ment, because it bears the anther on its margin ; which 

 organ when tiic blossom is expanded is a slender, long, brown 

 body, attached from its base for about half its length to the 

 margin of the filament or upper lip of the nectarium, the 

 upper half remaining free. The style is a fleshy club-shaped 

 body, rising far above the anther, rounded at the point, where 

 it has generally a sphacelated appearance, but lias no well 

 defined stigma. 



In order to understand the true organization of the parts 

 of fructification, the flower-bud must be examined, as Mr. 

 Brown so often recommends, long before the expansion of 

 the corolla, for at the time this takes place the function of 

 fecundation has been some time over, and the organs are 

 become effete, and very much changed in their appearance. 

 If the flower bud be examined before it is half the size that 

 it arrives at when the blossom appears, the parts of fructifi- 

 cation will be found remarkably different from what they are 

 in an expanded flower. Removing carefully the calyx, the 

 laciniae of the exterior limb and those of the internal, at this 

 time very small, there comes in sight a large anther of an 

 oblong elliptical shape, with a deep groove down the front of 

 it. A short rounded filament supports this anther, and is 

 inserted at the posterior part near its base. The style at this 

 time is considerably shorter than the anther, is flat and 

 brought to a thin edge on one side, and is terminated with 

 a distinct, fleshy, lobular stigma, which is wasted at the 

 opening of the flower. At the posterior and lower part of the 

 anther and filament, but not extending half way up, is attached 

 the rudiment of what afterwards becomes the petal-like fila- 

 ment or upper lip of the nectarium, now rising far above the 

 anther, and carrying it up on its margin in an etfete withered 

 state ; the style also is extended greatly in length and loses 

 its stigma. 



The above observations were made on the flowers of 

 Canna coccinea, but are doubtless applicable also to Canna 

 lutea, and probably in degree to the other species. 



Flowers most part of the year. Communicated by the 

 Count de Yandes. 





