Mr. J. Miers on the Canellaceae. 345 



that entirely cncli)si's the pistil, the anthers consisting of 20 

 distinct lint-ar cells, which are affixed extrorsely upon it ; the 

 ovary is 1 -celled, and furnished with two opposite j)arietal pla- 

 ccnta^ which arise from the base, and upon each of these in the 

 middle are seen two n niforni ovules attached horizontally right 

 and left by a short and broad funiele in the sinus. The fruit, 

 like that of Drimys, is baccate and unilocular, containing four 

 slightly reniform, oval, rounded seeds, enveloped in a syrupy, 

 aromatic mucilage ; their external tunic forms a hard, shining, 

 black, crustaceous shell; and on the ventral face, just above its 

 sinus, is seen an ojkmi scar or hilar spot, indicating the point of 

 its sus})ension ; while a little above this is a small obsolete pro- 

 minence, corresponding with an internal point, to which the 

 enclosed nucleus is attached. On breaking this internal tunic, 

 the nucleus is found covered with a tolerably thick coating of a 

 loose tissue, in which is imbedded a short thread-like raphe 

 enclosing spiral vessels, proceeding from the point of attachment 

 just mentioned, and terminating in a thickening of the proper 

 integuments seen on the ventral face, just below the sinus. This 

 envelope of loose texture is quite analogous in its structure to a 

 similar development in the seed of Drhni/s, which I have else- 

 where more minutely described, differing only in its component 

 cells being filled with mucilaginous instead of fatty deposits; 

 the origin, course, and termination of the raphe being alike in 

 both cases : so also the innermost integuments are similar in 

 texture to those of Drimys, and though slightly agglutinated 

 together, are separable at all points except about the spot where 

 the raphe terminates in a broad, circular, dark, areolar thicken- 

 ing of their substance, where they are intimately connected. 

 The position and direction of the embryo, in very copious albu- 

 men, offer other striking points of resemblance; although the 

 embryo is much larger and more elongated in Canella, and more 

 minute in Drimys, still the cotyledons in the former, though 

 proportionally longer, are terete, and the extremity of the radicle 

 in both cases is closely contiguous to the point of suspension of 

 the seed. One feature, worthy of especial notice, is even more 

 strongly developed here than in Drimys : the direction of the 

 embryo does not correspond with the axis of the albumen ; nor 

 do the cotyledons tend towards the chalaza, but lie on the con- 

 trary side, in quite an excentrical position. I have observed in 

 Canella, as well as in Drimys, within the body of the albumen, 

 the remains of the embryo-sac extending from the cotyledonary 

 extremity of the embryo, surrounded by soft mucilaginous mat- 

 ter. From the above facts we may draw the same conclusion in 

 regard to the nature of these several tunics of the seed in Canella 

 that I have done in that of Drimys, — viz. that the external crus- 



