and during the winter retaining its deep green foliage, stained 
beneath with rich crimson. It is very easily cultivated, and increases 
freely by cuttings, or by seeds, which are produced in abundance. 
Our drawing was made in the Horticultural Society’s Garden, in 
September 1826. 
With respect to the genus to which it is referable, we think 
it cannot be otherwise than akin to the herbaceous Melastomacee 
of North America, upon which the genus Rhexia chiefly depends ; 
with these it agrees in having a persistent limb to the calyx, in the 
form and insertion of its anthers, in the form of the seed, in the 
texture and degree of cohesion of the capsule and calyx, and, in short, 
in all those characters which Mr. Don has assigned to Rhexia 
properly so called. 
As to the value of the above-named peculiarities, and of others 
upon which the genera of Melastomaceæ have by Mr. Don been made 
to depend, there is undoubtedly some difference of opinion. M. Achille 
Richard, in the Dictionnaire Classique d’Histoire Naturelle, for 1826, 
objects to them as insufficient, and as giving rise to new artificial 
divisions in the room of the old ones, that are natural. In the opinion 
of this learned Botanist, the whole order should be reduced to two 
enera; viz. Melastoma, which is distinguished by its fleshy 
ruit; and Rhexia, the fruit of which is dry and indehiscent. But 
we certainly can neither ‘adopt this mode of reasoning, nor avoid 
expressing our astonishment that M. Richard, who must be fully 
aware of the extreme inconvenience generally attending such a 
measure as that he proposes, should have been led in this particular 
instance to adopt it, especially as thé ground assigned for so doing is 
not for a moment tenable. All genera, and indeed all*the divisions 
of Naturalists, are necessarily artificial; and when one genus is called 
natural, and another artificial, all that can be meant by such expres- 
sions is, that the species of the one are less artificially combined than 
those of the other: this, we apprehend, is, at the present day; an 
universally admitted principle, to the proof of which we need not 
preceed ; and, considered in this view, there can be no question that 
» Richard’s proposition cannot be received. The fact, with respect 
to Melastomacese, as with Pomacez, and all similar natural groups of 
plants, seems to be this,— that as their species offer less distinct 
modifications of the parts of fructification than those of many other 
orders, the difficulty of subdividing them into those divisions that 
Naturalists call genera is very much increased, and when effected, 
the divisions or genera are obliged to depend upon less obvious 
differences than we have the power of assigning to those of natural 
orders in which less uniformity of organisation exists. In Melasto- 
maceæ, we may remark, that one character is to be found of which we 
suspect that use may hereafter be advantageously made: we allude 
to that cyathiform termination of the capsule which is so conspicuous 
in Sonerila, an undoubted Melastomaceous genus, which exists in 
various states in other genera, and which in Rhexia is wholly pee 
