202 MENTAL EVOLUTION IN ANIMALS. 



day ; in each case a secondary instinct would become blended 

 with a previously existing primary one, so producing a new 

 instinct with a double root or origin. 



Conversely, as an example of a primary instinct becoming 

 similarly blended with a previously existing secondary, let 

 us take the following : — 



The grouse of North America display the curious instinct 

 of burrowing a tunnel just below the surface of the snow. In 

 the end of this tunnel they sleep securely ; for, when any 

 four-footed enemy approaches the mouth of the tunnel, the 

 bird, in order to escape, has only to fly up through the thin 

 covering of snow. Now in this case the grouse probably 

 began to burrow for the sake of protection, or concealment, 

 or both ; and, if so, thus far the burrowing was probably an 

 act of intelligence. But the longer the tunnel the better 

 would it have served the purposes of escape, and therefore 

 natural selection would almost certainly have tended to 

 preserve the birds which made the longest tunnels, until the 

 utmost benefit that length of tunnel could give had been 

 attained.* 



Thus then we see that in the formation of instincts there 

 are two great principles in action, which may operate either 

 singly or in combination ; these two principles being the 

 lapsing of intelligence and the agency of natural selection. 

 In the previous chapter we were engaged in considering 

 instincts which are due to either one or other of these prin- 

 ciples alone ; in the present chapter we shall consider 

 instincts which are due to the joint operation of both prin- 

 ciples. 



Now it is clear at a glance that if even in fully formed 

 instincts we often find, as in the above examples, a "little 

 dose of judgment," it becomes difficult to estimate the im- 

 portance, either of this little dose of judgment becoming 

 habitual by repetition, and so improving the previous instinct, 

 or of its becoming mixed with the influence of natural selec- 

 tion. For, taking the latter case alone, if, as we have seen, 

 intelligent actions may by repetition become automatic 

 (secondary instincts), and if they may then vary and have 

 their variations fixed in beneficial lines by natural selection, 

 how much more scope may be given to natural selection in 



* The facts of this case have been told mc by Dr. Eae, F.R.S. 



