NATURAL SELECTION 233 



for consideration because they are the best documented and 

 most amenable to exact study. We wish, however, to make 

 one general comment which is applicable to the whole subject 

 of protection. Cuenot (1925, p. 335 and foil.) has very clearly 

 pointed out the difficulties involved in our assessment of what 

 may be regarded as ' protective.' (a) The existence and 

 efficacy of protection depend on observation on predator and 

 victims in the field, and exact observation of this kind is very 

 defective ; (b) the human evaluation of any protective device 

 may be fallacious, and can be shown to be so in specific cases ; 

 (c) owing to the enthusiasm of selectionists there is at present 

 a reaction against the cruder adaptive interpretations. There 

 is, however, enough evidence that particular devices are 

 directed against specific enemies. We cannot get rid of the 

 problem by a prejudiced disregard of these. 



Protective Resemblance. — Protective resemblance includes all 

 the methods by which animals secure their safety by their 

 similarity to other objects, whether the latter be living organ- 

 isms, particular inanimate objects or their natural background. 

 In this sense it includes mimicry ; but the latter is dealt with 

 in another section. 



There are, as is well known, three main kinds of protective 

 resemblance— simple homochromy or the resemblance of an 

 animal's colour to its background ; blending or deceptive 

 coloration ('camouflage'), which includes ' countershading ' ; 

 and what is sometimes termed assimilation, in which not only 

 the colour but also the surface modelling and the shape com- 

 bine to produce either a similarity to some inanimate object 

 or a blending of the animal with its background. 



Homochromy is in general a feature of whole genera and 

 families, indeed of whole faunas (e.g. desert and arctic animals). 

 In fact, Willey (191 1, chapter hi) regards cryptic colours as a 

 special case of a generalised primitive tendency and an adapta- 

 tion to a fundamental cryptozoic or hidden mode of life. 

 From this point of view we might admit the action of selection 

 in maintaining, in the majority of animals, a high level of 

 generalised protective colouring, while having little influence 

 on the specific manifestations of the general tendency. The 

 relatively few cases in which specific or racial colour differences 

 appear to be adaptive are considered later (p. 279). 



Many cases of homochromy are due either to individual 



