HUMAN GENETICS 34I 



in any strictly defined sense. In Lymantria we are only interested in the 

 sex genes and a few others. If two groups of Lymantria collected in 

 different places differ in respect of some or all of these genes we call 

 them different races. But in man we are interested in so many characters 

 that we can hardly ever find two individuals which do not differ in 

 some respects; we can never deal with groups which are uniform for 

 all the characters we can examine. The only hope of applying a race 

 concept is to limit the sphere of reference, and to do that we shall only 

 take account of certain sorts of characters. 



If we limit the number of characters severely enough, we can distin- 

 guish a few large groups of mankind, such that the different groups are 

 pretty sharply distinguished from one another and we do not find many 

 intermediates except obvious hybrids. Thus we can discriminate the 

 "races" of whites, negroes, American Indians, yellows, Australians, 

 etc. But these groups are themselves not at all uniform. For instance, 

 the genetic basis of colour seems to be different in different groups of 

 negroes; West Africans differ from whites in many polymeric colour 

 genes, so that in hybrids such as American mulattoes there is a segre- 

 gation into a continuous series of colour grades, whereas South Africans 

 seem to differ in only a few major genes, and hybrids show a fairly clean 

 segregation into dark and light types. ^ However, when we try to bring 

 such differences into the classification, and discriminate minor races 

 within the great groups of whites, blacks, etc., we find that it is ex- 

 tremely difficult to draw any dividing lines within the general chaos of 

 variations. 



We can set about the attempt in two ways. In the first place, we may 

 define a set of characters which we take to be characteristic of a certain 

 race. For instance, we might say that the Nordic race has long heads, 

 blond hair, and blue eyes. But this only has any sense if we believe that 

 a group of men having these characteristics exists now or existed at one 

 time. In the absence of any detailed knowledge about the past, the 

 justification for such a definition reduces to whether or not we can find 

 such a group at present. We must therefore tackle the problem in the 

 second way, which is to examine human populations and see how they 

 can be grouped. If we do this, we can in fact find groups which are 

 homogeneous in the sense that, although they show variation, this 

 variation is continuous and the group cannot be spUt up into smaller 

 groups which are significantly different from one another.^ But these 

 homogeneous groups are extremely smaU and insignificant. They are 

 1 Gates 1929. 2 cf. Morant 1928. 



