EXAMINATION OF THE THEORIES OF OTHER WRITERS. 271 



which the latter has given way when in conflict witli the 

 former. Lastly, I devoted a chapter to the consideration of 

 local and specific variations of instinct, showing liow these 

 constituted a kind of pahneontological evidence of the trans- 

 mutation of instinct. 



Such then is the a 2'>ostcriori proof of the two ways which, 

 either singly or in combination, must be regarded as those by 

 which all properly so-called instincts have been developed. 

 A diagram was given to show graphically how the sundry 

 principles concerned are related and inter-related with one 

 another. Here it was shown that when an instinct, whether 

 of single or blended origin, was perfected, it might vary or 

 ramify into modified forms, and even blend, or, as it were, 

 inarch with other instincts to produce a new growth. It is 

 difficult, or rather impossible, to trace the history of actual 

 instincts in this respect, from the fact that instincts are not 

 fossilized, and therefore leave no record of their transi- 

 tional states. But from all the evidence together — and 

 especially from what we may almost denominate the historical 

 evidence supplied by the facts of domestication — there can be 

 no reasonable doubt that instincts may not only have a double 

 root — one in the principle of selection, and the other in that 

 of lapsing 'intelligence — but also a more or less branching 

 stem, which (or the branches of which) may in some cases 

 become grafted with the stem or branches of other instincts. 



In estimating the comparative importance of the two 

 great factors in the formation of instinct, we had occasion to 

 differ on the one hand from Mr. Spencer, wlio attributes the 

 origin of all instincts to reflex action with little or no aid 

 from natural selection, and on the other hand with ^Ir. Lewes, 

 who goes to the opposite extreme of regarding all instincts 

 as cases of lapsed intelligence. It was shown, however, that 

 Mr. Spencer's view might be held to explain the rise of 

 doubtfully instinctive actions displayed by very low animals, 

 and that it is of much importance as an explanation of the 

 origin of Consciousness. The view, however, which I adopt 

 to explain the origin of instincts is substantially the same as 

 that which has been propounded by ]\Ir. Darwin, and which, 

 while recognizing both the factors which I have now so 

 repeatedly named — i.e., natural selection and lapsing intelli- 

 gence — whether singly or in combination, attributes most 

 importance to the former, especially if it be remembered that 



