ILLUSTRATIONS OF INDIAN BOTANY. 123 
the first living authority Mr. Robert Brown. He says in a letter to Dr. Graham referring to 
the plant which has called forth these remarks, “ In your plant the structure of the anther is 
indeed very remarkable and might well induce you to consider it a new genus; but it is right 
to add, that approaches to this structxre, and which serve to explain its analogy with the ordi- 
nary structure of the family exist in Garcinia, with which I suppose your plant would agree in 
its female flower as well as in fruit.’”’ From this concluding caution [ imagine that, before esta- 
blishing a genus on such grounds, he (Brown) would have ascertained the structure of the 
anther in the whole order, marked its variations, and then, and not till then, have determined 
on the propriety or otherwise of assigning a generic value to its variations: and [ can scarcely 
avoid thinking, that had sucha course been followed ia that instance a sectional value only would 
have been awarded. I confess that a less perfect examination of the order, than that which 
improved materials has now enabled me to effect, led me into a similar error, on which occasion, 
I proposed to subdivide the genus Garcinia into three distinct genera: Mangostana, Cambogia, 
and Stalagmitis (see Madras Journal of Science, vol. 4, page 304). This suggestion has not 
so far as Tam aware been yet adopted by any one, and [ trust it will not, as 1 now consider it 
wrong in principle, the variations in structure, there pointed out, not meriting a higher than sec- 
tional value in a genus so strictly natural. Influenced by this reduced estimate of the relative 
value of the several structural variations mentioned above, itis my intention on the present occa- 
sion to keep the old genus together, but divided into sections in accordance with them. I am 
induced to do so from observing that the variations are limited to the male flowers, and do not 
on any occasion extend to the female. For example G. Mangostana and G. cornea, are re- 
ferred to the same section, the former has 4-8 celled ovaries, and the latter usually 4, in G. Kydiana, 
Roxburgh describes the berry as being from 4 to 8 seeded, G. cowa from 6 to 8, and most of the 
others are described as having as far as 4, or 8 seeds, showing a general want of uniformity in 
this respect, variations, therefore, of the number of the cells of the ovary, cannot be admitted as 
generic, or even specific value in this genus. Should further acquaintance with the tribe show 
thdt in uniting Hebradendron or rather Cambogia, Lin, (for they are the same genus and the 
latter the more appropriate name) to Garcinia, I have erred, the error can be easily corrected, 
and in the mean time, my sections will afford the means of more easily determining the known 
species, and of referring to convenient places such new ones as may be discovered. For the 
present, nothing is more difficult than to make out from description the species of Garcinia. 
This is mainly owing to the male flowers, which afford by far the best specific characters, 
being too little attended to in characterizing them. Generally speaking, they are dioicous, 
and in collecting specimens care should be taken, to procure them of both sexes. ‘The foliage, 
except in a very few instances, do not afford good discriminating characters, and when it does, 
is usually accompanied by others which are more to be relied upon. : 
The following is the arrangement which I propose for the distribution of the species of 
Garcinia. 
Subgenus 1. WMangostana. Male: Stamens 4-adelphous, androphores, thick and fleshy, 
covered on all sides with anthers. Ovary rudimentary, supporting a large capitate, glabrous, 
abortive, stigma. Female—Stamens few, irregularly fascicled, usually imperfect. Ovary 4-8, 
or 10-celled. 
§ I. Anthers oblong, 2-celled, dehiscence, longitudinal introrse. 
: § II. Anthers flattened above, 1-celled, dehiscing by a transverse slit. 
Subgenus 2. Oxycarpus.* Male: Stamens monadelphous, androphores short, thick, fleshy, 
placed in the centre of the flower, usually 4-sided, anthers numerous, sessile or sub-sessile, capi- 
tate, with or without a sterile pistil—Female: Stamens 12-20, frequently imperfect, fascicled 
or monadelphous, forming a ring round the base of the ovary—ovary superior, 4-8 or 10 celled. 
§ I. Anthers 4-sided, with a polliniferous cel! on each side. 
* Cambogia would have been the more appropriate name for this subgenus, but the original Cambogia gutta erg hag 
been e-lavated to the rank of a genus eater new designation I have thought it better not to employ that name eee ae 2 
Botanists differing in opinion from me, should think that species generically distinct from Garcinia, in which case, Linnwus’s 
hame ought to be restored. 
