20 
CONTRIBUTIONS TO BOTANY. 
holes, opening into cavities of the internal condyle, which forms 
a globular expansion on that side within the cell, and upon which 
the hollow fungilliform seed is moulded. The embryo is situ- 
ated on the opposite or dorsal side, beneath an extremely thin 
coating of simple albumen, which is convex on that face, while 
it is extremely concave, very thick, and deeply cleft all over the 
opposite side, the thin integuments entering into its numerous 
anfractuosities. The embryo therefore partakes of the convexity 
of the dorsal side, its small terete radicle pointing to that part 
of the cell opposite to the style ; while its large cotyledons, of 
very delicate texture, are divaricately separated on the same 
plane, deeply laciniated all along their margins, and enclosed in 
cells of the albumen of corresponding form. In analyzing this 
seed, the greatest care is necessary in removing the thin plate 
of albumen which covers the embryo; and it requires a previous 
knowledge of its position in order to extract it entire, as the 
force required to break away the albumen, which is solid be- 
tween the sinuosities of the lacerations of the cotyledons, is 
likely to injure it, — a caution that is requisite in examining the 
seeds of the whole of the Heterocliniete. 
The authors of the ‘ Flora Indica ’ place this genus in a dis- 
tinct tribe {Cosctniece), on the ground that its petals are larger 
than the sepals, and that the structure of the seed is different 
from that of the Heterocliniece ; but in this they have considered 
the three inner sepals as petals, and they have relied upon 
Gaertner’s erroneous figure and description of the seed, not 
choosing to place faith in my more accurate analysis, to which 
they allude*. It wdll be seen that there is nothing in the struc- 
ture at variance with all the other genera of the Heterocliniea, 
except in the lacerated margins of the cotyledons : there are 
therefore no grounds to justify the retention of the Cosciniece as 
a distinct tribe, which view has been confirmed by the authors 
of the ‘ Genera Plantarum,^ who now reject it. The inner row 
of sepals, considered as petals by the before-mentioned botanists, 
differ in no respect in their form and appearance, except in size, 
from the more external rows — a circumstance of uniform occur- 
rence throughout the order. In Abuta we have exactly the same 
number and arrangement of floral envelopes as in Coscinium, 
and they have always been considered as sepals, which is proved 
by the existence of the ordinary form of scale-like petals in Tilia- 
cura, where all the other floral envelopes correspond with those 
of Abuta, We may therefore conclude that in Coscinium, as in 
* After the publication of the ‘ Flora Indica,’ I obtained, through the 
kindness of Sir W. Hooker, fresli drupes of Coscinium, by which my 
previous analysis of seeds given me by Prof. Lindley was completely con- 
firmed. 
