Eeactions and Peoducts in Interspecific Ceosses 175 



group, not as an individual, and its isolation is a process of production and not 

 accumulated progressively in any manner, from any process of survival, selec- 

 tion, or utility. It is purely a result of the reaction, and there is no reason why 

 it should be present other than that the gametic factors active in the production 

 of this line in their interaction produced a stable complex such that this was one 

 of the resultant products. 



The stock is also of interest in that it shows that other aspects of the isolation 

 problem are touched upon in an interesting way. The resultant combination of 

 different characteristics of the parent species has given habitudinal isolations 

 that are barriers to the breeding of the form in one direction at least. Its food 

 demands are for Solanum chrisotrichum, or S. hertwigii, upon which it feeds, 

 and upon which the only form with which it has been made to cross does not 

 feed, and there were other differences of habits and developmental rates, and 

 on the whole it was in many ways thoroughly isolated from its progenitors. All 

 of its character associations are the product of the act of origination, are in no 

 respect due to slow development or survival, and are entirely produced without 

 regard to utility in the form. Conditions of isolation seem to act in this useful 

 way only after they are fully established in the special life-relations into which 

 the processes of production have by chance happened to place it. It is not im- 

 probable that the same product under other conditions of Hfe might well have 

 the same attributes and qualities in all respects, which might not there serve at 

 all in the capacities to give the degree of physiological isolation that the observed 

 series showed. 



What constitutes fertility or infertility is clearly and only the capacity for 

 interaction of the gametes and inhibition of the parents from breeding, both of 

 which are the direct product of factorial composition, method, and rate of reac- 

 tion of the combining gametes. Whatever there may be of inhibition in repro- 

 duction is, in some instances at least, solely the indifferent results of the process 

 of origination ; but how generally this is true of the conditions existing in nature 

 is and must remain entirely a matter of opinion and interpretation. 



This is not the only instance in these experiments in which the problems of 

 fertility in species crosses has been elucidated by data derived as a by-product of 

 other work. In the crossing of undecimlineata and signaticollis, as described in 

 Chapter IV, there arose under certain conditions a race that was exceedingly 

 slow in its developmental reactions and equally long drawn out in its life-cycle. 

 This race was produced by the single series of interactions of the cross as a group, 

 with all of its attributes and qualities present and formed by the reactions pre- 

 ceding its appearance in an adult Pi array. It was, as shown by the test, of a 

 heterozygous nature and could by proper means be broken up, but in the usual 

 conditions of life it was as stable as the majority of species in nature. It would 

 breed, as far as the act of copulation was concerned, with either of the parent 

 species, but the different rates of development, or something that was the product 

 or associate of this difference, produced a huge mortality in the progeny of these 

 crosses. In this respect it is isolated, and while it might go through the reac- 

 tions of breeding, the differences in rate of reaction ia the gametes, or factors 

 associated therewith, directly eliminated the probability of more than sporadic 

 or small numbers of progeny. 



This race had another interesting aspect of its composition which in the con- 

 ditions of culture was of no moment, but in nature would have been in the form 



