Mr. J. Miers on the Canellacese. 34-9 



horizontally and Hn])ciiinposccl, one apparently from each pla- 

 centa, and tilling the whole space of the cell, where they arc 

 surrounded i)y a small (juantity of thin glutinous mucilage. 

 The seeds are black and shining, and arc constituted precisely 

 as in the species before described. 



2. CiNNAMODENDRON. 



This genus was proposed by Endlicher* for a Brazilian plant 

 first described by Nees and Martins, under the name of Canella 

 aa-illansf, since which, nothing has been recorded respecting it. 

 I have lately had an oj)portunity of examining that plant, and 

 find the details above referred to sufficiently correct ; but the 

 evidence I have now to communicate, regarding a second spe- 

 cies, establishes beyond doubt the validity of the genus, and its 

 close affinity to Cunella. This second species has so much 

 resemblance to Canella alba in its general characters, especially 

 in the similar properties of its bark, that the two have been 

 confounded together in commerce. It differs, however, in having 

 axillary liowcrs, while in Canella they arc terminal : they agree 

 with one another in their odoriferous and aromatic qualities, in 

 the number and form of their persistent sepals, in the union of 

 their extrorse stamens into a mon;ulel[)hous tube, in the number 

 and shape of their anther-cells, and in the size and shape of their 

 ovary and style ; but Cinnamodendron differs from Canella in its 

 rotate calyx and corolla, in the presence of an inner row of 

 petals, in a different stigma, in the number of its parietal pla- 

 centations, in the great number of its ovules, and its much 

 smaller and more numerous seeds. In the specimen I examined 

 of ^lartius's plant, I could not determine the number of its 

 placentce, as the flowers were in bud, and the ovary consequently 

 in a very early stage of development ; but in the Jamaica species 

 I have found that the placentations are generally 5, sometimes 

 reduced to 4, the number of ovules being very considerable : in 

 Canella, on the other hand, the placcntje are constantly 2, and 

 the number of ovules does not exceed 4, all of which, or some- 

 times only 2, are matured into seeds. 1 have remodelled its 

 generic characters in the following manner: — 



CiNNAMODENDRON, Endl. (Char, reform.) Sepala 3, orbiculata, 

 rotatim cxpansn., coriacea, margine ciliata, persistentia, sestiva- 

 tione imbrieata. Petala 10, biseriata ; 5 exteriora oblonga, 

 sepalis dimidio longiora, carnosa, expansa, sestivatione imbri- 

 eata; 5 intcriora membranacea, spathulato-oblonga, pellucido- 

 punctata, exterioribus alterna et vix Kquilonga, iisque erectiora: 



* Gen. Plant. 5458. f Nov. Act. Acad. Cses. xii. p. 18. tab. .3. 



