ADVANCE OF ANTHROPOLOGY. 485 



(1785-1846) lies at the foundation of ethnology, 

 but neither indulged in any special classification. 

 Broca, De Quatrefages, Haeckel, Huxley, and many 

 others suggested schemes, none of -which has been 

 found altogether satisfactory. The present tendency 

 seems to be to postpone further construction until 

 the criteria of race have been more thoroughly and 

 more critically studied. The practice of anthropo- 

 metry was greatly increased in exactness by the -work 

 of Quetelet, Gafton, and others; but there is still 

 need for careful criticism. Thus the zoological dis- 

 tinction between " variations " and " modifications " 

 has to be worked out in regard to racial distinctions ; 

 and the occurrence of " convergence " or " homoplas- 

 tic resemblance " familiar to the biologist, must be 

 carefully looked for. 



It seems fairly clear that in regard to physical 

 characters no reliance can be based on one character 

 by itself. Men cannot be classified by skull-char- 

 acters (especially if the observations be restricted 

 to adults) as crystals by their facets. The diagnostic 

 distinctions of races must rest on a combination of 

 characters. It seems also clear that speech and race 

 are anything but convertible terms, and that simili- 

 tudes in customs and belief afford no criterion of 

 genetic affinities. They are analogies, not homolo- 

 gies. 



Mr. Keane's picture of the chief branchings of 

 the human genealogical tree is briefly as follows: 

 (I.) The first ramification from the main stock is 

 that of the " generalised negro " (Homo cethiopi- 

 cus), branching off in various directions towards 

 Africa, Oceania, and Australia; (II. and III.) after 

 the negro dispersion the main stem throws off a 

 generalised Mongolo- American limb, which pres- 



