ILLUSTRATIONS OF INDIAN BOTANY.
61
among the Chinese as a stimulent and restorative " but by Europeans and Americans is considered nothing more than a demulcent, approaching liquorice in its qualities" Lindley.
The same author justly remarks, " this however requires further investigation, for we can- not believe that all the Chinese say, believe, and practice, is fabulous or imaginary." If we look back to the medical history of Sarsaparilla, now so much and deservedly esteemed as a remedial agent, though 30 years ago much contemned, the sensible properties and mode of administration of which greatly resembles Gensing, we will perhaps find an explanation of this difference of opinion. This supposition seems the more probable, as American writers compare Arabia nudicaulis , another plant of the order, with Sarsaparilla, and affirm it to be as valuable a medicine. These cases serve to show that we ought not hastily to reject popular medicines merely on the strength of rough chemical analyses, or because their operation is so imperceptible that they produce no very obvious effect on the human constitution until they have been administered continuously for some length of time.
Remarks on Geneiia and Species. The number of genera assigned! to this order by DeCandolle and some other recent writers on botany amount to thirteen. Meisner raises these to 17 after removing Adoxa, one of DeCandolle's genera, thereby adding 6 to the number indicated by the former. But to four of these he has appended a mark of doubt thereby intimating that he is uncertain whether they really belong to the order. Of the doubtful ones two are Indian, one beautifully figured by Mr. Griffith in Wallich's PI. Asiat Rariores, the other, described but not named in our Frodromus under the provisional appellation of Araliacea ? Kleinii. Of the remaining genera several are not I think tenable, the distinctions being very slight and not supported by marked differences of habit. Between Hedera and Pa> atropia I can see no sufficient difference. DeCandolle defines Hedera " Styli 5-10 conniventes aut in unicum con- creti" and Paratropia "omnia Araliae aut Hederae, sed stigmata sessilia, primo approximata efc disco epigyno immersa" distinctions by no means readily obvious in practice, at least so I find them, as two specimens, one taken from a reputed Hedera the other a Paratropia, when laid side by side on the stage of the microscope, I found so like, that I could scarcely tell the one from the other. Further distinctions are taken from the calyx, whether the limb is a little longer or shorter, which are variable marks and not to be depended upon. In Hedera the petals are described in our Prodromus as cohering at the point and separating like a calyptra, while in Paratropia they are said to expand. This also in the examination of a number of specimens I find equally unstable and valueless. The difference between A r alia and these, consists in its styles being free and devaricately spreading, ("styli 5 expansi devaricato-patentis" DC.) surely a very inadequate generic character, though it might serve as a sectional one to aid in dividing a large genus. On this however I do not insist, as I have no genuine Aralia to examine.
Gilibertia, of which I have a species only slightly differing from Roxburgh's G. palmata, perhaps a mere variety, differs from the preceding genera in the length of its style only. In this it is distinctly prominent and conical, projecting some distance beyond the disc, but in other respects it seems sufficiently to associate. Whether the difference indicated, merits the distinction of elevating those plants in which it occurs to the rank of a genus, I am unable to say s yet, as it has already b°en so employed and is readily obvious in practice, I offer no objection, merely observing that G. JYalagu, the authority for which is Rheede's plate, Hort. Mab. 2-26, certainly does not belong to the order, but is a species of Leea. Sciodophyllum, the generic character of which, as given by DeCandolle, is " Omnia Aralias sed petala apicibus in oalyp- trae for mam cohaerentia " This reduces it to Hedera, as defined by us, but DC. gives Hedera free petals, which I find sometimes the case, sometimes not, a specimen now before me of H. He lex var chrysocarpa DO, having free petals, while the European plant seems to have them cohering. The character in short, is one of almost no value, and ought not to have so high an one assigned. The whole of these genera, as now defined, might I think with great advan» tage be reduced to one ; as genera grounded on such variable and inappreciable distinctions, can never be good ones, nor in any way tend to the advancement of science. Genera so purely artificial are misplaced in a natural system, where we look for natural ones, and the sooner they are discarded, and with them the doctrine which inculcates the non-existence of natural genera the better, as, it appears to me, nothing tends so much to undermine true science, as the maintenance of such principles. The existence of such a doctrine, confers on even the merist tyro,
