BY R. J. TILLYARD. 



629 



responding venation of a portion of the costal area of the fore- 

 wing of PsycJiopsis degans Guerin, (Text-fig. 18), in which this 

 difference is very clearly shown. It will be seen that the branches 

 passing into the costal area from the main subcostal vein (Sc) 

 are all of the nature of opinleU {ci'l), since they are all preceded 

 by trachefe. These veinlets are, however, supported by true 

 eross-vpins {x), which have arisen independently of the trachea- 

 tion. Also, the narrow areas between tSc, R, and Rs are 

 strengthened by true crogs-i'plns (a*). 



In the known fossils of the Psychopsidcv and allies, we find the 

 costal veinlets present as in recent forms. I^ut there are no 

 connecting cross-veins. Hence it is evident that these latter are 

 a later development, called into existence by the need for 

 strengthening the enlarged costal area of the wing. 



-^^<^^ 



Text-fig. 18. 

 Portion of enlarged costal area of fore wing of P.^ychopsis efer/an-'^ Guer., 

 ( x27), with the corresponding tracbeation (to the left) of the pupal 

 wing, ( X 60). C, costa; cf, trachet^ preceding the costal veinlets; 

 crl, costal veinlets; R, radius; Rs, radial sector; Sc, subcosta. 



Referrins: aoain to Text fio. 18, we notice a further difference 

 between the veinlets and the cross-veins. The former, like the 

 main veins of which they are branchlets, carry a regulai* 

 series of macrotrichia, ^\'hereas the latter have no macrotrichia 

 at all. This distinction holds throughout the whole of the 

 Holometabola, with only two exceptions, viz., the Raphidioidea 

 49 



