TTIE MORrilOLOGICAL COMrOSlTlON OF T'LAXTS. 



57 



banie time that the mcurYed edges of the foliar siu'faces 

 united. The arrangements of the tubular axis and its ap- 

 pendages, thus resulting, are still more instructive than those 

 of the solid axis. For while, even more clearly than in the 

 Dendrobium, we see at the point b, a continuity of structure 

 between the substance of the axis below the node, and tlie 

 substance of the sheath above the node; we see that this 

 sheath, instead of having its edges imited as in Dendrobium, 

 has them simply overlapping, so as to form an incomplete 

 hollow cylinder which may be taken off and unrolled ; 



and we see that were the overlappmg edges of this sheath, 

 united all the way from the node a to the node b, it would 

 constitute a tubular axis, like that which precedes it or like 

 that which it includes. And then, giving an unexpected 

 conclusiveness to the argument, it turns out that in one 

 family of grasses, the overlapping edges of the sheaths do 

 unite : thus furnishing us with a demonstration that tubular 

 gtructures are produced by the incurving and joining oi 

 foliar surfaces ; and that so, hollow axes may be interpreted 

 as above, without making any assumption unwarranted by 

 fact. One further correspondence between the 



t}^pe thus ideally constructed, and the endogenous t}^e, must 

 oe noted. If, as already pointed out, the transverse growth of 



