74 THE VARIATION OF ANIMALS IN NATURE 



function and occupation. Such units are generally known 

 as biological or physiological races. They have, of course, 

 been for a long time familiar to bacteriologists and have been 

 detected in Protozoa among which structurally indistinguishable 

 strains are found in different hosts. Similar ' host-specificity ' 

 accompanied by morphological differentiation is a well-known 

 phenomenon in various groups of parasitic Metazoa. The 

 whole problem of physiological differentiation involving such 

 phenomena as immunity, certain aspects of interspecific 

 sterility and graft-specificity has been recently reviewed by 

 Robson (1928, Chapter III), and Thorpe (1930, p. 177) has 

 given a survey of the special phenomenon of biological races 

 in insects, nematodes, etc. It should be noted (a) that it is 

 not always easy to distinguish ' physiological races ' from 

 those separated by habitat-preferences which may be 

 determined by other factors than physiological idiosyncrasy, 

 and (b) that ' physiological ' is sometimes used in a very broad 

 sense. Thus Fulton (1925) and Allard (1929) allude to the 

 stridulation of Orthoptera as physiologically differentiated. 



How frequent this phenomenon is it is not easy to say. 

 It may be that in every phylum the species are composed of 

 subordinate groups diversified in regard to their ' physio- 

 logical ' characters. The ground has not been sufficiently 

 explored from this point of view. A list of the features of this 

 order that seem in one group or another to be the basis of 

 racial diversification is sufficiently impressive to lead us to 

 believe that it must be of very frequent occurrence. 



While in practice it would be undesirable to give separate 

 names to the various physiological races within a species, it 

 should be noted that some botanists have definitely adopted 

 the practice of naming ecological subspecies and that Alpatov 

 (1924) has recognised similar subspecies (' truncicolae,' etc.) 

 in ants. 



Just as the taxonomist's species may contain divers struc- 

 tural, geographical and genetical subdivisions, it also seems 

 to contain elements that are diversified by habit, habitat- 

 preference, physiological reactions, food-preference and so 

 on. Such differentiation may or may not be accompanied 

 by structural differentiation and its occurrence must always 

 constitute an interesting starting-point for evolutionary inquiry, 

 as it invites the obvious query — do initial differences in food, 



