344 OUTLINES OF EVOLUTIONARY BIOLOGY 



together to share the expenses of a common advertisement and 

 thereby reduce the cost to each. ^Jpung birds have to learn 

 by experience which insects are goJeFto eat and which are not. 

 In making their experiments no doubt they themselves suffer, 

 but the subjects of the experiment are probably actually killed. 

 Obviously, then, if one experiment can be made to serve for a 

 number of different species of insects there will be a corre- 

 sponding reduction in the death-rate, and hence it is that we 



FIG. 176. A Synaposematic Group of South American Lepidoptera, all x -J. 

 (From a photograph.) 



A, Tithorea liarmonia ; B, Heliconius ethilla; C, Perrhybris (Mylothris) malenka. 

 g ; D, Dismorphia praxinoe, J ; E, Pericopis angulosa. 



find these groups of species all adopting the same type of 

 warning colour, and thus coming to resemble one another very 

 closely, although perhaps belonging to totally distinct families. 



We may illustrate this somewhat complex phenomenon by 

 reference to certain South American Lepidoptera which take 

 part in the formation of such a synaposematic group. In 

 Fig. 176 A, B and D represent butterflies belonging to three 

 distinct families, while E is a moth, as may be seen at once by 

 its thick body and the absence of terminal knobs on the antenna}. 



