6 DR. J. D. HOOKER ON THE STRUCTURE AND 



IVo forms of attachment are found amongst the genera vdth branched and much- 

 elongated rhizomes. In Uelosis the rhizome forms a tuber at each point of its attachment 

 to the various roots it meets with in its subterranean com'se, and a few vascular bimdles 

 from the root are rarely sent into it at these points ; but these do not appear to commu- 

 nicate directly with the previously existing vascular tissue of the rhizome, nor to become 

 blended with it : possibly, however, they may have been given off by it, or have been 

 independently formed in the rhizome; a point which can -only be determined by exami- 

 uinw the natru'e of the attachment at its first formation, and which I shall hereafter 

 discuss. In Langsdorffia the branch of the rhizome corrodes the bark of the roots it 

 encounters ; the fii'st contact in the case of L. rubiginosa being by means of woolly haii'S. 

 Both the rhizome and the root generally swell considerably, but often do not, and the 

 root sends long vascular branches, apparently covered with the cellular bark of the root, 

 right and left into the axis of the rhizome ; with whose vascular system, ho^'ever, I have 

 never found them to form an organic adhesion (see Plate II. figs. 10, 12, 13, 16, 17). In 

 this genus two or more species of dicotyledonous plants sometimes send their roots into 

 one tuber of an old rhizome, each penetrating at several points. 



In the Annals of the Vienna Museum (ii. 53), I find JBalanophorece arranged by linger 

 under three of the divisions, into which all parasites are separated by that author according 

 to the natiu-e of their parasitism ; they are the following : — 1. Parasites which form a 

 rhizome by which they adhere to the roots of plants, and from which the flower-buds rise. 

 Example, ScyhaUum. 2. Parasites which exercise a powerful specific action upon the root, 

 causing it to send vascular bundles into the rhizome, \A'hich hence becomes an organ in- 

 termediate in nature between the stock and the parasite. Examples, Balanopliora, Sarco- 

 j)hytc, Cynomorimn, Lopliophytmn ? Ombrophytmn ? 3. Parasites which form a rhizome 

 intimately attached by its vascular tissue to the root. Examples, Uelosis, Langsdorffia. 



It ajjpears to me that the above are rather distinctions of words than of facts ; and that 

 in so far as they are correct, any one of the tlu'ee definitions is more or less applicable to 

 all the species : for all form rhizomes, all owe theii" adhesion to then- power of exerting a 

 specific action upon the roots from which they derive thcii' nourishment, and except in the 

 case of Lopliophytitm (and perhaps of Ombropliytiim, which I assume to have the same 

 mode of parasitism as Lophophytum), all more or less present the appearance of the vas- 

 cular bimdlcs of the root being enclosed in the cellular tissue of the parasite. Eurther, if 

 my observations are correct, both Uelosis and Langsdorffia should be transferred to the 

 fii'st class ; for there is certainly no distinct union of their vascular bundles with those of 

 the root, nor do their rhizomes appear to send any bundles towards the root; on the 

 contrary, the appearance is perfectly distinct of the root sending its branches into the 

 i-hizome. Langsdorffia indeed is described both by Richard and (apparently following 

 him) Martins, as sending forth root-fibres from its rhizome ; but I not only fail to discover 

 these on any of the very numerous specimens I have examined, but I find this appearance 

 to be produced by fibres being given off from the roots of the plant on which the j)arasite 

 grows, which fibres become included within the rhizome (Plate II. fig. 11). 



The di&erences therefore that prevail amongst the modes of parasitism of Balanophorew, 

 ai-e of degree only : the power of erosion and of forming an organic adhesion is the main 



