326 



scmwGE, 



[Vol. IX., No. 317 



to this scientific method, which, as he justly con- 

 siders, can alone give to such a work a permanent 

 value. In the introduction, however, which occu- 

 pies about a third part of the first volume, he has 

 allowed himself more freedom, and has entered 

 into many disquisitions which will interest the 

 general reader, and will doubtless evoke much 

 discussion and some dissent. He treats of the aim 

 and limits of linguistic science ; of the relation of 

 speech to thought ; of the origin of language, in- 

 cluding the great question of the unity or plurality 

 of beginnings ; of the development of speech ; of 

 its material and formative parts ; of the proofs of 

 kinship among languages ; of their classification, 

 according to the various systems which have been 

 proposed by philologists ; of the elements of 

 speech, — the root, the word, the sentence; of 

 articulate sounds (phonology) ; of the expression 

 of thought by writing, and of the influence of 

 writing on the development of language. This 

 list of topics is much abridged, and gives only an 

 imperfect idea of the many subjects on which the 

 author touches in this important introduction, in 

 which he has condensed the conclusions of long- 

 continued study and profound analysis. 



In his classification he has sought to combine 

 the ethnological and philological methods, and 

 thus to link his earlier ' General ethnography ' 

 with the present work. The attempt was a nat- 

 ural one, but cannot be said to be altogether suc- 

 cessful ; and it is easy to see that the author him- 

 self, whose candor throughout is transparent, was 

 finally not altogether satisfied with it. In the 

 classification of races he selects (as in his * Eth- 

 nography ' ) the hair as the best criterion. He 

 divides all mankind primarily into two classes, — 

 the ' woolly-haired ' {ulotrichi) and the ' smooth- 

 haired' (lissotrichi). Each of these classes is 

 again subdivided into two divisions. The woolly- 

 haired class comprises the ' tuft-haired ' {lopho- 

 comi) and the ' fleecy-haired ' (eriocomi) ; while 

 the smooth-haired races comprehend the ' straight- 

 haired ' (euthycomi) and the 'wavy-haired' (eu- 

 plocomi). Other high authorities, including St. 

 Hilaire, Bory de St. Vincent, and Huxley, have 

 adopted the hair as the best primary character- 

 istic for distinguishing the races. But while the 

 epithets drawn from it are excellent descriptive 

 terms, they are found in practice, like those de- 

 rived from the shades of color and the shape of 

 the head, to be far too wavering and uiicertain to 

 serve the purposes of a true scientific classifica- 

 tion. Such is the conclusion of Pi-ichard, Peschel, 

 Quatrefages, Wilson, and other able ethnologists 

 who have tested these methods. 



To this opinion Dr. Miiller's own matured views 

 plainly tend. Though he formally preserves 



throughout his work — evidently for the reason 

 that has been suggested — the four classes distin- 

 guished by the hair, he practically deserts this 

 classification for that which his studies and phil- 

 osophical insight have convinced him to be the 

 only satisfactory and proper one, — at least for a 

 philological treatise, — namely, the genealogical 

 classification, based on the distinction of linguistic 

 stocks. These stocks are, in fact, in comparative 

 philology, what the elementary substances are in 

 chemistry, — the sole and siifficient ground of a 

 true scientific classification. The question of the 

 origin of these stocks, or linguistic families, is too 

 extensive and too much contested to be here con- 

 sidered ; but that their distinction and deter- 

 mination constitute the primary element and 

 foundation of linguistic science is a definite con- 

 clusion, for which the high authority of Dr. 

 Miiller may now be claimed. 



The main body of the work consists of careful 

 analyses of the phonetic and grammatical systems 

 of all the languages whose sounds and grammar 

 are known. In most instances — and, in fact, 

 wherever compositions in the language are found 

 — specimens of the text are given, with inter- 

 linear translations, and with annotations explain- 

 ing every grammatical peculiarity. Such transla- 

 tions are, of course, the best test of the author's 

 knowledge of the language. The labor required 

 to master so completely the intricacies and pecul- 

 iarities of this large number of idioms — from 

 the monosyllabic Chinese and Anamese, with 

 their variety of tones and positions, to the multi- 

 tudinous inflections of American tongues — 

 must have been enormous ; nor would mere in- 

 dustry have been sufficient, without large experi- 

 ence, and what may fairly be termed linguistic 

 genius. The first volume comprises the languages 

 of the woolly-haired races, and is devoted almost 

 entirely to the African tongues. The single ex- 

 ception is the Mafor language, spoken on the north- 

 west coast of New Guinea. The Mafor people are 

 not more woolly-haired than many other tribes 

 of Melanesia. But as the latter speak ' mixed 

 languages,' mainly of the Malaisian type, they 

 are relegated to the ' Malayan race,' which is in- 

 cluded among the smooth-haired races. Thus the 

 classification by the hair breaks down on its first 

 application ; and we cannot be surprised that the 

 author, hampered at the outset by his earlier 

 ethnological theories, is glad, as his work proceeds, 

 to escape from them, and restrict himself entirely 

 to the genealogical classification. 



The second volume opens with an interesting 

 description and comparison of the very peculiar 

 and in some respects highly organized Australian 

 languages, which are shown conclusively to be- 



