248 MENTAL EVOLUTION IN MAN. 



much more by individual change. In languages of this kind 

 the principle of agglutination has been so perfected that 

 the original composition is more or less obscured, and the 

 resulting words therefore admit of being themselves twisted 

 into a variety of shapes significant of finer grades of meaning, 

 in the way of declension, conjugation, &:c. Or, to state the 

 case as it has been stated by some philologists, in agglutina- 

 tive tongues the welded elements are not sufficiently welded 

 to admit of flexion : they are too loosely joined together, or 

 still too independent one of another. But when the union 

 has grown more intimate, the structure allows of more artistic 

 treatment at the hands of language-makers : the " amalgama- 

 tion " of elements having become complete, the resulting alloy 

 can be manipulated in a variety of ways without involving its 

 disintegration. Moreover, this principle of inflection may 

 extend from the component parts to the root itself; not only 

 suffixes and prefixes, but even the word which these modify, 

 may undergo inflectional change. So that, upon the whole, 

 the best general idea of these various types of language- 

 structure may perhaps be given by the following formulae, 

 which I take from Hovelacque.* 



In the isolating type the formula of a word is simply R, 

 and that of a sentence R-f-R-{-R, &c., where R stands for 

 " root." If we represent by r those roots whose sense has 

 become obscured so as to pass into the state of prefixes and 

 suffixes significant only of relationship between other words, 

 we shall have a formula of agglutination, Rr, Rrr, rR, rRr, 

 &c. Lastly, the essence of an inflecting language consists in 

 the power of a root to express, by modification of its own 

 form, its various relations to other roots. Not that the roots 

 of all words are necessarily modified ; for they often remain 

 as they do in agglutinating tongues. But they may be 

 modified, and " languages in which relations may be thus 

 expressed, not only by sufiixes and prefixes, but also by a 



* This method of representation was devised by Schleicher, who carries it 

 further than I have occasion to do in the text. See Memoirs of Academy of St. 

 Petersburg, vol. i., No. 7, 1859. 



