16G Dr. A. Jefferis Turner's Observations on 



bar between 8 and the end of the cell. In the fore- 

 wing there is a small median cell, but no trace of an 

 areole, G, 7, 8, 9 are stalked, and the anal veins anastomose, 

 16 running into lc. The tibial spurs are obsolete ; the 

 palpi moderate, porrect ; the frenulum well developed, and 

 the $ antennae bipectinate to apex. 



This concludes my jresent study of the neuration of the 

 Cossidae. I have not attempted to figure every genus, 

 but, so far as I know, I have not omitted any important 

 deviation of structure. As a result I have convinced myself 

 that this is a natural and compact family not divisible 

 naturally even into subfamilies. There is, it is true, a 

 considerable and very interesting degree of variation in 

 several directions, but all these lines of development are 

 linked together by forms of intermediate structure. 



I have also convinced myself that the Cossidae have 

 retained the most ancient form of neuration among the 

 existing families of the Lepidoptera Heteroneura, and that 

 from this neuration that of all the other families can be 

 easily derived by a process of reduction, the stages of 

 which can be readily traced. The study of this family 

 has therefore appeared to be of fundamental importance, 

 and it is this that has encouraged me to consider it in 

 detail. But to establish my conclusion it is necessary to 

 study also the neuration of these other groups, more 

 particularly of those families that may be considered of 

 primitive type, or at least to contain genera of primitive 

 type. Naturally this survey cannot be undertaken in an 

 exhaustive manner within the limits of a short paper. 

 I can do no more than select one or two of the more 

 primitive genera in the case of each family, paying par- 

 ticular attention to those families which agree with the 

 Cossidae in the primitive character of possessing three anal 

 veins, la, lb, and lc, in the hind-wing. These families are 

 the Torlricidae, Tineidae, Castniadae, Zygaenidae, I/ima- 

 codidae, Psychidae, and Pyralidae. 



The Classification of the Lepidoptera. 



Before considering the relationship of the Cossidae to 

 other families it is necessary to make a few remarks on 

 the classification of the Lepidoptera. It is now generally 

 agreed that the primary division of this order is into two 

 very unequal groups, (a) those with closely similar fore- 



