143 
Some New Theories of the Greek Ka-Perfect. 
/ 
a ground for this explanation Thiersch and Ahrens cited ovxen and 
firfKsri. But the x here belonged originally to the ovx in all probability, 
so that ovxsri is for ovx-en (not for an earlier ^ov-eri), while prfxeri is 
formed after the analogy of ovxen. 
It is scarcely necessary to add that a theory involving such an extraor¬ 
dinary phonetic phenomenon as the bodily insertion of x, a change not 
only not elsewhere exemplified in Greek, but also without analogy in other 
languages, is unworthy of further consideration to-day, when the inviola¬ 
bility of phonetic laws is so insisted upon by even the less rigid inves¬ 
tigators. 
Curtius, in his earlier work, the Tempora und Modi, followed Thiersch 
and Ahrens, but in his Verbum der Griechischen Sprache , he propounds the 
following explanation of the -hoc- perfect. He designates the -hoc as a 
“stammbildendes Element.'’'’ He further assumes that nominal stems were 
first formed with this suffix; and from these nominal stems verb-stems, just 
as the verb-stems in -va- 9 - rv-ro-,-avo~, -to-, -6xo~, are, by Curtius, in their 
origin, assumed to be nominal stems. Curtius further says, ‘‘ Assuming that 
there was a nominal stem *fia-xa (Ion. *(3rj-xa) then from this we might 
have a reduplicated (3e-f3a-xa (Ion. (3s-ftr)-xa ), and such reduplicated 
stems might, in the period when the verbal forms were still undetermined, 
establish themselves here and there in the perfect by the side of the shorter 
forms in -a” 
To this theory of Curtius Brugmann presents the following objections: 
1. Where are these nominal stems in -xa? The language does not seem 
to have had them. We should expect them to exist in some abundance; 
but we find only a very few. We may cite ^ijxr/ and 6coxby, but from 
%r} n V we could hardly directly explain the form re-Seixa as the vocalism 
of the two words is different, one having rj the other ei. Nor can we satis¬ 
factorily explain the aorist eSrjxa from ^rj K V. For Srjxr] means £ chest, 
vessel,’ and the testimony of the Sanskrit dha-ka, which has the same 
meaning, seems to indicate that in the Ursprache the word had this spe¬ 
cialized sense of the root dhe -. It is difficult to comprehend how a 
denominative aorist or perfect in the sense of ‘ put’ should be formed from 
a’noun meaning ‘ holder.’ 
2. Curtius speaks of a form /3e0jxa establishing itself by the side of a 
form * /3s(j?ja at a time when the verbal inflections were still undeter¬ 
mined. Brugmann with propriety inquires when Curtius conceived this 
to have been. Was it subsequent to the period of the Ursprache? If so, 
he should remember that the verbal inflections were then no longer un¬ 
determined. The substantial unity of the verbal inflection of the separate 
languages shows this. Or does Curtius mean the period of the Ursprache? 
If so he should consider that we can not pretend to explain a formation 
which has confessedly developed upon specific Greek ground, by referring 
to the Ursprache, or any conditions which may have prevailed in it. 
To these two leading objections of Brugmann might be added another, 
