sular flora, 6 genera are enumerated, only two of which are not found in Europe Photinia and Eryobotria, since the publication of our work one (Alchemilla) has been added to the continental flora and two to the Ceylon one, namely, Agrimonia and Alchemilla, and what is remarkable the species are identical with European ones. To these may be added, I believe, two of the sub-order Amygdaleae from Ceylon of which I have specimens, but the fruit too young to enable me with certainty to identify the genus, though, I think it may with considerable confidence be referred to Blume's genus Polyodontia; as, in addition to the similarity of the flowers, the leaves have the two glands on the under surface near the base. In Nepaul this order is comparatively abundant. From that country there several species of Spiraea, many of Potentilla, one or two of Geum, several of Cerasus, one or two of Prunus, one of Pyrus &c., and from the same tract of country Royle's Prinsepia, a genus which Messrs. Royle and Lindley refer to Chrysobalaneae, but which, so far as I can make out from his not very perfect figure and character strictly belongs to Amygdaleae, and but for the baccate fruit might pass very well for a Prunus, whence I am disposed to infer that these two genera are very nearly allied, and that Lindley is premature in departing from the generally received opinion, that Chrysobalaneae are but a section of Rosaceae. That Prinsepia does not belong to Chrysobalaneae is shown by its style being terminal not arising from the base, and by its pendulous, not erect ovules. It differs from Cerasus and Prunus by having a baccate, not drupaceous fruit. Dr. Lindley proposes to remove the section Sorbifoliae from Spiraea, and form for them a distinct genus under the name of Schizonotus. His reasons for this separation are not stated by himself but Meisner assigns the following brief character. " Stamina disco calyce ubique adnato inserta, ovaria 5 connata stipitata. Caet spiraeas" one species only I have examined Sp. (Schizonotus) Lindleana? and find that he (Meisner) has in that definition missed the only character of any value, namely, the position of the ovules. In Spiraea they are superposed, sometimes only two, sometimes several, but in Schizonotus Lindleana there are several 10 or 12 all pendulous from the apex of the cell of the ovary. This structure added to the very different habit seems to afford good grounds for the formation of a new genus. In it the ovaries are not stipitate.
POLYODONTIA, Blume.
Calyx inferior, campanulate, 6-toothed, deciduous : petals 6-minute, inserted on the margin of the calyx : stamens numerous, (12-18) about equal, inserted with the petals: ovary free, 1 -celled, with 2 pendulous ovules: style one : stigma peltate : drupe reniform, dry, 1- seeded: embryo exalbuminous, inverse.
Leaves alternate, entire, exstipulate, usually furnished beneath, with two glands near the base : racemes axillary and lateral, solitary, or several together, tomentose, flowers small, 1 bracteated. (" Genus hocci vix ali Amygdala differre videtur" Blume.)
P. arborea, (Bl.) Leaves alternate, oblong, entire, exstipulate, bi-glandular, beneath near the base : racemes axillary and lateral, solitary, or several ; tomentose. — Blume, Bijdr. p. 1105.
Hab.—Alpine forests in Java.
P. ? Ceylanica, (R. W.) Leaves from elliptic very obtuse at both ends, to sub-orbicular, glabrous, when dry, of a rusty brown colour beneath, racemes axillary, solitary, (always ?) about the length of the leaves, covered with short appressed hairs : flowers small, petals 5 reflexed, externally very hairy round the margin.
Ceylon in forests above Numbady.
My specimens were taken from a tree by the roadside, near the highest point of the ascent towards Newera Ellia.
P. ? Walkirii, (R. W.) Leaves oblong ovate, sub-acuminated, glabrous above, and with the young branches tomentose beneath, afterwards becoming glabrous, racemes axillary, solitary, petals and rachis densely clothed, with short rusty coloured tomentum.
Ceylon.—Communicated by Colonel Walker.
This species seems nearly allied with the Java one, but differs in having the ramuli and leaves tomentose. It certainly appears distinct from the preceding, but as I. have only a solitary specimen I am not enabled to give more perfect distinguishing characters.
EXPLANATION OF PLATE 85.
- 1 . Potentilla Kleinian, a small flowering plant—natural size.
- 2. An expanded flower.
- 3. The same partially dissected.
- 4. Stamens.
- 5. The ovary cut vertically, showing the attachment of the carpels to the surface of the receptacle.
- 6. A detached carpel.
- 7. The same cut vertically.
- 8. Cut transversely.
- 9-10. Portions of a leaf magnified.
