282 PRODUCTION OF RACERS AND SPECIKS IN IvEPTINOTARSA. 



and breed true to their type. These differ in many characters from the par- 

 ents, and this new combination of characteristics is, as my cultures show, 

 handed down from generation to generation with undiminished intensity and 

 unchanged arrangement. These behave in heredity, especially in crossing, 

 exactly as do independent species. 



As far as I can judge, my variations from decemlineata are as pure and 

 strong as De Vries's mutants from CEnothera, holding their own even in 

 crossing with the parent species. The experiments with pallida are strong 

 evidence in favor of the origin of species by rapid change, better, perhaps, 

 than that afforded by De Vries's plants, because pallida, given the necessary 

 start, needs neither selection nor attention to take care of itself, and my cul- 

 tures would, I have no doubt, have spread widely in nature, as they began to 

 do, had I allowed them to continue. 



It is probable that, in view of the evidence which is herein produced, there 

 will be raised the objection to the origin of species by rapid evolution that, as 

 I have shown, these rapidly developing extreme variations occur rarely — 

 very rarely — and are usually single individuals, which are able to increase 

 in numbers only w^hen cared for in cultures, and hence can not be of any 

 marked utility in the evolution of species, owing to the immense improb- 

 ability, as shown, of their being able to get a start from a single individual. 

 This objection is valid only in as far as it concerns single individuals. I 

 have given evidence in the case of the Cabin John Bridge specimens, and 

 in the section on place variations it is shown that under fluctuations in the 

 environment there are produced sometimes a considerable number of these 

 extreme and permanent variations. Given a fair numerical start, pallida was 

 able to take care of itself, so that the point is not that it is impossible, but that 

 it must get a fair start numerically in order to succeed, and, although decem- 

 lineata shows several of these extreme variations, they occur at so infrequent 

 intervals that I doubt if under present conditions any of these forms could 

 become established in nature. Whatever we may believe concerning the future 

 of these variations in the evolution of the genus Leptinotarsa, my cultures of 

 decemlineata show in a clear and unmistakable manner the inadequacy of 

 selection to create new elemental species and the fact that such do arise by 

 sudden transformation. By selection we can create races, but we must main- 

 tain themi by the same process, else they revert to the mediocre of the parent 

 species. Races, too, differ in one or two characters from the parent element- 

 ary species in a complex of characters which often maintain their fixity without 

 the aid of selection. Races and elementary species, therefore, differ in several 

 attributes, but I think the difference is not in kind, but only in degree. They 

 stand apart, and apparently there is no intermediate ground, yet our observa- 

 tions are all too meager and our attention too much distracted by the striking 

 variations. It is my firm conviction that we shall later discern plenty of 



