442 DARWINISM chap. 



]\Iuch of the mystery of instinct arises from the persistent 

 refusal to recognise the agency of imitation, memory, observa- 

 tion, and reason as often forming part of it. Yet there is 

 ample e\ddence that such agency must be taken into account. 

 Both Wilson and Leroy state that young birds build inferior 

 nests to old ones, and the latter author observes that the best 

 nests are made by birds whose young remain longest in the 

 nest. So, migration is now well ascertained to be effected by 

 means of "vision, long flights being made on bright moonlight 

 nights when the birds fly very high, while on cloudy nights 

 they fly low, and then often lose their way. Thousands 

 annually fly out to sea and perish, showing that the instinct 

 to migi^ate is imperfect, and is not a good substitute for reason 

 and observation. 



Again, miieh of the perfection of instinct is due to the 

 extreme severity of the selection dui'ing its development, any 

 failure invohdng destruction. The chick which cannot break 

 the eggshell, the caterpillar that fails to suspend itself properly 

 or to spin a safe cocoon, the bees that lose their way or that 

 fail to store honey, ine^dtably perish. So the birds that fail 

 to feed and protect their young, or the butterflies that lay 

 their eggs on the wrong food-plant, leave no oftspring, and 

 the race vrith. imperfect instincts perishes. Now, during the 

 long and very slow corn-se of development of each organism, 

 this rigid selection at every step of progress has led to the 

 preservation of every detail of structure, faculty, or habit that 

 has been necessary for the preservation of the race, and has 

 thus gradually built up the various instincts which seem so 

 marvellous to us, but which can yet be shown to be in many 

 cases still imperfect. Here, as everywhere else in nature, we 

 find comparative, not absolute perfection, with every gradation 

 from Avhat is clearly due to imitation or reason up to what 

 seems to us perfect instinct — that in which a complex action 

 is performed without any previous ex]^)erience or instruction.^ 



1 Weismaun explains instinct on similar lines, and gives many interesting 

 illustrations (see Essays on Heredity). He holds "that all instinct is entirely 

 due to the operation of natural selection, and has its foundation, not upon 

 inherited experiences, but upon variations of the germ." Many interesting 

 and difficult eases of instinct are discussed by Darwin in Chapter VIII of the 

 Orif/in of Species, which should be read in connection with the above remarks. 



Since this chapter was ^\Titten my attention has been directed to ilr. 



