588 PliOF. E. B. POULTOW : NATL'^fiAL SELECTION 



mimetic resemblance is carried, and tlie very different methods 

 employed, in a nearly related set of insects. 



The Hymenoptera, which form the models of these insects, 

 have not beeQ worked out in many of the instances. In the case 

 of JEJsthesis ferruffineus, however, two Aculeate Hymenoptera 

 of very different form, although belonging to the same large 

 group — the JEumenidce, which inhabit the same region (Australia), 

 have a colouring and pattern which are superficially very 

 similar to those which the Longicorn has attained; while a 

 Dipterous insect, an Australian species of Dasypogon (^Asilidce), 

 also appears to fall into the group. The appearance of the 

 three latter insects is represented in Plate 41, the dipterous insect 

 being shown in fig. 5 a, the wasps in fig. 5 b, Ahispa australis 

 (Smith), and fig. 5 c, Eumenes LatreiUei (de Sauss.). These three 

 figures were drawn from specimens in the British Museum, kind 

 permiission Laving been accorded to me to figure them for the 

 purpose of this memoir. Puture study will j)robably add many 

 species to this very characteristic group. 



We thus find that wasps and allied forms are resembled by 

 species of many groups of insects, and the resemblance is attained 

 in all kinds of different ways. 



The numerous mimetic resemblancestothe aggressive, abundant, 

 and well-defended ants supply an even better illustration. In 

 the majority of cases the whole body of the mimetic form is 

 moulded from the ancestral shape, which is still exhibited by 

 its non-mimetic allies, into that which is characteristic of an ant. 

 In some groups this means a large amount of alteration, in 

 others less. In this case, too, the resemblance extends to species 

 which are altogether outside the Insecta ; for many small species 

 of spiders closely mimic ants. In the family of Attidce alone, 

 and such resemblances occur in several other families of spiders, 

 Greorge W. and Elizabeth Gr. Peckham state (Occasional Pape^rs 

 of the Natural History Society of Wisconsin, vol. ii.j 1892, 

 Milwaukee ; see also a paper by the same authors in vol. i., 

 1889) that about a hundred ant-like species are known froin 

 various parts of the world, and that they are " very much more 

 numerous in South America and in the Malay Archipelago than 

 in any other countries," viz. in the very countries in which other 

 examples of mimicry are especially abundant. The spider with 

 its two-fold division of body is often made to assume the appear- 

 ance of an ant, with its three-fold division, by a constriction which 



