Mutations 71 



b. 3. Crumples few, 

 scarcely narrowing 

 into the stalk, 

 almost grasslike. 13. CE. sublinearis. 



II. Leaves sessile, short and broad, 



almost heart-shaped, crumpled. 14. CE. nandla. 



How the mutants were produced in the garden. Most 

 of the types previously described were found growing 

 wild near their parent species, CE. Lamarkiana. 



De Vries wished to determine whether these mutations 

 could be produced from seed of (E. Lamarkiana planted in 

 the garden (Fig. 21). Four series of experiments were 

 performed, lasting through five to nine generations in 

 which thousands of individuals were grown and studied. 

 A description is here given of one of these experiments. 1 

 The others were very similar. The pedigree culture foegan 

 in 1886, when seed was planted in the garden from 

 nine plants found growing wild. These nine plants 

 constituted the first generation. The second generation 

 flowered in 1889. This generation consisted of fifteen 

 thousand seedlings of which ten were distinct mutations 

 five la ta and five nandla. There were no intermediates. 



The third generation of ten thousand plants produced for 

 the first time in pedigree cultures a plant of CE. rubrinervis, 

 along with three plants of CE. lata and three of CE. nandla. 



The fourth generation of fourteen thousand plants 

 yielded a higher percentage of mutants. These were 

 as follows : oblonga 176 ; lata 73 ; nandla 60 ; albida 15 ; 

 rubrinervis 8 ; scintillans 1 ; and gigas 1 . 



1 De Vries, "Species and Variation, their Origin by Mutation," pp. 

 549-575. 



