40 DR. J. D. HOOKER ON THE GENERA AND SPECIES 



do these quite accord with one another, but only in what I assume to be mere individual, 

 and not specific characters. 



Prof. Liebmann sums up the differences between the Mexican and the Brazilian indivi- 

 duals, as residing in the more globose female capitulum, shorter stem, more shortly pedi- 

 celled pei'ianth, twisted style, binate palese of the male receptacle, which are also clavate 

 and cUlated at the base, white papillose (male) perianth, longer filament, globose synema 

 and globose poUen. It is also added, that the anthers are 2-celled, and dehisce differently, 

 leavmg a triangular opening between them, and that the filaments are free immediately 

 below the anthers. 



With regard to these points, I find the capitulum if anything more depressed in the 

 Mexican plant and in Liebmann' s accurate figure of it, than ia Richard's drawing of the 

 Brazilian, or than in most of my specimens either of the Brazilian or Colombian jjlant. 

 The stem (peduncle) varies extremely in length, from ^ an inch to 8 inches, and con- 

 siderably even on the same rhizome. The perianths of Mexican specimens are much 

 longer than those of Liebmann's figure, and they are of the same length as those of my 

 BrazUian specimens, though shorter than in Richard's or Martins' figures. The styles of 

 the Mexican plant are very slightly twisted, and that from left to right, not the opposite 

 way, as represented in Prof. Liebmann's figure ; and there is the same twist in Mr.Pm-die's 

 and in some of the Brazilian specimens. The perianths seem constantly papillose, though 

 varying in degree with age, drying, and other less obvious causes. Globose pollen is the 

 common form in the genus. 



Of the remaining distinctive characters, I have occasionally found binate palese of the 

 described shape both on L. hypogcea and tomentosa; the filament in Liebmann's figure is 

 so extremely short, that it appears impossible to draw a character from it ; the synema 

 varies in form, according to its age, and that represented in the figure of T. 3Iexicana 

 entirely agrees with Brazilian individuals ; and finally, the anthers of all, though 4-celled 

 in their early and perfect state, become 2-celled previous to dehiscence, by the contraction 

 of the septum. 1 therefore feel justified in referring the Thonningia Mexicana to 

 Laugsdorjfta hypogcea. 



The parasitism of Laiigsdorffia is remarkable : the dichotomously branching rhizomes 

 appear most frequently to corrode, as it were, the bark of the roots they encounter, which 

 they even sever, and then enclose the end that remains attached to the parent plant : the 

 root swells considerably at the junction, and appears to send prolongations of wood into 

 the rhizome of the parasite, which run along its axis for several inches ; but though 

 there is an intimate union between the wood of the root and the cellular tissue of the 

 parasite, there seems to be no blending of their vascular systems. The rhizome also 

 invariably swells at the junction, but does not branch from that point, as is often the case 

 with Selosis. Both Richard and Martins represent rootlets as given off from the rhizome 

 at a considerable distance from any parasitic union ; but I do not fijid such in any of my 

 specimens, nor have any other Balcmophorece rootlets, though at the junction of root and 

 parasite similar rootlets to those figured by Martius are often given off by the root, and 

 tbese being partially enveloped by the parasite, appear to proceed from it. Martius and 

 Langsdorff further say that the plant grasps other roots by means of these fibres, and 



