106 THE VARIATION OF ANIMALS IN NATURE 



In birds there is a very noticeable tendency to form geo- 

 graphical races (cf. Troglodytes musculus, fig. 12), and this is 

 probably connected with the tendency of migratory species to 

 return to the same spot to breed. Exceptions occur to Rensch's 

 rule that habits condition race-formation. Thus Chapman 

 (1923, p. 252) states that Buarremon brunneinuchus, though it 

 ranges from Mexico to Peru and is essentially sedentary in 

 habits, ' shows no appreciable variation which can be corre- 

 lated with any given area.' This is all the more striking when 

 it is realised that a species, B. inornatus, has been evolved in and 



Fig. 10. — Variation in the Finch, Buarremon. a and c, B. brwmeinuchus ; bandd, 

 B. inornatus from the Chimbo Valley and Los Llanos, Ecuador. 



(After Chapman, 1923.) 



is restricted to a single valley in Ecuador. The case of the 

 Common Heron (p. 105) has already been discussed. We 

 believe that these instances must be referred to some inherent 

 inability to vary. 



As far as the recorded facts go, Rensch's rule holds for 

 mammals, though some exceptions should be noted. Roosevelt 

 and Heller (191 5, p. 570) show that the Steinbok (Raphicervus 

 campestris) is remarkably uniform throughout its range and 

 is not separable into geographical races. Christy (1929) 

 finds that the African Buffalo (Bubalis coffer) is undifferentiated 

 over all its range, while the Congo Buffalo (B. nanus) has many 

 local races. The remarkable differences in local variation 



