Page:Miscellaneousbot01brow.djvu/431

 RAFFLESIA ARN'OLDf. 413

I may liereafter have an opportunity of entei'ing fully [-232 into the question whether Kajjleslacccn and Balanophoreo'

received several names, as vasa porosa, puncfata, vasiform cellular tissue, dotted ducts, &c., and which are so conspicuous in the majority of arborescent Phocnogamous plants, have never been observed in any part strictly belonf^ing to these parasites. But even adniittinc: the non-existence of the large vessels liere referred to, their absence will hardly be regarded as a sufficient reason for the union into one class of the two families in question, especially when it is considered —

First. That conformity in vascular structure, even when accompanied by peculiarity of tissue, does not always indicate, much less determine, botanical affinity. This is strikingly exemplified in Conifercc and Winteranea>, two families which, though so nearly agreeing in the uniformity and peculiarity of their vessels, and in both of which the large tubes referred to are wanting, yet differ so widely from each other in their organs of reproduction and in their leaves, that they may be regarded as placed at opposite extremities of the scale of Dicoti/ledoiies.

Secondli/. That uniformity of vascular structure is not always found in strictly natural families. Thus many tropical woody^ climbers exhibit remark- able peculiarities of vascular arrangement not existing in the greater part of the families to which they respectively belong, but which peculiarities appear to have no inlluence whatever in modifying their reproductive orfrans.

Thus also in Myzodcndron^ the whole woody tissue consists of vasa scalari- forinia, a peculiar structure, and very different from that of all the other genera belonging to Loranthacecc, to which this genus has been referred, and to which, though it does not absolutely belong, it is nearly related. Even this peculiar structure of the stems of Myzodendron admits of considerable modifications in the different species of the genus, which is strikingly exemplified in comparing the loose vascular tissue with large and singularly constructed medullary ravs of M. brachi/stuchyum and quadrijlorum witii the more minute vessels and ex- tremely narrow rajs of J/, imnctulatnm.

I may also notice that in TUlandsia nsneoides, as well as in the nearly related species of that genus, the capillary stems are destitute even of spiral vessels, though in Brorneliaccce generally the ordinary vascular system is found.

Whatever may be the state of vessels in the fully developed parasites belong- ing to RaJlesiaceiP, it appears to me that at least RaJ/fesia in its very early

��1 Mi/zodendron of Banks and Solander, from /ly^sw or fxv^u) sier/o, and <}kv^oovy lias been changed to Misodendron by De Candolle and all following systematic writers; no doubt merely from a mistake as to the intended derivation. Mi/zo- dendron, hitherto referred to Loranthaceoi, to which it is certainly closely allied, especially through Aiiddap/uie of Poeppig, appears to me to have characters sufficient to distinguish it as, at least, a suborder or tribe {Myzodoidrccc), namely, the structure of its ovarium, in which it approaches to Santalacece, having three ovula suspended from the apex of a central placenta, only one of which ripens; the entire absence of floral envelope in the male; the singular feathery appendages of the female flower and fruit compensating in the dis- persion and subsequent adhesion of its seeds, which are destitute of that viscidity existing in those of the parasitic Loranthacece. ; and lastly, the embryo being undivided, with its dilated and exsertcd radicle enclosed in a semi- transparent covering, a continuation of the membrane lining the cavity of the albumen in which the cmbrvo is lodged.

�� �