108 APPENDIX. 



20. Tkiodanis rupestris 90G, this is very 

 near to the Campanula flagellaris of Kunth, 

 perhaps of this Genus. His Mexican C. perfo- 

 liata appears somewhat different from ours by 

 flowers geminate and ternate ; yet Torrey would 

 have all these to be varieties ! which is the type 

 then and the prnnordial form? 



21. Helenium DiscovATUM 942, compare it 

 with H. rnexicanum of Kunth, too shortly noti- 

 ced bv him. 



E\D OF THIS FT.ORA- 



