THE REVISION OP' THE SPHINGIDES. 45 



spirits in lepidopterologj^ in shifting the term " family " to a lower 

 category than that to which it was originally applied. We do not see 

 what good it serves, etc." They divide the Sphiuf/idae into (A) 

 SpJiiniiiilof asL'iitanopJiorae, and (B) Spliiwiidae aeiiianojiliorae. These 

 are two most real and natural divisions, but they pretend to be 

 neither subfamilies, families, nor super-families. By precisely the 

 same device of giving it its full title, Spldniiidae-^finannphorae- 

 chocrocainpinae, we may pretend to deprive the Choerocainpinae of any 

 right to be a subfamily, and call it B2. That Semanophorae has not a 

 recognised ending is a quibble. Keally we have : — 



Superfamily : Sphingides. 

 Family : Semanophorid.e. 

 Subfamily : Chceeoc ampins. 



The systematic portion of the Bevision is, of course, the largest 

 and principal part of the work. We have already said that broadly 

 we thoroughly agree with it, and where, in details, we do not, the 

 points are such as to bear different interpretations, especially if 

 approached from different aspects. In dealing with a somewhat 

 isolated group like the Sphinges, and in endeavouring to arrange them 

 phylogenetically, the temptation is strong to find earlier and later 

 forms, and arrange them as if descended from one another in the 

 manner in which genealogists arrange human pedigrees in a tree, 

 without making proper allowance for the fact that, in the human 

 trees, only the leaflets represent still existing forms. In a tree of the 

 Sphinges, or other group, it is erroneous to take, say, an oak, and put 

 some species as representing the trunk, others the branches, and others 

 the twigs and leaves. All represent leaves. Better than an oak-tree 

 would be a box, yew, or laurel, trimmed as we sometimes see them 

 into plum-pudding form ; each species is a leaf or twig on this. We 

 realise on this more clearly the great difficulty there is in saying 

 whether two twigs side by side are sprung from a single twig which 

 was a leaf-bearing one last year, or whether we must follow each twig, 

 perhaps to the main stem, before we find a common origin. The 

 gardener's shears may represent the struggle for existence. Either of 

 two species, as we now have it, is as far from the common ancestor 

 (measured by generations) as the other, and, though we call it a more 

 generalised form, because it has varied less or m fewer particulars, it is 

 very probably, in some particulars, as specialised, or even more 

 specialised, than that we regard as highly evolved. No species then, 

 strictly speaking, represents the ancestor of another, all that can be 

 said is that it preserves more simply some or other feature of that 

 ancestor. All branches of the tree would be equally long, but some 

 would only have one leaf at the end, others would be very bushy, but 

 the leaves on it would all be terminal. That all present forms are of 

 equal age is not quite true, to assume one generation a year would be 

 correct for the Sphingides in the temperate regions, but in the warmer 

 areas there may be several, and it also happens that, in the tropics, 

 vitality is more abounding on the one hand and the environment more 

 varied and varying on the other. We might, therefore, expect to find 

 less specialised forms proportionally more numerous in the cooler 

 areas. This seems in accord with the Palaearctic or Nearctic distri- 

 bution of such forms as Sphinijulus, Smerinthus, Ceratomia, Dara- 

 2)sa, Proserpinus, and Deidawia. This is, as we shall revert to again. 



