296 Habit and Instinct. 



vague as to rob it of all valid biological significance. By 

 correlated variations should surely be meant those which 

 have been shown to occur in some direct or inverse ratio 

 to each other. When the correlation of any two variations 

 has thus been definitely established, a study of either of 

 them will enable us to infer the condition of the other. 

 This is how Darwin used the term : * and we should aim 

 rather at increased than at diminished exactness. 



Bearing all this in mind, it is worth while to consider 

 how far truly instinctive activities may be fairly considered 

 to be (a) of direct utilitarian value ; QS) correlated in the 

 narrower sense with useful characters; or (c) merely 

 incidental, without traceable relationship to other useful 

 characters, though involved in the congenital nexus. 



There can be little question that a large number of 

 instinctive activities in birds fall under the first class, and 

 are of direct and obvious utilitarian value. The in- 

 stinctive activities connected with pecking and feeding ; of 

 locomotion, whether of running, swimming, or flying ; the 

 sexual instincts and that of nidification, those associated 

 with incubation and the rearing of young ; aU these, and 

 they form perhaps the most remarkable and stable of 

 instinctive activities, are of decisive use, and are readily 

 explicable as the direct outcome of natural selection. 

 To the same class belong, it would appear, the protective 

 instincts of hiding, crouching, or shamming dead; the 

 aggressive activities such as the hissing of geese or 

 ostriches, and the scolding of young moorhens and other 

 birds; perhaps the scratching of the ground by fowls 

 and their allies ; the secondary sexual activities, dancing, 

 display, singing, together with those associated with recog- 

 nition marks, such as the flirting of the tail of moorhens, 



* "Animals and Plants under Domestication," second edition, vol. ii, 

 p. 311. 



