NO. 1310. NORTH AMERICAN TJIYSANOPTERA— HINDS. 105 



the wings may be fully developed, reduced to short pads not reaching 

 beyond the thorax, or even entirely absent. Intermediate conditions 

 are rare, though I have found a few specimens in which the wings 

 were about one-half their normal length and cntircl}- functionless. 

 These three conditions may occur even in the same species {C'hiro- 

 thrips inmiicatm Haliday). When the wings are reduced, the little 

 pads are rounded or oval in shape and are laid closely upon the 

 thorax. The fore pad is larger, bears a few small spines, and covers 

 the spineless hind pad completely. No fringes are present, but the 

 fore pad has a distinct scale. Trybom, who has made quite an exten- 

 sive stud}^ of this subject (425), recognizes eight classes into which 

 these insects may be divided according to the varying conditions of 

 the wings. 



1. Both sexes entirely wingless. 



2. Males and some of the females wingless. 



3. Males entirely wingless, but females with normally developed 

 wings. 



4. Long winged and wingless individuals of l)oth sexes occur. 



5. Males and a majority of females with reduced, but a number of 

 females with normally developed wings. 



0. Both sexes always short winged, 



T. Long winged as well as short winged individuals of })oth sexes 

 occur. 



8. Both sexes always long winged. 



The appearance of a long winged generation following several 

 which have short wings is strongly suggestive of a similar condition 

 among the Aphidfe. In at least some species of Thysanoptera where 

 this condition obtains the summer generations develop long wings 

 while the fall generations are almost entirely short winged, so that 

 nearly all the hibernating females have onl}^ wing pads. Long and 

 short winged forms commonly alternate in the same sex, but short 

 winged and entirely wingless forms of the same sex are not kno^vn. 

 When only one sex is wingless it is the male. Wing pads are usuall}^ 

 rather difficult to see, but their presence or absence can be deduced 

 from the structure of the thorax, even though they are themselves 

 invisible. 



ABDOMEN. 



The form of the abdomen varies from cylindrical to elongate-ovoid. 

 In Terebrantia the segments are nearly C3'lindrical in cross section, 

 while in Tubulifera the abdomen is flattened, giving the cross section 

 an elliptical outline. The terminal segments especially are difl^er- 

 ently formed and characteristic of the suborders. The abdomen is 

 always composed of ten segments, of which the second to the seventh, 

 inclusive, are similarly formed in nearly all cases, while the others are 

 varial)lc and bear the most distinctive characters of the abdomen. 



