244 [September 



teristic of Hymenoptera than of Diptera. I know no Dipteron that 

 comes anywhere near the very elongate and attenuate form of the 

 Hymenopterous Pelecinus polycerator $ Drury. 



I do not wish to be understood here as doubting or denying the fact, 

 of there being often a very striking resemblance between insects be- 

 longrinor to different Orders and different families of the same Order, 

 but only the assumption that is made, that of two similar forms, A and 

 B, it is B that imitates A, not A that imitates B, and the inference 

 drawn therefrom, that the group to which A belongs is superior to that 

 to which B belongs. Because an ^Ejeria is named hombiformis, it is 

 concluded that it is JSgeria that imitates Bombus, not Bombus that 

 imitates ^geria; but when a Dipterous genus is named Bittacomorpha 

 from the Neuropterous genus Blttacus, the corresponding conclusion 

 that the Dipteron imitates the Neuropteron is passed over in silence. 

 The Dipterous Toxophora segeriifonnis Westw., as its name denotes, 

 imitates an ^gen'a ; but according to Dana's theory, it must be the 

 ^ge.ria that imitates the Toxophora^ not the Toxophora that imitates 

 the jEgeria. All that we can safely say in this matter is what Latreille 

 said long ag-o, viz. that " Nature seems to work after a certain limited 

 number of patterns, which she reproduces with modifications in widely 

 distinct classes and orders;" (quoted Westw. Intr. I. p. 326, note); in 

 other words, to drop all metaphorical language, that there appears to 

 be a genetic connection between widely removed species of the same 

 subkingdom. That this iteration of peculiar types is sometimes con- 

 fined to single species, is proved by the fact of the long, uniquely- 

 shaped tail in the hind wings of a North American moth, Attacus Inna 

 Linn., being exactly reproduced in the hind wing of a North American 

 butterfly, Hesperia (^gonilaba) proteus Godart, the other Attacus hav- 

 ing no vestiges of any tail and the other Goniloba having only a short 

 rudimental one. The great truth, which was foreshadowed years ago by 

 the illustrious French entomologist, is also deducible from a fact which 

 Prof Dana has well insisted on, viz. that in the several Classes and 

 Orders of Annulata there exist definite limits of size, within which 

 each is confined, and which differ materially in the different Classes 

 and in the different Orders. Still more obvious is this law in the case 

 of the inferior subdivisions, such as families, subfamilies and genera ; 

 and the lower down in the series we go, or in other words the closer the 



