114: 



which the stamen is altogether absent, whereas s' — exception- 

 ally long — is present. If one admitted that the stamen is 

 merely the left portion of .i', then there would be no reason 

 why it should let go .y', on the contrary it would on account 

 of its petaloid expanding tie it faster. It is just when considering 

 the stamen as an independent organ that the space between 

 X and ,y' appears quite natural. Cases of the stamen wanting 

 are by no means of rare occurrence in Canna, that on the 

 contrary it may be quite free standing between .s' and x, is 

 seen in a specimen which I have carefully preserved. At the 

 close of this paragraph may be added that according to Penzig 

 (IX, p. 375) in a flower of Calathea (allied to Canna) permanent 

 adherence of the stamen to the antisepalous staminode has 

 been observed. 



Assuming the above to be the correct view, no further argu- 

 ment about the diagram of Canna appears necessary, as it com- 

 pletely tallies with that of the typical Monocotyledones and 

 especially of the Zingiberaceae. 



In connection with the frequently stated affinity between 

 Cannaceae and Zingiberaceae it may be to the purpose to point out 

 that in a normal Canna-flower there is only half a stamen 

 fertile, all the rest being sterile, whereas in the Zingiberaceae 

 one inner stamen being fertile, another as a vascular bundle 

 is concealed in the synema (labellum) and the rest are reduced 

 to sterility. A greatei" difference consists in the innerwall of 

 the flowertube being formed by the staminodes in the Zingi- 

 beraceae and in the Cannaceae by the upper portions of the 

 carpels. 



We have next to review the coalescences of the parts of the 

 flower, its torsions and other idiosyncrasies which augment the 

 intricacies of Canna. In the first place I would direct attention 

 to the transverse section of the ovary. This is, as might be 

 expected, three-celled (fig. 17 A and B) '). In A, which has been 



1) The obovoid vesicles which make the surface somewhat rough, are made up 

 of thin-walled cells, containing much water. On the ripe fruit they have become 

 weak thorns. 



