GENETICS AND RACE 459 



of fairly narrow-nosed people of short stature around the wedge of narrow- 

 nosed taller people projecting into the forest. 



Shaxby has shown that skin pigmentation grades closely from the Sudan 

 to near the Arctic Circle, where Scandinavia yields the Nordic type, as it is 

 called. Similar points might be made in respect of other characters ; thus 

 the pattern of the main mass of mankind may be said to be one of transitions 

 in some respects between certain standards in Africa, Europe, Eastern Asia, 

 Papuo-Melanesia . 



A scheme based on transitions more or less under environmental influences 

 is, however, not much more satisfactory than a purely geographical classi- 

 fication, for we cannot but be impressed by the fact that almost every 

 population consists of disparate elements that reappear or persist side by 

 side in a population generation after generation. We cannot treat an 

 ordinary population as a unit to be described by giving means and standard 

 deviations for each character. Those figures often are mere abstractions. 

 We need to try to see how bundles of characters are grouped together, what 

 bundles occur and seem to be transmitted as entities, and how the propor- 

 tionate numbers with different bundles vary from district to district. For 

 they do differ, and we can understand this better if we remember that each 

 of us had, in theory, 32,768 ancestors about the time of the discovery of 

 America, and 1,073,000,000 about the time of the Norman Conquest. As 

 marriage was largely localised and few rural areas with persistent inter- 

 marriage had a population of 32,000 in the fifteenth century, we realise how 

 much branches of genealogical trees must intertwine, and so how possible 

 it is for an element, a group of characters, that got into a locality long ago 

 through a good number of individuals to go on century after century in spite 

 of some intermixture with individuals from outside, provided it has not to 

 work against a Mendelian dominant. Needless to say, the bundle of 

 characters need not be and is not exhibited by every member of the local 

 group, nor is any claim made that all members of the group are of strictly 

 localised descent. We are dealing only with proportions of a population. 

 Another necessary caveat is that the interpretation of differences between 

 localities has to be done with great reserve unless we know from skeletons 

 a fair amount about the back history of the district's population. Fortu- 

 nately, as regards Egypt, we have the knowledge, from the work of Elliot 

 Smith, Morant and others, that there has been a persistence of a bundle of 

 physical characters from predynastic days till ours. In our own country, 

 in some areas once inhabited (c. 1900 B.C.) by the beaker-making people, 

 the characters which distinguish their skulls are still found in certain cases 

 among the modern population. In some areas of special isolation quite a 

 number of people may carry and transmit a bundle of characters that seems 

 associated with the very earliest population known from skeletal remains in 

 Britain. 



The recurrence of these bundles is more than can be accounted for by 

 any estimate of the probability of recombinations occurring in the course of 

 intermixture ; persistence seems the more likely hypothesis, and the linkage 

 of characters in a bundle is fairly obviously a feature. 



We are, then, dealing with bundles of characters inherited more or less 

 as such, diverse bundles often co-existing side by side. Even an inter- 

 breeding population therefore need by no means form a unit, and averages 

 may mislead seriously. A pure race, with essentially uniform bundles of 

 characters in all its members, probably does not exist ; indeed it is better 

 not to use the term race at all in view of its purified misapplication in 

 political discussions as well as of the inherent biological difficulties attached 

 to the use of this word. 



