100 CARNEGIE INSTITUTION OF WASHINGTON. 



and economy in the operations. In addition, a small but commodious 

 record building has been erected and equipped, and also concrete cellars 

 with heavy double covers, provided for the installation of the recording 

 instruments, protecting them from the weather and reducing the instru- 

 mental error as well as the dangers of damage to the instruments. In 

 all respects the plant, as now constructed, gives greater security and is 

 far more efficient than the previous arrangements. 



The winter of 1914-15 showed another possible mode of action of 

 the desert complex upon its inhabitants, for the cold, wet conditions 

 inflicted (in all of the cultures) a heavy loss, roughly estimated at the 

 present time at about 90 per cent. This loss was uniformly of the 

 extremes in all the hibernating populations, so that the different extreme 

 types isolated (for tests of this year) in the last generation of 1914 were 

 eliminated. This is in accord with man}'- other experiences at the 

 Desert Laboratory during the course of these investigations, and I 

 have had the same results in nature in experiments in the tropics in 

 Mexico and at Chicago. It thus seems that the eliminating action in a 

 population, together with its relation to the mean or modal group, is 

 rather constant, regardless of the nature of the factors of elimination. 

 In this process there is an element of conservation rather than one 

 of diversification. In all of these experiences in elimination I find 

 that a freshly isolated divergent group is, rather uniformly, entirely 

 eliminated, while one that has been isolated for two or more genera- 

 tions is not eliminated, as it has apparently undergone some population 

 adjustment and established a mean that is able to meet the diverse 

 eliminating factors to which it is subjected. Interesting data have 

 been collected, in the course of the experiences at the Desert Labora- 

 tory, upon which it will be possible to base an experimental analysis of 

 this problem. There can be no reasonable doubt that these relations 

 are of the highest importance in the establishment of species and groups 

 in nature and in their distribution and ecological relations. In spite 

 of the heavy eliminations, no losses of any moment were sustained in 

 the cultures. 



In 1912, 1913, and 1914 report was made upon the alteration of the 

 water relations in L. decemlineata. The tests which were made in 

 the winter of 1914-15 show almost the same results as hitherto; the 

 stock was returned to Chicago for hibernation. Tests to determine 

 whether this alteration is reversible have shown, as the result of breed- 

 ing the Tucson lines at Chicago for two generations before hibernation, 

 that a small percentage may be able to pass the northern winter 

 (0.1 to 0.5 per cent), but in subsequent seasons these have not been 

 able to develop a race that has the cold-resisting capacity of the original 

 stock. Apparently, the alteration is slightly reversible, but more 

 slowly, and possibly not to the original condition. Tests of the inheri- 

 tance of this alteration in 1914-15 gave the same results found in 

 previous years. 



