MIMICRY AMONGST THE BLATTIDJE. 363 



Fletcher who has watched this insect in a state of nature in 

 Ceylon, tells me that when it is flying it looks very like a con- 

 spicuous Aganstid moth, Mimeusemia ceylonica Hmpsn The 

 resemblance is certainly not very striking when the dried insects 

 are seen side by side in a cabinet, but no field-naturalist will 

 attach very much importance to that, and in any case Mr Fletcher 

 does not maintain that the resemblance is detailed and accurate 

 but merely generalised. ' 



The power which the females of species of Perisjyhaeria and 

 1 seudoglomeris have of rolling themselves up into spherical balls 

 when alarmed is well known, and on account of their convex 

 torm and black shining colour, they undoubtedly bear an ex- 

 tremely close resemblance to the pill-millipedes which are so 

 abundant in the tropics. But here again I doubt if any parti- 

 cular species of millipedes are copied. It is certainly a fact that 

 whilst two species of Perisphaeria were not infrequently met with 

 in Sarawak, both rather small, black species, I never once found 

 a millipede corresponding in size or colour to them. It is by no 

 means certain that the pill-millipedes are distasteful animals— on 

 the contrary, it is quite probable that they are palatable but well 

 protected by their hard integuments and power of rolling up into 

 a hall. Ihe same habit is shown by many terrestrial Isopoda 

 but no one considers that the Isopods mimic the Millipedes or 

 the Millipedes the Isopods. The similarity of habit and form is 

 attributed to homoplasy, and I see no reason why the same habit 

 evolved °° ches Sh0uld nofc aIso have been <l uite independently 



Having now passed in rapid review the principal genera of 

 tflattidas which show a more or less generalised resemblance to 

 insects ol other orders, it only remains to consider in greater 

 detail the genus Prosoplecta Sauss., nearly all the members of 

 which present a remarkably close and detailed resemblance to 

 definite specific models amongst the Coleoptera, so far as these 

 have been discovered. With but two exceptions the species of 

 1 rosoplecta present an appearance which is conveniently sum- 

 marised as Coccinelliform ; that is to say, the outline of the body 

 is oval verging on spherical, the form is markedly convex the 

 integuments are smooth and nitid, the tegmina are corneous 

 with obsolescent venation and do not extend beyond the apex of 

 tne abdomen, the legs and antennas are short and, finally the 

 insects are gaily coloured. It is scarcely necessary to point out 

 that the Coccinelliform type is found amongst other families of 

 Coleoptera besides the Coccinellidas ; it is found, for example 

 amongst the Cassididae, Chrysomelid*, and Galerucid*, whilst 

 many of the Scutelleridae, a family of Hemiptera, also present 

 much the same facies. The two species, P. coccinella Sauss. and 

 1. bipunctata Br., are, in spite of the name of the first, far less 

 Coccinelliform than the other species of the genus, and may 

 certainly be regarded as more primitive. The form is more 

 depressed, and I am inclined to suppose that these two species 



m 



