444 Stout: The Origin of Dwarf Plants 



seen in the well-known cases cited by De Vries of non-isolable races, 

 eversporting varieties, and inconstant species, indicative of irregular 

 and sporadic inheritance and expression of ever present tendencies 

 which in these cases rigid selection fails to isolate. 



That there are cases of variation that are cumulative is evident. 

 De Vries especially has given data on two such cases. In a race of 

 Chrysanthemum segetum, having an average of twenty-one ray 

 florets in the terminal inflorescence in a crop of 1,500 plants one 

 plant was found with four lateral flower heads with twenty-two 

 ray florets. Seed from this plant gave a progeny of 423 plants 

 one of which had a terminal inflorescence of thirty-four rays. Seed 

 from this plant gave a mixed progeny with one plant having sixty- 

 six ligulate florets, three of which were among the tubular disk 

 flowers. Seed from this plant gave progeny with florets ranging 

 from thirty-three to one hundred and one with a few completely 

 double flowers. 



In respect to these results De Vries says (1901 , p. 526) : " Es wird 

 sich jedem Leser die Frage aufdrangen: ist dieser Uebergang ein 

 allmahlicher oder ein stossweiser gewesen? Mir scheint das 

 letztere der Fall zu sein, aber es hangt dabei viel ab von der 

 Bedeutung, welche man den Wortern giebt. Jedenfalls geschah 

 die Umwandlung nicht im Laufe der Jahrhunderte, wie es die 

 Selectionstheorie anzunehmen pflegt, nicht einmal brauchte es 

 dazu Jahrzehnte. Drei Jahre geniigten, und solches in einer 

 Cultur von nur wenigen Quadratmetern Umfang." 



De Vries considers that this result is due to the reappearance of 

 a latent character. It is not clear what the latent character is in 

 this case. The species already possessed ray flowers in the outer 

 circle of flowers. The development of a so-called double-flowered 

 race consisted of a change of tubular flowers to ray flowers. Fur- 

 thermore, as wholly double flowered plants are sterile, the double 

 flowered plants are obtained continually by planting seed from 

 plants not fully double flowered. 



A similar case is seen in the development of Linaria vulgaris 

 peloria. Individuals of certain races produce rarely a single peloric 

 flower, an evidence De Vries considers of a semi-latent character 

 which seldom becomes active. Seed from two such plants gave 



