120 THE VALUE OF COLOUE 



either hypothesis voHtion has nothing to do with 

 the growth of resemblance, but that it is believed 

 to be brought about by the survival in successive 

 generations of those individuals most like the 

 model or most like one another. The death of 

 individual A or B as a result of the tasting is 

 no difficulty. Far more individuals of A, B, C, D, 

 &c., would be killed by experimental tasting 

 if they had different patterns than if they had 

 the same, and this is advantage enough to cause 

 a strong trend in the direction of resemblance. 



How far does the constitution of this wonderful 

 combination — the largest and most complicated 

 as yet known in aU the world — convey to us the 

 idea of Mimicry working along the lines supposed 

 by Bates or those suggested by Miiller ? Figures 

 1 to 52 of Mr. Marshall's coloured plate ^ represent 

 a set of forty-two or forty-three species or forms 

 of insects captured in Mashonaland, and all ex- 

 cept two in the neighbourhood of Salisbury. 

 The combination includes six species of Lycidae ; 

 nine beetles of five groups all specially protected 

 by nauseous qualities, TelepJioridae, Melyridae, 

 Phytophaga, Lagriidae, Cantharidae ; six Longi- 

 corn beetles ; one Coprid beetle ; eight stinging 

 Hymenoptera ; three or four parasitic Hymeno- 

 ptera {Braconidae, a group much mimicked and 

 shown by some experiments to be distasteful) ; five 

 bugs (Hemiptera, another group in which unpalata- 



■ Trans. Ent. Soc. Land., 1902, plate xviii. See also. 517,- where 

 the group is analysed. 



