REVISION OF THE NORTH AMERICAN SPECIES OF 
TEPHROSIA. 
B. L. ROBINSON. 
Waite the greater part of the difficulties of the genus 
Tephrosia lie, happily, beyond the geographical limits of the 
North American flora, yet the dozen species which inhabit 
sandy regions in our southern states are sufficiently variable in 
their foliage and similar in their floral and carpological structure 
tolead to diverse views on their specific limits and proper arrange- 
went. Miss Vail’s recent monograph? of these species, although 
sounding in long and detailed descriptions, fails to bring out 
early the primary divisions of the genus. The introductory 
*nopsis— which, although arranged in perfect accord with the 
meal sequence of the species in the paper, is styled an“ arti- 
‘: key” —is based chiefly upon the length and density of 
a features which are too variable and confluent 
aay diagnostic characters of the first rank. It seems to 
stem a that the sympodial and monopodial structure of the 
| Ciently nits difference much more evident and, indeed, suffi- 
“Shige to divide our species at once into two well- 
ts, eg 7 After this first division the size of the flow- 
emplo cg the inflorescence, and nature of the foliage may 
In mi hae greater diagnostic effect. 
Craceg er nation of species it is believed that Miss Vail’s 
is only separated artificially from her C. Small. 
ky adduced are chiefly the “truly prostrate ” habit of 
the eee and its narrower, more numerous leaflets. 
: to eam of the leaflets, this is (according to: Miss 
; Paely no ver, in C. Smallii, and 5 to 12™™ in C. floridana, 
Rid 'y Striking difference. In number the leaflets are 
na 
The distinctions 
AS to 
x“ V. ail) 
: "A tevis, nC. Smallii from 3 to 11, while in Mr. Nash’s 
>to on f ; 
Ps 25336, the North American species of the genus Cracca. Bull. Torr. Bot. 
ees 
193 
