AFFINITIES OF BALANOPHOREiE. 17 



appear minutely wi-inkled on the surface. The long single style of Mystropetalon termi- 

 nates in a clavate or capitate, and evidently 3-lohed stigma. Sarcoplnjte has a sessile, 

 broad, discoid stigma. The style of Cynomorimn is more complex than in any other 

 species, and terminates in a 2-lobed stigma ; it is provided with two vascular cords and a 

 central groove occupied by stigmatic tissue : — a detailed account of it wiU be found in the 

 remarks upon C. coccineum. 



In all the above-mentioned plants the cellular tissue of the ovary is very loose, consist- 

 ing of oblong utricles, usually furnished with cytoblasts, and without any vascular tissue 

 in its walls (except in the style of Cynomoriwm) : there is, however, a manifest approach 

 to vascular tissue in the woody cells of the superior perianth of Thonnmgia, and perhaps 

 also of Langsdorffia. 



Ovule. — This is invariably solitary, and pendulous from the summit of the cavity of the 

 ovary. In both 3IonostyU and Bistyli its insertion is so near the very centre of the cavity, 

 that I cannot detect any deviation in its position from the axis of the ovary ; nor in the 

 Distyli do I find it to be placed nearer to one of the styles than to the other. 



The earliest appearance of the ovule of Balanophora is as a solitary cell, protruded from 

 the wall of the ovary : its subsequent stages I have followed to some extent in B. involu- 

 crata, though, oxNdng to the rapid sphacelation of the cellular tissue of the ovary imme- 

 diately after opening it, and the extreme minuteness of all the parts, the analysis is one of 

 great deUcacy, and proportionately liable to error. 



Plate V. figs. 11 & 12. represent an opened ovary of B. involucrata, sho^^'ing a very 

 young ovule, consisting of a delicate hyaline sac suspended almost immediately below the 

 insertion of the style, and containing two free spherical cells, each fuU of fluid and covered 

 with opaque spots, which are probably cytoblasts. I found it impossible to detach the 

 ovule, or to view it, except in situ, and by transmitted light. The formation of cells pro- 

 ceeds with great rapidity within the sac, but I was unable to trace their evolution. The 

 resemblance between the cells thus developed, and those in the embryo- sac of ordinary 

 ovules, is obvious, and it suggests the possibility of the ovule being reduced to an embryo- 

 sac. I could obtaia no clue to the period at which impregnation is effected, nor to the 

 particular action of the pollen-tubes, which I never foimd within the cavity of the ovary 

 or ovule* : nor could I trace on any part of the surface of the ovule, any indication of a 

 chalaza, raphe, or foramen, at which impregnation is probably effected. After the OYvle 

 has swelled, so as to fill the cavity of the ovary, it adheres by means of its membranous 

 coat to the walls of the ovary; at which time it consists of a dense opaque mass of 

 cohering hexagonal cells. 



The ovule, as thus described, does not materially differ from that of Viscum, as described 

 in Decaisne's admu-able memoir on that plant (Memoires de 1' Academic de BrvixeUes), 

 except in being more simple ; the ovule of Viscum consisting of an embryo-sac covered by a 

 delicate cellular membrane (the tercine of Mirbel), and the greater portion of the substance 

 of its nucleus being undeveloped. Regarding Balanophora as presenting the most reduced 

 form of onile, Loranthacece are a step higher, and from these the passage is dii-ect to the 

 naked nuclei of Santalacece and theu- allies, of Cornea, CaprifoUacea, Bubiacece, JJmbelli- 



* But which M. Hoftneister has obserred in the ovule itself of Cynomorium. 

 VOL. XXII. D 



