by " interpretations of the face of nature " with the accompanying; 

 inexact methods and superficial considerations. It was upon 

 the safe basis of the first-named conceptions, and by means of 

 the methods entailed that de Vries has so successfully grappled 

 with the problems involved in the investigation of the part played 

 by discontinuous variation in evolution. 



In view of the amount of orderly and well-authenticated evi- 

 dence now at hand, it may be assumed as demonstrated that 

 characters and groups of characters of appreciable physiological 

 value, originate, appear in new combinations, or become latent, 

 in hereditary series of organisms in such manner as to constitute 

 distinct breaks in descent. 



This is the main thesis of the mutation theory : the saltatoiy 

 movements of characters, regardless of the taxonomic value o\ 

 the resultant forms. That the derivatives might be considered 

 as species by one systematist, and varieties by another is quite 

 incidental and of very little importance. The main contention 

 lies in the claim that characters of a definite nature appear, and 

 become inactive suddenly, and do not always need thousands of 

 years for their infinitely slow external realization, or for their 

 gradual disappearance from a strain. 



Of course the principal corollary of the mutation-theory is 

 that the saltations in question do result in the constitution of new 

 species and varieties. As a matter of interest it may be stated 

 that all of the .systematists who have seriously examined the 

 adult mutants of the evening-primroses cultivated in the New 

 York Botanical Garden have held the opinion that certain ones 

 were to be considered as species and others as varieties. 



Furthermore, these conclusions are confirmed when the char- 

 acters of the mutants are subjected to statistical methods of 

 investigation. In the observations of Dr. Shull, which will be 

 presented more fully before the Botanical Society of America, it 

 has been found that qualities of the mutants, susceptible of 

 measurement, depart definitely and clearly from the parental type 

 and fluctuate about a new mean, and do not intergrade with the 

 parental form. The am[)litude of fluctuation about the new 

 center is greater than that of correspondent parental qualities. 



