AFFINITIES OF BALANOPHORE^. 9 



conceded by Goeppert and all observers, and which is easily demonstrated. Hence, 

 besides the difficulty of reconciling the theory of two origins for the vascular tissue in 

 one plant to any knov^ii law, we must also break through the weU-established law, that 

 the formation of buds is a repetition of the process of germination. 



2. M. Goeppert accurately describes the vascular bundles of the inflorescence as origina- 

 ting in the buds whilst stiU enclosed within the rhizome ; but whereas he figures and 

 describes them as having free terminations, I find them to become confluent with the 

 vascular bundles of the rliizome. To any one versed in the dissection of vegetable tissues 

 it can be no wonder that this is a point almost incapable of demonstrative proof in the solid, 

 opaque tubers of BalanophorecB, which generally turn of a deep brown when fii'st cut, and 

 become black in spirits ; whose tissues cannot be torn ; and in which the vascular bundles 

 of the peduncle are so delicate, and run in such sinuous courses, that it is impossible so to 

 bisect a plant that these bvmdles shall be traced continuously from the inflorescence to 

 the base of the rhizome : I have, however, repeatedly found that the appearance of a free 

 termination to the bundles is produced by cutting them obliquely across. A long mace- 

 ration of the parts, and a careful picking away of the cellular tissues, are the only means 

 I have found available for proving their confluence by direct observation ; but at the same 

 time I must confess that, whilst carrying on these dissections in various species of Bala- 

 nophora, and in Bhopalocnemis, I have repeatedly changed my opinion, and indeed have 

 on some occasions been almost convinced of the truth of the contrary view to that I have 

 finally adopted, so deceptive are appearances. 



In the tuberous Helosidew, and in Cynomoriitm and Sarcophyte, the bundles of the 

 rhizome are so unsymmetrically arranged, so much smaller in diameter, and so much 

 more tortuous, that I have hitherto been unable to trace this confluence in them ; whereas 

 in both species of Helosis, and in Langsdorffia, which present the most perfect develop- 

 ment of a cylincU'ical rhizome, the origin of the vessels of the peduncle in those of the 

 rhizome is perfectly evident, and requnes little skill in dissection to demonstrate. 



3. With regard to the anatomical difl'erences stated by M. Goeppert to exist between 

 the vessels forming the vascular system of the rhizome and of the peduncle, they 

 certainly do not exist in all the species. Tliis is however quite consistent with Gceppert's 

 analysis being perfectly accui-ate, for it is to be expected both that the vessels of the 

 perennial rhizome should differ from those of the annual peduncle, and that from the 

 form and direction of development of these organs being essentially different (the one 

 chiefly increasing in breadth and the other in length), their vessels would be different also. 

 In Balanophora the cellular sheath enclosing the vascular bundles is the same in the 

 rhizome and peduncle (except that the individual utricles are longer in the latter) ; in 

 both cases the cells are colourless, void of solid contents, and with few dots or markings 

 on their walls ; thus always contrasting strongly in appearance with the adjacent paren- 

 chyma, which abounds in wax. (See Plates IV., VI. &c.) 



In the fuUy-formed flowering specimen oiB. involucrata (Plate IV. fig. 1), I find no 

 .greater differences between the vessels in the rhizome and those in the peduncle than 

 might be expected in organs so dissimilar in age and proportions. Plate IV. fig. 14. is a 

 transverse, and 15. a vertical section of a vascular bundle fi-om below the capitulum, com- 



VOL. XXII. c 



