Table of Greek Correlatives, with Explanations. 341 
demonstrative, yet it is generally weakened into that of a mere arti- 
cle-—The forms téder» and 16% are also used as relatives.—Gs, al- 
though placed in this column, is from 63, 4, 6, used as a demonstra- 
tive. Comp. Rem. 6. below. 
Rem. 3. The lengthened forms, as tovovr0¢, tovodt0s, Tyhixodtog, ty- 
vixavTa, TooOauTéxig, are more emphatic than the others. 
Rem. 4. Many of the forms in this column are also strengthened or 
rendered emphatic, by the addition of 02; as 602, roodade, toode, ty- 
hixdade, tyyixdde. Add also totoadi, tovovtoct, tydexovtoat. 
The IIL. column is complete. It consists of interrogatives, which 
are a very peculiar class of words. An interrogative sentence is not 
a full or complete proposition in itself, but is an imperfect proposition 
or assertion, which is offered to another to complete or fill up. The 
interrogative element in Greek has three forms; viz. initial t, which 
is found alone in the first and seventh series, initial x and 7, which 
are found together in each of the other series, and probably once ex- 
isted also in the first and seventh series. Of these forms we shall 
treat separately.—(1.) The oldest of them is ¥, which is retained in 
the Ionic dialect, and is found abundantly in the other Indo-European 
languages. This letter is justly supposed to have a natural appropri- 
ateness to perform the function of an interrogative. Hence we are 
not to regard this column as formed from the second column, but as 
having a coetaneous origin with it. Itdiffers from it not by inflection 
er derivation; but, if I may so speak, by correlation. A leading 
letter of one organ is exchanged for a leading Jetter of another organ, 
each having its own natural significancy.—(2.) A second form is ini- 
tial 7, which seems to have usurped the place of ~ in all the Greek 
dialects except the Ionic. As this form has arisen, not by the com- 
mutation of one letter for another of the same organ, but by the 
commutation of a letter of one organ for the corresponding letter of 
another organ, it seems to require some illustration and confirmation. 
The following are examples of the interchange of k and p: Sanse. 
, Lat. agua, water; Sanse. pangtsha, Arab. khams, Mol. xéune, 
ia. quingue, Welsh pump, Ir. kuig, five; Gr. émouer, Lat. sequor ; 
Gr. jjzag, Lat. jecur ; Gr. dei, Lat. linquo ; Gr. tos, Lat. equus ; 
Gr. aéaw, Lat. coguo ; Gr. ét0ga for tésvaga, Lat. guaticor 3 ; Gr. débxos, 
Lat. lupus ; Gr. onxés, Lat. sepes ; Gr. oxihe, Lat. spolia ; Gr. ogy- 
xis, Lat. vespas; Oscan pitpit, Lat. quidquid ; Ir. keann, Welsh 
pen, the head. There can be no doubt that there is a physiological 
ground for this change, in the similar state of the organs of speech.— 
(3.) A third form is initial t, in the first and seventh series, which 
seem to be formed from the first column by correlation. ‘That * was 
even here the original form seems to be shown by the analogous 
