ENDOGENS. 



11 



Hydrocharide^:. 



In Hydrocharis Morsus-ranw the greater part of the ovules are attached to the dissepiments, none however to their inner edges ; but a considerable part arise from the midribs and wall of 

 the ovary, over the surface of which they are irregularly scattered, so that this family as regards placentation suits well as being the epigynous form of Butomacece. In Stratiotes abides there are no 

 ovules on the dissepiments as they arise only at the junction of the wall of the ovary with the dissepiment, so that the placentation may be regarded as really dorsal, as it is in Hydrocharis. The 

 ovule is, as described by Prof. Agardh (Theor. Syst. Plant, p. 39), anatropal, ascending with the raphe dorsal, with some few exceptions in which it is more or less lateral, thus corresponding very 

 nearly with Butomus. 



Palmaceje. 



This family agrees either in structure or habit with so many others, that it has given rise to the suggestion of very different comparisons, and until very recently I felt an uncertainty as to 

 their nearest affinity* While engaged in the inquiry, Mr. Bennett favoured me with an opportunity of examining the specimens of Thrinax in the British Museum, and I now am satisfied that they 

 are very near the Aracece, the ovule of Thrinax being very much like that of Dieffenbachia among the latter. It is erect from the base of the ovary and incompletely anatropal, so that the foramen 

 appears like a small aperture on its side close to the base. I find also on a further examination of the germination of Aram, that they approach very nearly in the mode of the growth of the 

 cotyledon, the opening for the emission of the plumule being very small and terminating above in a furrow precisely as in Phoenix ; I have no hesitation in regarding the latter as exorhizal. The 

 position of the carpel in Thrinax, and also of the fertile carpel in Phoenix and Calamus, is variable, very nearly in the same manner as in Sparganium, nor have I met with any other genera among 

 the Endogens which correspond so nearly with Sparganium in this character, and I should therefore regard Palmacece as connecting Typhacece with Liliacea and Aracem. 



Melanthace^e. 



In Veratrum album the primary radicle is exorhizal, and in all probability it will prove to be so in the few remaining families included in the Exorhizal Section, the germination of which is 

 at present unknown. (V. Trans. Linn. Soc. Vol. XXII, p. 402.) 



