STUDIES IN CEANOTHUS. I77 



ino- form in a given district. Where the seedlinors survive 

 in great numbers, cross-fertiHzation being made certain 

 by the swarms of insects attracted to their fragrant flow- 

 ers, a continual crossing takes place, not only between 

 the original forms, but between the hybrids and their 

 parents on either side. 



Many of the species accepted in this paper are likely 

 to prove too closely related as their forms become better 

 known. No one, so far as I am aware, has yet made any 

 systematic attempt to collect over the whole area the 

 forms of even the nearly allied Eastern species. The prob- 

 able hybrid origin of several of the species raises a ques- 

 tion of some importance, which may perhaps be easiest 

 answered, as in the case of the willows, by propagating the 

 distinct forms and hybridizing them artificially. 



In the pages following the original descriptions of nearly 

 all the species, with the exception of C. Aniericamis and 

 its well-settled synonyms, are republished verbatim, in the 

 hope that future students of the group may find them as 

 great a convenience as they would have been to the 

 writer. The modification which descriptions undergo by 

 increasing knowledge of their variations is often very 

 great, and it seems to me the duty of monographers to 

 give the original diagnoses in addition to their own, so 

 that their readers may be in possession of the data neces- 

 sary to form some sort of independent opinion, without 

 the necessity of acquiring a considerable library. 



For kind assistance and favors, I have to thank the 

 curators of the herbaria of Harvard, Kew, Columbia 

 College, Philadelphia Academy of Sciences, U. S. Agri- 

 cultural Department, Missouri Botanic Garden and the 

 Boston Horticultural Society. For specimens and notes 

 of distribution, etc., I am indebted to C. G. Pringle, 

 Edward Palmer, J. G. Lemmon, S. B. Parish, J. W. 



2d Ser., Vol. IV. ( 12 ) June 2, 189-t. 



