examined flowers, and concluded both were wrong, and that the flower had 2 sepal s and 2 petals, and therefore adopted Choisy's name as the better of the two, though we thought both bad. I have now ascertained that Willdenow was the nearer right of the three, but that we are all wrong. The examination of flower-buds before anthesis invariably showed a perianth of eight parts, while flowers after anthesis as regularly showed one of four, and these the exterior calycine series, the interior or coroloide ones having disappeared — being then neither apetalous, nor spurious, from the absence of its calyx, I have judged it advisable to give a new name, and have selected one equally expressive of past and, possibly, present blunders, which will not, I conceive, require further alteration.
§ III. Leaves of the Perianth 4 — (sepals 2, petals 2 ?- or sepals 4, petals 0 ? ) Apoterium, Blume.
9. C. Moonii, (R.W.) Ramuli terete, leaves long, linear lanceolate, (8-12 inches long, 1½ broad) acute, floral ones about oval : flowers short pedicelled, forming numerous small umbellate panicles : panicles axillary, solitary, congested towards the ends of the branches, forming together large leapy terminal panicles : floral envelopes 4, the interior pair the larger : fruit — .
C. longifolium, Moon's MSS. not Willdenow.
Ceylon. — Eastern Korle, Moon.
This plant seems only to have been found by Mr. Moon — and the specimen before me is not such as to admit of a very perfect character being made from it. The very long lanceolate leaves of the older branches (the larger ones at least a foot in length) and the smaller almost oval floral ones, mixed with the large terminal, very many flowered, panicles, seem to indicate, that the tree when in full flower, must be one of surpassing beauty.
10. C. Burmanni, (R.W.) Young shoots, quadrangular, tomentose, leaves oval or slightly obovate, spathulate, sometimes sub-emarginate, panicles small, axillary, few flowered, pedicels opposite, usually longer than the peduncles, furnished at the base with a small deciduous bractea : perianth four-leaved, fruit small, globose, or very slightly oval.
C. calaba, Lin. partly. — Burm. Thes. Zeyl. tab. 60.
Ceylon.
Burman's figure I consider the typical form of this species and is a good representation of the plant so far as it goes, the fruit only being required to render it nearly perfect, but there is now before me another form which I am uncertain whether to consider a species or a variety. The forms of the leaves are nearly the same, but more decidedly oval and much smaller, the inflorescence also differs a tittle ; in both it is imperfectly umbellate panicled, but in this the flowers are much more numerous and the peduncles and pedicels shorter, the structure, however, of the flower is the same: the fruit I have not seen, and while it remains unknown I prefer considering this form a variety rather than a species.
j3 parvifolium, young shoots quadrangular, tomentose: leaves from oval to obovate, spathulate, tapering towards the base : peduncles axillary, several flowered, rarely one flowered, pedicels opposite, short, furnished at the base, with a minute deciduous bractea : perianth four- leaved: fruit — .
I have a second variety from Mergui intermediate between the Ceylon ones. It has the larger leaves and small fruit of the first with the short umbellate panicles ef the second; and has in addition the bracteae, which
in the Ceylon plants are very small, almost inconspicuous and early deciduous, as long as the pedicels and as permanent as the flowers.
7 bractiatum. Leaves oblong oval or obovate, very obtuse, often emarginatc at the apex: panicles much shorter than the leaves, axillary, pedicels short, almost concealed by the lanceolate, acute, pubescent bractea: fruit globose.
Mergui.— Griffith Herb. Nos. 439 and 595— Flower and fruit. Apoterium sulatri. ? Blume.
11. C. tetrapetala, (Roxb.) " Leaves short petioled, oval, lanceolate, very finely serrulate : umbels axillary : corolla 4-petaled." Roxb. Fl. Ind. 608.
Moluccas. — Of this species I know nothing beyond what is expressed in the above brief character. — It how- ever belongs to this section, as he considers all the leaves of the pereanth petals, and characterizes the genus as having an 8 petaled corolla.
Species imperfectly Jcnown.
12. C. surega, (Buch. Roxb.) " Leaves linear, oblong, flowers verticelled below the leaves." Roxb. Fl. Ind. 2-608.
13. C. ? lanceolarivm, (Roxb. Fl. Ind. 2-608.) " Twigs square, leaves lanceolar, obtuse, lucid, finely veined" (acute at the base, Blume) Roxb. 1. c. C. lanceolatum. Blume, Bijd. 1-217.
Muritias. — Neither Roxburgh nor Blume have seen the flowers of this plant, our knowledge of its genus is therefore only conjectural.
14. C. poly ant hum, Wall. L. 4844.
15. C ? marginatum, 4845.
16. C. relusum, 4846.
17. C. pulcherimum, 4848.
18. C. amcenum, 4849.
19. C. mesucefolium, 4850.
20. C. longifolium, 4851. " Certe non kujus generis." (Wall.)
Kayea— Wall. Sepals 4, petals 4. (scarcely longer than the calyx) Stamens numerous, filaments united at the base; anthers orbicular, cells curved round the edge of a broad circular connectivum. Ovary free, 1 -celled, with several (4) erect ovules attached to the base. Style one, stigma 4-lobed. Fruit—
A very large handsome tree, young shoots slightly 4-sided — leaves opposite, approximated towards the ends of the branches, coriaceous, linear, lanceolate, acute. Panicles axillary and terminal, many flowered, congested near the ends of the branches, and forming by their union large terminal panicles.
K.florilunda, Wall. PI. As. Rar. 3, page 5, tab. 210.
This genus is very nearly allied to Calophyllum, agreeing with it in habit, inflorescence, and generally, in the structure of the flowers. But in this the connectivum of the anther is dilated so that the pollen cells, in place of being appi'oximated and parallel to each other, are remote and bent, being curved round its edges : and the ovary, in place of one, has four erect ovules. The two together form very sufficient grounds for keeping the genera separate. Kayea in short, stands in the same relation to Calophyllum that Gyjiotroches does to Garcinia. The cells of the ovary of Gynotroches have 3 ovules, while they are solitary in Garcinia. The ovary of Kayea has 4 ovules, while in Calophyllum it is solitary.
One plant appertaining to this order, yet remains to be noticed, though rather imperfectly known, as it e v j.
