134 THE GROWTH OF GROUPS 



reared in nature together with the parent species it was 

 not delicate, indeed it endured the frosts of winter rather 

 better than the parent species. Although, as shown, we 

 may explain away each of these preventive factors or 

 even invert them in order to support the opposite thesis, 

 there is no gainsaying the fact that, so far as it was possible 

 to examine the group decemlineata, in nature, the mutant 

 pallida was not meeting with success among it. 



It is evident, therefore, that argument is not of much 

 use in deciding the question of the efficacy of sports in 

 general. It can only be decided by observation. But 

 Mr. Tower's declaration that it is " a real impossibility " 

 for a mutant to become established cannot justly be made, 

 I think, as a result of his observations upon decemlineata. 

 He observed a quarter of a million of the beetles of a 

 certain species, that is to say less than a thousandth part 

 of the number born even in one year, and from his obser- 

 vations draws a conclusion not only concerning the other 

 999 thousandth parts, but also regarding the much vaster 

 mass that will no doubt be born within the next few 

 centuries. Real impossibilities stand for all time. 



Let us keep in mind the fact that among the quarter 

 of a million examined, there was no established group of 

 the kind pallida ; but let us suppose that such a group 

 had been found, a group that covered seven or eight 

 square miles of country and comprised perhaps a million 

 individuals ; would this have been considered as evidence 

 by Mr. Tower that groups arise from mutants ? Evidently 

 not, for he had evidence of this kind in the case of the 

 rubicunda group at Toluca, but was not satisfied with it. 

 The mutant rubicunda was born of multitceniata in 

 Chicago, and at the same time there was a group of 



