ON CERTAIN EPIGYNOUS MONOCOTYLEDONS. 163 



the inner segments of the perigone, a little to the side, since only 

 one of its cells is destined to be perfected ; opposite to it stands 

 the labellum, the development of which, in the very early con- 

 ditions, outstrips the two lateral innermost parts of the flower. 

 After this all remain very backward in their development, and 

 only attain their large size in comparison with the anther, at a 

 later epoch. With regard to the anther itself, it appears first 

 as a little globule or vesicle, and a furrow is gradually formed 

 on its anterior side. It deserves mention, that before the twist- 

 ing and lateral development of the parts, the tooth which occurs 

 on one side of the full-grown labellum is the oldest part of this 

 organ, and stands in the middle of the young flower before the 

 anther. The stigma comes to light last, as a little body exca- 

 vated on its upper surface, and it rises up with the other organs 

 of the flower. I must here again declare against any fusion of 

 originally distinct organs ; the lower parts of the floral organs do 

 not begin to elongate until a later period, and are confluent from 

 the first. 



Two species of Canna which I have been able to observe in a 

 young stage, exhibit exactly the same conditions. The peta- 

 loid filament only acquires at a late period the expansion which 

 it exhibits in the full-grown condition, and the same is the case 

 with the rest of the inmost organs, which in the earliest condi- 

 tion appear as little globules, just like the anther. The inflo- 

 rescence here again agrees with the mode of ramification of the 

 plant ; the flower-branches describe a three-membered spiral 

 round the stem ; from each member of this spiral arise 2 to 3 

 flowers, or only 1 or 2 flowers occur with 2 to 3 bracts. The 

 primordial leaf [vorblatt) we see occupy so important a place 

 in Calathea, is here regularly suppressed in the little partial 

 branches, which indicates that each 2-3 flowers correspond to a 

 single flower, if we regard each single lateral bract of the latter 

 as belonging to another flower. 



With regard to the import of the parts of the flower of this 

 family, the view of Lestiboudois and Lindley, according to which 

 the more or less foliaceous organs situated inside the outer two 

 circles of the flower are stamens, appears to me the most correct, 

 when we take into account the history of the development. The 

 outer two circles of the flower are quite developed at the time the 



11* 



