234 THE VARIATION OF ANIMALS IN NATURE 



accommodation (produced by reflex action on the pigment cells 

 of the skin by various sense organs) or to the deposition in the 

 skin of pigments extracted from food-material. With these we 

 have no concern, except to point out that in all probability 

 we have not sufficiently realised that more cases of homochromy 

 are due to the former cause than we are at present prepared to 

 believe. 



As is well known, there are some remarkable cases of 

 assimilative resemblance to inanimate objects (stick insects ; 

 Kallima), and we should do well to bear in mind Cuenot's 

 warning that these are not to be lightly dismissed out of a 

 reaction against the enthusiasm of ardent selectionists. 



In commencing a critical study of this subject there are 

 two general points to note : 



i. One of the first things that attract our attention is the 

 capricious incidence of protective resemblance. One 

 cannot help speculating why it is brought to such a 

 high state of perfection (e.g.) in Phasmids and yet is 

 nearly entirely absent (e.g.) in land molluscs. The 

 ready answer that we must seek the explanation in 

 differences of habit not only begs the question as to 

 the origin of habits (p. 300), but ignores the very 

 real difficulty that a whole group of animals, like the 

 Gastropoda, of high adaptability, exposed to numerous 

 enemies, living in habitats in which protective 

 resemblance might be advantageously developed, and 

 possessing in the shell a notoriously plastic external 

 covering, have exhibited very few convincing cases 

 of this phenomenon. 



2. Though there are abundant cases of protective resem- 

 blance of one kind or another, there are numerous 

 instances of animals which are not thus protected, 

 are either fairly or markedly conspicuous and are not 

 known to be noxious or protected by some special 

 habit. It seems that there is a general tendency to 

 a cryptic coloration, and that in special cases this is 

 brought to a high state of perfection. We are a little 

 inclined to suspect that the latter is related to special 

 kinds of habitats (e.g. deserts) which have a homo- 

 geneous facies, and that where the background is 



