[Vol. 9 



254 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



and has been reported also from as far north as Washington and 

 as far east as Colorado. This species is the most widely dis- 

 tributed of the genus and has apparently spread as a weed along 

 lines of transportation. Except for this species no members of 

 the genus are known from Colorado. 



Evidence at hand goes to show that Caulanthus originated in 

 the interior region of southern California. Several reasons may 



for this belief: (1) 



the 



number of species. (2) The most primitive species of th 



here. (3) 



from th 



the most recent species pi 



SUMMARY OF CONCLUSIONS ON CAULANTHUS 



1. Probably long, terete pods are to be considered primitive in 

 Caulanthus and the shorter or compressed ones as derived. 



2. The entire stigma is considered primitive and the degree 

 of lobing is thought to be indicative of the degree of specializa- 

 tion. The most deeply lobed stigmas are found in those species 

 that are most specialized in other respects. In C. lasiophyllus 

 and its allies, however, this rule does not hold, since in this group 

 the stigma is quite uniform and but slightly lobed. 



3. The "cell pattern" of the septum is found to be of great 

 importance as a generic character in Caulanthus. In C. lasio- 

 phyllus a type similar to that seen in Thely podium is believed 

 to have been independently developed. 



4. The narrow, crisped and channeled type of petal is consid- 

 ered primitive for this genus. The flowers of the ancestral spe- 

 cies were probably purple in color. 



5. The most primitive type of inflorescence found in Caulan- 

 thus is lax and racemose. 



6. Amplexicaul cauline leaves are believed to represent the 

 primitive type and from them the petioled basal- or stem-leaves 



have been developed. 



7. The presence of trichomes is an indication of specialization. 

 They were probably not present in the first species of Caulanthus. 



8. Guillenia of Greene could be maintained as a genus without 

 offense to phylogeny but it is found inadvisable to separate it 



from Caulanthus. 



9. The annual habit is held to be primitive. 



10. Stanfordia of Watson is not thought worthy of generic 

 rank and is merged with Caulanthus. 



