THE LEAVES OF VASCULAR PLANTS. 273 



composed of a petiole, dilated at the apex into two lips, 

 the lower large and very concave, open above by a 

 circular orifice, which is hard and provided with three 

 nerves or longitudinal wings on its outer border ; the 

 upper lip is smaller, flat, and acts as a cover to the cup, 

 which is often half filled with a slightly sweetish liquid ; 

 but I do not know if it is secreted by the plant, or 

 produced by the entrance of the rain. 



Hitherto we have seen examples of cavities opening 

 externally, but there are other leaves which present 

 them entirely closed ; such are, for example, the cylin- 

 drical and hollow leaves of several species oi Allium, and of 

 some kinds of Ornithogalum. These leaves are traversed 

 throughout their whole length by a remarkable cavity : 

 it may be believed that this is either a true hollow, 

 produced by the tearing of the internal cellular tissue, 

 or that it is a tube formed by a petiole dilated into 

 a leaf, folded into a tube, and having its edges and apex 

 united. Thus, in the same manner as the sheath of the 

 Cyperaceae seems analogous to that of the Graminea?, 

 except that it has its edges united, — in the same way it 

 may be said that the fistular leaves oi Allium differ from 

 those of Sarracenia only by this uniting together of the 

 edges. The strvicture of the leaves of Iris tends to 

 confirm this last hypothesis. 



Finally, there are some leaves which present several 

 cavities, which there are some reasons for regarding 

 as simple gaps : such are the quadrilocular leaves of 

 Lobelia Dortmanna, and of Isoetes palustris. 



VOL. I. 



