Cooke : Ox the Classification of Insects. 



69 



same principle in operation, which seems to have been the guiding 

 one in the arrangement of the Coleoptera. The British butterflies are 

 so few in number of species, that we have nothing to compare with 

 the magnificent genus Ornithoptera of Java, Sumatra, and Borneo, 

 which takes the lead in a general collection. 



5. Teichopteea, — This order is intended to contain the insects 

 commonly known as caddis-flies only — in fact the species monographed 

 by M'Lachlan under this title. He has recently revised, and I think 

 improved, the arrangement of the British species, beginning with the 

 genus Neuronia and ending with the minute species, that is to say the 

 family Hydroptilidae. It would be difficult now to find any departure 

 from the principle of classification advocated in this paper. 



6. Stegoptera. — Xewman includes the caddis-flies alons; with the 

 Neuroptera-Planipennia under the title proposed. If it be permitted 

 to restrict the name to the latter insects as a separate order, the title is 

 surely worth retaining in preference to Xeuroptera-planipennia, which 

 was proposed only as a division of the Linucean order. The separation 

 of these insects into a distinct order is not a proposal of mine ; I only 

 advocate its adoption. M'Lachlan has monographed the British 

 species, but if we are to deal with them on the same principle as with 

 the Trichoptera, the arrangement will want revising. 



This order terminates the first division of the class Insecta, and I 

 think there may be observed a regular gradation from a hard-bodied, 

 strong, and well-defended class of insects, to a soft-bodied, weak, and 

 defenceless one. 



Division II. Order 1. — Neuroptera Odoxata. — It will be 

 observed that I propose to separate the Xeuroptera of Linnisus into 

 four distinct orders— two belonging to the first division of the Insecta, 

 and two to the second, according 'to the nature of the metamorphosis. 

 Fabricius proposed the Odonata for the dragon-flies, and I do not like 

 to drop the title of Neuroptera, as it is done in the two orders in the 

 first division : hence the combination. I believe, also, that Fabricius 

 considered the Odonata to form a distinct order, therefore there is 

 nothing new in the proposition which I make ; it is only an attempt 

 to restore that which has been allowed to drop. The reason for so 

 doing is this : it has become very evident of late years that the sexual 

 characters of insects, the genitalia and their armature including the 

 anal appendages, are of high value in classification, and also as a test 

 in the distinction of species. Now, the male genitalia of the dragon- 

 flies differ from all other insects, so far as I am aware, in being 

 situated, not at the tip of the abdomen, but at its base underneath- 



