ILLUSTRATIONS OF INDIAN BOTANY. 
Leases 
215 
In addition to these I have specimens of what appears to me two new species, 
he one corresponds with DeCandolle’s W. crinerve in having three nerved leaves, but the 
Sale are quite cylindrical, and the leaves in place of being oval and about two inches long 
are ovate lanceolate, tapering to a point, and from 6 to 8 inches long, 
This species is allied t 
M. amplexicaule aud M. Heyneanum, but I think quite distinct from both. The lateral nerves 
are marginal, 
Memecylon cordatum, (Lam,) iri Sigg or sub- 
sessile, broadly ovate-cordate at the bare, very ob- 
tuse, or rounded at the apex; elude axillary, sub- 
capitate or corymbose, branches terete. Lamark Encl. 
4 p. 89. Ill. t. 284 f. 2. 
Ha as aie rT. 
sare leaves are as broad at the base as they are long, 
metimes quite sessile, sometimes furnished with a 
short petio 
Memecy lon jambosioides, (R. W.) Branches eylindri- 
cal, gla ches us: leaves ovate, lanceolate, acuminated ; 
3. nerved; the lateral pair of nerves abe arginal, united 
with smaller transverse parallel veins: flowers numer- 
ous, forming axillary capituh Aa palicelled. 
Hab.—Courtallum We SeCe 
aves are almost entirely 
those, on my specimens, are upwards of 10 inches long. 
The flowers, before expansion, form deuse nearly ses- 
The other species, which approaches 
long narrow linear lanceolate blunt Se gece leaves ; about six inch 
Aves coher, in habit, differs 4 in having 
sile axillary clusters, and the pedicels do not afterwards 
materially elonga The fruit is glabrous, polished, 
judging from preserved specimens, a dry berry.—Seed 
one, exalbuminous, cotyledons foliaceous, exceedingly 
convolute 
Memecyton rai ifolium, (R. W.) Branches, terate 
leaves, confined to the ae Rt ty — — 
attenuated below, blast 
cles short from oes scars 0 of fal e 
numerous, umbellate, or sub-capitate, pediedlis om vey 
Habitat. aJengle about Courtallum. 
This species only differs from I. 
form of the Pipl in age wd are broad in Pipa er 
to their lengt s long by on half 
road, or hal th that size; while i in coed ong are from 
a to four inches long and scarcely half an inch 
oad, tipsy to the petiol below, but ‘rant pointed. 
The flowers and inflorescence are the same in both. 
EXPLANATION OF PLATE 93, 
- Memecylon ramiflorum, m.) M. yi 
Pe cing W. and A.) vblgiec 4 Mota natural si. 
2. An expande 
3. Stamens, but the acl ers very incorrectly repre 
sented, oval, dehiscing their whole length in _, of 
2-celled abor ve, the base ending in a membranou 
a of the connective. 
4. A an fa 
5. The ovary shortly seqtd the en “ the petals cut 
vertically, seboalt ng the ascending 
6-7. A berry nt shiniatare cut ‘canapecaaly, ie to show 
the numerous s 
LXII.—MELASTOMACEA. 
A vast tropical order, containing according to Meisner’s list, 100 Sgr a four of 
tribe of Melastomac 
Memecyleae, which he considers a tribe only 
maceae and not a order, This 
union of these two families i is, it appears as me, scarcely authorized by the Saractes of the flow- 
ers, and is not supported by the frait and = w hich differ “emegede in the two orders, 
to DeCandolle’s enumeration there area 
ie g 
mber, [ believe, has Sits: very 
ecies,which n 
roa teed augmented since the siabstieatiiin of fas work, Of that number aisha America 
ing a rote ad sma!l number for Asia, Africa, and New Holland. 
In our species only are des 
cribed, a number which m 
more recent 
sions have not euadetially extended, though they have afforded ample evidence of the aificalty 
of distinguishing the species of this order, through sand tendency to vary their 
for example on comparing many very dissimilar looki 
A more intimate sagosinlancs, hamenss a tet arog pA 
distinguish them as species 
ae furnish us with characters by which to augment t the num 
rms. I fi 
mens, no good marks by which to 
e parts 
