342 Linnean Society. [Nov. 7> 



growing-point, and this he finds to be obviously the regular struc- 

 ture. The flower appeared in one instance not to be terminal, but 

 to be produced from the axil of the second scaly leaf, while the buds 

 in the axils of the leaves with laminae were leaf-buds. On the sub- 

 ject of affinity, Mr. Clarke thinks that the Nymphseal alliance, as 

 usually limited, has no very near relationship except with Ranun- 

 culacea ; but that much analogy exists between it and some Endo- 

 genous families, and that it may also be connected with Crypto- 

 gam(B, through Ceratophyllum and Chara, the embryo of Ceratophyllum 

 showing a difference between its second foliaceous appendages as 

 compared with those that follow, in analogy possibly with that of 

 Nelumbium. 



The paper was accompanied with a series of illustrative drawings. 



Read further, "Notes ouCephalotea and Belvisiaceee." By Benjamin 

 Clarke, Esq., F.L.S., &c. 



In these two notes Mr. Clarke gives a general account of the 

 structure of the remarkable plants on which the families are founded. 

 The ovule of Cephalotus he describes as erect, anatropal, with a 

 dorsal raphe, and a large and somewhat two-lipped foramen at the 

 base, on the inner side of the funiculus. The torus in the ripe 

 fruit is described as broad and conical, almost filling the calyx ; 

 during the stage of flowering it is almost flat, the carpels appearing 

 to be attached to the flat base of the calyx. As the fruit advances 

 in growth the conical torus forms between the carpels, to the sides 

 of which they are attached ; and after the carpels are fallen off, it 

 forms a rirn immediately above the part to which they were attached, 

 not unlike the expanded style of Sarracenia in miniature, the rim, 

 however, being comparatively much more contracted. The rim has 

 six angles, which alternate with the attachments of the six carpels, 

 and from the place of attachment of the carpels, or immediately above 

 them, are formed six small filamentous processes. Mr. Clark considers 

 this small peltate process as analogous with the expanded termination 

 of the style in Sarraceniacea, and consequently regards the nearest 

 afiinity of Cephalotece as being with that order. In other respects he 

 considers it as very nearly allied to Francoacece and Ranunculaceee, 

 and thinks it shows some analogy with Aristolochiacece in the glands 

 of its calyx. It approaches Rosacece in its perigynous stamens, and 

 in the position of its raphe, if the ovule pendulous with raphe next 

 the placenta be regarded as an equivalent character. 



In his note on Belvisiaceas, Mr. Clarke describes the ovary of 

 Napoleona as five-celled ; the ovules as two — four in each cell, when 



