It is curious that all plants hitherto found parasitical on roots, have no green leaves; to this, marked exceptions exists in Cuscuta and Cassytha, such true-leaved parasites being found only on the ascending axis; this rule is so permanent, that species of certain genera, such as Burmannia, the bulk of which are not parasitical, have no leaves. The mode of attachment of all parasitical plants is I think the same, otherwise I should suspect the above difference to point to a marked one in the nature of the fluid derived from the stock: thus leafless plants might be supposed to induce no particular change in the fluid they imbibe, while the others might be supposed to elaborate their own from that of the stock.
There is another very remarkable circumstance connected with the most typical leafless parasites, in their very frequent limitation to the genus Cissus, on which perhaps all Rafflesiaceæ and Cynomorieæ are exclusively found.
My chief reason for supposing Sarcocodon to be Monocotyledonous, or rather Endogenous, is the ternary division of its parts, and if my supposition be correct, it tends to establish, if indeed other ample evidence did not exist, the great permanence and consequent value of this numerical character.
And with respect to Sarcocoidalis I shall adopt the same opinion, if I find on enquiry that a binary number, and imperfection of the female as compared with the male, are more characteristic of Endogenous than of Exogenous growth. This same genus I consider in both these characters to allude to some analogy with one or more Acrogenous divisions.
The establishment of the order of Rhizanths, as well as that of Gymnosperms, I consider as a retrograde step in Botanical science. It is totally opposed to all sound principles of classification, and is a proof that, in the nineteenth century, arbitrary characters are still sought for, and when found are obstinately maintained.
Even in the arbitrary character, which is considered as destructive of all their other claims to ordinary vegetable rank, there is no unison whatever, for Rafflesiaceæ have ordinary ovula, while Sarcocoidalis very extraordinary.
The amount of testimony proving their analogy in germination to be with Acrogens, must be very strong before I am convinced that plants with perfect ovula as Rafflesia, etc. germinate from an indeterminate point, the existence of an aperture in the coats, points in the most marked manner to some part representing a radicle. With the exception perhaps of Sarcocoidalis, these plants differ in no respect whatever from other Phænogamous vegetables; we have instances of the same parasitical growth, and instances of the same apparent want of a radicle or homogeneousness of embryo, and in the structure of the parts of the flower there is tolerably absolute general identity.
It may be worthy of remark, as tending to prove the soundness of Mr. Brown’s views with regard to the affinity of Rafflesia with Aristolochia, that a certain large and fleshy flowered species of the latter genus has the same putrescent smelling flowers.
In Rhizantheæ, as proposed by Endlicher, we have an assemblage of discordant characters; we have plants associated, differing in the number of their parts; we have some of comparatively simple roots associated with others of decidedly complex organization; we have Rafflesia in which highly complex female parts exist, associated with Sarcocoidalis, in which these are very simple. But besides the objection of combining discrepancies on the strength of one agreement, the establishment of divisions upon such pretexts is objectionable in another point of view; viz., that of making a transition of structure on one point, instead of in several.
We might as well form into one division all the ternarily formed Dicotyledons, and into another all those Monocotyledonous plants with evident distinction between the calyx and corolla.