249 



give no insight into possible seasonal differences; they were 

 all collected between the end of Jnne and the beginning of 

 August. 



Lastly, I may add, that I have examined numerous roots 

 of this plant, without finding any kind of mycorhiza; but this 

 also, no doubt, could scarcely be expected to be found. 



2. BorapMla. 



The three species of this group agree closely in the following 

 features: — il) Epidermis with undulating walls, (2) glandular 

 hairs, (3) marked difference between the palisade-tissue and the 

 spongy parenchyma (the spongy parenchyma varying from cells 

 which are slightly branched, but arranged in irregular meshes 

 and rows, with larger intercellular spaces [S. stellaris) — to 

 much branched ("stellate") cells in the species S. nivalis and 

 S. hieraciifoUa ; compact spongy parenchyma immediately 

 beneath the epidermis does not occur), (4) hydathodes exactly 

 like those in the section Nephrophyllum. The differences are 

 as follows: — 



Hairs: — 



Two kinds (i. e. both glandular hairs and marginal non- 

 glandular hairs): S. nivalis and S. stellaris. 

 Glandular hairs only: S. hieraciifoUa. 



Spongy parenchyma: — 



(1) very loosely arranged, consisting of unbranched and 

 branched cells, in rows of irregular length, the main 

 direction of which is parallel to the longitudinal axis 

 of the leaf: S. stellaris. 



(2) all the cells of the spongy parenchyma branched: 

 »S. hieraciifoUa and nivalis. 



It can therefore be seen that the leaf-anatomy of this 

 section corresponds closely with that of the section Nephro- 

 phyllum. It is in reality impossible, on the basis of our knowledge 



