AND THEIR GERMINATION. 



409 



occurs in Alismacece, a near ally of the HydrocharidecB ; besides which, they may not im- 

 probably prove to have the embryo of Triuridem, which would lend some further support 

 to such a comparison. And if the Bromeliacece should prove cndorhizal, such a sulidi- 

 vision of the epigynous families would not be unnatural, as they connect Hydrocharidea 

 and 3Iarantacece. 



The following is added as a provisional arrangement of the Endorhiza; and their allies, 

 including especially nil those families in which the radicle is enlarged ; and the remainder 

 of the class Endogens thus separated will, I believe, be found to form by themselves a 

 very natural section. But it must be confessed that no entirely natural arrangement 

 can be expected to be made imtil the structure of the embryo and its germination in all 

 the families is definitely ascertained. 



c 

 a. 



2 ° 



EXPLANATION OF THE PLATE. 



Tab. LXVIII. 



Fig. 1. Nuphar lutea. The radicle at the time of its first protrusion, before the first leaf appears. 



Fig. 2. The seed of the same after the first leaf has appeared. This leaf is rudimentary, having no 

 lamina. 



Fig. 3. The same, after the second leaf has formed, and a secondary radicle protruded, which has a coleo- 

 rhiza. The same coleorhiza appears on the secondary radicles, which succeed. 



Fig. 4. An embryo of a 2fan«icA(-//ia, the membranous sheath containing the plumule having been laid 

 open to show the position of its first and second leaves. 



Fig. 5. A longitudinal section of the same, showing that the radicle is perfectly adherent to the surround- 

 ing tissue, and that in the ungerminated embryo no trace of a canal is formed for it. 

 VOL. XXII. 3 II 



