276 The Pedigree Families. 



eight absolutely smooth-leaved individuals whose seed 

 I harvested in autumn. There were no mutants. 



In the following year (1890) I sowed the seed in 

 a bed of about 3 square meters, and in the summer pulled 

 up all the biennial plants as well as the younger ones of 

 those tliat had produced a stem and those whose leaves 

 were more or less crumpled. There were left, in Sep- 

 tember, 25 smooth-leaved individuals which however 

 flowered late and set comparatively little seed. Still, I 

 got about 25 ccm, of seed from them and was able there- 

 fore to continue the experiment in 1891 on a more ex- 

 tensive scale. It occupied, in fact, 14 square meters and 

 mutants appeared again. There were two examples of 

 lata, an annual one and a rosette, two feeble rosettes of 

 cUiptica and another eUiptica which was so delicate that 

 though it developed a stem, it soon died. 



From this time I ceased to pay attention to the special 

 peculiarity of this family, i. e., its smooth leaves, and 

 used it for other purposes. The smoothness of the 

 leaves however maintained itself (in spite of the fact 

 that the plants were left exposed to the visits of insects) 

 in the majority of the plants, the remainder being ordi- 

 nary Lamarckiana. The experiments in question were 

 discontinued in 1894, since when I have again got the 

 family to breed true by employing artificial fertilization. 

 But it did not mutate any more after that. 



The experiments referred to above were three in 

 number and will be briefly described here : I shall have to 

 deal with two of them in greater detail later on. The 

 object of the first was to breed a tricotylous race. This 

 object was attained, although the first two generations 

 were not so rich in tricotylous specimens as were those 

 of the similar experiment referred to above, with O. 



