438 ETHNOLOGY. 



the same inconsistent alteration or anomaly frequently takes place also in 

 tlie family of Ural-Altaic languages. [For furtlicr developments, see the 

 Notes at the end of this article.] 



Thus we find in the Dakota or Sioux language, hEpaq, (second son of 

 a family,) and hApay, (second daughter oi 1^i•^im.i\y -,) ciij, elder 6ro</ier,cUi), 

 elder sister-,'^ cii^ksi, son, cUyksi, daughter, &c. Also, the demonstratives 

 koi}, that, and kii^, this, the, (the definite articles,) seem to come, in some 

 respects, under this head. 



To investigate the grammatical structure of languages from a compar- 

 ative point of view is, however, but one part of the work of the philologist ; 

 the other equally essential part consists in the study of the words them- 

 selves, the very material of which languages are made. We do not, as 

 yet, intend to touch on tbe question of Dakota words and their possible 

 afQnities, but reserve all that pertains to comparative etymology for some 

 other time. The identity of words in different languages, or simply their 

 afSnity, may be either imuiediately recognized, or rendered evident 

 by a regular process of philological reasoning, especially" when such 

 words appear, as it were, disguised, in consequence of certain alterations 

 due to ti'me and to various vicissitudes, whereby either the original 

 vowels, or the consonants, or both, have become changed. Then, also, 

 it frequently happens that one and the same word, when compared in 

 cognate languages, may appear as different parts of speech, so that in 

 one of them it may exist as a noun, and in another only as a verb, &c. 

 Moreover, the same word may have become gradually modified in its 

 original meaning, so that it denotes, for instance, in one of the cognate 

 languages, the genus, and in another, merely the species of the same thing 

 or idea. Or it may also hapi)en that when several synonymous expres- 

 sions originally existed in what we may call a mother language, they 

 have become so scattered in their descent that only one of these words 

 is found in a certain one of the derived languages ; while others again be- 

 long to other cognate tongues, or even their dialects, exclusively. 



The foregoing is sufficient to account for the frequent failures in es- 

 tablishing tlie reUitionship of certain languages in regard to the affinity 

 of all their icords. 



On this occasion it will be enough to mention, in passing, as it were, 

 one or two of the most frequently used words, such as the names of 

 father, mother, &c. 



In regard to these most familiar expressions, we again find a sur- 

 prising coincidence between the tongues of Upper Asia (or more ex- 

 tt'nsively viewed, the Ural-Altaic or Tartar-rinnish stock of lauguagesj 

 and the Dakota. 



Father is in Dakota ate ; in Turco-Tartar, ata ; Mongolian and its 

 branches, ctsii, etsige ; in the Finnish languages we meet with the 

 forms attje, atd, &c.; they all having at {=et) as their radical syllable. 

 Now, as to mother, it is in the Dakota language ina ; and in the Asiatic 

 tongues just mentioned it is ana, aniya, inc, eniye, &c. 



