ANTEXNiE OF LEPIDOPTERA. 37 



Sesiids and Sphinxes the resemblances are more than superficial ; 

 they are as marked in the microscopic as in the macroscopic charac- 

 ters, and it is difficult to believe that such forms could arise unless 

 there was some genetic relationship. 



The Specialized MACROFRENATiE. 



The Specialized ]Macrofrenata3 includes " certain moths and all 

 skippers and butterflies. In these insects the anal area of tlie hind 

 wing is reduced, containing only one or two anal veins." This divi- 

 sion is again divided into two subdivisions : the Frenuluin-conservers 

 and the Frenulum-losers. The first subdivision contains those moths 

 in which the frenulum has been retained, while the second contains 

 forms " in which the frenulum has been supplanted by a greatly 

 extended humeral area of the hind wings." In a few members of 

 the first group we find no frenulum, l)ut it has evidently been lost 

 late in the life of the species, as closely allied specialized forms still 

 retain it. "Among the Frenulum-losers," on the other hand, "the 

 loss of the frenulum occurs while the race is still in a very general- 

 ized condition, no trace of a frenulum being found among these in- 

 sects, except a rudiment in the inost generalized forms {Bombyx, 

 Cidnmis).'' Thus the two subdivisions really represent two distinct 

 lines of descent v/hich separated far back in the history of the races, 

 and are not simply arbitrary groups based on the presence or absence 

 of a single character, as has been understood by some authors. 



The Frexulum-coxservers. — Among the frenulum-conservers 

 we find moths with highly specialized antennae From the very fact 

 of their great develo})mcnt we might fairly expect to find consider- 

 able variation even in small groups ; and such is the case. A feature 

 whicli is characteristic of the whole group is the specialized condition 

 of the chitinous surface of the clavola. The perpendicular planes 

 on the surface are not everywhere continuous, and usually the gen- 

 eral surface slopes up to the top of the plane on one side and pre- 

 sents an appearance such as is represented in Figs. 87 or 33. In the 

 filiform antenme of each family there is not so much difficulty in 

 finding characteristic structures, but when the highly pectinate forms 

 are studied the problem becomes more difficult. The greatest varia- 

 tion occurs among the Geometrina. There we find somr forms 

 scarcely more specialized than many of the jNIicrofrenuta? ; and 

 others, as highly specialized as any of the Frenulum-conservers. 

 The following table will serve to separate the antennae of this group 



TRANS. AM. ENT. SOC. XXIII. FEBRUARY. 1896. 



