500 



ON TRIPLE HYBRIDS. 



characters the lata hybrids were strikingly intermediate between 

 the two parents. Even the bent tips of the young leaves were to 

 be seen, and at once revealed the father. But the most interesting 

 feature was that of the pollen. 0. lata is purely female, while 0. 

 strigosa has the ordinary supply of pollen. The hybrids showed 

 some pollen and a relatively small number of fertile grains. These 

 were, during ordinary weather, insufficient for fertilization, even 

 if the pollen of many flowers was brought upon one stigma. But 

 there were individual fluctuations, and so I succeeded in getting 

 self-fertilized fruits on at least one specimen. 



Oenothera lata x 0. biennis Chicago. — This cross was made in 

 1905, both parents being annual. Two specimens of 0. lata were 

 used. I sowed their seed in different lots in the three succeeding 

 years and had the following results: 



The laeta and the velutina were of the description already given. 

 I sowed the self-fertilized seed of 1906 in 1907 and 1908 and had 

 four lots of the second generation. The lots of the first included 69 

 and 139 individuals, all of which were laeta; the sowings of the 

 other strain extended over 38 and 158 samples, all of which were 

 velutina. Both of the twins thus complied with the rule of constancy 

 in the second generation. From the lots of 1907 1 cultivated in 1908 

 a third generation comprising 70 children of laeta and 90 of velutina; 

 both lots were once more uniform and similar to their parents. 



The lata individuals of this cross were almost similar to those of 

 0. lata itself. However, they had the smaller flowers of 0. biennis 

 with the self-fertilizing position of the stigma, and the long lobes of 

 the father. Their flower buds were intermediate between those of 

 the two parents. In 1906 and 1907 the anthers seemed wholly barren; 

 but in 1908, during the very warm last days of July, they yielded 

 some pollen, which was used partly for self-fertilization, partly for 

 crosses. 



0. lata X Hookeri. — This cross is the most interesting of all the 

 crosses with 0. lata, since its lata hybrids are almost as rich in pollen 



