The Bionomics of South African Inserts. 523 



is the character which protects them (" Essays on Natural 

 Selection," 1875, p. 94). In answer to a letter in which 

 I drew his attention to Mr. JMarshall's record of a large 

 Curculio found in the crop of a guinea-fowl (see p. 350), he 

 wrote, Feb. 5, 1901, "The large Malayan Anthrihidm Q.ve 

 intensehj hard. The guinea-fowl proves nothing, as these 

 beetles are almost all arboreal, and their chief enemies are 

 smaller birds. Their protective colours may save them 

 from the larger insectivorous birds, their hardness from 

 the smaller." The mimicry of Malayan CurculionidfB, 

 AnthrihidcV, and Brenthidie by Longicorns cannot be 

 doubted. The cases are too numerous and the details of 

 the resemblance too precise to admit of any other explana- 

 tion. In South Africa, on the other hand, Mr. Marshall 

 shows that only the first group is mimicked, and of this 

 he has sent me a very beautiful example. Experiments 

 are greatly wanted, especially in Borneo, where all three 

 groups abound. In addition to their hardness Mr. Shel- 

 ford shows that the larger Bornean Curculios are defended 

 by their great strength ; they can even cause intense pain 

 to man by clasping the fingers with their legs and digging 

 the proboscis into the fiesh. Such defences as hardness 

 and strength depend for their success on the size of 

 enemies ; for even hardness could not avail against an 

 enemy large enough to swallow the beetle whole, so that 

 it could be ground down in the gizzard, or the interior 

 slowly extracted by digestive fluids gaimng access by the 

 joints and other apertures. Defence by a sting, a nauseous 

 taste or smell, or unwholesome qualities, is effective against 

 enemies of all sizes and all degrees of strength, although fail- 

 ing against occasional specially-adapted foes. It is possible 

 that these considerations may enable us to understand why 

 it is that certain Bhynchopliora are remarkable among Cole- 

 optera for combining a cryptic colouring with sufficient 

 iuimunity to render them feasible models for mimicry. 

 The ordinary methods of active defence among verte- 

 brates — the power of biting or pecking, of kicking or 

 tearing with hoofs or claws — together with the passive 

 resistance of a spiny or hard external covering, are 

 almost invariably associated with cryptic colouring and 

 modes of life favouring concealment. The probable ex- 

 planation is that all such methods of defence must fail before 

 large and important classes of still stronger enemies or 

 foes with cunning sufficient to circumvent the passive 



