282 



groups. Tliese two marginal bauds commence above tbe mar- 

 ginal hairs of the leaf and widen continuously towards the apex 

 to the sacrifice of the middle band. Therefore stomata are 

 quite absent from the leaf-blade below the point at whicli the 

 marginal hairs begin , they all occur towards the apex of 

 the leaf. 



Børgesen states (I. c, p. 225) that the stomata are more 

 abundant upon the upper surface, but yet, at the same time, 

 records that it has eight, while the lower surface has twelve, 

 per unit of surface. The latter statement unquestionably cor- 

 responds better with my observations. Lazniewski's statement 

 that the stomata are entirely absent from the exposed leaf- 

 apices of many rosette-plants does not at all agree with the 

 conditions found by me in this species, which has all its sto- 

 mata especially placed [in the most exposed parts of the leaf. 



A wax-covering is present in the form of small grains of 

 irregular form upon the apical, exposed parts of the upper side 

 of the leaf. 



The description given above of the mesophyll of S. Coty- 

 ledon exactly suits that of the present species. The palisade- 

 cells are very slightly differentiated, are barrel-shaped, and the 

 tissue merges below imperceptibly into the spongy parenchyma 

 with its more isodiametric, unbranched cells (Fig. 26). 



The palisade-cells are more or less obliquely placed 

 towards the apex of the leaves. Lazniewski (I. c.) has found 

 this to be the case in many rosette- plants, and connects it 

 with the peculiar way in which light falls upon such a rosette 

 with its obliquely erect leaves (this feature was first pointed 

 out by Pick). 



At the base of the leaf the difference between the palisade- 

 tissue and the spongy parenchyma is even further obliterated, 

 nor does any obliquity occur there; the whole mesophyll is 

 homogeneous, exactly like that found by Lazniewski in many 

 Alpine rosette-plants. 



