144 PLANT HYBRIDIZATION BEFORE MENDEL 



remains in the present instance to discuss the contributions of 

 Louis de Vilmorin (1816-1860), and of his son Henry (1843- 

 1899), to investigations in heredity and in hybridization. 



The first experimental effort, since the work of Sageret, to find 

 a definite numerical relation in the transmission of characters 

 from a cross was the work of Louis de Vilmorin, carried on with 

 Lupinus hirsutus from 1856-1860, and reported upon by his son 

 in 1879. (7b.) This species affords the advantage of being gen- 

 erally self-fertilized, and has ordinarily blue, but also frequently 

 rose-colored flowers, there being no other color or intermediate 

 shade. The plants used came from seeds of these two varieties, 

 from commercial lots, kept pure by rogueing out all plants not of 

 the desired color. It was Vilmorin's conception that, in a self- 

 fertilized plant such as lupine, there was introduced a great ad- 

 vantage in the study of heredity, since each individual was the 

 descendant of a single plant of the preceding generation, and not 

 of a number of ancestors, doubling itself at each stage, as in the 

 case of plants where two individuals are involved in seed re- 

 production. 



"It may then be admitted," says Vilmorin, "that the seed sowed the 

 first year of the experiments, in 1856, reckoned a series of at least fifteen 

 ascendants, which have given flowers constantly of the same color, blue 

 for some, rose for the others." (7b, p. 6.) 



No crosses were made, but records were kept for four years of 

 the different kinds of plants derived from each sowing. Out of the 

 progeny produced each year, instead of planting all or a consid- 

 erable number, but one representative of each color was planted, 

 as a rule, so that large numbers are not available. The fact that 

 both the blue and the rose-colored plants for the most part broke 

 up into blue and rose for each year indicates that each strain was in 

 the hybrid or heterozygous condition. 



In forty cases during the five years, the rose-flowered plants 

 broke up into blue and rose ; in three apparently, and in the other 

 cases possibly, there appeared to be a 3 : 1 ratio of rose to blue. 

 In thirty-six cases in the same period, the blue-flowered plants in 

 turn broke up into blue and rose ; in six of these cases, the ratio 

 was close to 3: 1. It is evident that Vilmorin's experiments need 

 repetition, since a clear breaking-up of both blue and rose-flowered 

 plants into blue and rose again would not be expected. A few 

 cases of rose and a few cases of blue bred true. To V^ilmorin, it 



