272 The Pedigree Families. 



fact that it was allowed to cross freely with the parent 

 species clearly indicate the operation of some definite 

 hereditary process. Whether the ten plants of the first 

 year (1887) had a common origin cannot of course be 

 decided a posteriori, but it cannot be regarded as other 

 than extremely probable. 



I now come to the cultures in my experimental gar- 

 den For these I collected, as I have already said, the 

 seeds of some examples of O. laevifolia in the field at 

 Hilversum towards the end of the summer of 1887. 



These seeds produced in my garden in 1888 over 

 200 plants of which about 60 % were annual ; some of 

 them were O. Lamarckiana and some O. laevifolia. I 

 selected the seed from the strongest and most precocious 

 examples of the crop for the cultivation of the family, 

 keeping, of course, the two species separate. 



In 1889 three beds of 10 square meters were sown 

 with the seeds of the 0. Lamarckiana plants partly in 

 the hope of getting new forms and partly with a view to 

 an investigation, to be described later on, into the con- 

 ditions determining the annual or biennial habit. 



Like the two other extensive sowings in the La- 

 inarckiana-iRmily this one also proved to be rich in mu- 

 tants. I wish^to call especial attention to this fact be- 

 cause I believe that the relative scarcity of mutants in the 

 other branches of this family can be attributed to the 

 small extent of the individual generations in these cul- 

 tures. 



I got altogether 41 mutants whose appearance in the 

 various years and cultures I shall now put in the form of 

 a genealogical table. This is constructed on exactly the 

 same plan as the previous ones. 



