogenen Yerbalformen, Curtius' Stud, m 377 £ B. M a n g o l d , De dieotasi Homérica, imprimis verborum in -aio, ibid. vi 139 ff. F. D. A l l e n , The Epio Forms of Yerbs in cèco, Transact, of the Amerio. Philol. Assooiat. rv (1873) pp. 1 ff. J. W a c k e r n a g e l , Die episohe Zerdehnung, Bezzenberger's Beitr. rv 259 ff. I n a m a , Degli aoriati greoi, Bivjsta di filol. n 249 ff. L. M e y e r , Grieoh. Aoriste, Beri. 1879. A. Zi o k i e r , De oausis duplicis formae aoristi Graeoi, 1865. Th. N ö l t i n g , Über den genetischen Zusammenhang des Aoristus I I mit dem Perfeotum II der grieoh. Spraohe, Wismar 1843. T h e A u t h o r , Über einige grieoh. Präteritalformen mit a vor der Personalendung, Bezzenberger's Beitr. u 245 ff. L. D o e d e r l e i n , De aoristis quibusdam seoundis, Erl. 1857. S»

36

Formation of the Tensé Stem.

§485.

§ 485. The first point to realise is that there never was any real difference between the P r e s e n t stem and the Strong Aorist. There is no difference, for example, between the imperfect Skr. d-bha-t Gr. s-tptj (ybha'to "W. S c h u l z e , Zwei verkannte Aoriste (ta^ov und ¿ W ] , Kuhn's Zeitsohr. x x i x 230 ff. E b e l , Reduplicierte Aoriste im Griech., ibid. n 46 ff. G. C u r t i u s , Der erste Aorist des Passivs, ibid. I 25 ff. J . "W a o k e r n a g e l , Der Passivaorist auf -¿yr, ibid. x x x 302 ff. "W. K ü h n e , De aoristi passivi formis atque usu Homerico, Harburg 1877 and Güstrow 1878. " W a l k e r , Greek Aorists and Perfeots in -*a, Class. Review, v 446 ff. H a t z i d a k i s , Zur Präsensbildung des Neugrieohisohen, Kuhn's Zeitschr. x x v n 69 ff. A l b a n i a n . G. Meyer, Das Yerbum snbstantivum im Albanesisohen, in M. Hertz zum 70. Geburtst., 1888, pp. 81 ff. I t a l i o. C o r s s e n , Zur Bildung der Präsensstämme, in Beitr. zur ital. Spraohkunde pp. 475 ff. C l u d i u s , Über die Bildung des "Verbi sunt, Günther und Wachsmuth's Athenäum n (Halle 1817) 136 ff. J. D a r m e s t e t e r , De ooniugatione Latini verbi dare, Paris 1877. Ph. T h i e l m a n n , Das Yerbum dare im Lateinischen, Leipz. 1882. F. F r ö h d e , Die lat. Präsentia auf -Ilo, Bezzenberger's Beitr. in 285 ff. K. F. J o h a n s s o n , Nägra ord om de latinska verbalbildningarne med n i presensstammen, Akadem. afhandlinger til prof. S. Bugge, Christiania 1889, pp. 21 ff. C h . P l o i x , Des verbes latins en sco, Mém. d. 1. Soo. d. lingu., vi 399 ff. K. S i t t l , De linguae Latinae verbis inoohativis, Arohiv f. lat. Lexikogr. I 465 ff. C. P a s c a l , I suffissi formatori delle oonjugazione latine, ßevista di filol. x i x 449 ff. R. T h u m e y s e n , Über Herkunft und Bildung der lat. Yerba auf -io der 3. und 4. Conj. und ihr gegenseitiges Verhältniss, Leipz. 1879. C. P e t e r , Über die schwachen Yerba der lat. Sprache, Rhein. Mus. m 95 ff., 360 ff. M. B r é a l , Yerbes dérivés latins, Mém. d. 1. Soc. d. lingu. vi 342 ff. F. d e S a u s s u r e , Sur une classe de verbes latins en -eo, ibid. m 279 ff. C. P a u l i , Gesohichto der lat. Yerba auf uo, Stettin 1865. 0 . I. F e h r n b o r g , De verbis Latinis in uo divisas desjnentibus, Stockholm 1889. C. P a u c k e r , Die verba denominativa auf -are, Kuhn's Zeitschr. x x v i 261 ff., 415 ff. R. J o n a s , De verbis frequentativis et intensivis apud comoediae Latinae soriptores, (i) Posen 1871, (n) Meseritz 1872. Idem, Gebrauch der Yerba frequentativa und intensiva in der Siteren lat. Prosa (Cato, Varrò, Sallust), Posen 1879 und 1884. C. P a u o k e r , Die verba frequentativa, Kuhn's Zeitsohr. x x v i 243 ff., 409 ff. W ö l f f l i n , Die Verba frequentativa und intensiva, Arohiv f. lat. Lexikogr. IV 197 ff. Idem, Die verba desuperlativa, ibid. II 355 ff. G. C u r t i u s , Über die Spuren einer lat. o-Conjugation, Symbols philoL Bonn. 1864 pp. 271 ff. = Kleine Schriften Ii 133 ff. W ö l f f l i n , Die verba desiderativa, Arohiv f. lat. Lexikogr. I 408 ff. G. C u r t i u s , De aoristi Latini reliquiis, Kieler Leotionsverzeichn. 1857—58 = Curtius'

show, disclose, s-att]

37

Formation of the Tense-Stem.

§ 485.

(y^sta-

inform')

(^dreugh-

'deceive')

the

Skr. budhd-nta

aorist

and the aorist Skr. a-stha-t

'stand'); between the imperfect Skr. Gr. e-yXvq>f

Gr.

d-druha-t

gley.bh- 'split, incise') and

Gr. t-nv&e-to

(\/~*bhey,dh-

'wake,

Stud, v 429 ff. C o r s s e n , Kein Aoristus II im Lateinischen, in Beitr. zur ital. Sprachk. pp. 538 ff. F. G. F u m i , Sulla formazione latina del preterito e futuro imperfetti, Progr. del R. Lieeo Chiabrera in Savona 1875-76. E e l t i o . D ' A r b o i s de J u b a i n v i l l e , Etude sur le présent du yerbe irlandais, Mém. d. 1. Soe. d. lingu. y 237 ff. "Wh. S t o k e s , The Neo-Celtic Verb Substantive, Trans. Phil. Soc. 1885—87, pp. 202 ff. Idem, The Old-Irish Verb Substantive, Euhn's Zeitschr. xxvni 55 ff. W i n d i s oh, Das ir. praesens seoundarium, ibid, XXVII 156 ff. Idem, Das ir. ¿-Präteritum, Euhn-Schleioher's Beitr. vui 442 ff. T h u r n e y s e n , Das sogen. Präsens der Gewohnheit im Irisohen, Idg. Forsch. I 329 ff. L o t t n e r , Traces of the Italio imperfeot in the Keltio languages, Trans. Phil. Soc. 1859, pp. 31 ff. T h u r n e y s e n , Zu den ir. Verbalformen sigmatisoher Bildung, Euhn's Zeitsohr. XXXT 62 ff. — Further on page 4, footnote. G e r m a n i o . A m e l u n g , Die Bildung der Tempusstämme durch Vokalsteigerung im Deutschen, Berl. 1871. P e t e r s o n , Vom Ablaut mit bes. Rücksicht auf den Ablaut des starken Zeitworts im German., Lund 1877. A. M o l l e r , Die reduplioierenden Verba im Deutschen als abgeleitete Verba, eine etymol. Untersuchung, Potsd. 1866. H. L i c h t e n b e r g e r , De verbis quae in vetustissima Germanorum lingua reduplicatum praeteritum exhibeant, Nancy 1891. G. B u r g h a u s e r , Idg. Präsengbildung im German., Wien 1887. J. von F i e r l i n g e r , Zur deutsohen Conjugation (Präsentia der Wurzelolasse, Zur westgerm. Flexion des verb, subst.), Euhn's Zeitsohr. XXVII 432 ff. H. E e r n , Over eeinige vormen van 't werkwoord zijn in 't Germaansch, Taal- en Letterbode v 89 ff. J. S c h m i d t , Die german. Flexion des verbum substant. und das hiatusfSllende r im Hoohd., Euhn's Zeitsohr. xxv 592 ff. W. W i l m a n n s , Die Flexion der Verba tuon, gän, stän im Ahd., Zeitschr. f. deutsch. Alterth. xxxiTT 424 ff. S k l a d n y , Über das gotisohe Passiv, Neisse 1873. Egge, Inchoative or »-Verbs in Gothic, Amer. Journ. Phil, vn 38 ff. S i e v e r s, Zur Flexion der schwaohen Verba, Paul-Braune's Beitr. vm 90 ff. K ö g e l , Die schwaohen Verba zweiter und dritter Classe, ibid. ix 504 ff. T h e A u t h o r , Die got. Imperativform hiri und die Denominativa von consonantisohen Stämmen, Morph. Unters, iv 414 ff. B a l t o - S l a v o n i c . G. U l j a n o v , Znacenija glagolnych osnov v litovsko-slavjanskom jazykë (meaning of verbal stems in Lithu-Slavonic), Russkij filol. vestnik xxiv 105 ff., xxv 41 ff. 0 . W i e d e m a n n , Das litau. Präteritum, ein Beitrag zur Verbalflexion der idg. Sprachen, Strasab. 1891. L e s k i e n , Die Präsensbildungen des Slavisohen und ihr Verhältnis« zum Inflnitivstamm, Aroh. f. slay. Philol. v 497 ff. M i k l o s i o h ,

38

Formation of the Tense Stem.

§485.

notice, learn'). Often the same form is imperfect in one language and aorist in another; the forms ¿-§ene-t (y^gen'gignere') and *6-d$ce-t (v^ defile- 'bite') are imperfect in Sanskrit (Ajanat ddafat), and aorist in Greek (iyivtTO siiaxf). Or the same form is both, in one and the same language; Skr. d-pa-t 'drank' is imperfect of the pres. pd-ti, and aorist of the pres. piba-ti. "What the meaning of a given form was, whether imperfect or aorist, depended on its relation to others. See Delbrück, Ai. Yerb. p. 16, Ai. Tempuslehre p. 5. For our purpose, then, the stems of the present and the strong aorist go together; and where it is advisable to refer to the difference in the kind of action implied, we shall use the terms imperfect-present and aorist-present. Some of the forms which in grammars of this or that language are called Future Indicative were originally Conjunctive; for example, Lat. eri-s age-s. These will be found under Conjunctive (§§ 910 ff.). In form they belong to the Present. In the same place will be found the Idg. series of forms built up with the suffix -sio-, as Skr. da-sydmi Lith. dä'-siu 'dabo'. The -sio- stands on the same level as -so-sko- and other formative suffixes used in the present tense; it is probably made up of -s(o)- + -¿o-, as -nj,o- is of -n(o)~ + (§ 743). Thus these futures are treated under the Present Tense. With the Present also should strictly speaking be classed the s-Aorist. Its characteristic s cannot be separated from the s which is so common in present and regular in future stems; and its whole inflexion follows the same principle as the present. The s-aorist would properly go with Class X I X of Present Stems (cp. §§ 655,656). A separate chapter is given all the same to this Das Imperf. in den slay. Sprachen, Sitznngaber. d. "Wien. Akad. Lxxvn 5 ff. O. W i e d e m a n n , Zar Stammbildung der Yerben auf -tiqti, Aroh. f. slay. Philol. x 652 ff. W. Bur da, Ein Beispiel der Prfisensstammbildung mittels ta im Slayischen, Euhn-Sehleicher's Beitr. VI 392. Miklosich,

"Verba intensiya im Alt slowenischen, ibid. I 67 ff. Idem, Aorist [in Old-Slovenian], Sitzungsber. der Wien. Akad.

LXXXI

Einfacher

100 ff.

§á 4 8 5 - 4 8 7 .

Formation of the Tense Stem.

39

Aorist (§§ 810 ff.), but merely to assist in getting a general view of this large group of forms. As to the proethnic Perfect, as *dedorke = Skr. dadárSa Gr. dsäopxf (v° derb- see"), it is distinguished from present forms by its grade of ablaut in the singular indicative active, by some special personal endings in the indicative, and (if we include the verb infinite) by a peculiar formation in the participle active. The remaining forms of the perfect system, with which we must include the pluperfect, have exact counterparts in the system of the present, and nothing but its use can tell us whether a given one of these forms is perfect, present, or aorist; even the reduplication with e is not confined to the perfect (§ 471 pp. 15 f.). Often the kind of action denoted is so little obvious, that grammarians doubt whether to class certain forms under Perfect or Present Stem (cp. Delbrück, Altind. Yerb. 122 f., Whitney's Sanskrit Gram. § 868, Curtius Yerb II 2 24 f.). It is clear that notwithstanding these points of contact between the two classes, a special chapter must be given to the perfect, on account of the peculiarities which it has. "We therefore divide Yerb Forms, from the point of view of the Formation of these Tenses, into three groups: I. P r e s e n t (including I m p e r f e c t - and A o r i s t - P r e sent). II. s-Aorist. III. Perfect. § 488. Before we proceed to our subject in detail, two distinctions must be explained which are usually made, and to which some attention must be given in discussing Tense Morphology. These are (1) the distinction between Primitive or Primary verbs, and Derivative or Secondary verbs (Denominative or Deverbative): (2) that between Root-Determinatives, and TenseSuffixes, or the elements used in forming a tense stem. § 487. First — Primitive and Derivative Yerbs. Primitives, such as *es-ti 'est' and *a§e-ti 'agit', are contrasted with two classes of derived verbs: (1) a class which in the formation of

40

Formation of the Tense Stem.

§487.

the stem is wholly verbal, as much as are the primitives; as Sanskrit Desideratives and Intensives (ní-m-$a-ti ne-ni-yd-te from náya-ti 'leads'), and Inchoatives in Latin (gemiscd from gemo): (2) those which clearly contain a Noun Stem, called Denominatives; as Skr. gatu-yá-ti 'procures access from gütú-s 'access', Gr. noi/naívw 'I tend' from notfiijv 'herdsman', Lat. planta-t from planta. (1) The formation of Desideratives, Inchoatives, Intensives, Iteratives, Frequentatives, Causatives and the rest is in principle absolutely the same as that of the so-called Primitive verbs connected with them. There is a distinction, however, in the meaning of the present tense; in these verbs the present had a second special meaning in addition to that of time. This distinguished them from the primitives, which had a simpler meaning in the present, and the formation with some special meaning became a more or less fertile type. But they were not originally derived from their primitives for the express purpose of conveying this new meaning; the new meaning, indeed, does not always date from the proethnic stage at all, but belongs to a later period, and it often has nothing to do with the form, but depends on other circumstances. This is the case with Lat. inchoatives in -sco (§ 674). Thus it is clear that we cannot use this different meaning as a principle of classification; our aim is historical, and we base our classification on the conditions which prevailed in the proethnic stage, and as far as possible on etymology. W e must then be content to point out the special meaning where it is of any importance, and, wherever it is possible, to explain how the meaning came in. (2) It is less easy to classify the second group of forms, and to find out how far indicative stems, which we see combined with personal endings into a word, are purely verbal, and how far they are wholly or partly nominal. I f we could only see which were based on noun stems and which derived from verbs, this would of course be the main principle of distinction. There is no manner of doubt that the Idg. languages had

§487.

Formation of the Tense Stem.

41

not only denominative verbs with an additional suffix between stem and personal ending — such as Skr. gatu-yd-ti apas-yd-ti ('is active', from dpas- 'work') Lith. pdsako-ju ('I recount' from pasaka account'), which have -io- between stem and ending — but also others where the personal suffix was added immediately to the noun stem. Such forms are Lat. plantd-s planta-t etc. from planta, O.H.G. salbd-s 'thou anointest' salbd-t etc. from salba ointment', Lith. j&'sto 'he girds' justo-me from j&'sta 'girdle', Aeol. xt(Ji&-(iiv 'we honour' from xl/.ia 'honour'. Also Skr. mdrga-ti 'tracks, traces' from mOrga-s 'path, track', phala-ti 'bears fruit' from phdla-m 'fruit', Gr. 9e^it-xo 'grew warm' from e-xpcuofie 'was useful' beside /paio/^tm, Skr. jiva-ti Lat. wvi-t O.C.S1. Mve-tu 'lives' from jl-vd-s vT-vo-s zi-vu 'alive'. With very good reason, all forms with a thematic vowel, and therefore all presents formed by -o- -no- -to- -io- etc., have been explained as noun-stems with added personal endings (so, for example, Curtius Yerb I 2 14 f., 161, 239, 296): as specimens take Skr. dja-ti 'drives' Gr. ayci Lat. agi-t with Skr. ajd-s 'driver' Gr. ayo-g Lat. prod-igu-s; Skr. pana-tS 'buys' with pana-s 'wager, stipulation' Lith. pelna-s 'profit' (I § 259 p. 212), Skr. vena-ti 'yearns' with vena-s 'yearning', Goth, fraihni-p asks' with Skr. praSnd-s 'question'; containing -%no- -eno- -ono- (Class XIY): compare Skr. kxparia-te 'he acts pitifully, begs' with Icfpayd-s 'pitiful, poor', Gr. d-yyavet 'sharpens' with fhjyavo-v 'something to sharpen with, whetstone', Goth. us-lukni-P 'opens itself' with us-lukn-s 'open', Lith. kiipinu 'I heap up' with kiipina-s 'heaped'; Skr. ve$ta-te 'turns round' with vesta-s 'bond, noose', Gr. e-^Xaors 'grew, sprouted' with fiXaoTos 'bud, sprout'; Skr. pdya-ti "stinks' with puya-rn 'ill smelling discharge, matter'. Even some non - thematic and primitive stems have the same kind of relation to noun stems. For example take Skr. dhjr$nu-mds 'we are brave' and dhfsnu-§ 'brave'. The rootextending suffix -a-, in *bhuu-a- *bhu-a- (Lith. biivo Lat. -bat), *tr-a- (Skr. trd-sva imper. 'preserve, save', Lat. in-trU-mus tra-ns) it seems necessary to identify with the feminine suffix -a-, compare Skr. ji-jydu 'he has overcome' (fut. jyd-

42

Formation of the Tense Stem.

§487.

-sya-ti etc.) Gr. Ion. jSe-jStiy-rat (aor. piti-aaro etc.) with fem. Skr. jyfl- jiya- 'power, superiority' Gr. pia from y/^qei- (Skr. jdy-a-ti ji-na-ti and others). So also -es-, which extends the root in *u-es- 'clothe' (Skr. vas-tS Gr. eni-sorai and other words) must be the same as the neuter suffix -es-, and the tenseformative -as- in Skr. d-jOrif-ur 'they have grown old' the same as -as- the neuter suffix (Gr. yrjpag). Many other proofs will meet us in the course of our enquiry. It need hardly be said that these denominatives or nounverbs did not all appear at the same time. The different types of formation belong to very different periods; and in the earliest strata, e. g. in verbs of Class II such as Skr. dja-ti Lat. agi-t, their noun origin was forgotten even in the proethnic language. But of what verbs, then, can we be certain that when their stem was fused with a personal pronoun it was a verb and not a noun? Of none at all. Even where the stem is the bare root, reduplicated or not, as in *es-ti SO-TI, *sta-t OTRJ, *bhibhajrti Skr. bibhe-ti, the stem may be regarded as a nomen actionis or agentis (cp. the Root Nouns, I I §§ 159 ff., pp. 478 ff.). In the formation of those verbs which are traditionally called Denominative there is nothing to distinguish them from what are classed as primary verbs. Lat. planta-s is just like intra-s hiCL-s, Aeol. ert/ua-fitv like ¿¿ga-/uev szlt]-fj.ev, Lith. j&'sto like blj5s Undo. Even the present formation with -jois nothing peculiar to the denominative class. W e see in Skr. apas-yd-ti pftanO-yd-ti Gr. ovofiaivw etc. the same present secondary suffix -jo- as we see in reduplicated forms such as Skr. dedii-yd-te Gr. yapyaipio (Class XXYII), in forms such as Skr. gfbha-yd-ti, pass. trd-ya-tS, Gr. ipw for Impou for *laa-i.o- (Class XXYIII), and in futures such as Skr. tq-s-ya-tS ved-i$-ya-ti (Class XXX). Lat. plants (for *planta-%o) Skr. pftanO-yd-ti are related to Lat. planta-s Lith. j&'sto just as Lat. intrd (for *intrd-id) Skr. tra-yd-tS to Lat. in-tra-s Skr. tra-ti trd-sva, as Skr. dSdiS-yd-te to dSdi$-fS, and as Skr. fut. vGdi$-yd-ti to aor. d-vSdi$-ma.

§487.

Formation of the Tense Stem.

43

That the term Denominative Verbs cannot be restricted to one special mode of inflexion is clear from many other instances where verbs have been derived from nouns by simply imitating the inflexion of any Primary Yerb. Primary verbs in -6$ (Causatives, and Intensives or Iteratives) were the model for Skr. mantraya-te 'he takes counsel, advises' from mdntra-s, and Goth. fuUjan O.C.S1. plUni-ti 'to fill' from full-s plunu. In Gothic, primary verbs like af-lifnan were the model for ftdlnan 'to become full' from full-s; in Lithuanian, kitpin-ti etc. were the model for such derivatives as Unksmin-ti 'to make cheerful' from lifiksma-s, and virstii vifsti etc. for gelstii gelsti 'to become yellow' from gelta-s. These and similar re-formations will be discussed in § 793. They were due to the fact that there were nouns from the same root as some of the primary verbs, and from these they were believed to be derived. Then real denominatives were formed and used along with these apparent ones. Thus in our classification of verbs, which depends first and foremost upon differences of inflexion^ no use can be made of the traditional distinction between Primary and Denominative. Even if the term Denominative were to be restricted to its common application it would be misleading. The feeling of a speaker for his language can give no help here. Often it cannot be made out whether the speaker regarded a given form as Denominative or not; his feeling often changed according to suggested associations; and if feeling of this sort were made the standard, we should often enough be led to class with Denominatives verbs which were only so by false analogy, and to class as Primary some which were undoubtedly derived from a noun. If again we took as our standard not the feeling of the speaker, but the actual formation of the words, we should be no nearer to getting a settled boundary line. It is easy to say, let those verbs be called denominative which contain noun formative suffixes, thus showing their noun origin, words that is like Gr. rl/udw from rT-fitj, noi/ualvto from itoi-firfv, or Goth, futtnan from fulls (ground-form *p]-nos). But not

44

Formation of the Tense Stem.

§§ 487,488.

to mention that this criterion excludes verbs derived from root nouns, little is gained by this mode of classification; for the task of historical grammar is not so much to analyse the forms and to describe their etymological structure, as to discover their origin and growth. In numberless instances doubts arise as to the correctness of our terminology. The commonest example is that of two classes of verbs running together, a primary and a denominative; e. g. in Greek, verbs in -¿io and verbs in -e-i6 both become -¿a; in Germanic, verbs in -is -¿id and those in -e-i6 -i-i6 both became (Goth.) -ja; in Lithuanian, verbs in -¿id and verbs in -O-mi (-a-id) both became -au (inf. -y-ti). Here the question whether a given verb is primary or denominative is absurd, because it may quite well have been both. For instance, Lith. bradau bradyti 'to wade about' may be derived both from brada, subst. 'wading' on the analogy of justau justyti 'to gird', a denominative from j$s-ta 'girdle', and from bredu 'I wade' on the analogy of -manau

-manyti,

the old "primary" ¿id - byeform of menu 'I re-

member'; and Greek rgonsu) may come from rgono-g on the analogy of vooviw : voa-ro-g, and from XQITIM on the analogy of qiogi(o ( =

Skr. bhardya-ti):

.

But however faulty our grammatical terminology may be, we cannot afford to dispense with it altogether in a book like this. I shall keep the term Denominative for verbs derived from nouns in the later periods, when the verb stem was still more or less felt to be originally a noun; for instance, Skr. gCLtu-ya-ti, Gr. rl/naw, and L a t .

planta-t.

§ 488. Turn we now to the distinction drawn between R o o t - D e t e r m i n a t i v e s and S u f f i x e s or other elements used in forming the T e n s e Stem. What is usually understood, or may be understood, by the term Root-determinative has been set forth in I I § 8 Rem. 2 pp. 20 f. A reference should be added to Curtius, Greek Etymology2 pp. 59 ff., and Fick, Worterb. I V 3 44 ff.1 1) Another work, systematic, and valuable in spite of muoh bold eonjeoture, is Per Persson's Studien zur Lehre von der Wurzelerweiterung

§488.

45

Formation of the Tense Stem.

These elements may appear in any part of the verb. For instance, from Idg. *rS-dh- 'take counsel' come Skr. d-rOdha-t radhnfi-ti

raddhvd

rddhya-tS

ratsya-ti,

raradh-a,

d-rdtsi-t,

rdddhd-s

etc.; from Idg. *sr-eu- sr-u- 'flow* come Skr.

sravisyd-ti,

susrdv-a,

srutd-s

etc.

But

they

are

srava-ti,

sometimes

found only in present or aorist forms, and disappear in the rest; as Lat. per-cello for *-cel-do beside perf. -culi, Lith. vir-du

'I boil' beside

pret. viriavi

inf. vir-ti,

O.C.S1.

zi-vq

'I live' beside aor. zi-chu inf. zi-ti.!) Again, present formativesuffixes, to use the stock phrase, spread beyond their own proper area both in the original language and later. These two reasons make it impossible always to keep Root-Determinatives distinct from Present Formative-Suffixes; the origin of both, by the way, is equally obscure. The tense which we call Present was almost always the foundation for the whole structure of the Verb and its associated noun forms; and the spread of root determinatives over all the verbal system is due to the same principle which from Skr. pi-nva-ti 'fattens' makes the perfect pininva and the participle pinw-ta-s, and makes Skr.

d-ytmk-s-mahi

jungo

jiingiu

(\/^jeug-

L a t . jHnx-%

L i t h . jiink-siu

from

yunkte

'iungere').

There is something else which shows the impossibility of carrying out the usual distiction between Determinatives alid ordinary Inflexions. In discussing the inflexion of the present in primary classes of verbs, it is too common to find the first syllable of a form taken for the uninflected kernel of it. Because in *bhereti 'fert', the syllable bher- is this kernel, that is, the root, therefore in *treseti (Skr. trdsati Gr. rpsa) the syllable tres- is called the root; then, because there is not the same syllable in Skr. tar-ald-s 'moving to and fro, uni Wurzelvariation, Upsala 1891. This has reached me too late for anything more than occasional use. With his treatment of the main questions of principle as set forth on pages 202 and following, I agree. 1) In Lat. vi-vo too the ^o-suffix was once confined to the present. vìxì victum are re-formates, for 'visi *vi-tum. See Osthoff, Paul-Braune's Beitr. vm 274; Stolz, Lat. Gr.* p. 383.

46

Formation of the Tense Stem.

§488.

trembling' Gr. vg-s/xo) Lat. tr-etnd Lith. tr-imii 'I tremble', -es- is called a "determinative", whilst in Skr. vas-te 'clothes himself* GT. eni-eotai (\fey,-,

in Lith. au-nii Lat. ex-uo) -es- is not so called be-

cause these verbs are looked.upon as parallel to forms like *es-ti. But inasmuch as *tres- and *ues- run right through the whole system of their verbs, they have become "roots". And there is no more reason for separating Skr. 1 st sing, tr-ase v-ase from 1 st sing, yaj-ase %tyj-ase than for separating (say) *bhu-o (Lat. -bd O.C.S1. 3 rd pi. bq)

Skr. d-hv-a-t

Gr. s-nX-s from *bher-o

(Gr. (psQ-w) Skr. d-vid-a-t. "We always hear of an "s-suffix" in such words as Skr. yaj-ase; but why? Simply because the ending -asS is not the first syllable of the word. The S of *pl-e- 'fill* (Skr. prd-si

the Soot;

Gr. nXfj-to Lat. -ple-s) is called part of

but it is the same S which we have in *m^n-S-

Gr. e-fiavt] Lith. mbi-e), *tak-e-

(Lat. tac-S-s O.H.G.

dag-e-s),

where it is called Inflexion. And the "determinative" -dh- is called inflexional in Gr. , but not in e-Sga-fro-v i-dag-do-v, or ax-So-fiat. The question whether a verbal element, which can be analysed no further, is or is not a separate syllable has, it is true, some importance; for it influenced the grouping of the forms in the speaker's memory, and this affected the developement of a language in many ways: e. g. the root in Lith. v-ejii O.C.S1. v-ijq 'I wind, turn' ( = Skr. v-dya-ti weaves'), since it formed in itself no syllable, did not follow the course taken by the other verbs in -¿id (Class XXXH) in Balto-Slavonic. But this cannot justify the making a distinction, as is so often done, between things which are clearly connected. Dealing as we do with the parent language, and from this point investigating the growth of the Verbal System, we must discuss together Skr. v-asS and yaj-asS, Greek nk-rj-TO s-^l-i] and f-fiav-tj ¿-{tak-rj.

If, as it seems right to do, a special Present Class is given to *Ss-ti 'is' (Skr. as-ti, y'"'es-), another to *yems-ti 'vomits' (Skr. vdmi-ti, y/^uem-), and a third to *bhse-ti 'chews up, devours* (Skr. psd-ti, \nbhaxs- seen in bd-bhas-ti), it is only consistent to distinguish each of the following as another class of Present

§§ 488,489.

Formation of the Tense Stem.

47

Stems: — a u- : e^-class for Skr. sr-dva-ti Gr. Q-s{f)st 'flows' Skr. d-su-srS-t from y^ser- seen in Skr. si-sar-ti, for Skr. dr-dva-ti 'runs' d-du-dr-uva-t from \/ n der-, seen in Skr. dr-a-ti dr-ama-ti 'runs', and others; an »»-class for Skr. dr-ama-ti Gr. i-dg-afto-v from the above mentioned der-, for Gr. rg-e/^iw Lat. tr-emo Lith. tr-imii *I tremble' from y/^ter-, seen in Skr. tar-ald-s 'moving to and fro, trembling' tr-dsa-ti 'trembles', and others; a class (probably connected closely with the w- : ey,class) for Skr. ji-va-ti Lat. vT-vi-t O.C.S1. zi-ve-tu 'lives' from y^gej-, seen in Avest. gay-a 'life' jy-aiti- 'life' Gr. £jf (for *gi-e-), O.C.S1. zi-ti 'to live', for Avest. ni-Saurvaiti 'defends' Skr. dh&rva-ti 'harms' bhdrvati 'chews, destroys' etc. In the same way we come to a p-class, a M-class, a £-class, and so forth. But this principle will not be consistently carried out, for two reasons. First, in these and many similar classes which might be made only a few examples occur, and thus for our period such formative elements as these can hardly be said to have any real productive power. Secondly, any attempt to make such a classification complete would lead us into labyrinths of root-analysis which would properly be without the scope of a compendium like the present. Roots with this kind of Determinatives, then, which we do not place in any separate class, we shall generally assume to be incapable of further analysis; and thus we place (say) Gr. rp-sfica in the same division as vtfiw and ysfia. § 489. The formation of the Moods, the stems of the Injunctive, Conjunctive, Optative, and Imperative, will follow that of the Tense Stem (§§ 909 ff.). It must however, be here pointed out that the elements which are generally regarded as mood-formative are sometimes etymologically the same as in the indicative. Injunctive and Indicative forms, of course, cannot be separated. And it is beyond all doubt that the short Conjunctive vowel (Gr. -«- -o-), as in *es-e-ti Skr. dsati Lat. erit (indie. *es-ti 'est'), Horn. a-Xs-xat (indie. cU-ro 'sprang'), is the same as what is called the thematic vowel in the Indicative (as *a§-e-ti Skr. ajati Lat. agit).

48

The Present Stem.

§§ 489,490.

Further, I hold that the conjunctive vowel -a- in Lat. f e r a - s etc. is the same as -a- found after weak root-forms in the indicative (Classes X and XI), and also the same as the a. which forms feminine nouns (§ 487 pp. 41 f.); thus Lat. f u - a - m u s belongs to the same class of word3 as the Indie. Lat. -bd-mus (for *f%-a-mos) and Lith. biiv-o-me (§ 578), and that Lat. poscat for *porsca-t, the ihdic. O.H.G. f o r s c o t 'demands', and the Skr. fem. ppchd 'question' (common ground-form *p%k-s1c(l-) in point of etymology must all go together. So also the Italic conjunctive -8- is to be identified with the Indicative -e- (Classes X and XI), and so forth. In all these cases it were proper to keep together whatever forms are etymologically akin. But if we did so, a student who is used to the practice observed hitherto, of arranging forms according to their function, would hardly be able to find his way. So I prefer to give this up, and simply call attention to etymology and structure where it is convenient to do so. THE PRESENT STEM. IMPERFECT PRESENT AND AORIST PRESENT.1)

§ 490. The classes of the Present Stem are very commonly divided into two groups: (1) Thematic, or verbs in -o (Bopp's First Main Conjugation); and (2) Non-thematic, or verbs in -mi (Bopp's Second Main Conjugation. The first group has in the Indicative -o- or -e- just before the personal ending; but -5 is the ending of the 1 st person singular. These vowels were distributed amongst the persons of the singular and plural (we may leave the dual out for the present) in very much the same way as they are in Greek; -e- in the 2nd person of both, and the 3 rd singular, -o- in the 1 st persons (but 1 st sing. pres. act. -6) and in the 3 rd plural: 1) For works bearing on this subject, see footnote to page 33.

48

The Present Stem.

§§ 489,490.

Further, I hold that the conjunctive vowel -a- in Lat. f e r a - s etc. is the same as -a- found after weak root-forms in the indicative (Classes X and XI), and also the same as the a. which forms feminine nouns (§ 487 pp. 41 f.); thus Lat. f u - a - m u s belongs to the same class of word3 as the Indie. Lat. -bd-mus (for *f%-a-mos) and Lith. biiv-o-me (§ 578), and that Lat. poscat for *porsca-t, the ihdic. O.H.G. f o r s c o t 'demands', and the Skr. fem. ppchd 'question' (common ground-form *p%k-s1c(l-) in point of etymology must all go together. So also the Italic conjunctive -8- is to be identified with the Indicative -e- (Classes X and XI), and so forth. In all these cases it were proper to keep together whatever forms are etymologically akin. But if we did so, a student who is used to the practice observed hitherto, of arranging forms according to their function, would hardly be able to find his way. So I prefer to give this up, and simply call attention to etymology and structure where it is convenient to do so. THE PRESENT STEM. IMPERFECT PRESENT AND AORIST PRESENT.1)

§ 490. The classes of the Present Stem are very commonly divided into two groups: (1) Thematic, or verbs in -o (Bopp's First Main Conjugation); and (2) Non-thematic, or verbs in -mi (Bopp's Second Main Conjugation. The first group has in the Indicative -o- or -e- just before the personal ending; but -5 is the ending of the 1 st person singular. These vowels were distributed amongst the persons of the singular and plural (we may leave the dual out for the present) in very much the same way as they are in Greek; -e- in the 2nd person of both, and the 3 rd singular, -o- in the 1 st persons (but 1 st sing. pres. act. -6) and in the 3 rd plural: 1) For works bearing on this subject, see footnote to page 33.

49

The Present Stem.

§490.

compare 2nd sing, erpsot-t; (pegt-at t(ptgt-o (for the indie, pres. act. cp. Goth, bairi-s), 2nd pi. (pegs-re ((ptpt-ze ). P r . Ar. *jan-ti Idg. *qem-ti, see § 493 p. 51. Imperative: Skr. ga-dhi ga-hi Avest. gaidl. 3 rd pi. Skr. a-gm-an gm-dn st Avesi. g"m-en. Skr. 1 dual ganvahi regular for *gqi-u(I § 225 p. 193, § 229 p. 195), only with changed accent. -n- (for -m-) passing by analogy into other weak persons: Skr. 1 s t pi. d-gan-ma 2 nd pi. gan-td gdn-ta beside ga-td, ginstead of j- in Skr. d-gan gdn-tu (Avest. jantu), j- instead of g- in opt. 3 rd sing. Avest. jam-ya-p O.Pers. jam-iyd (Skr. gam-ya-t), see I § 451 p. 334.

§498.

59

Pr.Ar. *ai-ti, Idg. *ei-tl, see § 493 p. 51. 3 rd sing. Skr. e-ti Avest. ae-iti O.Pers. ai-tiy, 3 rd pi. Skr. y-ánti Avest. y-?inti, imper. Skr. i-hi Avest. i-di i-di O.Pers. i-dty. By re-formast tion: Skr. 1 sing, i-tni instead of e-mi. Preterite 1 st sing. Skr. ay-am O.Pers. ayam i. e. ay-am, 3 rd sing. Skr. üí-t Avest. cli-jj 3 rd dual. Skr. üí-tüm Avest. ai-tem. Conjunctive: Skr. áy-a-ti á-ya-t Avest. ay-a-p (cp. indie. Skr. dy-a-te, Avest. imper. ay-a conj. ay-a-p opt. ay-oi-p). Skr. kse-ti Avest. sa$-iti 'lingers, dwells', 3 rd dual Skr. ksi-tás 3 rd pi. Skr. ksiy-ávti, conj. Skr. ksáy-a-t: Gr. Horn. tv-xi¿-fitvo-b' 'well built'. — With thematic vowel Skr. ksiy-d-ti. y/^fch'U- 'hear: Skr. 2nd sing. Sro-si, 1 st sing, á-érav-am 3 rd sing, á-érd-t, 2nd pi. áru-ta and following the singular iró-ta Avest. srao-ta, Avest. 2nd pi. mid. a-srii-dUm, Skr. imper. Sru-dhi; conj. Skr. 3 rd dual ¿ráv-a-tas, opt. Avest. 1 st pi. srvtma i. e. sruv-i-mcL: Gr. imper. y.lv-rt (cp. § 497 pp. 56 f.) IIspi-y./iv-fttvo-Q. — With thematic vowel Skr. éruv-a-m, Gr. xlvm (cp. § 527). y^deríc- 'see': Skr. á-daré-am Avest. dars-em, l 9t pi. Skr. d-dfé-ma, and also á-darS-ma following the singular; conj. Skr. dáré-a-t Avest. l 8 t pi. dares-a-ma (cp. indie. Skr. á-dará-a-t). — With thematic vowel Skr. 3 rd pi. á-dfé-a-n opt. dfs-e-t. Skr. á-grabh-am Avest. grab-em, 'I grasped', 3 rd pi. Skr. á-gfbh-ran. Skr. chand- appear': 3 rd sing, chánt-ti. y^bheid- 'findere': Skr. 1 st sing. á-bhSd-am 3rd sing. á-bhet; — with thematic vowel opt. bhid-e-t. Avest. mip(Skr. mith-) 'destroy': 3 rd sing, moist, conj. moip-a-p (cp. indie. Skr. méth-a-ti), opt. mip-ya-p. V^dheuQh- 'milk, give milk' (cp. Fick Wtb. I 4 73): Skr. 3 rd sing, dógdhi 3 rd pi. duh-ánti, mid. 3rd sing, dugdhé 3 rd pi. duh-ate -áte conj. doh-a-te, opt. duh-i-ta; — with thematic vowel á-duh-a-t opt. duh-e-t. V^jeug- 'iungere': Skr. 3 rd sing, mid. d-yuk-ta 1 st pi. d-yuj-mahi, Avest. 3 rd pi. yüj-En 1" pi. mid. yaoy-maide% with non-original strong stem; — with thematic vowel, Skr. á-yuj-a-t.

60

Present Stem: Class I — Skr.

ds-ti.

§498.

'wish, desire: Skr. 1 st sing, vdi-mi 3 rd sing, vds-ti 1 pi. td-masi, Avest. vasnfl vasti usmahi, conj. Skr. vdi-a-t Avest. vasafi (cp. indie. Skr. vdS-a-ti imper. vdi-a). — With thematic vowel Skr. ui-d-mana-s. y/^ueic-

st

Idg. *es-ti, see § 493 p. 52. Skr. sing. pi. s-mas s-thd s-dnti, Avest. sing, ahmi ahi asti pi. mahi (I § 558. 3 p. 414) sta henti, O.Pers. sing, amiy (I § 558.3 p. 415) ahy astiy 3 rd pi. hatiy i. e. hantiy; O.Pers. 1st pi. amahy with a- from the singular. Pret. Skr. 1 st sing. ds-am 3 rd sing, ds O.Pers. 1 st sing, a ham i. e. aham Avest. 3 rd sing, as (I § 647. 7 pp. 493 f., § 649. 6 p. 496), pi. Skr. ds-ma as-ta ds-an O.Pers. 3rd pi. aha i. e. aha, cp. § 481 pp. 29 f., also unaugmented Avest. 3rd sing, as 3 rd pi. h-en Skr. s-dn; on the 2nd and 3 rd sing. Skr. ds-%-5 ds-i-t, see § 574. Imperative: Avest. z-dl; Skr. edht for *az-dhi (I § 591 p. 447) instead of regular *dhi following the analogy of forms witli strong root. Conjunctive: Skr. as-a-ti ds-a-t Avest. atdh-a-iti awh-a-£ O.Pers. ah-a-tiy. Optative: Skr. s-ya-l s-iya-t Avest. Pr. Ar.

ds-mi

dsi

*as-ti,

ds-ti

h-ya-J). X ^ e d - eat': Skr. ad-mi dt-ti. So in all the weak persons ad-, as 3 rd pi. ad-anti 2nd pi. at-td imper. ad-dhi, obviously because such forms as *ta *dhi were not clear enough (cp. above, Skr. edhi). Conjunctive: *ad-a-ti *ad-a-t (cp. 2nd sing, mid. ad-a-sva Gr. sd-to Lat. ed-o Goth. it-a). On the relation between ad-mi and Lat. Sst Lith. est, see § 480 Rem. pp. 28 f., § 494 pp. 54 f.

Skr. dhdksi and others of the same sort, see § 493 p. 53. Skr. bhi-Sak-ti 'heals' (bhi- is a bye-form of abhi) was no longer recognised for a compound, hence 3 rd sing, a-bhisnalc R.-Y. x, 131. 5, following Class XV, and bhesajd-s 'healing'. y/^dhedo-, Skr. dha- da- Iran, da- (in Iranian the two stems ran into one, and it is no longer possible to distinguish their meaning exactly), see § 493 p. 53. Skr. a-dha-t dha-t a-da-t pi. d-dha-ma d-da-ma, Avest. da-fr da-ma O.Pers. a-da; on a in the plural, see § 495 p. 55; mid. Skr. a-dhi-ta a-di-ta,

§498.

imper. Avest.

Present Stem : Class I — Skr.

ds-ti.

61

Conjunctive: Skr. dhd-ti pi. mid. dha-mahe mid dä-ite (§ 933). Optative: Avest. d-yä-ß. \/~*stä-, see § 493 p. 55. Skr. ä-stha-t d-stha-ma (like a-dhä-ma, see above), Avest. paiti-stä-ß; mid. Skr. ä-sthi-ta. Conjunctive: Skr. stha-ti 2nd dual sthd-thas, Avest. mid. xstä-ite (§ 933). \ f d ä - 'separate, divide up' (Gr. '- etc. (the Author, Gr. Gr. 2 p. 47). — With thematic vowel t-xxav-o-v. R e m a r k . The Homerio 1 st sing, -ixzav and 3 r d sing, ixra are peculiar, a is certainly short only in O 432 Wat', met arSga xarexra Kv9iigmnt laSioini). Is it possible that the original forms were exrav txrs with Aeolic a, which would be re-formates of the same nature as 3 r d pi. EXJAY? Or is CXTU a re-formate like Skr. d-vrk beside vark (§ 499 p. 63),

and -txT&v due simply to the analogy of exru ?

I/bher- 'ferre': 2 nd pi. &si- 'destroy' = Skr. k$ay-: 3rd sing. mid. s-fih-ro: Skr. imper. k$i-dhi. Conjunctive (p&i-s-xai; in Skr. we should expect *May-a-te on the analogy of k$i-dhl (cp. indie, ksay-a-ti). Partic. 9ia> (sq>&tsv, 2 446) is smas, and like phrases), was a variant pole sum. pote is an adverb (aco. sing. neut. for *poti, or loc. in orig. -e, see III § 260 p. 160), op. bene sum, tuto sum. potisset potisse are for potis 'sset 'sse, cp. situst for situs 'st. But jvotes potest potestis come from pote es etc. So also possum possim (whence possem posse by complementary analogy) come from *potsum *potsim, pote-sum, pote-sim. It is doubtful, however, whether -e- disappeared by regular syncope, or whether potest: est suggested *potsum : sum (I § 501 p. 367). Ifed- 'eat': es est estis este, pass, estur (on -st- instead of -ss- -s- see I § 501 Rem. 2 p. 368); with thematic vowel edo edimus edunt, also edis edit etc. See § 480 Rem. pp. 28 f., § 494 pp. 54 f., § 498 p. 60. Optative: ed-i-m ed-i-mus instead 1) I § 110 page 105 should be corrected by this statement.

Present Stem: Class I — Skr. ds-ti.

71

of *d-l-, perhaps to distinguish this optative from the old optative of dö- give' (see below). I / d h e - 'place': Lat. con-di-mus con-di-tis credimus for *-fa-mos *-fa-tes: Gr. s-&t-/Liev, see § 493 p. 53. The forms -dö -dis -dit -dunt are thematic. IS dö- 'give': Lat. da-mus da-tis red-dimus -ditis: Gr. i-io-fxtv, see § 493 p. 53.') Imperative: ce-do (2 nd pi. ce-tte for *ce-date *ce-dite, I § 633 p. 474), see § 957. The old optative stem *d-i- (cp. Avest. 3 rd sing, d-yü-fi) is found in Ose. d a - d i d 'dedat'; to this the conj. d a - d a d "Lat. de-dat is related like Lat. ed-a-mus : ed-t-mus (see above). The old singular forms *dö-m *dö-s *dö-t are gone; we have instead dö däs dat. The last two represent the stem used in composition for the conjunctive, d-ä- (cp. -bds for *bhu-a-s indie, beside conj. fu-d-s, see § 578); and these created dö on the analogy of stö : stds, flö : fids etc. In composition, we see the same inflexion as legö h a s : ven-dö red-dö -dis -dit -dimits -diti -dunt. But undoubtedly -dimus -ditis are what *-damtis *-datis must regularly become, cp. fut. O.Lat. reddibö for *red-dabö. R e m a r k 2. The compounds of dhe- and do- were confused in Latin, beginning with the l 8 t and 2 nd plural; -di- = *-fa- *-dh»- and = *-da- *-d»-. Compare Darmesteter, De conj. Lat. verbi dare, Paris 1877; Postgate, Dare, 'to give' and -dere 'to put', Trans. Phil. Soc. 1880—81 pp. 99 ff.; Thielmann, Das verbum dare im Lat., Leipzig 1882; the Author, Liter. Centr. 1882 col. 1389 ff.

Whether the forms std-s sta-t from | / std- 'stand' are rightly placed here with the rest, as is suggested by Skr. (i-sthd-t and Gr. ¿-art] (§ 493 p. 54), is very doubtful because of std-mus sta-tis. One cannot see why an orig. *sta-mus (cp. da-mus) should have been altered (earrjitsv as compared with ¿¿Ofuv is quite a different thing, see § 495 p. 55); and so it 1) Br&il (M6m. Soo. Ling., vu 326) thinks he may regard as an un-

augmented preterite dat in Vergil's cratera anliquom quem dot Sidonia Dido (Aen. ix 266). Many points in Vergil's manner are in favour of Br6al's assumption (see Ladewig on Aen. I 79, ll 275, Kühner Ausf. Gr. ii 90).

Present Stem: Class I — Skr. âs-ti.

is preferable to refer the whole present of this verb sto to *std-i5; see § 584 Rem., § 706. This is supported by Umbr. stahu 'sto*. § 606. K e l t i c . ISes- 'to be', 1 ) 3 rd sing. 0 . Ir. is O.Cymr. iss is for *es-ti. 3 rd pi. O.Ir. it O.Cymr. int for *s-enti (II p. 196, footnote). The a- of the Irish proclitic sing. 1 st pers. am 2nd at (3 rd relat. as), plur. 1" ammi 2 nd adib (3 rd rel. ata), is from -e. The form am then had no -i at the end; and since it is usually written am with one m, it seems to have had m spirant, like Mid.Cymr. wyf. It must therefore not be derived from *esmi. The 2nd sing, at Mid.Cymr. wyt may contain the pronoun of the 2nd person, and may thus be explained as *esi-\-t-. Mid.Cymr. 1 st sing, wyf seems to be due to the analogy of the 2 nd sing. Is Ir. am the same? Others regard these forms as coming from the root ei- 'go. The 1 st pi. ammi Mid. Cymr. ym may be *esmesi. In the 2 nd pi. adib, -b is certainly an affixed personal pronoun, and -dithe ending of the 2nd pi. = -thi -the (ground-form *-tesi, the suffix re-formed on the analogy of the 1 st pi., see § 1014). This brings us back to an imaginary ground-form *s-e-tesi + sp-, which would be a re-formate following the 3 rd pi. *senti; and so perhaps the l 8 t pi. should be derived from *s-esmesi, a later contamination. Again, the Keltic ¿-preterite, as it is called, is partly of the same kind. In the 3 rd sing, of this preterite, the ending -t is said to represent the middle ending *-to (Strachan, Bezz. Beitr. XIII 128 IF., and Zimmer, Kuhn's Zeitschr. x x x 204 if.): e. g. O.Ir. as-bert 'dixit' Mid.Cymr. kymerth 'sumpsit* for *kymberth from |Sbher-. When -t ceased to be understood as a personal ending, the other persons which completed the tense were formed on the model of stems ending in -t: O.Ir. sing.

1) Compare Zimmer, Kelt. Stud, li 133; Stokes, The Neo-Celtic Verb Sub st., 43 ff., Kuhn's Zeitschr. x x v m 93 if. ; d'Arbois de Jubainville, Mém. Soe. Ling., v 239 f.; Strachan, Bezz. Beitr. x v 114 ff. In the text I follow chiefly information received from Thurneysen.

§507.

Present Stem: Class I — Skr. as-ti.

73

1 s t pers. -hurt

2 n d -birt,

3rd

-bartatur. =

plur. 1 s t -bartmar

2 n d *-bartid

Compare Lith. eitu 'I go' formed from el-t 'he goes'

Skr. e-ti § 6 8 6 R e m . 2, Gr. ¿Jo9ijv

from ¡-¿¡o'-fys

=

Skr.

d-di-ihcLs § 589. In forms like as-bert Strachan sees rootaorists of this class, Zimmer s-aorists (*ber-s-to). As a matter of fact, both these aorists may have been the source for some preterites such as these. To our Class I belong O.Ir. ro-St 'he took' for *-em-to, Mid.Cymr. gwan-t 'percussit, feriit'. § 607. Germanic. |/"uel- 'wish': opt. Goth, viljau pi. vilei-ma O.H.G. 2nd and 3rd sing, wili O.Icel. 1st sing, vilja. The strong stem (cp. Skr. vr-iytt-t vur-t-ta), like that of Lat. velim, is due to a confusion with the indie. *uel-(i)io- *uel-i- (O.H.G. willu

0.C.S1. veljq). I/~*qem- 'go,

Skr. yam-yd-m,

S e e § 4 9 3 p. 51, § 5 0 5 p. 69, § 9 2 8 . come':

opt. A . S . cyme

V es- 'to be', see § 493 p. 52. Goth, im,

is,

(bist bis), ist, erom

erum,

=

Goth.

*kumjan:

s e e § 4 9 3 p. 51. ist,

sijum

birum, erod

sium,

birut,

erud,

ero

sint; eru.

sijuj)

The indicative forms are siup,

sind;

O.Icel. em,

est,

First it m u s t

O.H.G.

him,

es (Run.

is),

be

mentioned

that the O.H.G. 2nd sing, bist bis belongs to a present to be described below in §§ 707 and 722, formed from iSbheu-, namely 1st sing. *bh%-ii5 2nd sing. *bhy,-l-si etc. (A.S. 1st sing. bio 2nd sing, bis 3rd sing, bid, O.Ir. biu etc.), and that the similarity of bis and *is ( = Goth, is) produced b-im b-irum b-irut.

1 s t sing. Goth, im

O.H.G.

(b-)im

for *immi

*ismi

=

Skr. ds-mi (I § 582 Rem. 2 p. 436); O.Icel. em instead of regular *im following the plural forms which begin with e, whence also the e in est and es. Whether the 2ud sing. Goth. is comes from Idg. *esi or *es-si (see § 984. 1), cannot be decided; O.Icel. est like O.H.G. bist has -t on the analogy of the preterite (§ 990. 3); on the very rare O.Icel. 2nd sing, es, see Noreen in Paul's Grundr. I 515 The 3rd sing. Goth. O.H.G. ist is for Idg. *es-ti; O.Icel. es (Run. is) A.S. O.Sax. is are doubtless the old injunctive Idg. *es-t; the 3rd pi. O.Icel. er-o is also injunctive (other explanations are suggested by J. Schmidt, Kuhn's Zeitschr. x x v 593; von Fierlinger, ibid.

74 440 footnote 2 ; Noreen, loc. cit.; Osthoft, Perf. 428 f.). In the plural, the weak stem is seen in Goth, s-ind O.H.G. s-int = Idg. *s-enti. O.Icel. ero eru is pr. Germ. *iz-unp (§ 1025. 1 ¿>), an injunctive in which the strong stem has taken the place of the weak (cp. Gr. taai, § 502 p. 66). As the ending of *izunp agreed with the so-called preterite-presents, such as Goth, mun-un (§ 508), on their analogy the 1 st pi. O.Icel. erom O.H.G. (b-)irum and the 2 nd pi. O.Icel. erod O.H.G. (b-)irut appeared. These forms then produced O.H.G. Frank. sind-un O.Sax. A.S sind-un. Goth, sijum sijujt are probably transformed from *iz-um *iz-up, caused partly by sind, partly by the feeling that the opt. sijau sijdis etc. should contain a stem sij-. — The optative has always a weak root: 1 s t sing. Goth. s-ijau O.H.G. s-i O.Icel. s-ja; on the inflexion, see § 947. — Partic. *s-und- = Idg. *s-nt- in Goth, sunjis 'true' for *sund-ja- = Skr. sat-yd- 'true'; also thematic *s-o-nt- in *sanp-u'true, truthful' A.S. sod O.Icel. sannr (cp. § 493 p. 53). XXVII

O.H.G. tuorn 'I do' (O.Sax. A.S. do-m) must be derived from V dhe-, along with the pret. te-ta and the subst. ta.-t (Goth, ga-de-di-) and others, but its vowel makes it impossible to derive the word from *dhe-mi. Perhaps it contains *dh-a(Class X , § 585), found in other parts of the verb as a conjunctive stem (Lat. con-da-m -da-mus); cp. Lat. 2 nd sini;-. d-cL-s 'thou givest' = conj. (red-)dan (§ 505 p. 71, § 937). R e m a r k . On O.H.G. si am stem 'I stand' and nam gem 'I go', see § 708. They certainly do not belong to this class of presents.

§ 608. Some Preterite-Presents may also be placed in this class. Goth, mun-un 'they think', opt. 1 st pi. mun-ei-ma : Skr. mid. 3 rd sing, a-ma-ta partie. man-and-s, [S men- 'think, mean'. Goth, rja-daiirs-im O.H.G. gi-turrun 'they dare', opt. Goth, ga-daurs-ei-wa O.H.G. gi-turr-i-m: Skr. partic. dhfS-dnd-s, |S dliers- 'dare'. Goth, vit-un O.H.G. wi$$-un 'they know', opt. Goth, vit-ei-ma O.H.G. wi$$-T-m8s: Skr. vet-ti opt. vid-yd-t. etc., see § 493 p. 52; the weak forms of this vera were present and perfect at the same time.

§§509,510.

Present Stem: Class I — Skr. ds-ti.

I f this view be right, Goth, mun-un

75

ga-da&rs-un

were originally injunctive, like Icel. er-o

er-u

vit-un

(§ 507 p. 73).

W e shall meet again with present forms among the preteritepresents (§§ 646, 887, 893). § 609.

Connected with Skr. ir-te

imper. Avest.

ar'-sva

Gr. og-ffo, which point to an Idg. mid. pres. *f-tai

(mentioned

¡ibove, § 4 9 7 , page 57) are A.8. 2 n d sing, ear-d

ar-d

thou art'

pi. ear-un

ar-on

meaning cp. Gr. op-cog-u, meaning 'I am'.

with ar-

— Idg. * f - .

ear-t

For

the

which in late Greek had also the

On the 2 nd sing., see § 990. 3.

R e m a r k . Germ, ar-was probably not a perfect stem, which would hare been or-. This is said to correct the note in Idg. Forsch. I 81.

You Fierlinger (Kuhn's Zeitschr. x x v n

436 if.) deduces

some other presents with weak stem and secondary from thematic forms with peculiar vocalism. fara

1 / p e r - (in Gr. nffjaoi 'I pass through' etc.),

'I fare, go',

he assumes an older present stem *pf-, Skr. 2

sing,

nd

§ 510.

1 st pi. *pf-m6s

(cp.

pdr-si).

Balto-Slavonic.

Lith. veizd-mi,

accent,

Thus, for Goth.

To Idg. *ueid-mi

'I see' belong-

and imperative Lith. veizdi veizd

O.C.S1.

vizdi,

— the imperative forms have non-original strong stem, and the O.C.S1. form has s instead of An undoubted re-formate -vtfdziu

see § 493 p. 52, §§ 949, 962.

is Lith. pa-vyzdmi

Idg.

*es-mi

'I am', see

§ 493

p. 52.

forms here to be cited are scattered I s t sing. Lith. es-mi O.C.S1. jes-mi\ verbs with -u then a 2

nd

see § 991.

cp.

The

over various

dialects.

Pruss. asmu),

3 r d sing. Lith. es-ti

asset

asse

es-t

essei)

O.C.S1.

(Pruss. ast

est)

The 1 st and 2 nd pi. may have taken esLith. es-me

O.C.S1. jes-mu Lat.

Lithuanian

and

person esml was made on the analogy of sukt : sukii.

in pr. B a l t o - S l a v . : estei)

pa-

on the analogy of thematic

Lith. esmu (like Lett, esmu

2 nd sing. Lith. est (Pruss. assai jes-tu.

instead of

'invideo', also used (cp. § 511).

-sins

and

jes-te. Gr.

es-te

Dor.

instead of s-

( P r u s s . asmai,

P a r t i c . Pruss. -sins ivr-sg

(p. 5 0

jesi,

O.C.S1.

dat.

astai

asti

-sentismu:

footnote).

—

Present Stem: Class I — Skr.

as-ti.

*s-o- in 0.C.81. 3 r d pi. sqtu partic. Lith. sqs sanczio 0.C.S1. sy septa.

*es-o- in Lith. 1 s t sing, esit 1 s t pi. esame

2 n d pi. esate

partic. esqs. It is not clear whether Lith. opt. (permissive) 3 rd sing, tese 'sit' is to be analysed te-sS (cp. Pruss. 2 nd pi. opt. sei-ti) or as t-esS. With augment pr. Balto-Slav. *e$-o-m etc.: iu O.C.S1. this occurs in the imperfects nesi-achu -ase -ase pi. -achomu -asete -achq, unthematic 2 nd pi. -as-te also found (so too t h e dual h a s both -asta -aste a n d -aseta -asete), see § 9 0 3 ;

in Lith. the preterite e- passed into the present, esu esi esa esame esate partic. esqs, see § 480 p. 28. — On Lith. 3 rd sing. yra, which comes from the root of Skr. tr-te Avest. are-sva Or. op-cup-a A.S. ear-3, see J . Schmidt in Kuhn's Zeitschr. x x v 595 f. The present of 1S ed- 'eat* was in pr. Balto-Slav. *ed-mi; for its e see § 480 Rem. pp. 28 f., § 494 p. 54. Lith. sing. 8t rd 8t l pers. emi 'I devour' 3 sing, esti est pi. l erne 2 nd este dual

1 s t edva

O.C.S1. sing.

2 n d esta (on edmi edme 1 s t pers. jairii 2 n d jasi

2 n d jaste 3 r d jad-r>tu (on

see I § 547 p. 4 0 1 ) ; 3 r d jastu

1 s t d u a l javi instead

pi. 1 s t jamu of *jadv6, see

I § 547 Rem. 3, p. 401). — Also thematic Lith. edu edi etc., Pruss. opt. 2 nd pi. idaiti O.C.S1. partic. jady jadqsta. Other presents of this class are found in one only of the two branches, Baltic or Slavonic, not in both: § 511. In L i t h u a n i a n there is a fine array of present forms of this class, but nearly all are defective and have only one or two persons left, chiefly the first and third. Some of them have come into this class quite late. Compare § 496 p. 56. Lists of m i - f o r m s may be found in Schleicher pp. 250 ff., Kurschat pp. 304 if., Bezzenberger Beitr. lit. Spr. 198 ff. (a few more come from the dialects). W e begin with those which may be regarded as representing Idg. originals. pa-velmi 'I will' 2 rd sing, pa-velt, refl. 1 st sing, velme-s: S k r . d-vf-ta etc., see § 4 9 3 p. 51.

ei-ml 'I go': Skr. e-mi etc., see § 493 p. 51. 2 nd sing. ei-si 3 rd sing, ei-ti ei-t. The strong stem passes into the

§511.

Present Stein : Class I — Skr. às-ti.

77

plural: 1 st ev-me 2nd ei-te; but 2nd pi. High Lithuanian eiste on the analogy of fate : eme, düste : dü'me. Old injunctives are ei 'let him go' {te ne ei 'let him not go') ei-mb 'let us go' (dual ei-vä). Imper. et-k — Lat. % (§ 957). Indie, now usually ei-nü, as Class X I I I (§615). Pruss. 2nd sing, ei-sei 3rd sing. Si-t l 8 t pi. ei-mai. Partie. Lith. ent- going' (in old printed books) doubtless stands for *i-ent-, first in compounds with prefix ending in a consonant, such as iss-ent- (I § 147 p. 132), cp. p. 50 footnote; but it is possible that it comes from *ie-nt~, Class X ; see § 593. — *%-o- appears to be contained in Pruss. opt. 2nd sing, jeis 2nd pi. jeiti. lek-ml 'I remain' 3 rd sing, lik-ti ISk-t: Skr. 2nd dual rik-tam 2nd sing. mid. rik-thüs, Vlejq- 'linquere'. r dudmi 'I lament' (regularly *raumi, I § 547 p. 401): Avest. 3*d sing. mid. raosta with irregular strong stem (cp. § 499 p. 62), I r e u d - rudere'; cp. Skr. rödi-ti pi. rudi-mas § 574. deg-ml 'I burn": Skr. 2nd sing, dhdksi, see § 493 p. 53. sedmi 'I sit' 3 rd sing, sest 1st pi. refl. sedme-s: Skr. 2nd sing. sdt-si, [/" sed sedere', see § 494 pp. 54 f. jü's-mi 'I gird' (beside fti'siu): Avest. 3 rd sing, yas-ti, stem jös-. Compare § 656. The imperatives de-k 'lay' (inf. de-ti, V dhi-) and dü'-k 'give' (inf. du-ti, V dö-), of which the latter must be compared with Lat. ce-dOf show the same formation as ei-k ( = Lat. f). O.Lith. du-di du-d 'give' doubtless = *dö-dhi. See I § 547 Rem. 1 p. 401, IV §§ 546, 957, 962. Some other presents of the same sort, to which there is nothing which answers outside the Baltic group, may here be named: bar-mi 'I scold' 3 rd sing, ap-bart, Vbher- (Lat. feriö), bar- for *bhf-; snSk-ti 'it snows', IX sneigh-; meg-mi 'I sleep' 3rd sing, mikt 2nd pi. mSkte; r&ug-mi 'I belch', K reyq- (Gr. ¿QsvyoLiai);J) serg-mi 'I protect, watch' 3rd sing, serkti; kosmi 'I cough' (Skr. kds-a-te). 1) On acoount of a form ridugmi, Wiedemann (Lit. Praet. 186) derived this word from a groundform *rey,g-mi (op. § 494 pp. 54 f.); which is very dubious.

78 All these verbs have in Lithuanian, beside this present formation, another with the same meaning, which in High Lithuanian is almost the only one. It is certainly no mere chance that in so many presents of the //«'-class, the bye-form is a verb in -iu with accentuated root (1st pi. -i-me Class XXYI § 727), czidudziu

as

sedgiu,

(czidudmi

sergiu,

kosiu,

'I sneeze'), stoviu

stjdziu (stovmi

{zydmi

'I

bloom'),

I stand'),

mSrdsiu

(mSrdmi 'I lie a-dying") etc. With these verbs in -iu, the 2nd and 3rd sing, ran together and became indistinguishable in form (2nd sing, -i for 3rd sing, -i for *-i-t), and it is probable that it was a wish to keep these persons distinct which first produced the non-thematic forms in most of these verbs. Perhaps on the analogy of sest(i) (sedmi) beside s'edziu was formed serkt(i) etc. Observe also tenk-mi inf. tikti),

instead of tenku 'I last' (pret. tekaii

and the 2 nd pi. gttbste

(g4lpste),

from gUbmi

'I help'

3 rd sing. g6lbt(i) gilpt(i), — for its s, compare that of ei-s-te § 511 p. 77. § 512. Slavonic. The form only without parallel in Lithuanian is sq-tu 'inquit', explained in § 493 p. 52. Class II:

Root +

Thematic Yowel forming P r e s e n t Stem.

the

§ 513. This class of present stems, invariably the largest in all Indo-Grermanic languages, falls into two divisions, according as the accent falls (A) upon the root syllable or (B) upon the thematic vowel. When the root carried the accent, it was of the strong grade (l 8t strong grade in the e-series), but weak grade if the accent fell upon the thematic vowel; e. g. (A) *bhty,dh-o= Skr. bodh-a-

Gr. nsv&-o-

)-/y-vv-vTo a n d L e s b . inf. o-siy-rjv (§ 643).

s-m-o-v

'I drank'

m-i7r

beside n7.-91 Class I ; -Av-bheudh(§ 513 p. 79). ijhity-o-v aor. 'I came* beside fut. sXsv6ofxai. xv,%t 'he hid' - < J 3 M :

beside xevd-w.

t-6y-o-v

'he held, had' oy-tTr,

beside ¿x~K>y

y/^segh-. i-nr-c-rn 'he flew' nt-i-adai beside nev-t-Tui. ftux-wv 'bleating, crying' beside tajxdofiai.

Partic.

With the secondary ending of the 2 nd sing, middle: -t-d-IJ^ beside I-AY-T-TO; ¿pgt&yc; tipedqg i. e *e-ur-e-thes y^yer- 'say' (sipca), see § 589. Whether these very were some of the original types which produced the series of aorists in -ttrjv, is of course doubtful; -thSs originally to have belonged only to non-thematic (§ 1047.2).

¡-oyfrom forms whole seems stems

R e m a r k 2. It ¡8 hard to classify forms in -u» -io->> and -vw -vo-v, along with which forms in I and ¡3 are common. There is nothing a priori

§528.

91

Present Stem: Class II — Skr. bhàr-a-ti sphur-ó-ti.

against assuming that these hare the suffix -jo- (Class XXVI), and that -i- fell out between vowels; indeed, this must be done for forms like Lesb. ipvuo (I § 130 p. 118). Titoput nif/ter beside nioum f'ntov may be illustrated by Skr. pi-yd-te, 9vw beside 9vio by Skr. dhu-yd-te, ivw beside i™ by O.Icel. ly-ja 'destroy, crush* (see § 707, and Osthoff, M. U. iv 12 ff.). But t and v may come from forms of Class I , by presents passing from this class into the thematic conjugation, cp. for example ni-i-fjtv thematic and TTZ-»L non-thematic, XTU> thematic and X V - T O not (also X V - T O , for Xva cp. Lat. !uo so-luo so-lvo), tS-gvm 'I roar' (also variant wgvio, cp. Skr. ruv-a-ti) but Skr. opt. ru-ya-t (Lat. ru-mor Goth, ru-na). Then these t- and «-forms would naturally be compared with ofoiro beside op-no, R - N O X O - V beside E - / G X U I , and other such, see § 497 p. 57. Another possibility is that the long vowel came in by analogy of other tenses, Xuu following Xv-mo, 9vw following &v-aio, just as we see Att. yrva (not *yiui) by analogy of yti-ou> (the Author, (Jr. Gr. 2 p. 31), and Lesb. ¿Stxyti (instead of diWii) by analogy of ¿Sixij-ma (§ 775). § 528. p. 51.

Italic.

mol-0:

Armeu. malem

O.Lat., \ftel-. lu-o so-lud

Lat. vol-0

so-lvo:

see § 493

etc., see § 523 p. 86.

tul-5

cp. Gr. Xv-io 'I loose* § 527 Rem. cp. ac-cerso

(with variant ningu-i-t,

(also vtt'. Skr. a-dl-dhS-t 'he looked' 1 st pi. dl-dhi-mas mid. pres. di-dhy-e pret. d-dl-dhl-ta; conj. dl-dhay-a-t. Skr. d-d%-dS-t 'he appeared' 3 rd pi. di-dy-ati imper. d%-di-hi di-dl-hl; conj. di-day-a-t', — with thematic vowel Gr. ¿i-^o-fiou 'I seek, strive' (orig. 'look out for something') for *it-fy-o-ficu (see § 469 p. 14, § 549). Skr. dt- and dhi- both became di- in Avestic cp. Avest. dadniti = Skr. dddhUti and dadati, § 540): di-3a$iti; — with thematic vowel imper. di-dy-a, cp. conj. di-dy-d-J>. Skr. vi-ves-ti 'works' 1 st pi. vi-vi$-mas, conj. 2nd sing. vl-ve§-a-s. iy-e-ti 'goes' only found in the 2nd sing. pret. aiy-S-$, Avest. 3rd pi. conj. yeyqn = Ar. *ii-a%-0-n (§ 473 p. 19). 1) Cp. § 465 p. 12, § 469 p. 14, § 739 on O.H.G. rtrgm and Gotb.

reira.

B r u g m a n s , Elements. IV.

7

98

Present Stem : Class III — Skr. bi-bhe-ti.

§538.

Skr. ju-ho-mi 'I offer, sacrifice' 1 st pi. ju-hu-mas 3rd pi. jé-hv-ati, conj. 2nd pi. ju-hav-a-tha, opt. 1st pi. ju-hu-yd-ma. st 1 pi. ju-hu-masi from hu- 'call'. 3rd pi. sù-sv-ati from su'press'. Sometimes a strong stem has got into the place of the weak (cp. § 499 p. 62), as Avest. 2nd sing. mid. ji-yae-sa from ji- 'live', Skr. 2nd pi. ju-ho-ta from hu- 'offer, sacrifice', 2nd sing, yu-yò-dhi 2nd dual yu-yo-tam from yu- 'keep off'. Remark, k in Skr. ci-ki-mi (l/gej-), and y in -Avest. ji-ya(-sa (l/get-) are taken from the perfect, where they were regular before o in the sing, indie, active (I § 445 ff. pp. 331 if.). In considering ji-gliar-U (§ 540), if its root belonged to the «-series, we must remember that one of the stems of this verb is jighr-, and gh was regular there ; so with ja-gar-ti from | / g e r - we must remember the stem ja-gr- (§ 560).

§ 538. R o o t s with other Y o w e l s . In Aryan, roots with a long a-vowel have generally in the reduplication a = Idg. e instead of i, when the weak stem in the root syllable had not %. Examples: Skr. dd-da-ti mid. da-t-te from y^do- 'give', jà-ha-ti pi. ja-hi-mas from Ar. zha'leave, give up'. But on the contrary èi-éd-ti imper. èì-ét-hi mid. éi-éi-tS from iSfco- 'whet, sharpen'. In the latter word we see the Idg. root-determinative t, which so often forced its way into the place of Ar. i = Idg. a (see § 498 pp. 61 f.) ; and this % is regularly echoed by » in the reduplicator ; compare èi-éì-hi with di-di-M from di- appear'. Skr. ja-hi-tam beside regular ja-hi-tam (see "Whitney, Sanskrit Roots, p. 204) has been altered by the influence of the mid. jf-hi-tS (§ 540), similarly ra-rt-dhvam by that of ri-ri-hi (rd- 'give'). So too the |S dhè- 'place' in Balto-Slavonic reduplicates with e, as Lith. 2nd pi. deste like O.H.Gr. da-ttha. These forms with e belong to Class Y, not like Gr. Si-Scopi ri-d'tj/Lti etc. It seems to me impossible to decide whether in Idg. the same present stem had both i and e in its reduplicated forms, as *dhi-dhe-ti and *dhe-dhé-ti, or whether e only came in by analogy of Class Y, and is of later date than the parent language. If the latter, then the influence of perfects with e in the reduplicator must by taken into account (§ 555).

§539.

99

Present Stem: Claas III — Skr. bi-bhé-ti.

Compare Gr. iXufri — *oi-oXa-th and Lesb. sikn&i = *as-oia-&i § 542. Under these circumstances, I cite Aryan and BaltoSlavonic forms both in Class I I I and Class V. § 639. P r . I d g . *bhi-lher-mi 'I bear' 1 st pi. *bhi-bhf-més 3 rd pi. *bhi-bhr-%,ti : Skr. bi-bhar-mi 2 nd dual bi-bhf-ihas 3 r d pi. bi-bhr-ati, Gr. 1 st pi. *ni-(pga-/n£v inferred from inf. ¿a-m zi-zan-a-nti Skr. d-fi-jan-a-t (§ 548). § 646. G e r m a n i c . O.H.G. se-sto-m 'sisto, I arrange, design', y/^sta-, with gradation lost, see § 539 p. 100. Whether O.H.G. te-ta O.Sax. de-da 'did' is an imperf. like Gr. ri-d-qv or an old perfect, remains doubtful; see § 886. § 646. B a l t o - S l a v o n i c . A few relics are the presents of y/^dhS- 'set, place' and do- 'give', but with e in the reduplicator (§ 538 p. 98). In pr. Balto-Slavonic the forms were *dhe-dh-mi and de-d-mi, which may be compared with Skr. da-d-mi and Avest. dazdl dasti (§ 541 pp. 101 f.). But they did not, as these did, arise only by the weak stem spreading into the singular, but from

104

§546.

Present Stem: Class III — bi-bhé-ti.

this and another cause together; the other cause was, that the 2nd sing, middle, which originally had the weak stem, had got an active meaning (see § 991 on Lith. desè-s dust O.C.S1. dasi). And since *dhe-dh-mi became *dedmi in pr. Balto-Slav. (I § 549

p. 402), the two verbs were confused in the present, and the same forms served for both (cp. Avest. dadaiti — Skr. dddhati a n d dddati,

§ 5 4 0 p. 101).

However, it was only in the meaning of 'I lay' that *dedmi survived for any time. Lith sing. 1st pers. dèmi for *dedmi, 2nd reflex, dese-s for *de-t-se-s, 3rd désti dèst, 2nd pi. déste. Now the verb is mostly thematic, de-d-ù dedì dèda etc. And rd demi 3 sing, desti too took è from non-present forms dejau desiu and the like, just as Gr. Lesb. àóiitijoi instead of à&xtio follows a&xij-oa,

and yeva

instead of *yém follows ysvaiu etc.

(§ 775). But in Slavonic we have degdetU — following the io-class (§ 733).

*de-d-ie-tu,

In the meaning 'I give', *dedmi was changed to *dodmi in

pr. Balto-Slavonic by analogy of non-present forms with *dó-, There is a reason why the vowel of the root got into *dedmi 'I give' and not into *dedmi 'pono*. It is that the difference between the vowel of the first syllable of the present and that of the other tenses was in *dedmi 'pono' only one of quantity, but in the other it was a difference of quality also; *dedmi: aor. *de-s- was backed up by such verbs as *tekò : aor. *ték-s(O.C.S1. tekq téchu), but there was no parallel for *dedmi: aor. *do-s~. Lith. sing. 1st pers. dumi, 2nd d&si for *dà-t-si, 3rd

d&'sti

dà'st,

pi.

1 s t dume,

2 n d duste

; dà'mi

dà'me

for

Now generally thematic, dudu etc. (also O.C.S1. dami dasi dastìi damu daste dadqtu ; dami damu have -m- for -dm-. Partic. thematic dady (da8t dqsta) like Lith. dà'dqs. As regards l dual Lith. duva *dùdmi

*dàdme.

Lett. dùdu).

O.C.S1. davè, see I § 547 p. 401. Remark.

The forms of the 2 nd pi. found in old Lith. books,

destit(e) and dustit(e) instead of déste and duste, were derived from the 3 rd sing, and pi. on the analogy òf tàri-t(ej : tùri, to distinguish more

clearly 2 nd plural from 3rd singular and plural.

§§ 547,548.

Present Stem: Class IV — Skr. aji-jan-a-t Gr. yl-yv-e-Tai.

Class IY. Reduplication ending in -t> or -U + Root -f- Thematic Yowel, forming the Present Stem. § 547. This class, like Class II, falls naturally into two sections, according as the root has the strong or the weak grade. The strong form, as in Class II, is the same as that of the non-thematic Conjunctive. Compare § 513 pp. 78 ff. § 548. A. Strong Root Syllable. In Aryan, this section includes a large class of forms, the Sanskrit Causative Aorist; an aorist formation which generally is found along with the present formed by -aya(§§ 795 if.). As to the varying quantity of the reduplicating vowel, see § 473 pp. 17 f. Skr. 3rd pi. mid. a-bi-bhay-a-nta beside bi-bhe-ti 'fears'. Imperative: mid. pi-prdy-a-sva beside d-pi-prS-t 'he satisfied, pleased', d-cu-cyav-a-t beside 3rd pi. a-cu-cyav-ur from cyu'to move, stir'. Skr. d-ti-tar-a-t Avest. ti-tar-a-J> from Skr. ti-tar-ti 'gets over or beyond'. Skr. pH-par-a-t from pi-par-ti 'fills', a-di-dhar-a-t beside 2nd and 3rd sing. dJ-dhar from dhar- 'hold fast'. Avest. bi-bar-Cimi (cp. Skr. conj. 2nd sing. bi-bhar-CL-si) beside Skr. bi-bhar-mi 'I carry'. Skr. a-ji-jan-a-t 'was born' Avest. zi-zan-a-Jt 3rd pi. zl-zan-a-nti, gen-. Skr. d-pi-pat-a-t, \Ppel- 'fly', a-si-sad-a-t, y/^sed- 'sit'. On the Irish conjuntive, used for the future, of which we have an example in gignid 'nascetur' for *gi-gen-a-ti, see § 544, page 103. Germanic. Apparently we have a form of this sort in Goth, rei-rai-p moves, trembles', connected with Skr. IS-ldy-a-ti 'wavers, trembles'; it may come from pr. Germ. *r%-rSi-5 (§ 469 p. 14, § 708). But this is not a certainty, because it has not yet been made out to what vowel series the root belongs (in Sanskrit we see a pret. d-le-lS-t, § 568).

105

106

Present Stem: Class IV — Skr. a-ji-jan-a-t Gr. yi-yv-e-rui. §§ 549,550.

§ 549. B. W e a k Eoot Syllable. Roots with i- and u- vowels. Avest. imper. di-dy-a (conj. di-dy-Ci-p), Gr. Mt,of.iai for *dt-$f,-o-ii(u beside Avest. didaeiti, see § 537 p. 97; Si-^tj-fiM (Class XI): M-£-o-ficu: Avest. didaeti = -Tti-n\rj-/xi: Skr. a-pi-pr-a-ta: Skr. pi-par-ti. Skr. ji-ghy-a-ti 'drives on beside hi-no-ti Class XVII; gh instead of h (I § 445 p. 331, § 454 p. 335) answers to k in the 3rd pi. ci-ky-ati, see § 537 Rem. p. 98. Also Skr. aorists such as d-Si-iriy-a-t from Sri- 'lean' (cp. d-ii-ire-t), d-ci-ksip-a-t from k$ip- 'throw', d-ri-ri$-a-t from ris- 'take hurt', d-Su-Sruv-a-t from Sru- 'hear', d-cu-krudh-a-t from krudh- 'grow angry', d-dU-dus-a-t from dtt$- 'grow bad, go to rack and ruin'. § 660. Roots with other vowels. Pr.Idg. *gi-gn-o, s/^gen- gignere': Gr. yi-yv-o-^iat Lat. gi-gn-o, cp. *§i-§en-o- § 548. *si-zd-o, y/^sed- 'sit': Skr. sidati instead of *sid-a-ti (I § 591 p. 447, § 593 p. 449, compare Bartholomae in Bezz. Beitr. xvn 117), Gr. "fw i. e. hizdo (I § 593 p. 449), Lat. sUd (I § 594 p. 450).!) *ni-nd-o 'I scold, blame' beside Skr. nad- 'shriek, roar* Gr. ovoo&e *ye blame, scold' for ovodro, whence oveiJog 'blame'.2) *si-st-o, \fsta- 'stare': Skr. tl-sph-a-ti Lat. si-st-i-t; *pi-b-e-ti, y/^po- 'drink': Skr. pi-b-a-ti Lat. li-b-i-t O.Ir. i-b-i-d, see § 539 p. 100.

1) Bechtel does not convince me that I am wrong in supposing the Idg. form to be *si-sd-o (Bechtel, Hauptprobl. der Idg. Lautlehre, 254). That alSffiat comes form alzS- or alsS- is unproven. Compare Idg. Forsoh. I 171 f. 2) This conjeoture (op. Osthoff, Perf. 394 f., and Bartholomae, Ar. Forsch. ii 84, Bezz. Beitr. xvn 116) seems to me more likely than that sugested by others (as Fick, "Wtb. I 4 96), namely, that Skr. nind-a-ti was formed from a \/ne%d- on the principle of Class XYI. The Skr. re-formation perf. ni-nind-tt etc. may be compared with perf. sid-atur fut. sid-isya-ti beside sid-a-ti.

§§ 551—553.

Pres. Stem : Class I Y — Skr. a-jl-jan-a-t Gr. yi-yv-e-rau 1 0 7

§ 651.

Aryan.

S k r . 3 r d pi. d-bi-bhr-a-n partic. bi-bhr-a-

from bl-bhar-ti 'bears',

-mOna-s p. 105).

cp. A v e s t . bl-bar-Wmi

ji-ghr-a-ti f r o m ji-ghar-ti 'smells'.

'strike, slay'. si-st-o, §

550.

548 Skr.

S k r . ji-ghn-a-te from

Skr. pl-bd-a-te 'becomes firm, strong',

S k r . ti-sfh-a-ti, opt.

(§

S k r . 3 r d sing, a-pi-pr-a-ta from pi-par-ti 'fills'.

han-

\Pped-.

A v e s t . hi-st-a-iti O . P e r s . mid. a-i-st-a-tCL: L a t . Ski-. 3 r d pi. mi-m-a-nti

from ml-ma-ti

'roars'

mi-ml-ya-t.

Another group of forms which comes in here is composed of such Skr. aorists as d-ffi-vgt-a-t from vart- 'vertere', d-cX-klp-a-t

from

'help',

kalp-

d-pi-spri-a-t

from

sparS-

'touch',

d-ci-krad-a-t from krand- 'roar'. A great many others were cast in the same mould as these; for instance, d-mi-mfna-t from mp-na-ti 'crushes'. § 662. G r e e k , yi-yv-o-FTCU, see § 550. ¡LU-JLIV-W beside ¡uiv-o) 'I remain, i'-o^-w beside e%-(o (*oty-io) 'I hold, have', y^segh-.

ni-nx-w

'I

fall';

whether

l was

original

(cp. S k r .

d-pi-pat-a-t, § 548 p. 105) is very doubtful; see § 473 p. 18. rt'xrw 'I beget' for *Ti-rx-w beside S-TCX-O-V, cp. the Author, Gr. Gr. 2 § 62 p. 74. ldx. from Vytiq- 'speak': Skr. d-vdc-a-t, Gr. e-ein-o-v eln-o-v inf. tln-siv (on Fun- for *ue-uq- see § 557). *se-sq-o- from V seq- 'be with, accompany': Skr. sd-Sc-a-ti, Gr. %-Cn-s-TO opt. '¿-On-o-i-ro inf. i-an-s-a&at. Skr. da-dh-a-ti 'places', Lith. de-d-ii, |/~ dhe-.

§ 682.

Aryan.

Skr. partic. ja-ghn-a-nt-,

Avest. 3rd pi.

ja-yn-e-nti conj. ja-yn-a-Ji: Gr. e-ne-

{¿o&t] and so forth, that is, from this grew the whole series of the "weak aorist passive". Compare O.Ir. 1st sing, -hurt rd by analogy of 3 sing, -lert, where -t is the middle personal

X) i((tfr>!s ileitis

i. e. *e-w-e-the8

follows Class I I B (§ 527 p. 90),

op. Skr. d-khy-a-t beside khy-a-ti, imper. jH-a beside jft-a-8ya-ti, etc. 2) Compare § 836, on e-xo^ea-»^; and § 840 on e-xQe/uao-fyi.

§590.

Present Stem: Class X — Skr.

131

dr-S-ti.

ending -to (§ 506 pp. 72 f.), and Lith. 1 st sing, eitit by analogy of 3 rd sing, et-t = Gr. tl-m (§ 686 Rem. 2). Dissyllabic stems with -3-, are rare in Attic (cp. t-yijpS-v for *tj-ha,X (Class XXXTT) as though belonging to Class X; e. g. Lesb. i\rj-(j.i 'I love, am wont' from qitXo-s, yogy-fit 'I carry' (Class XXXn), OTupuvat-fti 'I crown* from avsq>avo-q (Att. qui at for

) — Class X I Y , §§ 616 and 621, Class X X I X § 743 — so we have in Greek -n-io- instead of -no-. Lesb. xAiWw Horn. Att. nXtv(i) 'I bend, incline' for *xh-v-ito: Lat. in-cli-na-re O.Sax. Mi-nd-n Lett. sl%-nu § 603 p. 145. xgiwio xpivw 'I separate, choose out, decide', aiwofiai aivofiai 'I rob' (cp. Kretschmer, Kuhn's Zeitschr. x x x i 420). aXiva 'aXsi (aor. ¿Xivui), beside Lat. li-no etc., see § 598 p. 142, § 608 p. 148. o'rpjW 'I drive on for *o-Tpv-v-ipi from \/~ tuer- tur- (Skr. tvar-a-te 'hastens' O.H.G. clwir-u 'I turn quickly round, move', Skr. turana-s 'hastening'); with tru- : tur- compare Skr. hru-nd-ti: ju-hur-a-s beside hvdr-a-te, Avest. capru- Skr. catur- beside catvarand the like, (paivm 'I make appear, make visible, show' for *q>a-v-f,to: Armen. ba-na-m O.H.G. ba-nnu, see § 601 p. 144. Xuivw 'I gape' for *xa-v-f,(o with the aor. t-ya-vo-v, beside yd-tjxw xV'flV '• cp. Skr. M-na-s 'deserted, lacking' jl-ht-te 'yields, departs' (§ 540 p. 101). B y analogy of *XTEV-I W (xraVw) beside aor. f-xrtv-rra (sxruva) fut. *xTSvs(a)a (xTfvm), there were made in pr. Greek the aorist *'exhvau *ig>uvaa (exklva'¿(prjvu)and the future *xhvi(a)m * (xfavuJ ) from *xkivip> and *rpavip), and others in the same way. (

Remark, ipalvm shews that xXiva comes from *xHiBartholomae's doubts are unfounded (Stud. Idg. Spr., n 87 f.). The origin of the Att. ending -vvco is generally uncertain, as it may come from -vu>, -vf,a, or vfoa (§ 655). In any case, pairs of variants such as ¡ivva> Siva dvvm beside jStw Svm &vio produced I9vvui agrvvw beside l&vw agrvm, and then the analogy went further, and we have {¡Svvm xaypva etc. Compare Lith. keldunu § 615. § 612. pp. 148 f.

Italic,

ster-no

sper-nd

li-no falls

see § 608

§§ 612—614.

Present Stem: Class XIII — Skr.

151

mf-na-ti.

Other verbs with a weak grade of root: tollo ground-form *t\-no yf tel. si-no, origin obscure (cp. Osthoff, M. U. iv 133 f., Perf. 612). de-guno for *gus-nd, V gey,s-. Other verbs with strong grade of root, pello for *pel-nO (Umbr. a r - p e l t u 'admoveto'), beside Gr. mk-va-fiai, see § 602 p. 144. ex-cello for *cel-no, cp. Lith. kUno-ju § 606 Rem. p. 147. tem-nd may come from Idg. Hem- or Again cer-no, which is connected with Gr. y.givco and Lith. skir-iu, may be explained in two ways. If it contains the unextended root, it is on a level with pello etc. But it may have arisen in composition from *crind (I § 33 p. 34), in which case it will be analysed *cr-i-no and be more closely akin to Gr. xptvco. Lastly, pandO is doubtful. It is connected with Osc. p a t e n s i n s (Class XIY, § 622). If it comes from *pat-no (vol. II p. 161 footnote, Bartholomae, Bezz. Beitr. XVII 119), a must be derived from a on account of Gr. nsraoocu, and then the root had a weak grade. But pandd may belong to Class XYI; see § 632. § 613. Keltic. O.Ir. ser-nim 'sero' 3rd pi. -sernat (cp. "Windisch, Ir. "Worterb., p. 770 6). Perhaps sennim (sennaim) 'I drive, hunt' beside O.H.G. swimmu for *suem-no (§ 614 p. 152), cp. Mod. Cymr. chwyfaf 'I move, quiver' for *suem- (Thurneysen). § 614. Germanic. O.H.G. spur-nu 'I step, kick', and with strong-grade root syllable fir-spirni-t conj. -spirne: Lat. sper-no, see § 608 p. 148. A.S. mur-ne 'I trouble, grieve', cp. O.Sax. mor-n5-n, see § 605 p. 146.1) O.H.G. wallu 1) Forms like O.H.G. 1 s t pi. spurnames

infin. spurnan

partio. fir-

spurn an A.S. spurnan murnan are without fi-umlaut, by analogy doubtless

of the 2 nd and 3 r d sing. pres. and the plural of the pret. O.H.G. spurnum

etc., ep. O.H.G. inf. durfan beside darf dtirfum, and others. But A.S. has spornan as well as spurnan. Or had the "West Germanic originally forms of Class XYII beside those with -no-?

httnnu-m § 646.

Compare spurnum

with

152

Present Stem: Class XIII — Skr. mr-nd-ti.

§§614,615.

'I undulate, boil', ground-form *ul~no and willu 'I roll' O.Icel. veil 'I undulate, seethe' ground-form *uel-no. O.H.G. faUu 'I fall' Goth. O.H.G. kun-nan 'to know* partic. see § 608 p. 149. Goth, kun-na-nd-s O.H.G. kun-na-nt-i (indie, kann § 646): Skr. jd-na-ti, see § 598 pp. 141 f. O.H.G. chli-nu 'I stick, smear': O.Ir. gle-nim, |glei-, see § 604 p. 146. A.S. ¿i-ne O.Icel. gi-n 'I gape, yawn': O.C.S1. zi-ne-tU, see § 605 p. 146, § 608 p. 148, § 615 p. 153. Goth, kei-na (partic. kij-an-s) O.H.G. chl-nu 'I bud': Lith. gy-nu, see § 608 p. 149. O.H.G. swl-nu I disappear' (cp. Kretschmer, Kuhn's Zeitschr. xxxi 420), hr%-nu 'I touch, gain', gri-nu 'make a face, snarl', Goth, skei-na 'I appear, scl-nu O.H.G. backu Upper-G. pacchu 'I bake' pr. Germ. *bakkS for ground-form *bhdg-no (I § 214 p. 181, § 534 p. 391), cp. O.H.G. bahh-u 'I bake' Class II B Gr. qx/jy-m 'I roast' Class II A § 532 p. 94. O.H.G. spa-nu 'I attract, charm, drive on' (pret. spuon), ground-form * spa-no; also spannu 'I stretch, widen, I am in eager excitement' ground-form *sp9-nu-o Class XVIII (§ 654), beside Lat. spe-s spa-tiu-m O.H.G. spa-ti 'late'. Besides O.H.G. fir-spirnit and willu, other words have root syllables of the strong grade. O.H.G. quillu 'I spring, well up', gel-, O.H.G. swillu O.Icel. svell 'I swell, heave', O.H.G. scillu 'I resound, sound' O.Icel. skell 'I clatter, O.H.G. hillu 'I make a sound', O.H.G. gillu O.Icel. geU 'I yell, cry out'. O.H.G. sinnu 'I go, think' for *sind-nd cp. Goth, sandja 'I send', Mid.H.G. zinne I' burn for *tind-no cp. Goth, tandja 'I kindle'. I suggest that we class here verbs with -mm-, for -mn-, as O.H.G. swimmu 'I swim', cp. O.Ir. sennim § 613. Goth, fraih-na 'I ask' (perf. frah frehum, partic. fraihans) O.Icel. freg-n (fra fragum fregenn) A.S. frig-tie (frcegn frugnon frugnen) with pr. Germ, variation of x and j (I §§ 529, 530 pp. 384 if.), which was levelled down in different ways by different dialects: cp. Skr. praS-nd-s 'question' ISprek- (§ 607 p. 148). § 615. B a l t o - S l a v o n i c . than anywhere else.

Here this class is more creative

Lith. gy-nu p. 149. -ne-tu

'I revive, recover':

Lith. ly-na

Goth, kei-na,

'it rains': Lat. li-no

support':

in-clmare

O.C.81. zi-

Lett, sli-nu

'I lean

cp. Gr. Lesb. xh'-ww O.Sax. hli-no-n

(§ 603 p. 145, § 611 p. 150); Lett, si-nu

cp. Skr. si-na-ti verbs i

'binds, surrounds' V sai-;

doubtless

Lith. ry-nil rinetu

see § 608

(ibid).

'gapes, yawns': A.S. ~gt-ne {ibid).

upon,

'I

comes

from

swallow,

with

O.C.S1. si-ne-tu

(beside puv-ii)

' I rot', § 605 p. 146.

is

remain

'I knock down', iu-nil schtl-nu

*pljl-

*pljy-

Lith. ei-nu 'I go': a

unsafe

(see

§

'I snatch'

Lith. klu-nu

hanging',

griu-nii

(beside zUv-u)

*spiU-

1022).

'I sew'

gau-nu

(beside kluv-U) griuv-ii)

'I come to grief', O.C.S1. pli-ne-tu

Lett. 'spews'

and

Lith. au-nu

'put covering on the feet' \Seu-,

'sets in

(beside

(I § 60 p. 4 7 ) ,

Lett, splau-nu.

fu-na

Lith.

no doubt): cp. Skr. ju-nd-ti

(beside Lith. siuv-u)

like Lith. spidu-nu

si-t). 'knocks'

'I make rotten' cp. O.Icel.

quick motion, drives on, presses'. "I hook on to,

(sli-t,

'goes by'.

Lett, gti-nu

"I get' (-du- from gau-ti,

in the two Lettic

infinitive

'impetus').

Lat.

'I bind':

O.C.S1. ri-ne-tu

Lat. prod-lnunt

'shines', mi-ne-tu

Lith. pu-nu

the

devour';

sq 'starts' (cp. na-roji

comparison

for

153

Present Stem: Class XIII — Skr. mf-nd-ti.

§615.

plju-ne-tu Lett,

Lith. rdu-nu

au-nu

Lett,

'pull, tear, snatch out* (beside Lith. Lett, rau-ju)

rdu-nu

reu-,

and

etc. produced

Lith.

others (see Bielenstein, Lett. Spr. i 355). The analogy of rdu-nu: denominative

re-formations

beside the inf. kelauti karaldunu

infin. rdu-ti like

instead of karalduju

king' (from karala-s

keldunu

'to travel'

'king').

instead

(from kela-s

of

kelduju

kele-s

'way'),

beside infin. karalauti

'to be

Compare Gr. ifrvvio § 611 p. 150.

Slavonic gives a large number of «o-presents from verbs with other finals than -i and -w; planetU 'blazes up' for -ne-tu,

po-mq-ne-tu

mluk-ne-tu bunetu

'grows

'thinks upon, vrig-ne-tu dumb',

za-klenetu

awakes' for *bud-ne-tu,

'dries' (intr.).

'throws'

'shuts'

dvig-ne-tu

cp. Armen. sta-na-m

Gr. ara-va

(Vuerq-), *-klep-ne-tu,

'moves',

Also from one root in -a:

itself', akin to Pruss. stanintei,

for

*pol-

sta-ne-tu

such-ne-tu 'places

adverb of the pres. participle, Lat. de-stinare

§ 611 p. 149.

154

Present Stem : Class XIV — Skr. mj-nd-ti.

§§615,616.

In Slavonic the wo-suffix is not confined to the present stem; it appears elsewhere in the system of the verb, but then in the peculiar shape -nq-. Examples are aor. mi-nq-chu partic. pres. mi-nq-vu infin. mi-nq-ti sup. mi-nq-tU from mi-nq. -nq- is regular only in the infinitive and supine (except sta-ti from sta-nq). Remark.

The following I think is not improbably the history of

-nq-. Slavonic once had verbs in *-onq. (1 st sing, pres.), and their aorist ended in -on-su -qsu and their infinitive in -on-ti -qli, parallel to Lith. gyvenu 'I dwell* (fut. -$-siu infin. -in-ii) Tcilpinu *I heap' (fut. -j-siu infin. -iw-ti-), see § 624. Now in the present, -no- levelled out -ono-, which was only used with consonantal roots; but -ono- remained everywhere except in the present. Hence a compromise: an infinitive *vrigqti, for instance, would be transformed by analogy of vrignq, vrigne&i etc., and become vrignqti. Afterwards -nq- was extended to verbs from roots ending in a vowel, such as mi-nq,, and only sta-nq kept clear of this change (infin. ata-U). Compare with this § 624 at end, and Wiedemann, Arch. Slav. Phil, x 653 ff.

Lithuanian has no present stems with the suffix -no- from roots with a final explosive or fricative; instead of these the language has forms of Class XYI, such as bundit, as against Itinq in Old Church Slavonic. But there are a few in Lettic, all of them however with an interior nasal, which in most cases certainly belongs to the present suffix and not to the root: brinu 'I wade' for *brid-nu *brend-nu beside Lith. brendu and bredu (bridau brlsti), r&nu 'I find' for *riid-nu beside rMu — Lith. randu (radau rdsti), mif-nu 'mingo' for *menz-nu beside Lett, mqsu meigh-), linu 'I crawl' for lid-nu beside lidu = Lett, lendii (lindau llsti). The origin of this kind is obvious; the class with a nasal infix (Class XYI) has been contaminated with the -Mo-class, like O.C.S1. s^g-nq from v"seg- etc. (§ 636), and like Gr. h/unolvuj from Vleiq- etc. (§ 631). C l a s s XIY. Boot H — - e n o -

-ono-

forming the P r e s e n t Stem.

§ 616. It is quite clear that this class is derived from nouns; see § 487 p. 41, § 596. 6 p. 140, and below. It is note-

§617.

Present Stem : Class XIV — Skr.

mf-nd-ti.

155

worthy that the «-suffix is often extended by -¿o-; as Skr. isan-ya-ti beside (Yed.) isana-t, Gr. ofao&aivio beside okio&dvw, O.H.G. giwahann(i)u 'I recount5. See §§ 618 and 743. This is the same formation as Skr. vithuryd-ti from vithurd-s 'staggering, shaking', Gr. alalia from alolo-g, see § 770. § 617. Along with -eno- we find -pnd- -end-, inflected in the same way as denominatives from a-stems. To illustrate, take: Skr. pptand-yd-nt- 'fighting' beside Ayest. pesana-iti Skr. pytan-yd-ti beside Skr. pftana-m pftand- 'fight', Skr. bhandand-ya-ti 'shouts, cheers' beside bhanddna-s 'shouting' bhandand- 'shout'; Gr. eQvxavau) beside ¡(iwdva 'I hold back, bar, stem' (cp. Syyava 'I sharpen, whet' beside &ijyavo-r S-riyavrj 'whetstone', and Sanavaw 'I spend' beside ¿anavo-g 'extravagant' Sundvrj 'expense'); Lat. runcinare (cp. runcina 'plane') coquinare cdrindre farcindre; O.Icel. vakna 'I awake' pret. vakna-Sa, Goth. pret. ga-vakno-da beside pres. ga-vakna; Lett, stiprino-ju 'I strengthen' infin. stiprino-ti beside stiprinu (infin. stlprin-ti), gabend-ju 'I bring together' (infin. gabeno-ti) beside gabenii (infin. gaben-ti); and besides, the Lith. group of preterites, of which examples are 1 st pi. stlprino-me gabino-me, must be added. Seeing how clear is the denominative character of this fourteenth class, no doubt can be felt that all these verbs are derived from feminine stems. The nearest parallel is found in the verbs which will be discussed in § 769, Skr. priya-yd-te Goth, frijo, O.Ir. com-alnaim O.H.G. follom, and such like. That is to say, Skr. bhandano-yd-ti stands to bhandand 'shout' and bhanddna-s 'shouting' exactly as O.H.G. follo-m 'I fill' to folld 'fullness* and fol 'full', or as umntom 'I make wounded, I wound' to wunta 'a wound' and wunt 'wounded'. R e m a r k . The student must not suppose that I refuse to see the parallelism between I(VXARAA>: TGVXDRIO and NITRATO : TTCTVW ; O.Icel. vakna: Goth, ga-vakna and O.H.G. gind-m: O.Icel. gin; Lith. stiprin6ju : stiprinu

and lynoja : igna. See the end of § 598, and §§ 602, 605, 606. The origin of the a-flexion is different in the two sorts, but a-flexion in the one may well have influenced the other in different languages inde-

156

Present Stem : Class XIV — Skr.

mr-nd-ti.

§§ 618,619.

pendently. For instance, Greek verbs of the type of iqvxaraio may have been supported by the use of naram, or vice versa. Skr. bhandana-yA-te is not to be olassed with hrna-yd-nt-, a- quite isolated stem; we see this from a variant hrni-yd-mana-s (§ 600 p. 144). Suoh forms as *bhandani-yd~ti do not exist.

§ 618. I cite first forms which appear in more than one language. Here, as below with forms belonging to one language only (§§ 619 ff.), the extension with -io- must be cited too (§ 743). Lat. cruen-tu-s partic. of a present 3 rd sing. *cruini-f, Lith. kritvinu 'I make bloody' (partic. kriivinta-s = cruentu-s) from krii-vina-s 'bloody'. Armen. aroganem 'I sprinkle', Lith. sravinu 'I make flow', common ground-form *srou%n5, l/^sreu- (cp. Bugge, Idg. Forsch. I 451). Skr. injunctive isana-t 'let him set in motion, arouse, excite, quicken' and i§an-yd-ti, Gr. iaivta 'quicken, hasten, warm' for Ha-av-jjro. Goth, af-lifna 'I remain over' (pret. -no-da), Lith. lipinu 'I cause to adhere'. Goth, dukna 'I increase, grow' (pret. -nd-da), Lith. auginit 'I make grow, rear'. Gr. avaivw 'I make dry, wither' for *tiav(S-ocv-i,m, Lett. saUsinu 'I make dry'. Compare Alban. &an 'I dry, wither', for *saus-nio according to G. Meyer (Alb. Wort. 85, Alb. Stud. hi 43). Gr. TFQaaivw 'I make dry, dry up', Goth. ga-]>aursna 'I grow dry, wither' (pret. -no-da). § 619. Aryan. Avest. opt. 1 st pi. zaranaemcl and zaranye-te (partic. saranimna-) from zar- 'grow angry, ill', cp. partic. zaranu-mana- Skr. hp-ni-te § 596. 3 p. 138. Avest. pesana-iti 'fights' beside Skr. pftana-m pftano- Avest. pesana 'fight, battle' (cp. § 617 p. 155). Skr. typdna-te 'he behaves pitifully, prays' beside kfpand-s 'pitiful, miserable' kgpdna-m 'misery', isana-t 'let him set in motion and isan-yd-ti: Gr. iaivw, see § 618. Only with -io-: turan-ya-ti 'hastens, goes or makes

§§619—621.

Present Stem: Class XIV — Skr. mr-nd-ti.

157

to go quickly' from turdna-s 'hastening' (pres. tvdr-a-te 'hastens') cp. Gr. orpvvw § 611 p. 150; bhuran-ya-ti 'he is active' from bhurana-s 'active', and others. Compare Skr. p%ianO-yd-ti bhandana-ya-ti § 617 p. 155. There is nothing to decide whether this Aryan representes Idg. -pno- or -eno- (those who believe that becomes a in open syllables in Aryan will say, or either). isanyd-ti as compared with Qr. laivcu, so far goes, favours -yno-.

-anaIdg. o -onoas it

-eno- must be the suffix in Skr. bhdna-ti 'sounds, calls out', if this be derived from |Sbha-, and analysed bh-dnati; see p. 56 footnote. Perhaps the same suffix is used in some of those forms which are cited by Per Persson, Wurzelerweiterung pp. 70 ff., such as dhvana-ti sounds'. § 620. A r m e n i a n . In this language -ano- = Idg. -y.nois a very common present suffix. IK-anem 'I leave', aor. 3 rd sing, e-litf, \/^leig-. gt-anem ' I find', aor. 3 rd sing, e-git, lSy,eid-. fU-anem 'I spew, spit', aor. 3 r d sing. e-fuU. kl-anem 'I swallow', aor. 3 rd sing, e-kul. hat-anem 'I cut off', tes-anem 'I see', y/~>der%- (I § 263 p. 214). liz-anem 'I lick' for *lezanem, VleiQh-. -anem, like Greek -avw, is found in some forms which have another present suffix already. As for instance harcanem 'I ask' beside aor. harci, stem *pf(k)-sko- (§ 672), like Gr. aXvaxaviv beside dXv-oxa>; and very near akin to harcanem is Avest. per'sanypiti 'asks', if its -s- — Skr. -ch- (cp. Skr. prachana-m 'an asking') and not Idg. -1c- (cp. Goth, fraihna). -anim (cp. § 711) is a variant of -anem as Gr. -utva of -avio; e. g. mer-ani-m 'I die' (aor. mer-ay) like Gr. /tap-am« 'I make wither, decay', mac-ani-m 'I cleave to, hang on to, curdle', zerc-ani-m 'I free or save myself, run away'. § 621. G r e e k . In this language too -avo- = Idg. is very common. cUij>-av(o 'I earn', xvd-avio 'I honour, exalt'. xev&-av § 631. Lat. clang-d, cp. Gr. xXayy-avio and xXaCm for *-/.Xayy-f.co § 631 (pf. ydxXayya), O.Icel. hlakka 'I cry out' (-kk- for -nk-), beside Gr. xloj^a 'I cluck, caw' for **Xo)y-f.ddv(» xsiao/.tut § 631.

|ghed-,

'I understand,

'I

Lat. pre-hendo,

found',

conjecture'

for

Several languages give jo-inflexion to this type (Class X X I X ) . Examples: Gr. nriaaut nxirxm instead of *nriva-i,io, nXd'Qto for

166

Present Stem: Class XVI — Skr. yufij-d-ti.

§629.

*nXayy-)((i)- Lat. vinc-ii5, sanc-id (cp. sacer); Lith. jung-iu, Lett. mi/chu (beside mi/nu) 'mingo' for *minz-iu. See § 744. § 629. A r y a n . Skr. vind-d-ti Avest. vind-a-iti 'finds', [/" uej,d-; Skr. siftc-a-ti Avest. hinc-a-iti 'pours out', I/"seiq-; Skr. kpit-d-ti Avest. kerent-a-iti 'cuts', see § 628 where other examples are given. "We may also mention the following: Skr. iys-a-ti 'leaves over' beside iinas-ti; und-a-ti 'moistens, wets' beside undt-ti; umbh-a-ti 'holds together, holds in custody' beside 2 nd sing, unap; tymp-a-ti 'is satisfied' |/" terp-; bfh-a-ti 'strengthens* \/~bher§h-; Spith-a-ti from ¿rath- 'to become loose or soft'; Avest. mer'nc-a-its from marc- 'destroy' beside 2 nd pi. mid. mer'ng'-duye (§ 626). Sometimes in Sanskrit the accent is changed to the accent of Class I I A, as iumbh-a-ti and iumbh-a-ti 'adorns' (beside i6bh-a-te), partic. mid. tuftj-a-mdna-s (3rd pi. tuftj-ate Class XY, tuj-yd-te 'is struck, knocked'), dfh-a-ti beside dfh-d-ti 'strengthens' (beside dfh-ya-ti), pffic-a-ti mingles' (beside pfndk-ti and pi-pgg-dhi). With secondary strong grade vocalism (cp. § 628 p. 165): Skr. ¿ranth-a-te (gramm.) beside Jfnth-a-ti, Srambh-a-te 'entrusts (cp. ni-Spnbhd-s), anu-rafljati 'cleaves truly to, loves' (cp. raga-s 'colour, passion, love', Gr. pf'fw pcy/ux ^oysvg), Avest. 3 rd sing. pret. mor'nd-a-Ji for *marend-a-Jt (I § 94.3 p. 89) from mard- 'kill' (or does -ar- = -f-?); of the same sort may be Skr. vdnd-a-tS 'praises, honours' beside vdd-a-ti ud-yd-tS. Roots of the type peq- (§ 628 p. 165). Skr. spand-a-te 'throbs', beside Gr. o 'I tread something hard' arißago-g 'firm, pressed, solid' (cp. Lith. sUngiu § 637). sklfstü 'I flow apart' (sklindaü sklisti) beside sklld-ina-s 'full to overflowing' skleidziü 'I spread'; a pret. 3 rd sing, sklidu (sklidö) is also found, pointing to a present *sklind-it. Lett, stringstu 'I grow tight, dry up' (stringu stringt) beside Lith. string-u 'I remain hanging' (strigau) and streg-iu 'I crystallise, stiffen' (cp. § 628 p. 164). Lith. drfstü 'I grow bold' (drpaü dr\sti), dhers-. linkstü 'I bend' (linkaü linkti) beside Gr. Äsx-avy 'pan, fan' Aojo'-g 'crooked'; also Lat. lanx with nasal (for *\taq-Y). The model for these presents is

174

Present Stem: Class X V I — Skr. yunj-d-ti.

§§636,637.

seen in blista 'it darkens' beside blind-o l^bhlendh-, tfstti 'I stretch myself out' beside tis-au stem tens-, and the like. § 636. This formation is much rarer in Slavonic than it is in Baltic. O.C.S1. strig-q 'I shear, slave' for *string- ? see § 628 p. 164. grqd-q 'I come' (inf. grqsti) for ghpidh- or *ghrendh-: O.Ir. in-grennim, see § 628 p. 165. sqd-q 'I sit' (inf. sesti), V~ sed-, cp. Pruss. sindats syndens 'sitting' beside sidans sTdons = Lett, s'edqs. tyg-q 'I lie' (inf. lesti), V leqh-. trqsq 'I shake, shatter' inf. trqs-ti from tr-es-, unless it comes from *trem-so- (cp. Lith. trimii 'I tremble' Lat. tremd), see § 657. As regards grqi-q Iqkq 'I bend' prqd-q 'I spin' compare § 637. Sometimes extended by -io- (§ 628 p. 165). zqzdq 'I desire, thirst' for *sqd-iq (inf. zqdati) beside Lith. pa-si-gendu 'I miss' and geidziu 'I long for', glqzdq 'I look, gaze' for *glqd-iq (inf. glqditi) beside Mid.H.G. glinze 'I shine' O.H.G. gk$u 'I glitter'. See § 637. With nasal confined to the present system: ob-rqstq 'I find' for *-rqt-ia, inf. -resti aor. -rStu (for the etymology of this verb see § 687). There is another extension, with -no-, vyk-nq 'I grow used' doubtless derived from *vykq = Lett, jitku for *(j)unk-u, beside n6q 'I teach' (§ 635 p. 173). sqg-nq 'I long for' beside Lith. seg-u 'I fasten', cp. Skr. sa-saflj-a § 629 p. 166. krqnq 'deflecto' for *krqt-nq (cp. krqtiti 'to twist, turn'), beside Skr. kfnat-ti 'turns the thread, spins' k&rtana-m. sqk-nq 'I sink' beside Lett, siku 'I sink, fall' for *sink-u, Vseiq- (§ 628 p. 164). rqg-nq 'hisco' beside Lat. ringor (inf. ring-%) ric-tu-s. Compare § 637. § 637. Side by side with Lith. drimbil (ground-form *dhgmbh-o) and the like stand forms with e in the root syllable (cp. § 628 p. 165). drqs-u 'I am bold' (pret. drTs-au) beside dr\s-tu y/^dhers- § 635 p. 173. brendii (dialectic brindu for brendu) 'I wade' beside bredu (brid-au) O.C.S1. bred-q. lenk-iii 'I bend' (lenkiau lenkti) beside link-stii \Pleq- § 635 p. 173. trindu 'I am devoured by moths or worms' inf. trende-ti, with tride beside Skr. tfnatti tard-a-ti § 692. We may assume

§637.

Present Stem: Class XVI — Skr. yunj-d-ti.

175

that drjis-u for *drins-ii was coined to supplement dris-au on the analogy of renk-u: rinJcau, kertu : kirtau etc.; lenk-iu appears beside linksth on the analogy of grqz-iu 'I turn, twist' beside grfyztii 'I turn myself' etc. Slavonic verbs with q, grqd-q Iqk-q, and *krqt-q which appears to be implied by krq-nq, may quite well correspond to Lith. drimb-u or to Lith. drqs-u. ') Baltic en Slav, f is found in present stems from roots with ¿-vowels both extended and unextended. Lith. senkii 'I fall, sink' (of water) O.C.S1. sqk-nq 'I sink down' beside Lett, siku for *sink-u Skr. sific-d-ti \S seiq- (§ 828 p. 164). Lith. sprindziu 'I grasp with the hand' {sprqsti) O.C.S1. prqdq 'I spin' (prqsti) beside Lith. sprindi-s m. 'span' Lett, spraid-s place where one stands in a narrow compass' debes-spraisli-s 'vault of heaven' O.H.Gr. spreiten 'stretch out, separate, part asunder'. Lith. pa-si-gendit 'I miss' O.C.S1. zqzdq 'I desire, thirst' for *zqd-iq beside Lith. geidsiit 'I desire' Goth, gaidv n. 'lack' O.H.G. g%t 'eagerness, greed, avarice'. Lith. sting-iu 'I apply my strength to something' beside stlnkstii 'I congeal, get stiff' Gr. OTeifim (§ 635 p. 173). Lith. m$z-ii2) 'mingo' (misau m\szti) Lett. mif-nu for *menz-nS beside Lett, mi/chu for *minz-io (§ 635 p. 173) Lat. ming-o Lith. mlze f. 'cunnus' miz-iu-s 'penis', [/" meigh-. O.C.S1. glqMq (inf. gl^diti) and glqdajq (inf. glqdati) 'I look, gaze' beside Mid.H.G. glinze 'I shine, glitter' (pret. new formation glam) O.H.G. gltgu O.Sax. glltu 'I glitter' [S ghleid-. O.C.S1. rqgnq 'hisco' (rqgu 'jest', subst.) beside Lat. ringor ric-tu-s. If the Baltic forms stood alone, the explanation would be easy; we might say that the analogy of renk-: rinketc. produced senk- mens- beside sink- minz-; compare what is said above on drqsu. But this explanation does not suit

1) The fact that we find kr$t- and not crqt- is not sufficient to prove that the ground-form of krqt- is the weak grade *qrnt-. Such a form must have beoome Slav. *krintas *dhrns- becomes Lith. drinsand *qpit- becomes Lith. krint- (I § 285 p. 227). There never was a form *kirnt-, nor yet *gri}t-, which Bartholomae suggests as the ground-form of kret- (Stud. Idg. Spr., n 97). 2) Dialeetio minzu = *menzu (vol. I § 285 Bern., p. 227, is wrong).

176

Present Stem: Class XVII — Skr. ¡'-nS-ti.

§§ 637,638.

the Slavonic forms, because in Slavonic, before consonants, Idg. in becoms f, but Idg. y, becomes q (I § 219. 4 p. 186). R e m a r k . "Wiedemann's view (Aroh. Slav. Phil, x 652 f., Lit. Praet. 58, 168 f.) — that Idg. in and un before consonants become Slav, q and q, except in final syllables — can hardly be maintained in this connexion, because we have isto = Lett. Inhstas, lylco = Lith. lunka-s Pruss. lunka-n and smrid-q (see below). Nor is Streitberg's attempt satisfactory (Idg. Forsoh., i 283 f.). Perhaps the problem may be solved thus. We may suppose that originally in and un always became l and u; but that later, when in and un were again produced in any way before consonants, these became $ and q. "We may suppose that sink- first beeame *stk-, and afterwards, as the principle of Class XVI still remained aotive, the nasal crept into the stem anew; compare (say) Gr. Att. '¿vrvfii for *fcorvp which took the place of pr. Qr. "fervv/ii ( = Ion. &Vfy») for orig. *fta-vv-f*i (I § 565 p. 422). Similarly bqdq may come from *bhu-dh6 or *bhii-dO, and may have got its nasal only at a late stage of proethnic Slavonio; though it may equally well be derived from *bhyra-dh6 or -do attracted into the nasal class, or from *t)hy,on-dh6 or -do regarded as an extension of a form Hhurono (cp. § 701). Furthermore, for the 3 rd pi. smrid-qtu beside smrid-i-mu etc. we may assume that the old ending *-int(u) (cp. part, smrid-qt- Lith. smird- -»'«, which we see in the strong persons of the singular, seems to have pushed out Idg. -neu-, because of the analogy of the forms -va- : -va- (Class X I I ) , cp. § 480 p. 29, on vftsvaiow beside iifisvatm. Even if we supposed that -vv- represents Idg. -MM-, a weak grade, used along with -raw-, we should have to assume that the forms had followed -va- : -va-; and Avest. -rait- is not sufficient evidence for an Idg. -nU-. In the 3 rd pi., -w-avn (instead of *-vv-evn = Skr. -nuv-anti, see § 1021.3) seems to have become regular quite early; once there were in use such forms as *Tt-rf-avri *rt-vf-fvu = Skr. ci-nv-dnti (cp. § 638 p. 177); as to Ion. ayvvai Att. ayvvvrai, see §§ 1020. 2 and 1065. 2. Besides the forms mentioned in § 639 — op-w-fii, ag-vv-pai, ar6g-vv-f.it, NRAG-vv-ftai, Ta-vv-ftai, A-vv-fu, N-vv-fievou TI-VV-VTCU — there are yet others with weak-grade vowels in the root syllable. &ag-w-(iai in Hesychius (-«(>- = and &oQ-w-fiai (-op- = -f-) 'I leap, cover (of animals)' (I § 306 p. 241). xt-vv-ficu 'I move myself'. Cret. 3 rd sing. m-dix-vv-Tt = Att. im-dsiy.vv(Ti (on nt-, see the Author, Gr. Gr. 2 p. 219) y"deilc-. oiyvv/ui 'I open' Horn. w-(f)iy-w-vTo beside Lesb. inf. o-tiy-tjv, originally 'I make yield', beside O.H.G. wthhu 'I yield, give way'. t.dy-vv-f.u 'I mix' beside fut. fie%w, mejk- tneig-. 0f.10Qy-vv-f.it 'I wipe' for *tnfg-, \Smer§~. ax-w-fiut 'I am grieved, troubled', beside Goth, un-agands not fearing' og 'I fear'. An old form with strong root (third strong grade) is Horn, dyx-vx'-fisvo-g 'doing honour to, reverencing, greeting', see § 639 p. 178. Greek new formations with a strong root-form are ogsy-vv-ftt 'I reach, stretch out' y^reg-, ¿'eix-vv-fu 'I show' beside Cret. m-SU-vv-n, fevy-vv-fit 'I bind* V^j^Ug-, tir^y-vv-fa 'I fix' y/^pttkpa§- and others. Ion. ¿¿y.-vv-fu 'I show', coming, as we may conjecture, from a ydek-, but in use finally confused with 4 ¿Hx-vv-fti (cp. Fick, Wtb. I 66). ofi-vv-fii 'I swear' beside o,tto- (¿fio-aaai bfio-vrj-g), ollv-fit 'I destroy' for *ok-vv-/.u

182

Present Stem: Class X V I I —

r-nó-ti.

§643.

(I § 204 p. 170) beside oh- (okt-aaai), like ¿a/u-v?]-/ui beside iafta-, xafi-vco beside *a,ua- (§ 602 p. 144). The place of (Ion.) e'ivvfit 'I clothe' for *J~-sa-yv-/^i = Armen. z-genu-m (§ 639 p. 178) was in Attic taken by a new form '¿wvfti; see I § 565 pp. 422 f. The following are forms of the same kind: aftiwvf.it 'I quench, stop' for earlier &ivv/.u i. e. zdtivvfit (Hesych.) 1 ) from a stem *zq-es- \Sseg-, cp. aor. Horn, ojisa-oai; (¡¿evwfiat (gramm.) beside fid-ico 'pedo' aor. fti-saai for *pd-so-, earlier *ftzd-sa-, \Z~pezd- 'pedere' (cp. § 661). Further, ^oivvvfii 'I gird"-) beside laa-Ttjp Idg. j-Qs- (§ 656). On the model of these were made xogtvvv/iu 'I satisfy', nsravvv/iu 'I spread', Quivvvpu 'I strengthen', OT(tb)vvv/.u 'I strew, spread* and others; and the analogy of tj¡.«pi-tea -sa/.iai: afiii-/.u is represented by mwpsv/jv • ovvexi)v Hesych., compare m-w-zo-g 'enlightened, sensible' nivvaato nivvoi-g. This, along with vfj-nv-rio-g 'senseless, under age, minor and vrfmo-g (same meaning) for *vrj-nJz-io-q (I § 166 p. 147), is akin to Skr. pu-nd-ti 'purifies, clears up' (for the accent cp. Goth, hugs understanding, reason' beside Skr. Suci-s 'pure', § 907). But mw- does not come from *nv-w- (I § 48 p. 41); the ground-form was *pu-i-nu~, having the same determinative I as we see in Ital. *pu-%-io-s (Osc. p i i h i u i Lat. piu-s, see Bartholomae, Stud. Idg. Spr. n 185) Skr. pav-l-tdr-, and in Gr. nviQ Umbr. pir O.H.G. fuir 'fire'. It follows that *nivvfu: Skr. pu-na-mi == Skr. r-i-nva-ti Gr. ooivw: Skr. p-nva-ti Gr. oQ-vv-m (cp. § 596. 4 p. 138).

1) Hesychius has ¡¡civapfv' aftirrv/iev which is emended to ieirvfier This emendation is not necessary. There may quite well have been parallel forms, one in Class X I I and one in Class X V I I , as so often happens in Sanskrit. Then the form Zeirvpi in the text should be marked with an asterisk. 2) It is quite possible that Att. vno-la>rvra[i] C.I.A. I 77. 9 (second half of the 5 t h cent. B. C.) may represent the regular form (cp. l&ntroc, Meisterhans, Gr. 2 , p. 148).

§§ 643—646.

183

Present Stem: Class XVII — Skr. r-nS-ti.

Xa'Qv/.iai beside Xd^ofxat 'I take, seize', xttivviu beside xr«W> 'I slay* (for *HTSV-T:TO) are due to the analogy of TZVO/MU rivo/uat (for *Ti-vf-o-/Ltai): rtvvvrai

nvv/usvai, etc.

B u t xaivv/uat ' I surpass,

outdo' was formed from y.t/aa;tai because Js'tfaaftai has Sai-vv-uai (§ 707). On thematic forms in -vf-co see § 652. As regards those in -vv'w, as TUVVIO oiivvio Orpcovvv'co, found in the Homeric dialect and in Attic more and more often from the 4th century b. c. onwards, it is doubtful whether they represent pr. Idg. verbs in -nuyro, which may have been used side by side with

-nu-o

as in t h e

3 r d pi. S k r . ai-nuv-anti

Gr.

ay-w-dat

beside Skr. ci-nv-anti. They may equally well be a new formation peculiar to Greek. Prom rawficu ravvio yawf.iai, whose structure was less clear to the consciousness of those who used them than was that of iQ-vv-f.u ay-vv-fxi and words of that sort, were formed

ravvooou

analogy of sgvaaai

TtTavvarai iipvarai

fgvw 'I draw, pull'. hs-nHu-ti,

yavvaasrat sgvaascat

and

so forth on

Compare Skr. partic. ksiiu-td-s

f u t . a&nuvisya-te

from

the

beside t h e pres. slgv/.isvat aS-no-mi

ds-ta,

from

Urnu-tya-

from UT-tyWw-ti (§ 640 p. 179), and Greek itself ¿vva-To-g ¿¿vvr/(ju/uyr from Sv-va-ftai

(§ 602 p. 145).

F o r fu-vv-9-w (cp. Solmsen, Kuhn's Zeitschr. xxix 98), auov for *ai-oiov; cp. Skr. d-tvi-$-ur § 656.

From

y^preu-

(Lat. prmna): Skr. pld-sa-ti 'burns, singes', O.H.G. friu-su 'I freeze, am cold', cp. Lat. priir-iO. From y^leq- (Gr. ai-akx-eTv 'ward off'): Skr. rak-§a-ti 'guards, saves', Gr. alebo 'I ward off, help'. Connected with Skr. vdjaya-ti 'strengthens' 6jas- 'strength, power, might' Lat. augeO: Skr. uk-§a-ti 'grows strong, increases' partic. uk-$d-mOna-s (perf. vavdk$d) Avest. vax-sa-iti 'makes grow', Gr. a(f)tha> av%io 'I make grow, increase', cp. Lat. aux-iliu-m,') Gall. Uxello-dunu-m 'High-town' O.Ir. 5s uas 'above' (I § 517 p. 377), O.H.G. wahsu Goth, vahs-ja 'I grow' (pret. vdhs), Lith. auksz-tas 'high'.2) Gr.'¿xpm'I boil', which, along with Armen. epem 'I boil' (I § 561 p. 417), we may assign to the root of snm 'I see about, make right, arrange' (II. 11. 776 a/xtpl poog entxov ttgta) Skr. sdp-a-ti 'makes

a fuss about, carries on, sees about something'. From y ^ ter- (Skr. tar-ald-s 'moving to and fro, trembling' Gr. TQ-ifxw Lat. tr-emS § 488 p. 45): Skr. tr-dsa-ti 'trembles' (also tar-dsa-ti § 659), Gr. rg-t(6)u> 'I tremble, flee', O.C.S1. trqsq 'I shake, shatter' perhaps a re-formate instead of *tresq (§ 636 p. 174); with -s-, Lat. terred for Hers- (cp. Gr. STBQOSV syofirjOtv Hesych.).

Compare Skr. gr-asa-ti

hr-asa-ti

bhy-dsa-ti

Avest. v-atdha-itc § 659, Gr. |-t(o)w /M-f(or)w § 661, Lat. qu-ero-r § 662. 1) According to Brial's convincing explanation, Umbr. orer ose 'his (donis) maote' will fall in this place too. ose = pr. Ital. *aulese may be a vocative, which would make it necessary to start from an adj. *aukso-s meaning 'auctus'; it may also be an imperative like Gr. avte (cp. Lat. auxim). The first view is supported by Lat. macte, a vooative (F. D. Allen, Am. Journ. Phil., I 135 ff.). Pauli's explanation of ose (Alt. Stud, v 123) does not seem right to me. 2) On the relation between jteg- ajjg- ajfeg-, see Per Persson, Wtirz., 228.

§§ 658,659.

193

Present 8tem : Class X X — Skr. iq-sa-ti.

§ 668. In § 655 it was pointed out that these s-suffixes are probably connected with the noun suffixes -es- -as- -s(§§ 131 if.). A few more examples of this may be given: Skr. Gr.

Skr.

(§ 6 5 9 ) : L a t . decus.

S k r . -rdcas-

(§ 6 5 9 ) :

rdci§-

G r . ddog fitifoQ

(§ 6 6 7 ) :

Skr.

Armen. luci

(p. x h i ) .

Gr. xQt(f)ai;.

659):

daias-ya-ti

O . P e r s . patiy-

L a t . viso (§ 6 6 2 ) :

rucis-ya-s.

slu-chu (§

O.C.S1. oko gen. odes-e.

S k r . vdcas-.

'iao-g for ftra-fo-g

(§ 6 6 4 ) : S k r . kraviS-

(§659):

(§ 6 5 9 ) : S k r . sdhas-.

S k r . ik-sa-te

uk-sa-ti

O.C.S1.

S k r . bhu-sa-ti

S k r . ddk-sa-ti

S k r . sak-sa-nt-

Avest. vax-sa-itS

Skr.

tenor.

Gr. yls(f)oS.

bhuvas-.

(§ 6 5 9 ) ,

Lat.

tdnas-

S k r . Sro-sa-mUna-s

ojas-.

S k r . irdvas-

S k r . bhavas-axsaiy

Skr.

tq-sa-ti:

avhv:

(§ 6 6 0 ) :

Skr.

vedas-

O.Icel. inf.

hrjosa

Compare also S k r .

bhartsa-ti

(beside bhartsaya-ti) attacks sharply, rates, scolds', akin to Lat. fer-io, Lith. bar-iii 'I scold', and so doubtless derived from some such stem as *bhar-tas- (cp. sr6-tas- 'stream') or *bhar-dhasdaksd-

sak$a-

dak$a-ti

Gr. Ftrog,

gift').

are

The

connected

nouns

sdhas-

vdcas-

S k r . ¿lr-sa-

in

the

to Siras-,

-tqsa-

with

were therefore related

vaxsa-ite

(daias-ya-ti) to

'grace, which

vax-Sa-

sdk$a-nt-

*daiasvat-sd-

(cp. rd-dhas-

same hd-sa-

tqsa-ii

to

tdnas-

way to

as

havis-,

and so forth (II § 132 p. 190). S k r . bh&sa-ti S k r . hr-asa-ti S k r . yajas-.

(§ 6 5 9 ) : bhds- L a t . (§ 6 5 9 ) :

T h e s e are like

fas. S k r . yaj-as-l

hdras-. bhy-dsa-ti

(§

beside bhyds-

656): bhiyds-

(already mentioned in § 655, page 190). § 659.

d-tvi-sa-ta -§a-ti,

Aryan.

S k r . tq-sa-ti

d-ta-sa-t,

bhdsa-ti,

tvS-sa-ti

(tvS-id-s 'boisterous' Avest. pwae-§a- 'terror'),

plo-

Avest. vax-sa-iti

see

rdk-§a-ti,

iik-sa-ti

uk-id-mUna-s,

§ 657. Skr. ar-sa-ti f-§a-ti 'moves quickly, flows quickly', from ar- 'begin to move' (f-n6-ti). Skr. i-Sa-tB 'sets in motion, sends forth' Avest. aesemna- isaiti, from i- 'to send' (i-no-ti). Skr. ¿rS-sa-ti 'hangs to something, clasps* a-ili-sa-t, Avest. sraeSemna-,

\/~*Bei- 'lean' ( L a t . -cli-nd).

S k r . Sr6-sa-ti

"hears*

(redupl. iu-irti-Sa-tS § 667) Avest. sraoSemna-, i^Uey,- (2nd sing. B r a g m a n n , Elements. IV.

13

194

Present Stem: Class X X — Skr.

S 659.

tq-sa-ti.

cp. O.Ir. eluas 'ear' (I § 516 p. 377) O.Sax. hlus-t 'hearing' O.C.S1. sluchu 'hearing, faculty of hearing'. Skr. gho-sa-ti 'cries out, proclaims aloud' ( g h o s a - s 'confused noise, roar of a storm, cry of woe'), beside Goth, gau-non 'to cry for woe'. Skr. bhu-sa-ti 'applies himself to, takes trouble about', y/^bheu-

¿ro-si),

(Skr. bhava-ti).

Skr. sUrk-sa-ti

'is anxious about

something,

takes thought for it, or account of it', beside Goth, satirg-a 'care', which must be connected with O.Ir. sere 'love' or with Lith. serg-iu 'I protect'. Skr. mrak-sa-ti mfk-sa-ti 'rubs, strokes', beside mars-fi 3rd pi. mrj-dnti. Skr. aksa-te ground-form 'reaches' (§ 640 p. 179); from the same root, ndk-sa-ti 'reaches' beside ndi-a-ti Lith. nesz-u; with these must also be connected Goth, bi-niuhs-ja 'I search out niuhs-ein-s 'visitation, affliction', which come from *nux-s— Idg. (cp. gi-naiihan 'to suffice'). Skr. mok-sa-te 'gets free from something' (redupl. mii-muk-§a-ti § 667), beside muc-d-ti mWnc-d-ti. Skr. bhak-sa-ti 'enjoys, eats, devours', Avest. bax-sa-iti 'divides, receives for share', beside Skr. bhdj-a-ti. Skr. ddk-sa-ti 'acts so as to suit or satisfy somebody' mid. 'plunges, dips, ducks, is dexterous' (daksd-s 'dexterous') beside daias-ya-ti 'is at one's service, shows respect'. Skr. lak-sa-te 'marks', from lag- 'fasten on'. Skr. sdk-sa-nt'overpowering', from sah- 'to overpower'. Skr. injunct. ap-sa-nta 'they sought to get', beside Wp-no-ti (§ 600 p. 144), cp. the reduplicated tpsa-ti (§ 667). O.Pers. patiy-axsaiy 'I inspect', beside Skr. dk-s-i 'eye' O.C.S1. oko (gen. oces-e) 'eye', cp. the reduplicated *fyk-se-tai

Skr.

beside

tk-sa-te

(§

aS-no-ti

667)

and

the

Gr.

imper.

Avest. vax-sa-ite 'speaks', from vac- 'speak'. 'makes run' beside tac-a-iti 'runs' Skr. tak-a-ti. -iti 'learns', beside sacaye-iti 'teaches' Skr. cp. the reduplicated Skr. Siksa-te

aor.

Avest. Avest. ¿ak-no-ti

Avest. a-sixsa-nt-

oipsads. tax-sa-iti sax-sa-

'is able', (§ 667).

The preterite type belonging to this class is productive in Sanskrit, where, with roots which made final ks when s was added to them, it was used for the aorist; specially frequent when the root had f, i, or u. Examples: d-mfk-sa-t cp. above, mfk-sa-ti from marjd-sp)rk-sa-t from spark- 'touch', d-v%k-sa-t

from

195

Present Stem: Class XX — Skr. tq-sa-ti.

§§ 659,660.

'tear,

varh-

from

d-lik-sa-t

tear

out',

'lick',

lih-

from

a-dik-Sa-t

and

a-dhuk-sat

'show',

diS-

(the

d-duk-sat

latter an ad-formate of the type dduhat, cp. Gr. snsica instead of

*£ avbo, iipio see § 657 p. 192. xXao> "I break, break off' for *xXa-ow ground-form cp. M-xXao-Tcu, Skr. Sf-nti-ti 'breaks to pieces'; parallel stems xX-a- in partic. ano-xXag (Class X, § 582 p. 123), and xXa-S-. ygam 'esse' for *gf-spax-ffi-fiat) 'I shut myself in, fortify myself'. s-pXour-TO-v, pres. fiXati-ravc» 'I spring up, arise' (piua-tó-g 'shoot, bud'), orig. probably 'I get high' (used

214

Present Stem: Class XXIV — Skr. ce-ta-ti.

§§682,683.

of buds and shoots), beside {¡kwfr-po-g springing high, grown high' (I § 306 p. 242);

ij/uap-To-v Lesb. inf. afipgoryv

(for

*a^pari]v, I § 292 p. 233), pres. afiaprava) 'I miss, err', probably from d-ftag-ro- a-fifoa-vo- 'having no share' (from the root of fisg-og /noQ-o-g), cp. afiaqtlv ' ana.Qxa.veiv Hesych. (Curtius,

Verb ii 2 10 ff., and the Author, Sprachwiss. Abhandl. 160); on the present stems {¡kaoravw a/naprdva) see § 621 p. 158.

Att.

avvra

and

beside a-vvia

a-w-fxi

'I complete' (§ 639 p. 177)

partic. av-ijw-To-Q 'that cannot be completed', and similarly Att. apivo) beside agvw 'I create'.J) Of the numerous Yerbs in -n-ra) (Curtius mentions 48 of them), as rvn-xw 'I strike' axdn-Ttm 'I dig' nin-ro) 'coquo', those whose root originally ended in a velar have the first claim to a place in our group; such stems are nsnru) from s/^peq-, {¡\dnzu) beside Skr. marc-. However, it is possible to see the suffix -io- (Class XXVI) in every single one; and indeed the denominatives xaXemo) (from /aksno-g)

and aatQanxw

(from

aoTQunij) in all probability come from ^/aXsn-f.M and *daToaTt-(.(o (I § 131 p. 119). R e m a r k . I see no cogent reason for denying that ni may become 7tr (op. Kretschmer, Kuhn's Zeitsohr. x x x i 436 f.). All that can be said against deriving rvn-ra (op. rim-o-i) from *nm-ifo is that it may just as well contain the present suffix -to-. But we oannot derive from forms in -ito those whose root ended in -&(i) 'I make rot* (perf. ninvfra), beside nvo-v "pus": Lith. pu-d-inu pii-d-au 'I make rot' Lett, pa-pu-d-e 'fallow field' beside pttv-ii 'I make rotten'. /3qi-&m 'I weigh, press hard upon' (perf. (¡spgt&a), beside pptapo-g pap-v-g. § 694.

'I held', beside s-ax-s-q y/^segh-. xara-(iX-s-&er Hesych., beside O.Ir. gelid 'consumit' O.H.G. chela 'throat*. (pXey-E-d-m *I burn', beside . vf/t-s-9-o-fiat 'I pasture*, beside vsfi-o-fiui. reX-s-&a> 'I am', beside Tt'XXw. s-dx-s-9-o-v

xaTomlvsi

'I draw near', beside nsht-g nsXa-Goai. SuiM-d-fra) 'I pursue*, beside

beside

(§ frequentative strong grade from skél-du i-

t-3ap-&o-v

(§ 6 9 4

With -d-inu p. 2 2 3 ) ,

§§ 701,702.

compare Gr.

qch-ó-dvsi

beside

695 p. 224). The verbs in -d-au -d-y-ti, with meaning, often show a root syllable of the second (see § 790), as skal-dyti 'to split again and again 'I split myself skél-dinu 'I cause to be split'.

To the same dental group belongs the partic. II. pres. act. in - d a m a - s ; here the wi-suffix is the same as in vèlia-m a-s fut. vèszi-ma-s etc. (II § 72 p. 166), and had the original middle meaning. Therefore the form skél-dama-s, for example, which is now attached to the verb skelà, originally belonged to skél-du skél-d-Mu just as skél-dinu did. J O.C.S1. ja-dq 'I ride, vehor' ) beside inf. ja-ch-a-ti (§ 665 p. 198). 2 ) bqdq 'fio' may be derived from bheu-, by assuming *bhfi-a-dhS

*bhy-a-dd

(cp. L a t . -bam

for *bh%r(l-m)

or

*bhu-dho

which got a nasal in Class X V I (§ 637 Bern., p. 176); or even if we suppose that a present *bonq for *bhu-onó (Class X I Y , § 624 p. 162) was extended by -dhd or -do (cp. Lith. kattin-drinu 'to cause to be heated' derived from

*bhU-dd,

kaìt-inu

'I heat').

I. CLASSES X X Y I TO X X X I . PRESENT STEMS W I T H -io-.

§ 702. This suffix appears in the forms -io- -ie- or -i{o-ij,e-. Examples of -io- are Skr. hàr-ya-ti, Gr. /algoi for *XaP~lwi Goth, va&rk-ja, Lith. spir-iù se-ju O.C.S1. sé-jq. Of -»¿o-: Skr. mr-iyd-te Gr. tad--ia, Lat. suf-fio (for *dhu-iid) farc-id, O.Ir. b-iu (for *bhu-iiò), A.S. beò (also for *bhu-iid). We are reminded of -no-:-nfo(§ 596 p. 138); and the same double forms reappear in the noun-suffix -io- (I § 117 pp. 109 f., § 120 pp. I l l ff., I I § 63 pp. 122 fif., IH § 194 p. 74) 1) For the initial, op. Zubatf, Arohiv far slay. Phil., x m 623. 2) The derivation of Ved. yàda-mana-s yadura-s from yd- 'go* (Qrassmann, Worth., s. T. yad) is extremely doubtful.

§702.

Present Stem: Classes XXVI to XXXI — jo-presents.

229

which must be the same suffix as this of the verbs (compare such stems as Skr. pu-ya-ti 'stinks' pu-ya-m 'ill-smelling discharge, matter', § 487 pp. 41 f.). Another point in common between the two suffixes is this. In some forms of the verb-system we find a weak grade, -i-, or -1-. Examples are: -i-, Lat. 2nd sing, cap-i-sfrom cap-id, nd O.H.G. 2 sing, hev-i-s from heffu ( = Goth, haf-ja), Lith. 2nd pi. ilk-i-te from tik-iU; examples of -i-, Lat. 2nd sing. farc-t-s from farc-id, O.C.S1. 2nd sing, vel-i-si from vel-jq,2) This -?- is not found in the present system of Aryan or Greek; and it is more than chance that these very languages have discarded the weak forms of the same sort from their declension of noun stems with -io-. Details as to the Indicative Present will now be given. A r y a n and Greek as a rule have only -io- and -¿e- interchanged, as in the other thematic classes. E. g. Skr. hdr-yBL-mi hdr-ya-si hdr-ya-ti etc., like bhdr-OL-mi bhdr-a-si bhdr-a-ti;s) Gr. xctiga xaipeig yaigsi etc. like q>£QuM) beside Skr. is-ana-t, Gr. oha&-aivco beside o'Atad-uvea, O.H.G. gi-wah-annu beside Goth, af-lif-na

(Class X I Y §§ 616 if.); Greek miaam

of *7ttivo-iw) Lat. pim-io beside Lat. jung-o

beside Lat. pms-o,

jiing-iu

(Class X Y I §§ 627 if.); Skr. l-s-ya-ti

beside

i-sa-ti, Goth, vah-s-ja beside Avest. vax-sa-iti, beside Skr. tq-sa-ti Goth, -pin-sa, Skr. tr-as-ya-ti beside Skr. tr-asa-ti

(instead

Lith.

Lith. tq-s-iu Lith. tr-es-iii

Gr. TQ-i(a)-,

Gr. (fu-fxiv) 'cause to appear, make public, make known': Lat. f o r for *fa-(i)o-r, Lith. bd-ju 'I ask after, consider' O.C.S1. ba-ja 'fabulor'; still, these verbs may he derived from *bh-d-i5

1) Why, Idg. 8 in Lith.-Lett. becomes sometimes u and sometimes a (Lith. 0) is unknown.

§707.

Present Stem: Class X X Y I — Skr. hdr-ya-ti

drt-yd-te.

235

(cp. Skr. pass, bha-ya-te, not found in our texts), and their structure be the same as *tr-a-io (§ 735), compare § 495 p. 55. § 707.

Type B:

*§TY-IP-.

'die' *mr-iio- and *mp-io-: Skr. mr-iyd-te Avest. mer"-ye-iti, Lat. mor-ior (I § 120 p. 112), cp. below *bhu-iio*bhu-io-. l^der'tear, flay: Skr. dir-yd-te for *df-jfi-, Lith. dir-iii; type A, Or. SS!QM Lesb. SSQQW. y/~*$per-: Gr. ONAIPA 'I pant, struggle', Lith. spir-iii 'I strike with my foot, kick'. I/'sqel-: Gr. ay.dXho 'I scrape, hack' for , Lith. skilii (for *skil-iu) 'I strike a light, kindle'. I / ' m e n - 'think of, meditate': Gr. /.laivof-iai 'I am wild, enravished, mad', O.Ir. do muiniur 'I think or believe' (for *man-io*m%-io-), O.C.S1. min-jq 'I think'; to either {A) or (B) may belong Skr. mdnya-U 'thinks' Avest. l 8t sing, man-ya O.Pers. 2nd sing. conj. Vmer-

maniydhy

( I § 125 p. 116).

instead of *ghan-ya-tB

V ghen-:

S k r . han-yd-ts'is

struck'

(I § 454 Rem. p. 3 3 5 ) , O.C.S1.

zin-jq

'I cut off, reap'; of type A from this root we have Gr. &EIVW. Vgem-

'go':

S k r . -gam-yd-te,

Gr. pairo>,

L a t . ven-iO

(I § 204

p. 170, § 208 p. 174); veniS might also if we wished be classed as an example of type A. Vbheu'become, be' *bhy,-ij,o- and *bhu-io(so above we had *mr-iio- and *mf-io-): Gr. * and the like, found in the text of Homer and Hesiod (Curtius, Verb I s 304 f.), can be explained * n h f - i w (Lith. pldu-ju) and so forth. But there is practioally no objection to regarding them, as many scholars do, as oorrnptions for Aeolio forms of Class II, nlevw =

*7lXcf-U>. § 713.

T y p e B.

xahw>

oiou'po), ffxa'AAw, fialvofxou, fiulvw,

&via>, rt(o, , *pa£(o 'I do', qipd^w, XiOom/xev, xaao6u), ntbw,

Sim, Sni,

compare

xuuovzeg 'the dead' (then txavov got v from the present):

Skr.

§713.

Present Stem: Class XXVI — Skr.

hdr-ya-ti df$-yd-tg.

247

¿dm-ya-ti 'becomes still, is extinguisht' for *Hf%i-ie-ti (xawo differently explained by Kretschmer, Kuhn's Zeitschr. xxxi 428, 432; Fick, I 4 43). nrvpaj 'I make shy', cp. Lat. con-ster-nfl-re, avpw 'I drag' cp. aaipw 'I sweep' (with fj), axvXXw 'I tear to pieces, towse, worry' cp. Lith. skelii (*skel-iii) 'I split'; the v of this form needs explanation. Sim 'I beseech, fly, fear* doubtless for *di-tw: Skr. di-ya-ti 'flies'; of type A, Lett. dSi-ju 'I dance' (inf. dl-t); the forms Slsrt Sit rat and such like were associated with "URS "STCU, and this caused the formation of tv-dUoav Sitfiai and others by analogy of the parts of "r^u. (pgaoOM 'I enclose' for *(pQax-Lg-iO, akin to Skr. khdnj-a-ti 'limps', viu» 'I wash' groundform *nig-i6: Skr. pass. nij-ya-tS. ariCio 'I prick, pierce' for *, beside O.Icel. hnlt 'I knock against, hurt with a knock' Class II A. opv'oaw 'I dig' for *oova-t.w: Lith. rauk-iti 'I wrinkle', (A). dno-fivrtM 'I blow my nose' for *fivx-(.w: Skr. pass. muc-yd-tS 'is set free'; Lith. mauk-iil 'I scratch slightly, touch softly', type A. The theory that otw 'I swell', for does not belong to type A, is doubtful, in spite of an appeal to Lith. fo'd-ziu 'I smell'; it is also uncertain to which section belongs oaao/Lim 'I see', for *oq-io- (cp. I § 319 p. 258). It is risky to connect om/i«i with Goth, ah-ja 'I believe, surmise'. Forms with Idg. -iio-. iS-iw 'I sweat' is usually connected directly with Skr. svid-ya-ti O.H.G. swizzu. If that is so, ¿¿-tdlea is due to the analogy of denominatives in -i-io- and i&tw (Aristoph.) is a reformate like xovtca (§ 775). «riWco beside eofrw 'esse' for *ed + dho, cp. § 694 p. 223, § 765. A form * — Idg. *bhu-ij,o follows from (pT-rv 'sprout, shoot, scion' (pt-iv-s 'begetter', which must have been derived from it

248

Present Stem: Class XXVI — Skr. Mr-ya-ti drl-yd-te.

§§ 714,715.

as though the verbal stem were (pi- (§ 707 p. 235); a similar origin must be supposed for Lat. fT-tu-m cupt-tu-s and others (§§ 715 ff.). § 714. The identity of ending in nrpa^m eacpata (a is formed like xpijeouat: was this derived from to /prjog, or was it a formation like Skr. causal pyO-y-dya-ti? (cp. § 801). *xv-rj-¡, from xXa-rf- 'break off'. But it is doubtful whether nkrjaa 'I will fill' is *nXr/&-Ga> (cp. TfXtj-9(0 ninXyorat nXqorio-g) or nXij-ow (cp. nXrjro nsnXtjvtai), whether elfvaofiai 'I will come' is *cXev9-aopat (cp. ijXv-do-v fktvoTio-v) or ilsv-ao/uai (cp. ¿XrjXv-rs npoa-rjXvTO-e), whether Tivam 'I will make rot' is vfr-oco (cp. -nv-frm) or nv-acj (cp. Lith. puv-il). There is the same doubt in Lith. futures like plausiu from plau-d-ziu 'I wash' (Idg. *pleyrd-), sprdusiu from sprau-d-eiu 'I subdue' (Idg. *spreu-d-) g'esiu from g'e-du 'I sing'. As we know not in what period of Lithuanian these verbal classes arose, we are not compelled to assume that plausiu, say, comes from a supposed form *plautsid. The fact may be that ptau-siu is really future to plau-ju; and then, on the analogy of geidziii gelsiu gelsti, and others of this kind, plausiu was involuntarily associated with plaudMu plausti as its future. Compare §§ 688 ff.

§§ 751,752.

Present Stem: Class XXX — Skr. tq-s-yà-tè.

273

Forms with other present-signs sometimes make a «¿o-future in different languages; as Skr. indhisyati Gr. xAoiy?« Lith. junksiu.

See below, §§ 752 ff.

§ 751. The indicative with -sio- seems to have had in proethnic speech a participle attached, but no more (Skr. dd-sya-nt-,

Gr. Sw-awv,

Lith. dial, d&'sius for *d&siqs,

O.C.S1.

byiqsteje). In Sanskrit grew up a conj. with Ar. -H-, and an augmented preterite; and Greek developed an opt. with -i-. See §§ 753, 759. § 762. A r y a n , -sio- and -asjo-, but the latter is only to be found in Sanskrit (-¿s«/ir-). There is no example of a future in Old Persian; this is probably due to chance. In Sanskrit and Avestic this future was a living and productive type. It is used, true enough, less often in Yedic than later; but then in Yedic, injunctive and conjunctive forms were used with future meaning. To the exx. cited in §§ 748—750 may be added: Skr. vak$ya-ti Avest. vaxsye-ite beside Avest. vdk-ti 'speaks'; Skr. Skr. janisyd-ti Avest. partic. zqhya-mnabeside Skr. jdn-a-ti 'begets 5 ; Skr. bhantsya-ti bandhi$ya-ti beside badh-nd-ti 'binds'; rocisya-te beside roc-a-te 'shines'.

In Sanskrit we meet with specimens of this future made from presents of any kind (cp. § 750). mCLrksya-te (beside mraksya-te) from mars-ti Class I and marja-ti Class II 'wipes' (cp. § 494 p. 55, § 514 p. 81).

sidisya-ti

(beside

satsya-ti)

from stda-ti Class IY 'sits' (§ 550 p. 106). dadisya-te (beside da-sya-ti) from dd-dd-ti Class Y da-d-a-ti Class VI 'gives'; jahisya-ti

(beside

ha-sya-ti)

from

jd-ha-ti

ja-h-a-ti

'leaves,

deserts', jagarisyd-ti from ja-gar-ti Class Y 'wakes' (§ 560 pp. 109 f.). indhi$ya-ti from inddhe Class XY 'burns' V a%dh-. ainuvisya-ti from ai-n6-ti Class XYII 'attains'; jinvisya-ti from ji-no-ti Class XVII ji-nva-ti Class XVIII sets in motion, propels'. titiksi&ya-te from the desid. ti-tik-§a-te Class XXI from tij- 'to be sharp'. Jehydyisya-te from pass, khy-a-ya-te Class XXVIII 'is seen'. ftrugmann', Elements. IV. 18

274

Present Stem: Class X X X — Skr. iq-s-yd-te.

§§ 753—755.

Rather commoner in the later language is the future of denominatives in -yd-ti, Class XXXI,fas gopoLyisyd-ti from gopd-yd-ti 'guards' (gopa-s 'guardian'); and of present stems in -aya-ti (Causatives), Class X X X I I , such as vyayisya-te from vy-dya-ti 'enwraps, covers', dhCLrayisyd-ti from dhdr-aya-ti 'holds'. § 753. Sanskrit has an augmented preterite from the future stem, meaning on the point of ; as abharisya-t 'he was just going to take away, wished to take'. But this form usually stands as a conditional; and so Conditional it is called. There are a few scattered instances (in the Maha-Bharata) of Injunctive forms, implying wish; as 2nd pi. mid. bhavisya-dhvam. Similarly there are scattered Conjunctives ; as Yed. 2nd sing. karisyd-s. § 754. G r e e k . It is not quite certain that the Greek tf-future has any immediate connexion with the Aryan and BaltoSlavonic sio-type, as we have seen already (§ 747 p. 269). With -ao- (§ 748) and -so- -no- (§ 749), we find a third suffix, -oso-. § 755. (I) -ao-, a productive suffix in Ionic-Attic and elsewhere. Examples in §§ 748 and 750. There is an apparent anomaly in keeping a after sonants in the future arij-aut, as in the aorist sortjaa. This is most simply explained as being due to the analogy of ón%w schifa etc., consonantal stems. Compare I § 564 p. 421. Stems in Liquid or Nasal generally conform to Type I I (§ 757); but roots in p have -aco as well in the language of Homer and poets of the epic school: qvfrlgtiw (pres. (p&sipw 'I destroy') beside (pfrspéw -w. R e m a r k . "Why is it that beside a fut. iptegato there is no fut. *, as might be expected from finding ¡xt^a side by side with ì'xt^na ? This is explained without difficulty if we suppose txeija to be analogical, and due to Tiiva evince and the like (I § 563 Rem. 2 p. 419) ; for there were no such futures as *xre(vu> for *xTtrou>. "Waekernagel's view of ¿«.pa (Kuhn's Zeitschr. x x i x 127 ff.) is not oonvinoing, to my mind.

§756.

P r e s e n t S t e m : Class X X X —iSkr.

tq,-s-yd-te.

275

W a o k e r n a g e l would anyhow h a v e to m e e t the question whether, if tpftegau really comes from *tp»egoiu>, it must not h a v e kept - g o - under all circumstances, w h e r e v e r the accent l a y ; cp. viao/iai for *vivaiofiai contrasted with i'xruva for *ixTtvaa (The Author, Gr. Gr. a p. 61).

§ 756. -oo- forms futures from all sorts and kinds of stems, present, aorist, and perfect. Often there are parallel tf-aorists. (1) Horn, itdw-fity (beside Jw'-tfw) from ¿¿-¿w-/ui 'I give', Class III. t)i(W|ft> from dt-Jatfxw, Class X X I I I (aor. idi like òaxpé-oto etc.). To hastily reject this element in the matter would be all the more foolish, becsause it is clear as day that Primitive verbs have had influence over Denominatives in the futures òvo/iavèm -ài beside ugnai/», and xskscu instead of reXsoóea (§ 757 p. 277). As regards verbis in -lio and -wo, we have also to consider that the contained niouns often had -l-g and -v-s (cp. ìo/vm from ioxv-g); this may Jhave had something to do with it, and analogy may have finiished the work. How far this influence acted must remain unseittled while we have no exact statistics of -ua -DID and -Iw -vw. R e m a r k . ytltato ¿Jpaito oiyuiaj are to be kept distinct from fi&iiifiiiovTft; etc. because they come from -aia-¡,u>. ytlàio from yeXuio- (nom. yiXw?}), the strong form of yeXaa-, whenoe ytXa'a (§ 768 p. 282). ISgàm from iSfw;. $[ymw from a word parallel to Lat. rigor. These verbs in -wa-tui aire in all probability upgrowths of the separate period, when the languages were developing singly ; in this they resemble the Latin group exemplified by fulgur-is from fulgur (O.Lat. fulgus), and stand in contrast tto the -really old forms Or. reXeim rtXiw for *Titea-to> (§ 768 p. 282). The origin of the ending in SiViij Ttnyij for -/¡a is not • clear; cp. Horn. Siipiwr, Ion. (Archil.) Sitpfar, Pind. Siipìj. Compare "Waokerrnagel, Philol. Anz. 1887, p. 238; W. Schulze, Kuhn's Zeitschr. xxrx 269 f..

On the non-thematic present inflexion -G-fii (-a.if.tt) -rj-fit

-m-fii following Class X (instead of -am -sai -out) in A^eolic and Arcadian, see § 582 p. 123, § 589 p. 131. The type -(¡-/.a in our ^-denominatives came from the pre-Greek sitage; and in Greek itself its analogy produced -r^-fii and -u-fu. 1) Cp. yfvio instead of *yiw following ytv-nm and the rest ; the Aiuthor, Or. Or. 1 p. 31. Lithuanian : cp. pres. dèmi 3 r d sing, désti instead off dèmi disti following dé-siu dl-ti etc. (§ 546 p. 104).

§7776.

Present Stem: Class XXXI — Skr. deva-ya-ti.

295

§ 776. Before turning to trace the way by which the variious denominative endings became general types in Greek, we would quote some words of Siitterlin's. He says, "In the evejry-day language of inscriptions, analogy did not run riot as it cQid amongst the poets and orators, who were often forced to ¡adopt new words and terms, and depended partly on these for effect". (Zur Gesch. der Verba denom. im Altgr., l 5). (1) The type -aw, which could be made from o-nouns even in jpre-Greek times (§ 769 pp. 284 f.), did not spread so far as it nnight in forming factitives, because it was met by a countercurirent, the -oco class (4). Thus vsow 'I renew' may have cauised *vef&-iu> = Lat. novo O.H.G. niuwGm to drop out of use (cpj. p. 284 footnote). But in other directions -cuo was fertile; it served to denote disease or diseased appetite, the production of sounds, mechanical operations, and the like. Examples: ktnfgaco 'I have an eruption on the skin' (from Xsjiga eruption') andl similar words give rise to vSsgdio 'I have dropsy' from vdt{Qo-g 'dropsy'; o xi-vv-fiai; O.H.G. zunt(i)u 'I kindle' with Goth, tandja 'I kindle' beside Mid.H.G. zinden (strong verb) 'to burn, glow". Uncertain: Gr. ql-so) 'I overflow' beside ix-yXamo Lat. fl-e-re;

1) From this yeiwhich had become a root again before the end of proethnic Idg., a present was again formed by means of -ejo-; Skr. vy-dya-ti 'winds up, wraps up, covers, hides', Lat. vi-eo. 21*

324

Present Stem: Class XXXII — Skr. vSd-dya-ti.

§ 791'.

Goth, ga-nsja 'I cause' from \^nes-, originally 'I make to come forward'; O.C.S1. brija (brija) bri-ti 'to shear, shave' beside 8kr. bhur-ij- Gr. (pdo-o-c, and others. § 791. In all branches of our group, the verbal class now being discussed has two distinct meanings, botli of which must be regarded as holding for the original language. Each of them serves to contrast a verb with a simple verb from the same stem. First there is the Causal sense; the subject of the