named Raffles r v. 20? 



In these points the structure of Rafflesia remains to be ascer- 

 tained. In the mean time, however, if it be considered as a pa- 

 rasite, and as likely to agree with the other plants of the tribe in 

 the state of its embryo, it may be remarked, with reference (<> 

 the question of its affinities, that such a structure would approxi- 

 mate it rather to Asarince than to PuxsiflorecE. 



My principal and concluding observation relat< •> to the modes 

 of union between the stock and the parasite. These vary in the 

 different genera and species of the tribe, which may be divided 

 into such as are entirely dependent on the stock during the w bole 

 of their existence, and such as in their more advanerd Mate pro- 

 duce roots of their own. 



Among those that are in all stages absolutely parasitic, to 

 which division Rafflesia would probably belong, very great dif- 

 ferences also exist in the mode of connection. In some of those 

 that I have examined, especially two species of Balanophora* , 

 the nature of this connection is such, as can only be explained on 

 the supposition that the germinating seed of the parasite excites 

 a specific action in the stock, the result of which is the formation 

 of a structure, either wholly or in part, derived from the root, 

 and adapted to the support and protection of the undeveloped 

 parasite ; analogous therefore to the production of galls by the 

 puncture of insects. 



On this supposition, the connection between the flower of 

 Rafflesia and the root from which it springs, though considerably 

 different from any that I have yet met with, may also be explained. 

 But until either precisely the same kind of union shall have been 

 observed in plants known to be parasitic, or, which would be 



* Balanophora fungosa of Forster, and B. dioica, an unpublished species, lately sent 

 by Dr. Wallich from Nepaul, where it was discovered by Dr. Hamilton, and also found 

 in Java by Dr. Horsfield. 



2 o 2 still 



