BY R. J. TILLYARD. 297 



lobe of the hindwing is generally fairly large in these forms, thus 

 allowing a greater area of the hindwing to pass under the fore 

 than would be the case if it were normal in size. Thus it will 

 be seen that the type of coupling here represented is that which 

 I have already termed, in the Megaloptera and Trichoptera, 

 amplexiform. I propose to use this term to distinguish these 

 groups from those possessing a frenulum, without in any way 

 thereby indicating a belief that the amplexiform groups in the 

 Lepidoptera constitute a single line of descent. For it must be 

 evident that thei'e is no reason why the frenulum may not have 

 been lost independently along several distinct lines of descent, 

 leaving us at the present time with several isolated amplexiform 

 groups, each of which has its nearest relationship, not with the 

 other Amplexiformia, but with a different frenate group. 



As an illustration of this, we find, in the Ca^tiiiidce, a family 

 in which both the frenulum and the humeral lol)e are well de- 

 veloped. From such a group, either a typical Frenate group or 

 a typical Amplexiform group might be descended; the former 

 b}^ reduction of the humeral lobe, the latter by loss of the 

 frenulum. 



Considering that the condition of the wing-venation is of 

 greater importance than the structure of the wing coupling appa- 

 ratus, for the purposes of classification, I have already, in a pre- 

 liminary leport on this research (13), suggested that the primar}^ 

 division of the I^epidoptei-a into two Suborders should be based 

 upon the former instead of the latter. I therefore defined two 

 Suborders as follows: — - 



i. Suborder HOAIONEURA. 

 Venation of fore- and hind wings closel}- similar and of primi- 

 tive design. (Families Microjderyyidtt {senst. /at.), I'rototheoridcf', 

 and Hepialidce). 



ii. Suborder HETERONKURA, 

 Venation of hindwing reduced, and difi'ering widely from that 

 of the forewing. (All the other families). 



The reason for discarding the older terms Jugata and Frenata, 



