12 



than to difference in structure. The wing pads are folded back 

 upon the thorax, in which position the fore pads completely cover 

 the hind ones. The fore pads (Plate II., fig. 12) extend nearly 

 to the posterior edge of the thorax, and are marked by longitudinal 

 rows of short, stout spines, which vary in number. These spines 

 are wanting on the hind pads (Plate II., fig. 13), but both pairs 

 are covered with minute setae. 



About 98 per cent, of the adults which pass the winter are of 

 the wingless form, while from 90 to 95 per cent, of the first gener- 

 ation in the spring develop wings ; and the wingless form is far 

 outnumbered by the winged until late summer, when the propor- 

 tion declines, and in October only a small number, about the per- 

 centage that hibernates, of winged adults can be found. The 

 females continue to deposit their eggs, and the young larvje de- 

 velop and may be taken from the grass upon warm fall days until 

 snow covers the ground ; but only the adults survive the winter. 

 I have seen specimens revive after having been exposed to a 

 temperature of 21° below zero. At frequent intervals during 

 the winter I brought in grass from the field taken from beneath 

 snow, or entirely exposed, and in no case did I fail to find females 

 which became lively upon being kept a short time in a warm room. 

 Accidentally I found that these hibernating females could survive 

 for several days though submerged in a weak solution of potassium 

 hydrate, and they have even revived after being frozen solid in a 

 two per cent, solution of potassium hydrate ; but, so far as my 

 experiments went, freezing in pure water killed them. The hiber- 

 nated females become active in the spring as soon as the weather 

 is sufficiently warm to start the grass, and continue to deposit eggs 

 for from four to six weeks. The appearance of a number of 

 winged adults early in May marks the maturity of the first gener- 

 ation ; but, as the hibernated females are still active, there can be 

 no distinct line between the generations out of doors after this 

 time. Laboratory experiments show that there may be eight or 

 nine generations in a season. The length of the life cycle varies 

 from thirty days for the first generation to twelve days during the 

 heat of summer. 



I have sought in vain for the males of this species, for, although 

 I have mounted over a thousand specimens, have bred many more 

 in bottles in the laboratory, and have taken and examined large 

 numbers of them in the field, I have never seen any that I even 

 suspected were males. A series of experiments, begun in the 

 laboratory July 22, 1898, and continued into December, proved 

 that no males are developed in the fall generations. Experiments 



