This exposition of the characters of the genera of this order leaves no room to doubt, that the plant have figured is a true Michelia,
The genus Michelia under Blume's amended and simplified character, is one of easy recognition, but the determination of the species, owing to the great accession which has been made to their number, is now most difficult, not so much, perhaps, from their not affording adequate discriminating marks, as from their never having been subjected to a sufficiently comprehensive scrutiny to admit of their proper characters being elicited, by a careful comparison of' one with another. Nor. indeed is this to be wondered at, when we consider how seldom opportunities occur of examining them in their native places, and how few have yet found their way into cultivation. It is not without feelings of regret that I find myself forced, from the imperfection of my materials, to leave this task to another, or at all events to postpone the examination until some future opportunity, as the very few species of which I possess specimens, barely enables me to conjecture, what series of organs are most likely to furnish either good specific marks or sectional subdivisions. For the latter purpose the number of spathes or bracts enveloping the young flower bud may prove serviceable, namely, whether two or three. For example, in M. Champaca, there are two, one exterior, and early caducous, which, (for convenience) I call bractial,and one closely embracing the flower, which (for the same reason) I call calycine, while in M. Pulneyensis there are three, one bractial and two calycine. The number of petals will perhaps be found to furnish another set of good characters, though for the present that seems doubtful ; but the number in each verticel whether 3, 4, or 5, promises to afford excellent sectional characters, since it may be presumed, that that series of numbers will be constant in each species. These structural differences, aided by variations in the forms and surfaces of the leaves ; whether rough or smooth, glabrous or clothed ; the colour, kind, and degree, of clothing, of the spathes ; the form, size, colour, and whether smooth or warty, of the carpels ; and lastly, the number of seeds in each compared with the ovules, ought I think, to present such a combination of easily recognizable characters, as should leave but little difficulty in distinguishing a much more extended series of species, than we have any reason to believe appertains to this genus.
To what extent these hints for the discrimination of species will be found to avail in practice, it is difficult to say, but it seems desirable that they should speedily be brought to the test of experience, since there is no genus, of the same extent, in which it is so difficult to determine the species. Much attention, and the examination of numerous specimens will no doubt be required towards the determination of the value of characters taken from the corolla, but those taken from the spathes and fruit, will, I suspect, be more steady, and, by so much, more valuable, at the same time, that they are more easily ascertained.
Adopting the number of spathes as a sectional character in the manner mentioned above, the following distribution of species into two leading sections, will probably be found correct.
1st. Flower buds with one bractial and one calycine spathe.
M. Champaca, M. excelsa, M. Doltsopa, M. Kisopa, M. aurantiaca, M. fuscata, M. nilagirica? and three undescribed species in my herbarum frosn Ceylon.
2d. With one bractial and two calycine spathes.
M. Pulneyensis, M. lanuginosa (?) and perhaps some of the Magnoliae of DeCandolle's second section, which are referable to this genus.
The Nepal species with the exception of the last, T have referred to the first section on the authority of the following passage in Wallich's Tentamen Fl. Nepalensis, " they (the flower buds) are enveloped in two entire membranous rounded spathes, the outer one much sooner caducous than that within." Such is the case with the outer or bractial spathe of M. Pulneyensis, while the two interior ones seem to separate about the same time. M. lanuginosa, I have doubtfully referred to the second section, owing to two bracts being represented in the figure as if caducous about the same time ; while in the others, one only is figured. Should the sectional characters I have proposed be found applicable to all the species I have respectively referred to them, there can be no doubt of my plant being distinct from all, unless it should be found that M. nilagirica is incorrectly described, which I fear is the case, though from never having seen specimens, I do not feel myself at liberty to offer any opinion on the subject. I may here observe that the cluster of fruit represented in my figure does not belong to the true M. Pulneyensis, but to a new and distinct species, (namely, M. Rheedii of this work,) and
