194 Atkinson. 



differences in growth easily attributable to variations in moisture con- 

 ditions, etc., of the soil under this kind of culture, except for the scant 

 appearance of small reddish spots which, in one of the forms, also varies 

 according to external conditions. 



During July and August as the rosette stage became more and 

 more evident the new and larger leaves of the two parental types 

 began to present the specific form characteristic of each. The adult 

 rosettes in September and October differed by 5 or 6 characters. In 

 the following summer (1912) as they came into flower many other 

 striking differences were observed as a result of a critical analysis. 

 These characters were regarded as evidence strong enough to entitle 

 the two parents to distinct specific rank. Material of the two new 

 species was submitted to Mr. H. H. Bartlett of the Bureau of Plant 

 Industry, Washington, D. C. They were later published as Oenothera 

 nutans^) Atkinson and Bartlett, and Oenothera pycnocarpa Atkinson 

 and Bartlett. 



The hybrids which were obtained from the cross pollination of 

 1909, and brought to maturity in 1912 were the Fi of Oe. pycnocarpa 9 

 X Oe. nutans (I'. In the rosette stage some of these hybrids appeared 

 to be a blend of the characters of the two parents. But since a large 

 percentage of them were attacked by the downy mildew {Peronospora 

 arthuri Farlow), some of them so seriously that the leaves did not 

 reach their full development, a critical study of the rosette stage could 

 not then be made. Oe. pycnocarpa was more seriously attacked, but 

 enough of the rosettes were free to determine the morphological char- 

 acters. Oe. nutans on the other hand was immune. A number of the 

 most seriously diseased plants of Oe. pycnocar'pa, and of the hybrid, 

 died during the winter 1911 — 1912. 



In the summer of 1912 the Fi hybrids came into flower. There 

 were two distinct types with no intergrading forms. The petals of the 

 vegetative "blend" were strongly Like those of Oe. nutans but larger, while 

 those of the second hybrid type were exactly like those of Oe. pycno- 

 carpa. The vegetative characters of this type were not, however, a blend. 

 But since the rosettes of so many of the hybrids were so modified by 

 the influence of the Peronospora, the experimental work was repeated 

 for the purpose of a critical study of all stages. 



^) Bartlett, H. H., Systematic studies on Oenothera, III, New species from 

 Ithaca, N. Y. Rhodora, 15, 81—85, 1913. 



