BY R. J. TILLYARD. 233 



remaining Belmont fossils might reasonably be postponed to a 

 later date, although they were discovered earlier than the one 

 here dealt with. 



The characters of the wing ai'e such that it cannot be placed 

 in any known Order, either fossil or recent It is undoubtedly 

 allied to the Mecoptera and Protomecoptera, on the one hand, 

 and to the Trichoptera and Lepidoptera on the other. Its re- 

 lationships with the first two Orders mentioned are collateral, as 

 will be clearly seen when we come to compare it with Permo- 

 chorista from the same Beds. With the Trichoptera and Lepi- 

 doptera, on the other hand, its relationships are definitely 

 ancestral; a comparison of the fossil wing with Rhyacophila and 

 Micropteryx will show quite clearly that both these types can be 

 directly derived from the fossil by reduction, without a single 

 discordant character. That being so, we are now able to state 

 definitely, not only that the Mecoptera (and, of course, the Tri- 

 assic Protomecoptera also, though not yet found in Palaeozoic 

 strata) were in existence in Upper Permian times, but that the 

 Trichoptera and Lepidoptera were not then differentiated, being 

 represented by a common ancestral type closely allied to the 

 Mecoptera, such as we find in this fossil. 



The above relationships have suggested to me the name Para- 

 mecoptera for the new Order. Not only does this name indicate 

 the collateral evolution of the new type alongside the Mecoptera, 

 but it also allows future entomologists, should they so desire, to 

 merge the new Order into the Mecoptera, together with the Pro- 

 tomecoptera; so that a single Order Mecoptera, with its char- 

 acters defined on a wider basis, may one day be made to include 

 the three Suborders Paramecoptera (Permian), Protomecoptera 

 (Triassic;, and Eumecoptera (fossil and recent), these last being 

 the true Mecoptera, or Scorpion-flies, as at present understood. 

 For the present, however, the difficult task of the Phylogenist 

 in these groups can only be undertaken with clearness if these 

 new types, about which we know so little, are given distinct 

 ordinal rank, thus marking out clearly, and without any ambi- 

 guity, the venational characters of each separate type within the 

 complex of the Panorpoid Orders. That being so, I propose to 



