Stout: The Origin of Dwarf Plants 445 



out of a generation of two thousand one hundred plants, twenty 

 plants having only peloric flowers. Progeny of these peloric plants 

 gave ninety per cent, true to the type. In both Chrysanthemum 

 segetum and Linaria vulgaris peloria self sterility made it difficult to 

 grow pure line progenies. De Vries points out (1901, p. 564) that 

 while he considers these as cases of mutation they are of a type 

 quite different from that in Oenothera. In regard to the origin of 

 such variations that are hereditary the evidence is not at all con- 

 clusive that slight variations may not be inherited even in a cumul- 

 ative manner. 



That this is the case is further suggested by the difficulty of 

 drawing a definite line of distinction between species, varieties and 

 races. De Vries, who has written most fully on this point from 

 the standpoint of genetics, states that the best examples of varieties 

 are those showing latency of a single character which may be just 

 as constant as species (1901, p. 119); again he treats 0. nanella 

 (1901, p. 256) as a variety simply because somewhat similar dwarf 

 types recur in a great number of species. He also gives the general 

 view that varieties usually differ in one character (1901, p. 363); 

 he states in another connection that the chief difference between 

 improved races and species, even the smallest of elementary species^ 

 is the instability of the former and the stability of the latter (1901^ 

 p. 84); but we may note that inconstant species do occur (1901, 

 p. 270). 



It is clear that some cases of sporadic variation reproduce more 

 true to type than others. The so-called law of mutation that 

 "neue elementare Arten sind meist vollig constant, vom ersten 

 Augenblicke ihrer Entstehung an" (1900, p. 175) is itself a quahfied 

 statement. The difficulty of assigning a definite heredity in terms 

 of unit characters or unit factors to "newly born" characteristics 

 which appear in mutations has become apparent. This is especi- 

 ally recognized in part by the view of Gates that mutation is a 

 phenomenon of variability which is quite distinct from heredity. 



Thus far progeny has been grown from only one of the dwarf 

 plants of Hibiscus ocidiroseiis. One capsule of seed gave forty-six 

 plants, all but one of which were dwarf, and another capsule gave 

 twenty-seven dwarf, one robust and seven intermediates as judged 



