340 Table of Greek Correlatives, with Explanations. 



Iva, primarily i, q. Gr. ort or Lat. quod, and correctly derived by Pas- 

 sow from the old personal pronoun ?s, 2, — 8. Here i is lost in e {==ai;) 

 comp. Sansc. i-tas, thence ; Lat. i-nde, thence. — 9. Here i is lost in e 

 (=ai;) comp. Sansc. i-ha, here; Zend, i-dha, here; Lat. i-bi, there. — 

 10. Here i is lost in e {=ai ;) comp. Ef&a. — H. The form i], either, 

 or, (as if in this way, in that way,) probably for f; with subscript Iota, 

 an obsolete dative from the pronominal root I'. The form si if, (as if 

 in this way, under these circv/mstances,) probably with form of the 

 dative, from the same pronominal root. The vowel i in both forms 

 is lost in e {=ai.) For examples of the derivation of the conditional 

 conjunction from pronouns ; comp. Sansc. yadi, if, from yas, who; 

 it, if, neuter of the pronoun i ; Goth, ith, if, for ita, neuter of the 

 pronoun is; Germ, wenn, if, from wer, who. — 12. The forms £w, 

 yet, iSe and ridi, and, (comp. Sansc. i-ti, also,) we place here, although 

 with some hesitation. — 13. No form extant. — 14. No form extant. — 

 15. The form ia&xig from i'(Tog needs no illustration. — It ought here to 

 be observed that sv&ev, svQ-u, and probably the other forms in the 8th, 

 9th, and 10th series, are capable of the relative signification ; comp. 

 Germ, der and Eng. that, which are both demonstratives and relatives, 

 the ground of which is worthy of further investigation. 



Rem. 1. The lengthened forms, as ivjevd^Ev, (Ion. Iv&svTei',) ivrav- 

 ■9-a, (Ion. ivd-avia,) ivvavd-oT, (Ion. ipS-avjol,) are more emphatic than 

 the others. 



Rem. 2. Some of the forms in this column are strengthened or ren- 

 dered emphatic by the addition of Si; as lv&h8s, ip&dde. Add also 

 ivxev&evl, er&udt, svd^av&l. 



The II. column is not quite complete. It is, however, regular, beau- 

 tiful, and worthy of admiration. It consists of proper demon stratives, 

 an important and interesting class of words. They have sometimes 

 been called redditives, as answering or responding to the interroga- 

 tives. But this name is inappropriate, as it would seem to imply an 

 origin posterior to that of the interrogatives. The correlative ele- 

 ment is expressed throughout by t, a sound which is justly believed 

 to have a natural appropriateness to perform this oilice. Hence it is 

 found with remarkable uniformity in different families of the Indo- 

 European stock of languages, as we shall hereafter see. The irregu- 

 larity in the masculine and feminine forms of the pure pronoun is only 

 apparent, for the form of the oblique cases, rov, ttjc, rov ; xol, ttj, tco, 

 etc. plainly show that the nominative was originally r6, ttJ, r6 — It 

 ought here to be observed in respect to the pure pronoun, that it is 

 sometimes used as a relative, (comp. the remark on ei'T9^£j' and svS-a 

 above ;) and, in the genitive and dative singular, also as an interroga- 

 tive and indefinite ; and that although its primary sense is that of a 



