670 DR. J. C. MOTTRAM ON SECONDARY 



The table shows that 15 per cent, of large birds, 17 per cent, 

 of medium, and 15 per cent, of small often present sex differences. 

 In this respect there is little difi'erence, but of large birds 

 55 per cent., of medium-sized 43 per cent., of small 33 per cent, 

 present no sex difference. Examination horizontally also shows 

 that large birds less often present secondary sexual dimorphism 

 than the medium-sized or small birds. 



Instead of studying birds as a whole, certain families, chosen 

 at random, have been examined by themselves with respect to 

 the question of size. In barbets, jacamars, puff-birds, wood- 

 peckers, and toucans the Tables III. and VIII. show that the 

 smaller species present the most secondary sexual dimorphism. 

 In bomb ills and kingfishers, however, this association is not 

 found, though the reverse does not hold. 



Other families show the effect of size without any detailed 

 analysis; for instance, in the Anatidfe swans and geese present 

 little or no sex difference, and in the Falconida? it is only the 

 smaller species which show any sex difference. 



On the whole, theiefore, the consideration of size appears to 

 support the thesis. 



Table VIII. 



Notes. — Compiled from Monographs an the Capitonidseby C. T. &G.L. Marshall ; 

 Galbulida; and Bucconidse by P. L. Sclater; Alcediiiidie by R. B. Sharpe; 

 Picidffi by Alf. Malherbe. 

 For column headings, see Table III. 



Sociability. — This is another factor which should protect birds 

 from enemies to some extent. Numbei's may be able to resist 

 an enemy to which one would fall a victim. Sociable birds 

 should, therefore, show lesss secondary sexual dimorphism — as is 

 the case, see Table IX. 



