NATURAL SELECTION 243 



the belly ' countershadowed,' but also that of the tail and feet, 

 parts which cast so slight a shadow that the effect of counter- 

 shading must be minimal in its efficacy. 



We strongly suspect that ' countershading ' is not efficacious 

 in the sense originally propounded by Thayer and demon- 

 strated by his celebrated (if too plausible) models ; but we 

 think the subject requires further investigation. No satisfac- 

 tory alternative explanation of the pallor of the under-parts of 

 ' countershaded ' animals has so far been put forward. It is 

 just possible that it may be the expression of a ' physiological ' 

 gradient. 



Warning coloration. — Many of the exceptions to the rule 

 of protective coloration have been considered as examples of 

 warning colours. Familiar examples are seen in the black- 

 and-yellow livery of wasps or the brilliant colours of some 

 venomous snakes. There is little doubt that in the past this 

 principle has been pushed too far. It is a familiar fact that 

 many conspicuously coloured animals actually blend with 

 their background when seen in their natural surroundings, as 

 insisted by Longley (191 7). Apart from this reservation, 

 however, it is by no means easy to estimate the validity of the 

 warning colour theory. There are a good many striking cases 

 of brilliant colour associated with nauseous odour or some 

 special means of protection (stings, poison fangs, urticating 

 hairs, etc.). 



An objection has been made against the warning colour 

 hypothesis to the effect that a good number of non-noxious 

 forms are brilliantly coloured. For example, Gadow (191 1, 

 p. 2) has shown that there are in Mexico and Central and 

 South America ' a surprising number of harmless snakes which 

 resemble in their coloration the poisonous Elaps to a wonderful 

 extent.' These apparently contradictory cases have, of course, 

 been explained as due to mimicry. Gadow (I.e.) has tried 

 to evade this explanation, but his objections have been 

 subjected to a searching criticism by Sternfeld (191 3). The 

 general question of mimicry is discussed elsewhere, and we are 

 here concerned with the question whether the origin of ' warn- 

 ing ' colours is to be explained on the traditional lines. 



Gadow (I.e. pp. 2-3) has made the criticism in the case of 

 the poisonous Elaps that they are nocturnal and in the day- 

 time lead a hidden life, and that against their only serious 



