Dr. Horsfi eld's Lepidopierous Insects of India. 119 



far as regards the natural arrangement which it proposes, rests mainly on 

 the certainty with which the larvae can be referred to the perfect insects de- 

 scribed in it, he states with great particularity the precautions used in 

 taking coloured drawings of each of the caterpillars and of its chrysalis in 

 succession, which were numbered in accordance with a ticket affixed to 

 the imago of the same individual. He thus shows the almost impossibility 

 of any mistake having occurred on this most important point, and es- 

 tablishes the authenticity of his materials : a fact the more essential as on 

 them he has chiefly to rely, but little assistance being to be derived from 

 published works in which information is contained relative to the meta- 

 morphoses of exotic Lepidoptera. After noticing the principal of these, 

 he adverts with great and merited commendation to the JViener Verzeich- 

 niss, a work which, although published upwards of half a century since, 

 anticipated by full that period the discoveries of modern times; the 

 families of the Linnean Papiliones characterized in its pages exhibiting 

 not merely the rudiments of, but the actual, genera into which that very 

 extensive group is at the present moment divided. The leading divisions 

 of that work are derived chiefly from the larvse, and accord almost pre- 

 cisely with those which Dr. Horsfield had proposed to himself, previously 

 to his becoming acquainted with this excellent production of his prede- 

 cessors the " T/ieresja?ter," Messrs. Dennis and SchieffermuUer. 



Of the primary sections, or tribes, into which Dr. Horsfield regards 

 the Lepidoptera as naturally divisible, no characters are proposed at the 

 commencement of the exposition of the arrangement; they are merely 

 enumerated as the Papilionidce, Sphingidce, Bomhijcidcp, J^'octuida and 

 Phal(Bnida, Each of these tribes is then made to undergo a brief review, 

 with the object of exhibiting the sections, or stirpes, of which they are 

 respectively composed. As the stirpes of PapiUonide are noticed in a 

 more detailed manner in an after portion of the work, they may be 

 here passed over, until a rapid sketch has been given of those of the other 

 tribes. 



Among the SphingidcB the larvae are comparatively well-known ; 

 hence there is but little difficulty in recognising the existence of the fol- 

 lowing five types of form: 1. Larva vermiform, sluggish, somewhat 

 hairy, with a small retractile head, emd minute obscure feet; of this 

 stirps the genus Zt/gcBna may be regarded as the type : 2, Larva cylin- 



