1922] 



PAYS0N — STUDY OF THELYPODIUM AND ITS IMMEDIATE ALLIES 251 



one, have definitely two-lobed stigmas. It will be recalled that 

 the presence of trichomes, the assumption of the perennial habit, 

 and the two-lobed stigmas are considered recent characters in this 

 genus and in Thely podium. It would seem then that the group 

 of species with petioled stem-leaves is not primitive. This con- 

 clusion is strengthened by the evident fact that these eight 

 species are not closely related forms but represent at least three 

 different lines of development. They are interpreted as being at 

 the ends of the several lines of evolution, and so it seems cer- 

 tain that the petioled leaf has been developed independently at 

 least three times within the genus Caulanthus. 



The transition between stem- and basal-leaves that occurs in 

 some species of Caulanthus has already been referred to several 

 times. In such a species as C. heterophyllus, for instance, may 

 be seen a gradual transition from entire, amplexicaul upper stem- 

 leaves to toothed, elongated, and non-clasping basal-leaves. This 

 transition is believed to show the way in which the differentiation 

 has occurred and to show that the basal-leaves have originated 

 from the stem-leaves. It is significant that in the more primi- 

 tive species, such as C. amplexicaulis , C. inflatus, and C. Cooperi, 

 no differentiation is evident. It is also of interest to note that 

 whenever trichomes are present they occur only at the base of 

 the plant or are most numerous there. This basal region seems, 

 therefore, in some way to retain fewer of the ancestral characters 

 and to be the first to assume recent modifications. 



Duration of life. — The species of Caulanthus are mostly annual. 

 Two species only, C. major and C. crassicaulis, are short-lived 

 perennials, although some of the annual species may approach 

 the biennial habit. There seems to be, however, no true assump- 

 tion of the biennial habit as in Thelypodium. Since the two per- 

 ennial species may not be considered primitive for other reasons 

 it is supposed that the annual condition is the more primitive. 



GENERIC LIMITS OF CAULANTHUS 



Caulanthus, as here treated, contains many more species than 

 have been referred to it formerly and consequently embraces a 

 greater range of morphological variation. These changes have 

 made a formerly homogeneous group somewhat heterogeneous 

 and have involved a number of nomenclatorial transfers. These 

 two results are, in some ways, unfortunate but are inevitable if 

 the present theory of development is correct and if classification 



