MARSHALL'S GREAT MIMETIC SERIES 121 



bility is widespread) ; three moths (Arctiidae and 

 Zygaenidae, distasteful families) ; one fly. In fact 

 the whole combination, except perhaps one Phyto- 

 phagous, one Coprid and the Longicorn beetles, 

 and the fly, fall under the hypothesis of Mtiller 

 and not under that of Bates. And it is very 

 doubtful whether these exceptions will be sus- 

 tained : indeed the suspicion of unpalatability 

 already besets the mimetic Longicorns, and is 

 always on the heels I should say the hind tarsi 

 of a Phytophagous beetle. 



This most remarkable example which illustrates 

 so well the problem of Mimicry and the alterna- 

 tive hypotheses proposed for its solution, was, 

 as I have said, first described in 1902. Among 

 the most perfect of the mimetic resemblances 

 in it is that between the Longicorn beetle, Amphi- 

 desmus analis, and the Lycidae. It was with the 

 utmost astonishment and pleasure that I found 

 this very resemblance had almost certainly been 

 observed by Burchell. A specimen of the Amphi- 

 desmus exists in his collection and it bears 1 651 '. 

 Turning to the same number in the African 

 catalogue we find that the beetle is correctly 

 placed among the Longicorns, that it was cap- 

 tured at Uitenhage on Nov. 18, 1813, and that 

 it was found associated with Lycid beetles in 

 flowers (' consocians cum Lycis 78-87 in floribus '). 

 Looking up Nos. 78-87 in the collection and 

 catalogue, three species of Lycidae are found, all 

 captured on Nov. 18, 1813, at Uitenhage. Bur- 



