162 THE ENTOMOLOGIST. 



as these would seem, to some extent at any rate, to recapitulate 

 the evolutionary liistory of species, genera, &c., or at least to 

 point the way to a partial reconstruction of such history. 



But putting all this aside, and granting for the sake of argu- 

 ment that this is the most important — nay, far the most im- 

 portant — of all structural differences, it is still liable to abuse ; 

 for it cannot for a moment be conceded (or even claimed) that 

 this is the only mode of divergence, and parallel differences in 

 this respect may have been evolved after other divergences had 

 taken place, the dormant tendency having been retained, and we 

 might thus find parallel groups exhibiting the same different 

 forms of clasp, &c., even when these groups had in other respects 

 diverged somewhat widely from each other; and, on the other 

 hand, likenesses, or even identity of structure, in this respect 

 might be retained, especially in closely related species, even after 

 they were specifically separated by differences of habit, food- 

 plant, habitat, &c. In the first case we should find two or more 

 groups of insects, different members of which appeared, judging 

 by the genitalia alone, to be more closely related to members of 

 the other groups than to those of their own ; and in the second 

 case we should find insects, obviously from other reasons specifi- 

 cally distinct, " lumped " together, as a single species, on the 

 ground that they were not yet differentiated in this one par- 

 ticular. We have in fact no reason to suppose either that 

 differentiation necessarily takes place simultaneously along 

 various lines, or that it follows a certain routine order in the 

 various lines along which it takes place ; on the contrary, all 

 such evidence as we at present possess would seem to point in an 

 opposite direction. A further consideration tending towards the 

 same conclusion is the fact that it is possible in this matter, as 

 in others, that a similarity or even an identity of structure may 

 sometimes be reached along different evolutionary lines. Just 

 as I have wished to give the greatest possible weight to the 

 arguments of those who have advanced the strongest claims as 

 to the importance of these structures — a weight far greater than 

 in point of fact I consider them capable of supporting — so here 

 I feel my position to be so strong that I can afford to minimize 

 the results that might fairly be made to rest upon the argu- 

 ments that I have used, and to say only that it is manifestly 

 unsafe to rest any part of a classification, whether generic or 

 specific, upon the likeness or unlikeness, the difference or identity, 

 of the genital armature alone. I should wish to forestall the 

 possible objection, that nobody in point of fact does found any 

 part of their classification on the testimony of these structures 

 alone, by answering that though they generally appear to be un- 

 conscious of doing so in theory, yet some of the ablest exponents 

 of the system certainly do so in practice, even if they call in 

 other considerations afterwards in support of their results. If 



