Page:Illustrations of Indian Botany, Vol. 1.djvu/259

Rh in Salacia the flowers are for the most part congested in the axils of the leaves. Of this order Roxburgh seems only to have known five species, three of Hippocratea and two of Salacia, or Johnia as he, supposing them a distinct genus, called them. Under Wallich's hands the number has swollen to 19, but whether these will all prove distinct when thoroughly examined and compared may be doubtful, as some of them vary considerably according to circumstances.

It is not improbable that the species of Hippocratea here figured may be one of them, but as it was impossible for me to determine that from a name only, and as it evidently differs from all those described by Roxburgh, I could not hesitate about naming it. I have another species from Bombay, communicated by Mr. Graham under the name of H. obtusifolia. It does not however correspond with my specimens of that species nor indeed with any of the other Indian ones with which I am acquainted. From H. indica these three species are all known at a glance by their larger sized flowers; from each other they may be thus distinguished. H. obtusifolia, calyx fleshy, entire on the margin, petals triangular, tapering to a point. H. Grahamii, calyx membranous, fimbriated on the margin, petals obovate—spathulate. H. Arnottinna, calyx flashy, lobes obtuse, entire on the margin, petals suborbicular, unguiculate, reflexed. H. Richardiana of the Flora Senegambias, if really distinct, must be very closely allied to our H. obtusifolia, (I suspect they are the same species) the figure of H. paniculata of the same work, shows that it more nearly approaches my H. Grahamii, but is quite distinct.

The species of Salacia cannot, so far as I have yet been able to ascertain, be thus briefly and clearly distinguished by the flowers. The fruit seems to afford better distinguishing marks but is often wanting in preserved specimens. The inflorescence also gives several pretty good characters such as, whether the peduncles have one or several flowers, whether in the former case, there are few or many aggregated in the axil of the leaf, and whether they are longer or shorter than the petiol—according to these marks they may be thus distributed.

1.—Peduncles 1-flowered—(from an axillary tubercle).

A. Peduncles few—(1 or 2 to 6 or 8).

''S. Brunoniana. S. Roxburghii. S. prinoides. S. macrophylla, Blume, one from Ceylon, S. reticulata, R. W. one and lastly, S. senegalensis'' distinct, from Malabar, though for the present united with the Ceylon one, D. C.

B. Peduncles very numerous.

Two species from Mergui of which I have specimens are referable to this subsection—both, so far as I know, undescribed S. verrucosa, R. W. and S. multiflora, R. W.

These two subsections are perhaps too artificial to prove of much value in practice, as it is evident that change of circumstances may cause a species to pass from the one into the other.

2.—Peduncles several flowered.

A. Peduncles 2-3 cleft, bearing few subsessile flowers on the apex.

''S. pomifera. S. oblonga. S. Javanensis'', Blume. S. oblongifolia, Blume. S. melitotarpa, Blume.

B. Peduncles forked, each division terminating in a simple many flowered umbel, flowers longish, pedicelled.

To this subsection one Mergui species S. floribunda, R. W. belongs.

The new species of both genera of this enumeration may be thus designated and defined.

H. Arnottiana, (R. W.) Shrubby, twining, glabrous, leaves ovate, oblong, acuminated, coriaceous, remotely crenato—serrated on the margin; panicles small, few flowered, axillary and terminal : petals from broadly ovate obtuse to subreniform, unguiculate, carpels linear, spathulate, broadly  emarginate at the apex,

Hab.-Malabar.

The flowers of this species are perhaps, about the