ANTENNvE OF LEPIDOPTERA. 25 



tennal surface of a male Pyramta oxydalis, and shows the great 

 regularity of the areas so characteristic of this group of moths. 

 Among the TortricicUe we find that in most cases the areas become 

 elongated. Fig. 35, a portion of the antcMina of a male Dlcliella 

 sulfureana, shows this condition. This elongation of the areas is 

 carried much farther among the Tineids. Tinea dorsutrigella, Y'xg. 

 36, shows this extreme lengthening of cells and straightening out 

 of the cross-lines ; yet here we still find the plates continuous. The 

 function of these plates is doubtless to strengthen the thin layer of 

 chitin and give it greater rigidity. In those forms which have long 

 slender pectinations we find the longitudinal plates nnich thicker 

 and heavier, while the transverse ones are less developed and have a 

 tendency to straighten out or to disa})])ear. This is evidently better 

 adapted than the hexagonal arrangement to stiffen these slender 

 forms. Fig. 30 represents the dorsal aspect of a pectination of a 

 male Notolophus leueosfigma, and shows at the sides some of the 

 plates in j)rofile, while the accentuation of the longitudinal plates is 

 clearly seen in the middle of the figure. Note also that the longi- 

 tudinal plates extend slightly beyond the end of the pectination. 

 One cannot help being struck with the resemblance between the striae 

 or ribs of the scales and these thicker ridges on the chitin surface. 

 On the ventral aspect of the same pectination we find a modification 

 somewhat different, but yet along the same line. The abundance of 

 the sense-hairs and the other organs makes a lighter and more deli- 

 cate surface desirable, and here the transverse plates are either want- 

 ing entirely or transformed into longitudinal ones. In fact, we may 

 state it as a general rule that the development of chitin on the ven- 

 tral and sensitive surface is not carried so far as elsewhere. Fig. 31 

 represents this condition. In all these forms the general surface is 

 approximately level and the plates represent the only elevations from 

 the surface. Now we come to another development. F'ig. 32 re[)- 

 resents a portion of the dorsal surface of the antenna of a nuile 

 Automerls to; ab is the niesal line of the shaft, and the part to the 

 left is the extension of the surface upon the pectination. It will be 

 seen that the portion near the meson resembles the condition found 

 in Notolophm lencodlgma, with this difference however : in AutonierU 

 io the general surface, instead of being level, rises on one side of 

 each plate so as to make each area constitute a single slope with its 

 crest at the plate on the distal edge. The surface of the sloping area 

 is divided up into ridges with what was originally a transverse plate 



TRANS. AM. KNT. SOC. XXIII. (4) FEBRUARY, 1896. 



