VOL. III.] Azotes on (Enothcra. 249 



while the spring seedHngs come from seeds that did not germinate 

 the previous season, or perhaps from seeds ripened on the fall seed- 

 lings. These flowers open about sunset and are not fertilized in the 

 bud, for the pistil greatly surpasses the stamens. I have examined 

 hundreds of pods and have always found two rows of seeds in each 

 cell, eight rows in all. The seeds are round and pitted. 



CEnothera trichocalyx Nutt. Of this I have collected several 

 forms that vary with reference to the bud, the appearance of which 

 seems to be the chief difference between this and CE. albicaiilis. I 

 cannot determine to which species several belong, though the 

 Grand Junction CE. trichocalyx and the Denver CE. albicaulis seem 

 quite distinct. They all have lance-linear seeds, grooved where 

 they press against their companions, and often mottled with red. 

 I found the mottled seeds on the Grand Junction form of CE. tricho- 

 calyx and the Denver form of CE. albicaulis. In both, the seeds of 

 well developed pods have two rows in each cell. The plants from 

 Grand Junction have buds that are. conspicuously white vil- 

 lous and decidedly blunt; the tips are not in the least free This 

 seems to be the typical form, as I said before, of CE. trichocalyx. 



The form from Thompson's Springs, a station on the Rio Grande 

 Western in Utah, has villous buds that are acuminate but without 

 free tips. I have the same from along McElmo Creek, in south- 

 western Colorado. The form from Moab in Utah has smooth buds, 

 acuminate and with free tips. The form from Court House Wash, 

 on the road to Moab, has buds slightly villous, with tips acuminate 

 and partially free. These forms are all annuals or biennials. 



The Denver form of CE. albicaulis has sparingly villous pods, 

 acuminate and with free tips. It would appear that a specific dif- 

 ference between these two must be sought in some other organ. 

 CEnothera albicaulis is distinctively a perennial, but that might arise 

 from its situation. It is always found not far from water, while CE. 

 trichocalyx inhabits desert regions. 



In comparing the Denver CE. albicatdis with the forms of CE. 

 trichocalyx I find the leaves to be quite dissimilar, the former hav- 

 ing leaves that are either sparingly or deeply toothed and canescent 

 with appressed hairs; the latter having pinnately divided smooth 

 leaves with the segments narrow and linear. However, in looking 

 over Watson's Revision, I find that var. runciyiata and var. Calif or - 

 nica of CE. albicaiclis have pinnatifid leaves; so the difference in the 



