144 IV alter Petersen, 



' provided with,' and this is contrary to the use of -isko-, which 

 does not seem to have that meaning.^ 



6. Much more probable from every point of view than any 

 derivation so far mentioned is the one first advanced by Leo 

 Meyer, KZ. 6. 381 f., and adopted e. g. by Brugmann, Gr. 2. i-. 

 503. According to this theory -isko- is a conglutinate the last part 

 of which is the common I. E. -ko- suffix, while the first is the 

 comparative suffix, the weak grade to the strong -ies- -ios-,^ which 

 is also found in the superlative suffix Skt. -ista- Gr. -Kty-o- Goth, 

 -ista-. As to Vondrak's contention that for the Slavic -isko- the 

 long i beside -tsko- is an objection, Belie, Arch. f. si. Ph. 23. 179, 

 has shown how the i could be derived from contraction of an l 

 of the stem and the i. of the suffix -tsko-, and could then be 

 extended by analogy. Vondrak's second objection that the mean- 

 ing of the Slavic suffix is against connection with Gr. -ktxo- is also 

 of course not meant to affect the Gr. suffix itself nor I. E. -isko-, 

 nor is it serious even for Slav, -isko- when Brugmann's explanation 

 of the semantic development is more closely scrutinized. 



7. According to the latter (1. c.) we are to start from the use 

 of the comparative degree in adverbs in the sense ' coming rel- 

 atively close to the condition designated by the adjective,' e. g. 

 Germanic *al{)iz ' approximately old.' Adding to this the common 

 I. E. adjectival suffix -ko-,^ no change of meaning except return to 

 adjectival function would result, and O. H. G. altisc would have also 

 been originally ' rather old.' This being the case, -isko- would easily 

 be felt as a unity and abstracted as a single suffix. After this was 

 once done, it was easy to form derivatives from substantives with 

 the same suffix, e. g. Goth, mannisks : manna. In Greek, on the 

 other hand, the derivative adjectives were themselves substantivized, 

 VEavi<r/.o$ originally formed from *v£av6? ' young,' meaning ' rather 

 young,' was referred to vsavLa? after first having been substantiv- 

 ized as ' he who is rather young,' and so became a diminutive to it. 



^ Vondrak, however, would derive from it augmentatives and words 

 like biciSte ' Peitschenstiel ' : bic ' Peitsche.' 



^ A support for the contention that the -is- of -isko- is derived from the 

 comparative -i^es- -ios- is found in the habits of sandhi of Gr. -taxo-. When 

 e. g. heavy suffixes like -ev- in afj.qoQEvg disappear completely in the formation 

 of diminutives hke a^q^oqiexog (cf. § 12), the most rational explanation 

 seems to be that -taxo- has retained to a Umited extent the habits of the 

 comparative suffix, which, being originally a primary suffix, was added to 

 the root instead of the nominal stem. Cf. Brugmann, Gr. 2. i^. 656. 



3 Cf. Brugmann, Gr. 2. i^. 482 f. 



