UY E. .T. TILLYABD. 245 



Summing up the above evidence, it is quite clear that there is 

 not a single character in the venation of Belmontia which pre- 

 cludes its being accepted as the actual ancestor of Rhyacophila, 

 and of the whole Order Trichoptera, as now constituted. I do 

 not wish, however, to claim that this newly discovered genus 

 Belmontia is itself the actual ancestor of that Order; it is quite 

 sufficient to state that it is the first known representative of a nerv 

 Order Paramecoptera from ivhich the Trichoptera are undoubtedly 

 derived. I do not think that the Trichoptera originated in or 

 near Australia; most certainly the Rhyacophilidce themselves 

 did not. That being so, we must think of the Paramecoptera 

 either as having existed in other parts of the world also, in 

 Upper Permian times, or as having spread thither from Australia 

 during the Lower Trias. It would then be from some more 

 specialised and reduced type, within the Order, that the true 

 Trichoptera must have been actually derived. 



(ii.) Affinity with the Lepidoptera. 



Let us now turn our attention to the Lepidoptera Homoneura, 

 and compare Belmontia with one of the Micropterygidce. For 

 this purpose, I shall figure the pupal iving-tracheation as well as 

 the imaginal venation of the forewing of the genus Eriocrania. 

 (Owing to the reduction of the pupal wing-tracheation in all 

 Trichoptera, this comparison could not be made in the case of 

 Rhyacophila). This shows that the condition of the cubitus and 

 anal veins, together with the position of the arculus, is closely 

 similar in this pupal wing and in the fossil. In the region of 

 M, of course, Eriocrania, like all other Lepidoptera, lacks a 

 separate M 4 , and is thus more highly reduced than Rhyacophila. 

 The loss of one of the four usual branches of Rs is peculiar to 

 Eriocrania; the missing branch is present in its ally Mnemonica, 

 as well as in Sabatinca and Micropteryx. The forking of Sc into 

 Scj and Sc 2 is retained in Sabatinca, Micropteryx, and Mnemonica, 

 while Ri also is forked in the first and last of these three genera. 

 The radial cell is closed distally by the cross-vein ir in Sabatinca 

 and Micropteryx, and the primitive dichotomic branching of Rs 

 is preserved in both these genera. In the other genera of the 



