7226 Notices of New Books. 



perfection a drawing was made of them. The same individual which 

 had been submitted to the draughtsman was then separately confined, 

 watched with the most diligent care, and, as soon as it had passed 

 into the state of a chrysalis, again made the object of the pencil. A 

 determinate number was carefully attached to the drawing and to the 

 cage of the chrysalis. As soon as the perfect insect had appeared 

 and expanded its wings it was secured, set, and numbered in accord- 

 ance with the larva and chrysalis. During this period every possible 

 solicitude was employed to prevent mistakes. The original series, con- 

 sisting of the perfect insects and the chrysalides obtained by this mode 

 of proceeding, and numbered in accordance with the collection of 

 drawings made at the same time, is now deposited in the Museum of 

 the Honourable East India Company, and affords an authentic docu- 

 ment of the accuracy of the details regarding the metamorphoses of 

 Javanese Lepidoptera, which will be offered in the course of this 

 work." 



Dr. Horsfield does not claim the merit of being the first to consider 

 the larva as important in the natural classification of Lepidoptera, but 

 he is the earliest English author who calls our attention to the invaluable 

 work of Denis and Schiffermuller, whose maxim and motto of " One 

 eye to the butterfly and another to the caterpillar " he cites with the 

 cordial approbation it deserves. 



It does not, however, seem to me that either the authors of the now 

 celebrated Vienna Catalogue, or Dr. Horsfield himself, carried out the 

 obviously important principle which both profess to admire. Denis 

 and Schiffermuller commence by adopting the Linnean primary groups 

 as final ; and Dr. Horsfield by restricting his divisions to five, in 

 accordance with MacLeay's singularly clever but most artificial hypo- 

 thesis, thus grouping the whole of the Lepidoptera into Papilionidae, 

 Sphingidse, Bombycidae, Noctuidae and Phalaenidae, and entirely ignoring 

 as divisions such groups as Geometra, Pyralis, Pterophorus, Tortrix, 

 Tinea and Psyche. Neither is the group Papilionidae associated by any 

 common character of the larvae. Tlien tlie work before us makes no 

 attempt whatever to pursue the hypothetical arrangement here super- 

 ficially indicated. The accomplished author of the idea is now no more, 

 and to other hands has been allotted the important task of working 

 out the details, which, I say it without any feeling of disrespect towards 

 the founder, are now unfettered by any attempt to force nature into 

 that harness which none of us can compel her to wear. 



The first volume, already noticed, treats of the Papilionidae and 

 Sphingidae : the second, now before us, of the third tribe under the 



