26o PRODUCTION OF RACES AND SPECIES IN I.EPTINOTARSA. 



acquainted with his material in all its natural phases; and second, by long- 

 continued experimental investigation. In the study of evolution in Leptino- 

 tarsa I have carried on the two lines of investigation stated. The first line of 

 inquiry has been presented in the first chapters of this paper. In this chapter, 

 in which will be taken up the second line of attack, it will not be out of place 

 to review briefly and bring together the general results of the first inquiry, 

 since it is these conclusions which form the starting-point for the experi- 

 mental portion of this research. 



The genus Leptinotarsa everywhere lives in situations where the earth is 

 covered with perennial grasses, with a greater or less growth of herbaceous 

 plants, some of which must be Solanacese, upon which the beetles feed. They 

 are typical grassland forms. This fact explains their general distribution 

 over the American continent — their presence in some parts and absence in 

 others. To the conditions of existence of a grassland habitat they are singu- 

 larly adapted. The general climatic conditions of grasslands are frequent 

 although often small precipitation during the growing season, which keeps 

 the superficial soil moist and the lower stratum of air humid through evap- 

 oration, followed by a dry season of greater or less duration. I have shown 

 how the beetles are limited by these very factors of soil and atmospheric 

 moisture in their general and local distribution, and how their life histories, 

 reproduction, aestivation, and physiology are adapted to these conditions. 



But the grasslands are not all alike; some are more, others less humid; 

 some are warm, others cold ; hence a particular trend in species evolution in 

 the genus shows closely related species occupying adjoining but dififering 

 environmental complexes; and we inquire whether these species have come 

 about by slow modification and extinction or b}^ sudden transformation. We 

 have examined the variations, fluctuating, place, and geographical, and get 

 no satisfactory evidence as to the method of evolution, too much supposition 

 being required to connect variations with evolution. This study of distribu- 

 tion and variation gives no information concerning the method of evolution, 

 but it shows conclusively, first, that species formation has been definite, the 

 successive and related species having developed in contiguous habitats as the 

 genus evolved ; and, second, that variations are in the same direction as spe- 

 cies formation. Evolution has been definite and not promiscuous; it is 

 orthogenetic. 



Throughout there appears a growing body of evidence tending to show 

 that selection does not create nor even preserve extremes of variation, but 

 that it acts as a conservative factor to eliminate extremes and to confine the 

 species close to the mode. Correspondingly, there is found in increasing 

 volume evidence that new races and characters develop rapidly, but always in 

 conformity with the general orthogenetic trend of evolution of the group or 

 genus. 



