Table of Greek Correlatives, with Explanations. 339 



Each word in this table, it will readily be seen, consists of two 

 parts ^ viz. the correlative element, wliich it has in common with 

 other words in the same column, and the modifying element, which 

 it has in common with other words in the same row or series. The 

 tabular mode of exhibition which we have adopted, suggests a natural 

 mode of discussing these words ; which is, first, to treat of the cor- 

 relative element in each column, and then to treat of the modifying 

 element in each series. These elements, moreover, are to be consid- 

 ered severally, both as to their logical import, and as to their gram- 

 matical form. By this regular but thorough process, we hope to con- 

 dense much into a little space, and to arrive at important general 

 principles. 



The I. column is incomplete, but very important and interesting, as 

 exhibiting venerable reliques of an ancient and nearly obsolete pro- 

 noun. It contains determinatives, by which we intend weak or un- 

 emphatic demonstratives, like Eng. he, she, it, or Lat. is, ea, id. As 

 the correlative import and form is often obscured, it may be useful to 

 consider each member of this column separately. — 1. The existence 

 of a very rare pronominal root i or i', in the sense of Eng. he, she, it, 

 or Lat. is, ea, id, is now admitted by the best Grecists, as Buttmann, 

 (Ausf. Sprachlehre, B. II. p. 413.) and Passow, (Ilandworterb. s. v.) 

 Its existence is also confirmed by the analogy of the Sanscrit and 

 other Indo-European dialects, as we shall hereafter see. But it is un- 

 certain, whether the vowel of this root should have the rough breath- 

 ing, ?, which accords best with the pronoun of the third person, ov, ol, 

 £, or the smooth breathing, t', which accords best with some of the de- 

 rivatives, as 'ixTog, log, etc. and with its form in cognate languages. The 

 form of the pure pronoun was probably ?>, t, or t'c, t', like rig, rl,* — 2. 

 The form hegog, other, (as if more or other than he,) without doubt 

 belongs here, although the i is lost in e [=ai ;) comp. Sansc.^ i-taras, 

 another, Lat. i-terum, again. — 3. 'iaog, like, equal, (as if so great,) 

 which we sometimes find as a correlative to &g, see Acts II : 17. — 4. 

 log, one, (as li of this kind, such,) a form occurring in Homer for the 

 numeral slg. The numeral for one, in most of the Indo-European 

 languages, is of pronominal origin; as Sansc. eka, one, compounded 

 of the demonstrative e and the interrogative ka ; Zend, aeva, one, 

 connected with Sanscrit pronominal adverb eva, only; Goth, ains. 

 Germ, ein, Eng. 07ie, Gr. £/', Old Lat. oinos, Lat. vnus, all derived 

 from Sansc. ena, this. — 5. No form extant. — 6. No form extant. — 7. 



* Max. Schmidt, (Comment, de Pronom. Graeco et Latino, p. 15.) endeavors to 

 distinguish between ? demonstrative and relative, and i' the third personal pronoun, 

 but I think without success. 



