OF WASHINGTON, VOLUME XVIII, 1916 159 



behind this chitinization is added indication that they do not be- 

 long with the Anterodorsals. 



The practically constant proportional differences in length of 

 the epicranial setae in all but the most specialized leaf-mining 

 species coupled with their arrangement into easily definable 

 groups, make it possible for us to homologize the setae and to 

 correlate even the most trifling differences in their positions. 



An important fact which is not brought out by Dampf, is that 

 changes in position of the setae are due to modifications by growth 

 or contraction of the chitinous areas upon which they occur. 

 This is shown by the fact that the setae of any given group always 

 remain in that group, and by the relative position of the groups 

 themselves under obviously different and easily recognized changes 

 in the head areas. For example, where there is an enlargement 

 of the head surface there is not a corresponding spreading out of 

 all the head setae, but only of those comprising the Posterodorsal 

 group. We find also in several normally round heads (among 

 most of the free-feeding Micros) a crowding forward of the setae, 

 indicating an enlargement of the posterior part of the head at the 

 expense of the frontal area. The most numerous changes in fact 

 are due to modifications of the epicranium back of the area occu- 

 pied by the Anterodorsal, Ocellar and Subocellar setae. And 

 consequently, changes in the relative position of the Posterodorsal 

 and Lateral setae are more frequent and striking than among the 

 other groups. In the round feeding larvae of the Lithocolletinae, 

 however, the development is more in the opposite direction, the 

 Anterodorsal, Ocellar, and Subocellar areas are larger in propor- 

 tion to the Posterodorsal, and there is a consequent spreading out 

 of the setae of the former, the Lateral group is thrust further 

 back and the Posterodorsal group is restricted to a smaller area. 



All changes, however, are not confined to group movements. 

 There are also differences in the relative position of the setae 

 within a given group, chiefly differences in distance, but also, 

 among the Anterodorsals especially, in the alignment of the setae. 

 In both the most striking changes are noted among the genera of 

 any given family, but here again we find certain tendencies which 

 aid in the identification of larger groups; for example, Ad-2 and 

 Ad-3 are rather closely approximate in Stenomidae or Oecopho- 

 ridae, and in the Tortricoidea the three Anterodorsal setae form 

 a very obtuse triangle often, in fact, lying in almost a straight line 

 with L-l, while in other families their alignment tends more to- 

 wards a right angle. 



The positions of the punctures appear at first hand much less 

 constant within families than those of the setae, and offer greater 

 difficulties, for they are often invisible except under the highest 



