424 NEW SPECIES OF CRESCENTIA. 
may ultimately prove to be distinct from those which are 
already described, it has at least afforded me an opportunity 
of minutely examining its structure; and, as the genus has 
not yet received a fixed “local habitation” in the Natural 
System, I may be allowed to make a few observations on its 
affinities. Notwithstanding that the fruit is fleshy and inde- 
hiscent, if the pulp and seeds are scooped out of it when 
ripe, the internal surface of the pericarp presents the follow- 
ing appearance :—A well marked suture is seen dividing it 
into two portions which stand anterior and posterior to the 
axis of inflorescence, while another which is less distinctly 
marked, crosses this, and is no doubt the midrib of the peri- 
carpal leaves. This structure at once refers the genus to the 
dicarpose group of Dr Lindley’s monopetalous plants, and 
its unimbricated calyx, unsymmetrical flowers, and ex- | 
albuminous seeds, unattached to placental hooks, place it in 
the Bignonial alliance of that group; and it is consequently 
with the Orders contained in it that Crescentia has the most 
numerous resemblances. The Orders of this alliance are 
Pedaliacee, Bignoniacee, and Cyrtandracee; andit isin Big- — 
noniacee that Dr Lindley has placed, apparently provision- — 
ally, Crescentia; but it seems to be very different from the 
normal genera of that Order, in its 1-celled ovary, four pari- 
etal placentz, fleshy indehiscent fruit, and wingless seeds. 
os: rs also in the anomalous structure of its calyx, —— 
: although that of Spathodea is somewhat analogous; and the — 
didynamous character of Crescentia differs very materially from — ; 
2 that of Bignoniacee, the posterior pair of stamens in the former : 
being the longest, while in the latter, the anterior pair are 
on ge than the posterior. With Bignoniacee it agrees in 
, and approaches it somewhat further through Eceremo- 
, which has a 1-celled fruit, and parietal placentæ 
economy of its SES Crescentia is more ch 
