OF THE OVARIUM OF SIPHONODON. 137 



this torus being produced, as in NymphcBa, Victoria, &c., into a styliform and stigma-like 

 body. Each cai'pcl bears two ovules on each of the ten marginal placentae, making twenty 

 ovules in all. The annulus is the free upper portion of the five confluent carpeUary leaves ; 

 the five ridges are the lines of junction of these ; the five stigmata are each double, 

 formed by the terminations of the confluent placental margins of the adjacent carpels, as 

 in PapaveracecB and many other Orders. 



There is one point, however, to wliich further allusion is necessary, as possibly in part 

 explaining Mr. Griffith's views ; and this is, the above-mentioned line of loose cellular 

 tissue that extends from the base of the central column to the ovarian cavities, and which 

 is met at the ovarian cavity by the true stigmatic tissue. It is very possible that this line 

 indicates the existence of a stage in the early state of the ovarium in which the carpeUary 

 leaves were not completely closed ; for though I feel satisfied that the ovules of this plant 

 are at no period independent of the carpeUary leaves, and are indeed formed in cavities of 

 those leaves and from theu" margins, it does not foUow that there may not have been a free 

 opening to these cavities, or one closed only by a very lax tissue. It is indeed held by some 

 botanists, that aU carpeUary leaves are congenitally open, and close more or less com- 

 pletely afterwards ; an opinion which is not as yet absolutely proven, though I cannot 

 but think that the open ovary of Conifer ce* and its aUies is a strong argument in its 

 favour. This however, if true, by no means reconciles Mr. Griffith's observations, or his 

 theoretical view of the structure of the ovary of Siphonodon, with my own. 



The structure I have described in Siphonodon suggests a different view of the affinities 

 of this obscure genus than those which have been doubtfuUy adopted by Griffith, though 

 in the absence of ripe seeds it wUl be difficult to establish these, and I shaU not therefore 

 attempt to do so now. 



In submitting this very singular plant, then, as a strong proof of the validity of those 

 laws of carpeUary placentation which it has been supposed to have subverted, I cannot 

 refrain from expressing my admiration of the learning which Mr. Griffith has displayed 

 in his discussion of the view he somewhat hastUy adopted ; — of the guarded manner in 

 which he expresses his opinion ; — of the fuU weight he gives to every structural point that 

 seems to him to miUtate against it, and of the candour with which he states every adverse 

 argument that suggests itself to him. Though I beUeve his observations and conclusions 

 to be erroneous, it must be recoUected that the plant is a very anomalous one, its parts 

 exceedingly smaU, and that my experience assures me that specimens preserved in spirits, 

 such as I examined, are in many respects much better for determining structural points 

 from than U^^Jlg ones are. Mr. Griffith's paper fm-ther aboimds in acute observations on 

 many other points in the structure of Siphonodon, to which I have not aUuded ; and it 

 contains, in a note, a short abstract of the only accurate account hitherto published, so 

 far as I know, of the development and true nature of the ascidia of Nepenthes'\. 



* It appears more consonant with the known laws of vegetable morphology to regard the Coniferous ovary as an 

 arrest of the usual tendency of ovaries to close, than to suppose the ovaries of most PhaBnogams to be congenitally 

 closed, and that of Conifers open. 



t Some years ago I prepared drawings of the development of the pitchers of Nepenthes, from plants in the Royal 

 Garden at Kew. These confirm Griffith's observations in everj' particular, and prove the pitchers to be modifications of 

 excurrent midribs. Each pitcher commences as a gland at the anterior apex of the conical mamilla, which represents 



VOL. XXII. T 



