1884.] on the Barwinian Theory of Instinct. 143 



exhibited the instinct, and carefully noted the attitude in which 

 they feigned death. Some of these insects he then killed, and he 

 found that in no case did the attitude in which they feigned death 

 resemble the attitude in which they really died. Consequently we 

 must conclude that all the instinct amounts to is that of remaining 

 motionless, and therefore inconspicuous, in the presence of danger ; 

 and there is no more difficulty in understanding how such an instinct 

 as this should be developed by natural selection in an animal which 

 has no great powers of locomotion, than there is in understanding 

 how the instinct to run away from danger should be developed in 

 another animal with powers of rapid locomotion. The case, however, 

 is not, I think, quite so easy to understand in the feigning death of 

 higher animals. From the evidence which I have I find it almost 

 impossible to doubt that certain birds, foxes, wolves, and monkeys, 

 not to mention some other and more doubtful cases, exhibit the 

 peculiarity of appearing dead when captured by man. As all these 

 animals are highly locomotive, we cannot here attribute the fact to 

 protective causes. Moreover, in these animals this behaviour is not 

 truly instinctive, inasmuch as it is not presented by all, or even most 

 individuals. As yet, however, observation of the facts is insufficient 

 to furnish any data as to their explanation, although I may remark 

 that possibly they may be due to the occurrence of the mesmeric or 

 hypnotic state, which we know from recent researches may be induced 

 in animals under the influence of forcible manipulation. 



The instinct of feigning injury by certain birds presents a peculiar 

 difficulty. As we all know, partridges, ducks, and plovers, when they 

 have a brood of young ones, and are alarmed by the approach of a 

 carnivorous quadj-uped, such as a dog, will pretend to be wounded, 

 flapping along the ground with an apparently broken wing in order 

 to induce the four-footed enemy to follow, and thus to give time for 

 the young brood to disperse and hide themselves. The difficulty here, 

 of course, is to understand how the birds can have acquired the idea 

 of pretending to have a broken wing, for the occasions must be very 

 rare on which any bird has seen a companion thus wounded followed 

 by a carnivorous quadruped ; and even if such observations on their 

 part were of frequent occurrence, it would be difficult to accredit the 

 animals with so high a degree of reasoning power as would be re- 

 quired for them intentionally to imitate such movements. When I 

 consulted Mr. Darwin with reference to this difficulty, he gave me a 

 provisional hypothesis by which it appeared to him that it might be 

 met. He said that anyone might observe, when a hen has a brood 

 of young chickens and is threatened by a dog, that she will alternately 

 rush at the dog and back again to the chickens. Now if we could 

 suppose that under these circumstances the mother bird is sufficiently 

 intelligent to observe that when she runs away from the dog, she is 

 followed by the dog, it is not impossible that the maternal instinct 

 might induce her to run away from a brood in order to lead the dog 

 away from it. If this happened in any cases, natural selection would 

 tend to preserve those mother birds which adopted this device. I 



