8 BULLETIN 826, U. S. DEPARTMENT OF AGRICULTURE. 



same line of development as the xiphidina but developing these spec- 

 ialized spines during the same period in which the wings and antennae 

 have become reduced. Still another subtribe, the Pentalonina, 

 shows a very peculiar wing venation. This is not so much a primitive 

 wing as a more specialized one. It is placed, therefore, as one of the 

 highest subtribes of the Aphidini. 



There remains yet for discussion the tribe Setaphidini. This, it 

 seems evident, belongs with the Aphidinae. In regard to the an- 

 tennae and the wings it is quite highly specialized but in regard to the 

 cornicles, cauda, and anal plate this statement can not be made. The 

 natural position of this tribe is somewhat doubtful. Its ancestors 

 evidently separated from the aphidian line before the prominent corn- 

 icles of the Aphidina, Macrosiphina, etc., appeared and yet the species 

 are more specialized in many ways than are members of those sub- 

 tribes. It would appear that the lines separated after that of the 

 Greenideini, for the cornicles are not hairy. Yet this separation must 

 have taken place a considerable time before that of the Aphidina and 

 Macrosiphina. The tribe is placed, therefore, as indicated in the 

 diagram (fig. 1). 



The subfamilies, other than the Aphidinae, include the most spec- 

 ialized members of the family. By far the most primitive of these 

 subfamilies is the Mindarinae. This subfamily, as has been indicated, 

 is a remnant from the past, giving some idea of the ancestors of the 

 Eriosomatinae and the Hormaphidinae. The wing structure is partic- 

 ularly worthy of study. The wing of no other living aphid is like it, 

 but this peculiar structure is abundantly met with in fossil forms. 

 The media, it is true, is more reduced than in certain members of the 

 Aphidinae, but this is of very little importance as compared with the 

 wing's peculiar structure. The form also feeds upon conifers and this 

 is undoubtedly a primitive habit. The cauda and anal plate are 

 unlike those met with either in the Eriosomatinae or the Hormaph- 

 idinae. 



The sexual forms are interesting. They have become sufficiently 

 specialized toward the Eriosomatinae to have lost the wings, but they 

 retain the beak, at least in most individuals, and feed. The ovaries 

 of the oviparous female also are developed so that a number of eggs 

 are laid. 



The two remaining subfamilies are the most highly specialized of all 

 aphids. 



The Eriosomatinae are, in many ways more specialized than the 

 Hormapliidinae, but in other ways they are more primitive. The 

 whole Eriosoma line separates at once on the sexual forms. These 

 are small, apterous, and beakless. Throughout their life they take 

 no nourishment, and the ovaries of the oviparous female become 

 atrophied, so that only one develops and of the eggs therein only one 



