1925] PALMER, SYNOPSIS OF NORTH AMERICAN CRATAEGI 9 



of specimens in cultivation seems to indicate that these characters, as 

 thus liberally defined, generally breed true in seedlings. Exceptions 

 may be expected in the case of hypothetical hybrids. It would therefore 

 appear that the number of stamens and color of anthers furnish one of 

 the most definite criteria for distinguishing between species, although 

 perhaps conclusive only when combined with other important though 

 variable morphologic characters; and they can scarcely be ignored, 

 especially in the construction of keys, unless we are willing to adopt 

 a much broader conception of a few variable and composite species. 

 This of course involves a controversial question upon which there is 

 the widest latitude for difference of opinion, and which is beyond the 

 province or purpose of this paper. It is perhaps safe to say, however, 

 that no competent student of the genus today would be willing to con- 

 sider returning to the score or so of species recognized by the botanists 



of a few decades ago. 



A brief synopsis of the fruit characters is given in the tables following 

 the name of each species. These have in almost all cases been taken 

 from the original descriptions and it should be remembered that a 

 considerable deviation from some of these specifications may be expected 

 in individual plants under varying ecological conditions. At least a 

 wide acquaintance with many of the species in the field and observation 

 of specimens under cultivation has led me to this conclusion. In a few 

 cases where it seemed evident that errors had occurred in the original 

 descriptions or where the examination of a series of specimens showed 

 too wide a departure from these descriptions, corrections have been made. 

 The question mark in parenthesis (?) following name of species and author, 

 indicates that specimens have not been seen or that the position of the 

 species in the tables is not definitely determinable from the description 

 drawn uoon. 



WITH 



AND RANGE 



Crataegus 



abbreviata Sargent in Jour. Arnold Arb. ill. 187 (1922).— Crus-galli. 



Texas (type locality: Brazoria), 

 abdita Beadle in Biltmore Bot. Studies, l. 75 (1902).— Flavae.— Florida 



(type locality: River Junction), 

 abjecta Sargent in Proc. Acad. Nat. Sci. Phila. lvii. 636 (1905). 



Intricatae. — Pennsylvania (type locality: Sellersville). 

 aboriginum Sargent in Rhodora, v. 163 (1903) = C. rotundifolia var. 



aboriginum Sarg. 

 abstrusa Beadle in Biltmore Bot. Studies, i. 66 (1902).— Pulcherrimae. 



Florida (type locality: Tallahassee), 

 accincta Sargent in Proc. Acad. Nat. Sci. Phila. lxii. 154 (1910). 



Crus-galli. — Pennsylvania (type locality: McKees Rocks). 



