regular formation of the superlative, -yās- being subsequently § 106. shortened to -is-, but -is-ta- changed into -is-tha- (§ 59, 1); thus e.g. máh-ištha-, yáv-istha-, lágh-is-tha-, gár-iš-tha-, kšód-is-tha-, etc., sthéstha-, sphéštha-, gyéštha- are=*sthā-is-tha-, *sphā-iš-ṭha-, *ģyā-iš-tha-; présṭha- either stands for pre-istha-, in which case the i of -is- would have disappeared in the e, or, as I think more likely, an earlier √pra is underlying, and it must be divided *pra-is-ta- (cf. the comparative, § 104, with this superlative); in bhúyišṭha-, beside the comparative bhú-yãs-, -yis- has arisen for -is- by an unusual splitting-up of i to yi.

Greek. Suffix -To-=Sk. -ta- is common in ordinal numbers, thus πρώ-το- (first), τρί-το- (third), τέταρ-το- (fourth), πέμπ-το(fifth), ex-TO- (sixth), eva-To- (ninth), déκa-To- (tenth), eikoσ-tó(twentieth), etc.

-TO-=-ta-, when added to -10-=-yans-, forms superlatives to comparatives in -tov-=origl. -yans-, e.g. Káк-1σ-TO-, ẻλáx-10-70-, ἥδισ-το-, ἔχθ-ισ-το-, μέγισ-το-, πλε-ϊσ-το-, etc. (cf. § 104).

The reduplicated form of the suffix, -ta-ta- = Gk. -Tа-to-, appears as the regular superlative formation beside the comparatives in -ta-ra-; e.g. κουφό-τα-το-, σοφώ-τα-το-, γλυκύ-τα-το-, χαριέσ-τα-το-, φίλ-τα-το-, φιλαί-τα-το-, εὐδαιμον-έσ-τα-το-, λario-Tα-TO-. These superlatives were formed, as the adduced examples show, corresponding to the parallel comparative forms in -TEρo-, which may be compared (§ 105). -τερο-,

Latin. Suffix -to-, -tu-, Sk. and original -ta-, rarely serves as superlative-suffix in Latin, where -mo-=Sk. -ma-, and -ti-mo-, -si-mo-, f.f. -ta-ma-, is preferred; e.g. quar-to- (fourth); quo-to(which in order or number), stem and √quo-, original ka-. After-is--yans- is found not -to-, but only the representative of original -ta-ma- (q. v.).

2. Suffix -ma- and ma-ta- here and there in Gk.

I have not before me any perfectly certain example of suffix -ma- used to express the superlative in the original IndoEuropean; however, as -ma- appears in three divisions of the

$107.

§ 107. speech-stem in this function, it must be presupposed as existent in the original-language. Perhaps sapta-ma- (seventh) and akta-ma- (eighth) must be ascribed to the fundamental-language, because in these numbers this method of forming the ordinal is almost universal.

Sanskrit. Suffix -ma- forms the superlative in ava-má(undermost, next, last), from áva- (as prep. 'from'), a pronominal-stem (demonstr.); madhya-má-, mádhya- (mid-); para-má(furthest, last, best), pára- (removed, excellent); ādi-má- (first), ādi (beginning); sapta-má- (seventh), saptán- (seven); ašṭa-má(eighth), ašṭa- (eight); nava-má- (ninth), návan- (nine); daça-má- (tenth), dáçan- (ten).

Greek. The superlative suffix -μo-, f.f. -ma-, is not frequent, it forms eßdo-po- (seventh) alone, from πтά, with a remarkable softening of TT to BS in the stem of the word.

In éẞdó-μa-To- (seventh) we find suffix -ma-ta-, as in Keltic, which is seen unmistakably in πú-μа-то- (last) also.

Latin. Suffix-mo-, f.f. -ma-, is a favourite means of forming superlatives, e.g. sum-mo- from *sup-mo- (highest), sup-, cf. sup-er, comparative; infi-mo-, cf. comparative infe-ro-; mini-mo-, cf. min-ōr-; moreover the ordinals pri-mo-, septi-mo-, deci-mo-.

In plurimo- (most), earlier plusimo-, plourumo-, ploirumo-, pli-si-mo-, a f.f. *pra-yans-ma- seems to be underlying; between s and m the auxiliary vowel u occurs (cf. s-u-m, § 43), later i (§ 43); thus by the usual shortening of -yans- to -is-, arose *plo-is-u-mo-, i.e. ploirumo-, and by coalescence of oi to ei, i (as in dat. abl. pl. of the o-stem, e.g. nouïs, noueis from *nouois), *plisimo-; in plourumo-, later plurimo-, y may have been lost (as in minus for *min-yus), so that this form points back to a form *plo-yus-u-mo-, like plūs, plous, to *plo-yus.

§ 108. 3. Suffix -ta-ma- occurs in Sk., Zend, Gk., Lat. and Gothic, and dates therefore from the common original-language.

Indo-European original-language. Although the suffix -tama- was in existence, we can hardly point to any

word-stems provided with this suffix; a form nava-tama- (per- § 108. haps beside nav-yans-ta-) may be merely conjectured.

Sanskrit. -ta-ma- is the regular superlative formation beside the comparatives in -tara-, thus e.g. púnya-tama-, ka-tamá(one of many, interrog.), ya-tamá- (which of many, rel.) çúki-tama-, agnimát-tama-, vīçati-tamá- (twentieth) from vìçáti-, etc. Suffix -tama- occurs also after comparatives in -yans-, and superlatives in -istha-, e.g. gyéstha-tama-.

Greek. Wanting.

Latin. The suffix original -tama- occurs somewhat seldom added directly to the root or stem of the adj.; but it is the regular means of forming superlatives, in which case it is added to the comparative in -yans-; f.f. -ta-ma-, i.e. Lat. -tu-mo-, -ti-mo-, after gutturals -si-mo- (§ 77, 1, d) is added to the root in maximo- (greatest), i.e. *mag-timo-, cf. mag-is, mag-nus; op-timo-, op-tumo- (best); ul-timo- (last); in-timo- (inmost), etc. Suffix -timo-, or more probably -simo- (cf. mac-simo-), is directly added to adj.-stems only when they end in r or l, stems ending in a vowel lose that vowel; thus ueter-rimo- for *ueter-simo-, from *ueter-timo-, stem ueter- (old; cf. torreo for *torseo and the like), pulcher-rimo-, stem pulchero- (beautiful); facil-limo- for *facil-simo- from *facil-timo-, unless indeed these forms be for *ueter-is-timo-, *facil-is-timo-, whence *ueterstimo-, *facilstimo-, *uetersimo-, *facilsimo- (cf. uellem from *uel-sem).

From the f.f. *-is-tama-, the combination of the comparative suffix -is--yans- with the -tama- of the superlative, arose next -is-tumo-, -is-timo-, retained in the archaic soll-is-timo-; from -is-tumo-, -is-timo-, arose -issumo-, issimo-, by regular assimilation, e.g. doct-is-simo-, etc. Cf. mag-is-ter, min-is-ter, which show the combination of the comparative suffix -yans-+-tara(cf. § 105), and therefore stand parallel to superlative -yans-+ -tama-, and the corresponding Sk. superlative formations. Cf. the suffix -timo- in other functions, mari-timo- (maritime), stem mari- (neut. mare, sea); fini-timo- (neighbouring), fini- (finis masc. bound); legi-timo- (lawful), stem lēg-, lēgi- (lex, fem. law).

$109.

4. Stems of Numerals.

STEMS OF FUNDAMENTAL NUMBERS.

Simple numbers 1-10.

1. Indo-European original-language.

The stem for the first numeral cannot be determined with absolute certainty, because the several Indo-Eur. languages differ greatly in the way they express the number one. Various stems in Indo-Eur. indicate the notion of the No. 1, but yet they are all alike formed from vi; ai-na- has most in its favour, because it serves to indicate the number one in the two European divisions of the Indo-European, and is likewise found in the Aryan in another function.

Sanskrit. éka-, probably a stem-formation by means of suffix -ka- from pronominal suffix i, or, what is not very different, a compound of ai from i with pronominal √ka.

Greek. Nom. sg. m. eîs, i.e. *év-s, ntr. ěv, gen. év-ós, fem. pía (cf. Leo Meyer, Kuhn's Zeitschr. v. 161 sqq., viii. 129 sqq., 161 sqq.; ́id, vgl. Gramm. der griech. u. lat. Spr. ii. 417 sqq.). Stem ev-, f.f. san-, is considered as standing for sam-, and this sam (cf. Lat. sim-plex, sem-el, sin-guli), after losing a final a, as corresponding to Sk. samá- (similar, like; a superlative of demonstr.-stem sa-). This view is especially supported by fem. uía, which probably stands for **μía, i.e. sm-ya-sam-yā- (a ya-stem, fem. only, as frequently). It cannot well be doubted that ev- contains the pronominal-root sa-; however, we hold that the final n is a later formation in Greek, cf. stem Tí-v-= original ki- (pron. interrog.), because a change of suffix -ma- to n is unexampled. So we conjecture for masc. neut. a stem sa-n-, developed from sa-, but recognize in fem. a f.f. sa-myā, thus varying from that of masc. and neut., i.e. a superlativestem in -ma-, fem. -myā, from same √sɑ.

Latin. Early Latin oi-no-, whence u-no-, f.f. ai-na-, is, like

Sk. pronominal-stem e-na- (this), a stem in -na- from demonstra- § 109. tivo-pronominal √i.

2. Indo-European original-language dua- or dva-; Sanskrit dva-; Greek Svo-=dva-; Latin duo--Svo- (duō-bus like ambō-bus is probably caused by the nom. dual duō, ambō, f.f. dvā, ambhā).

3. Indo-European original-language tri-; -i- is a suffix and tar, tra, the root (cf. the ordinal); tr-i- is therefore probably shortened from *tar-i-, or perhaps from *tra-i- by loss of the a; the assumption that tri- is an archaic weakening from tra- is, I think, less likely; Sanskrit tri-, the fem. having the stem -ti-sar-, in which Bopp conjectures a reduplication, and which he imagines to have arisen from *ti-tar-; Greek Tp-; Latin tri-.

τρι-;

4. Indo-European original-language katvar-; Sanskrit katvár-, shortened katúr-, fem. katasár- (clearly formed after the analogy of 3, according to Bopp compounded with it, in which case ka= 'unum'); Greek TÉTTAP-, TÉσσαP-, for *TETFαρ-, KETƑаρ-, Dôr. TéтOρ- for *TeτFoр-, with 7=original k (§ 62, 1), Boiôt. πéτтаρ-, Hom. and Aiol. πíσup-, f.f. katur-, π=к (§ 62, 1), and σ before v=T, as in σú for Tú; Latin quatuor-, the best authenticated spelling quattuor is unsupported etymologically (inscriptions have also quattor, quator).

5. Indo-European original-language kankan-, an obviously reduplicated form; Sanskrit páńkan-, with p=k (§ 52, 1); Greek Tévтe, Aiol. πéμже, π and т=к (§ 62, 1); Latin quinque.

6.

Indo-European original-language. By combining the Zend khšvas on the one side with the Greek, Latin and Gothic form with final ks on the other, the resulting f.f. would be ksvaks for Indo-European (Ebel, Beitr. iii. 270; Zeitschr. xiv. 259 sqq.), which likewise seems to be reduplicated, perhaps original *ksva-ksva-; Sanskrit šaš, probably immediately from *kšakš for *ksa-ks (§ 55, 2); Greek and Latin with dissimila

t