18 ON SOME SPECIES OF GRACCA 
iii. 866 he adds Cracca bicolor Boiss. (which is apparently a Ben- 
thamantha), and associates it with C. Greenmanii and what he now 
calls ‘ Cracca villosa cinerea Li. (Kuntze). * Dr. Millspaugh points 
i i 
ie 
sitiastad at the same place and on the same day, belon 
B. mollis Alef. or a closely allied form. These, like Schott’s 
no. 865 (from Yucatan), are labelled ‘‘ Tephrosia spicata Torr. & 
Gray”; the two seem identical, and his 163 (from Merida) 1s 
also a Benthamantha ; the two last numbers are not cited by Dr. 
Millspaugh.t f 
Whether the plants here enumerated can be retained as distinct 
species is open to doubt, as various puzzling intermediate forms 
exist between them. Dr. Kuntze (Rev. Gen. i. 164, 165) points out 
the inconstancy of certain characters relied on by previous authors 
to separate the species, and reduces. all the plants which had up 
to that time been described under Cracca Benth. to varieties of 
C. caribea. 
In the National Herbarium there is a specimen from J acquin of 
his Galega caribaea. The leaflets are small, 19-21 in number, 
oblong or lanceolate-oblong and aristate-mucronate, sericeous- 
villose, and not, as Jacquin states in his description, glabrous. The 
plant is figured and described in Jacquin’s Selec irp. Amer 
as Tephrosia cinerea. The leaflets in Jacquin’s specimen are only 
5-6 mm. long; in Hahn’s plant they are 1-1-2 cm. ; in West India 
i 1 
* The modern American invention of trinominals seems to demand strenu- 
ous protest, 
We note that Dr. Millspaugh (I. c. p. 364) proposes a new name—Bauhinia 
Cavanillei—tor Pauletia inermis Cav.; he has appaneatty overlooked our identi- 
fication of this plant with B. ungulata L. (see Journ. Bot. 1897, 232.) 
| 
