104 
Selections. 
[no. 3^ NEW SERIES^ 
plant I have seen, the sepals are large and leaf -like, completely 
covering the fruit. (It is probably I'rophis spinosa, Roxb. — Epicar- 
purus Timorensis, Dene.) Thwaites. 
We are indebted to Dr. Arnott for the following notes on the 
genus Epicarpurus and its allies : — 
Blume, in his Bijdr., p. 488, has established the genus Epicar- 
purus for a plant he calls E. orientalisy and for which he cites 
Rheede, Hort. Mai., vol. i. t. 43 ; this last is universally allowed 
to be Trophis aspera. Blume, in the edition I possess, and the 
only one I ever heard of, gives us no information ac to the relative 
size of the two cotyledons ; but M. Decaisne, in his ' Herb Timo- 
rensis Descriptio,' p. 171, says, " M. Blume indique, dans son 
Bijdragen^ les cotyledons de son Epicarpurus comme etant inegaux." 
In the E. Timorensis, which Decaisne describes and figures in that 
memoir, the cotyledons are represented unequal, but he adds that in 
" Trophis aspera. Wall. L, n. 4,640," the cotyledons are foliaceous 
and equal. 
I do not know precisely what plant M. Decaisne had before him, 
but in all that I have examined under the name or similitude of T, 
aspera, the cotyledons are nearly as described by Roxburgh in his 
Flor. Ind. vol. iii. p. 761, and represented in a drawing in the 
E. I. C. Museum, tab. 118, viz., " cotyledons two, very unequal, 
the largest being nineteen-twentieths of the whole embryo, and 
one side divided half-way through into two lohes : the small cotyledon 
is hid between the lobes of the larger one.'^ If M. Decaisne has, from 
its smallness, overlooked the one cotyledon, and mistaken the two 
lobes of the greater one for two equal cotyledons, the difference be- 
tween his, Roxburgh's and my observations will be accounted for. 
At all events, I consider that T. aspera (and so marked is that 
species that I have seen no other confounded with it) must be held 
as the type of Epicarpurus of Blume. 
In his * Bijdragen,' p. 507, Blume suggests that his Urtica spi- 
nosa is another species of Epicarpurus : Decaisne adds E. Timoren- 
sis, and says that Alhrandia of Gaudichaud also belongs to it. Gau- 
dichaud's character of Albrandia, in Freycinet's Voy. p. 709, is too 
