23O WILLIAM LAWRENCE TOWER. 



banks of the Calumet River near Chicago, and therefore char- 

 acteristic of, and in their activities, adjusted to the conditions 

 of the place of origin. 



These stocks have two yearly generations: pass the winter 

 in the adult condition, hibernate in the soil, emerge in the 

 spring, as soon as the temperature is high enough; and begin 

 breeding to produce the first summer generation which matures 

 in July; this first summer generation breeds at once, giving 

 the second summer generation maturing in August or early 

 September. These do not breed, but after feeding undergo 

 preparations for, and enter, hibernation during September or 

 early October, where they remain until the arrival of awakening 

 conditions in the following spring. 



In nature this reproductive rhythm is never pure nor regular; 

 many of the overwintering population remain alive, or are late 

 in emergence from hibernation, and are able, therefore, to cross 

 with the first summer generation, giving hybrids between two 

 differing conditions in the rhythm of reproduction, and all 

 materials direct from nature are liable to be in one way or an- 

 other heterozygous with respect to this rhythm. Pure lines, or 

 stocks in which the rhythm of reproduction was homozygous, 

 have been used in these experiments and no others. 



The second summer generation does not mature its gametes, 

 and after feeding undergoes changes preparatory to hibernation 

 that are common to insects; the elimination of waste products 

 and the reduction of the water content of the tissues, lowering 

 the freezing point thereof, so that the animals are not killed by 

 the low temperatures of the winter months, an operation common 

 to many plants and animals in resting conditions. Upon emer- 

 gence in the spring, or for breeding, this water reduction is rapidly 

 compensated for by water derived from the food and that 

 absorbed by the tissues from the atmopshere or hygroscopic 

 water. 



The location chosen for the experiments at the Desert Labora- 

 tory of the Carnegie Institution of Washington was on the edge 

 of the broad flood plain of the Santa Cruz River at the foot of 

 Tumamoc Hill, a deposit of gravel and boulder beds with a 

 thick surface layer of adobe soil, of uniform texture and com- 



