85 
As the result of a re-examination of the 
seed of Heritiera utilis* it seems desirable to transfer this species 
to the genus Jarrietia. Bentham and Hooker’s classificationt 
irregularly crowded anthers and albuminous seeds being stated 
y them to be characteristic of Sterculia and J arrietia, whilst 
parts which have all the appearance of cotyledons, but on 
has the appearance of being exalbuminous with thick plano- 
Convex cotyledons. It may, perhaps, be questioned w ether 
the single ‘character of the presence or absence of albumen is 
whether Argyrodendron is not generically distinct from the 
former. Until the genera of Sterculieae have been mono- 
Tarrietia utilis was discovered in the Gold Coast Colony by 
Mr. H. N, Thompson, and was described in 1908.t cl- 
mens collected by Mr. Thompson bore unifoliolate leaves only. 
Tn the following year the species was described independently 
m Ivory Coast specimens by Dr. A. Chevalier, under the 
name Cola proteiformis,§ the specific name being given in allu- 
22 to the polymorphy of the leaves, simple, trifoliolate and 
digitate leaves occurring on the same individual and sometimes 
;7en on the same branch.|| Gold Coast specimens with unifo- 
liolate, trifoliolate and septemfoliolate leaves were received 
Hee MS. PF. “Chipn in 1912, Finally in January of the 
* Kew Ball. 1909, p. 348; H. N. Thompson, Gold Cost, Report on 
Forests, t. 3. 
t Gen. Pl. vol. i. p- 215. 
+ Kew Bull. 1908, p. 257. (Triplochiton utile.) 
* Vég Ut. Afr. Trop, Frang., vol. ¥. p. 250 (1909). 
I] le. wi, Pp. 60, fig. 9. 
