192,5) PALMER, SYNOPSIS OF NORTH AMERICAN CRATAEGI 7 



country from which the seeds or plants were derived. North America 

 was usually considered sufficiently definite and all of the eastern parts 

 of the continent might be included under the general designations 

 Carolina, Virginia and Canada. These early names were in many cases 

 taken up and variously interpreted by later writers. Some of them, 

 through study of the type specimens where they exist or of the plants 

 still in cultivation have been found to have precedence over later names 

 and more careful study will probably reveal other such cases. It is 

 possible too that duplicate descriptions of some identical plants have 

 been published under different names by later authors working inde- 

 pendently. 



The species have as far as possible been referred to the natural sections 



or groups to which they belong, those recognized being the twenty 

 wholly or partially arborescent groups of Sargent's Manual of the 

 Trees of North America and two additional shrubby ones, Uniflorae 

 and Pulcherrimae. Perhaps two or three other small groups should be 

 included in a fuller treatment of the North American species, at least 

 in the writer's opinion. The distinctions between these groups are usually 

 clear but in some cases there is an apparent merging of characters and 

 it is difficult to say to which of two sections certain species belong. 

 Where the group was indicated by the author of the species his treatment 

 has usually been followed when the species comes within the recognized 

 sections, although in a few instances there was some doubt in the writer's 

 mind as to the correctness of the reference. Where the original de- 

 scription does not make the group clear and no specimens were available 

 for examination I have seldom ventured to supply it and such species 

 are necessarily excluded from the synoptical tables. These, with other 

 debatable names that could not definitely be assigned to synonymy 

 and all others of which I have not seen specimens, are indicated by a 

 question mark in parenthesis. 



The geographical range is given usually by states and provinces 

 except in the cases of a few widely distributed species where it did not 

 seem necessary to enumerate all political divisions. Where a wider 

 range is indicated than that given in earlier publications it is in all 

 cases based upon specimens examined, most of which are in the Arnold 



Arboretum Herbarium. In giving type localities, details have sometimes 

 been supplied from the type specimens where these were available. 



The synoptical tables were designed merely as a temporary substitute 

 for a more complete key to the genus, the idea being to bring together 

 species having common characters into small groups so that in connec- 

 tion with the geographical range only a few would have to be considered 

 in identifying material. Like ordinary keys they are of course expected 

 to be used only as guides to the fuller descriptions of the manuals. 



In these tables the color of the anthers, number of stamens, glabrous 

 or pubescent character of corymbs at flowering time and general shape of 



