176 The Mechanism oe Evolution in Leptinotaksa 



of an added agent, isolating it from its relatives in breeding. It was the nni- 

 fonn experience with this stock that it came out of aestivation 1 to 3 weeks earlier 

 than either of the parents under the same conditions. This was uniformly true 

 where all had gone into aestivation in the early winter months voluntarily ; under 

 identity of conditions this special race was out first and well started in breeding 

 before the parent lines had started into action. Had this been the reaction in 

 nature, and I can discover no reason why it would not, the new race would then 

 be out and breeding in the spring before the others had emerged. Trivial con- 

 ditions of this sort in the composition of organisms are productive of not a little 

 of the separation of natural groups of organisms, and in the main it seems most 

 probable that characters of this sort have arisen as the result of the product of 

 the reactions productive of the origination of the group rather than in the indi- 

 rect ways of survivals and slow accentuations through long periods. 



In these experiments I have in the main eliminated, by the method of work- 

 ing, the sterility that comes through nonbreeding as the result of new conditions 

 by the provision of conditions such that the stocks from nature could breed in my 

 laboratory, and although there is much in the eif ects of these conditions of life in 

 their action upon the reproductive activities, I have eliminated as much of it as 

 possible, my operations being adequately complicated without the addition of 

 this group of factors. There is abundance of evidence that this relation to the 

 conditions of the medium are vastly important and govern the breeding activities 

 of all organisms directly in nature and in experiment. 



Probably the most common record in the accounts of the crossing of species is 

 that of the sterility of the F^ progeny and the inability to obtain Pg progeny of 

 crossings back upon either parent form. It has been discussed adequately in the 

 literature; tests by breeding and examinations of the germ-glands have shown 

 that there are slight or extensive derangements of the gametes, and there the 

 entire matter rests. In my experiments I have had but few instances in which 

 if Pi progeny were produced they were also not able to produce Pj and subse- 

 - quent generations. It was found in the peculiar race of undedmlineata men- 

 tioned in preceding pages that often, when it had given progeny in crosses, it 

 gave no second generation when these Pj were inbred. No reason could be dis- 

 covered except that often the males or less often the females would not develop 

 any mature germ-cells. Examinations showed that the gonads were not active 

 and were producing only a few or only defective germ-cells. In normal mate- 

 rials similar individuals are also found which have a relation to lines of sterility 

 in the race. In these hybrids the reactions were the same and the defective 

 nature of the gonads much like the conditions in sterile individuals of normal 

 stocks. There the matter rests for the present. I have no means of determin- 

 ing what the sterility is due to, whether to disease, inheritance, or incident 

 agents, but these cases are sterile independent of any conditions in the medium 

 that I have been able to discover. 



It is a constant experience in the crossing of these species by close attention to 

 the conditions of life, as temperature and food relations, that in the production 

 of Pj there are effected many times in such production a distinct increase in the 

 fertility, and often in practice the obtaining of Pj has depended upon these nice- 

 ties of adjustment in the conditions surrounding the organisms. One is led to 

 wonder to what extent the recorded sterility of P^ hybrids among animals, 

 especially in the crude conditions of zoological gardens, may not in reality be 



