Are Acquired Habits inherited? 305 



natural selection have produced the instinctive behaviour 

 of the Yucca moth and the Sitaris, what becomes of the 

 inherent improbability of the moorhen's dive ? 



If pressed to summarize my own opinion on this 

 vexed question, I should say, first, that there is but 

 little satisfactory and convincing evidence in favour of 

 transmission, but that variation does seem in some cases 

 to have followed the lines of adaptive modification, so as 

 to suggest some sort of connection between them ; secondly, 

 that there are many instincts, relatively definite and 

 stable, which may fairly be regarded as directly due to 

 natural selection, though here again, if we could accept 

 the view that adaptive modification marked out the lines 

 in which congenital variation should run, our conception 

 of the process of their evolution would be so far simplified ; 

 thirdly, that there are some peculiar traits, also seemingly 

 definite and stable, which can only be attributed to the 

 indirect effects of natural selection on the supposition 

 that they form part of the congenital nexus, and that they 

 have no intrinsic tendency to variation in any particular 

 direction ; and fourthly, that, in the present state of our 

 knowledge, it is best to accept provisionally the view that 

 they are thus indirectly due to natural selection. 



There can be no question, however, that when we con- 

 sider such instincts as those connected with nidification, 

 incubation, and maternity, together with those associated 

 with courtship, if indeed they be truly instinctive ; when 

 we see how individual effort runs in the same lines as 

 congenital co-ordination, as in the perfected flight of birds, 

 or the diving of water-fowl ; when we regard dispassionately 

 such a performance as that of the lapwing or duck that 

 decoys an intruder from her nest or young ; — we feel that 

 though the evidence for the transmission of acquired 

 habit is insufficient, yet some connection between variation 



x 



