124 MISSOURI BOTANICAL GARDEN. 



It is hoped that this key will aid botanists in the identifi- 

 cation, approximately at least, of these forms, and will lead 

 to an increased interest among collectors in the search for 

 these types throughout North America and Europe. Of 

 course the present key is very incomplete and it would be 

 desirable to add many forms to the list, but these can be 

 added as our knowledge of the forms becomes more accu- 

 rate through the medium of pedigree cultures. 



Even though the forms which will be discovered are found 

 to have been introduced or escaped from gardens, a com- 

 parison of the types from various stations as they occur 

 under such conditions is of interest, as I have already found 

 from my cultures of the Liverpool Oenotheras. The study and 

 comparison of such areas will lead to results of the greatest 

 interest and importance bearing on such problems as the ef- 

 fects of climatic conditions after long periods. The Oenotheras, 

 on account of their weed-like character, are of peculiar value 

 from this standpoint, because their habit of preempting 

 deserted areas and multiplying rapidly makes each such 

 patch or locality equivalent to a transplantation experi- 

 ment in which the success or failure of the various mutants 

 and segregates, and the changes they have undergone 

 under the conditions of the struggle, can be noted and 

 analyzed. 



It is a good example of the occasional vagaries of system- 

 atic botany that 0. grandifiora should have been so long 

 considered a mere " large-flowered variety" of 0. biennis, 

 while in reality it is at least as distinct from 0. biennis as 

 is 0. Lamarckiana. Though 0. grandifiora and 0. Lamarcki- 

 ana come in the same sub-group of the Onagra series, they 

 are distinct from each other in a great many characters, and 

 both appear undoubtedly to be good "Linnean" species. 

 What the relation of such large-flowered forms as 0. HooJc- 

 eri and 0. Drummondii is to 0. biennis and 0. grandifiora, I 

 am not at present prepared to indicate. The difference be- 

 tween the small-flowered forms, with short style , such as 

 0. biennis, and the large flowered forms with long style, 

 such as 0. grandifiora, is probably the most distinct line of 

 cleavage in the sub-genus Onagra. 



