286 The Pedigree Families. 



Let us now turn to the description of the second lata- 

 family. This arose in my garden in 1887 from seeds 

 which I took, in the preceding autumn, from a quinque- 

 locular fruit of an otherwise normal plant of Lamarckiana 

 growing in the field at Hilversum. Of these seeds not 

 many germinated and, of the plants which arose from 

 those which did, only five flowered in the first year in 

 my garden ; the rest were pulled up and thrown away. 

 Of these five, two were 0. lata and three O. Lamarckiana. 

 The seed of the two former was harvested separately 

 and sown in the following spring (1888) for the con- 

 tinuation of the experiment. 



In 1887 no other Oenotheras flowered in my experi- 

 mental garden besides the five individuals mentioned 

 above. The Lanwrckiajta-iamily was flowering at that 

 time in the Botanical Garden in a bed about 150 meters 

 from my garden and separated from it by shrubbery. 

 From the seeds of the two original lata plants there arose 

 in April and May 1888, 614 individuals of which 21 % 

 exhibited the lata characters. Towards the end of the 

 summer I found that about one-third of the plants had 

 remained in the rosette stage whilst two-thirds had de- 

 veloped stems. Lamarckiana and lata behave very much 

 alike in this respect. 



Of the annual I at as, which could be fertilized by the 

 rest of the plants up till the middle of September, I saved 

 the best 39 • whilst all the rest with a single exception 

 I pulled up and threw away. 



This exception was an example of 0. scintillans, the 

 first that occurred in my experiments. Its mother is 

 therefore one of the latas of 1887, its father one of 

 the three Lamarckianas of the same family. This scin- 

 tillans was biennial, flowering and ripening its fruits 



