290 



demonstrated in spite of much investigation; the leaves all 

 agreed with the description given. 



The cultivated specimen from the Botanic Garden alone 

 had very long internodes, but the structure of the leaf was 

 exactly like that of, for example, the specimens from either 

 Upernivik or Jan Mayen. 



We have now seen what is the leaf-anatomy of the different 

 species. It appears that the species belonging to any systematic 

 section show greater anatomical relationship mutually, than they 

 show with species of other sections. The structural features 

 which vary least are here also of the most value as a supple- 

 ment to the microscopical section-diagnosis. The structure of 

 the hydathode and the appearance of the radial walls of the 

 epidermis appear to be generally the characters which are least 

 influenced by external conditions and which with a certain 

 degree of constancy, remain uniform within the same section. 



There are three different types of hydathodes: — 



(1) Hydathode marginal, situated actually at the edge of 

 the leaf; no secretion of lime: Boraphila, Nephro- 

 phyllum. 



(2) Hydathode upon the upper surface of the leaf- 

 margin ; no cavity ; no lime : Dactyloides, Trachy- 

 phyllum. 



(3) Hydathode upon the upper surface of the leaf- 

 margin; cavity present; lime secreted: Euaizofiia, 

 Porphyrion. 



The other structural features, such as the thickness of the 

 epidermis and the structure of the mesophyll, which are more 

 easily influenced by external conditions, also are similar within 

 each section, as may be seen from what has been written 

 above. To this must be added, that Bonnier's investigations are 

 greatly in favour of the idea that the structure of the mesophyll 



