Page:Illustrations of Indian Botany, Vol. 1.djvu/208

98 the preceding sections of DeCandolle were untenable, not. only -keeps these genera distinct, but places seven genera between in his arrangement. The memoir itself I have not seen and cannot state his reasons for placing them so far apart, but it appears, from a remark of Griffith, that he was not acquainted with the fruit of Camellia, which may in some degree account for this unnatural separation. No species of Camellia has yet been found in the Peninsula of India.

After this article was written, and partly printed, I received Professor Meisner's Genera Plantarum* in which he gives in a tabular form, generic characters of all the genera of each natural order. From this work I have become acquainted with Cambessedes' characters of Camellia and Thea which, with a view to the fuller elucidation of the subject, I shall transcribe. Departing from the distribution adopted by Cambessedes he (Meisner) places them nearly side by side in the same section, under the following sectional character.

Sectional Character. — " Petals 5 more or less united at the base — Styles 3-5 cleft or parted, (that is composed of 3-5 styles united at the base or beyond the middle but not their whole length up to the stigmas.)"

— Calyx of 5-9 sepals, often in 3 series, the interior ones larger. Petals 5-9. Styles more or less deeply 3-5 cleft. Ovary 3-5 celled. Cells 5 or many ovuled ; ovules pendulous,, partitions thick, capsules woody, ovate, loculicidal, 3-4 valved.

— Calyx 5 parted, segments equal. Petals 5 9. Styles 3 cleft above the middle. Ovary 3 celled. Cells 4 ovuled, the two superior ovules erect the two inferior ones pendulous. Capsule 3 celled (often by abortion 2-1 celled) cells opening at the apex, one, rarely 2, seeded. Seeds wingless exalbuminous.

From a comparison of these characters it will be at once observed that the only point of structure not common to both is confined to the ovules, and the dehiscence of the capsules, which, in Camellia are said to be 5 or many and all pendulous, with loculicidal dehiscence, while in Thea they are described as limited to 4, namely, 2 ascending and 2 pendulous, and the cells opening at the apex. Mr. Griffith having ascertained that the dehiscence is the same in both, has removed one of the marks, and that the other can alone be admitted as a good generic distinction, I am far from thinking, though I confess, that in many instances even small genera are broken down on equally trivial grounds. But I consider the practice a tad one, and therefore fully agree with Mr. Griffith in thinking these two should be united.

The species of this genus are of difficult discrimination, and I begin now to doubt whether E. Wightiana and E. tristyla, are really distinct. The specimens from which the characters were taken certainly look different, but they were few in number. Considerable additions to my stock have, by supplying intermediate forms, rendered it probable that they are but varieties differing principally in luxuriance. I mention this with the view of directing attention to the subject. Assuming that they are distinct, I have now, from Ceylon, an intermediate form differing from both, in the leaves having the thick coriaceous texture of the one, and the large size of the other, while it differs from both in having hairy branches : this, on the supposition that the others are distinct, must form a species. I therefore call it

E. Ceylanica, R. W. extreme branches and costa of the leaves, hairy, leaves coriaceous, oval, acute at the base, retuse, and somewhat acuminated at the point, serrated and revolute on the margin, glabrous : flowers fasciculed, Suit small, crowned with the short persistent style and 3 stigmas.

Ceylon. — Colonel Walker.

My specimens are in fruit, and I have not seen the flowers. The leaves are from 1½ to 2 inches long, and about half the breadth in the broadest part : very hard and coriaceous in their texture, glabrous on both sides, except the prominent midrib, and finely serrated on the margin. The hairy branches seem to afford the best distinctive mark.

E. angustifolia which I have not found since my return to India, but of which I have received specimens, through the kindness of Mr. Griffith, from Mergui, appears certainly distinct.

already stated, I cannot agree with the able author in retaining all the three genera— Ganitrus is not taken up as a distinct genus in this arrangement. One error into which he has inadvertently fallen, I have met with and shall here notice. Dr. Meisner observes that we (Wight and Arnott) have in our Prodromus, referred the genus Lophira to Dipterocarpus. This is a mistake, originating in our having quoted Gartner's plate, No. 188, for Dipterocarpus without noticing Lophira which is figured in the same plate, but the genus Lophira is not once mentioned by us, whence I infer, he had not consulted the plate when writing his commentary.
 * " Plantarum Vaseularium genera, eorumque characters et affinitates, Tabulis Diagnostics exposita, et secundum ordines naturales digesta." I may here observe with reference to the Elœucarpeœ: as arranged in this most valuable work, which I shall very often have occasion to quote, that Professor Meisner reduces the order to a section of, Tiliaceæ and adopts DeCandolle's Accra. Hum and Elaocorjjus still retaining his characters. The genus Manama is also retained, with most of the species I have indicated referred to it. For the reasons