— 202 — 



in consequence of my latest observations I think the coastal 

 halophytic vegetation may be most naturally divided in the fol- 

 lowing way: 



A. The hydrophyte vegetation. 



I. The muddy and sandy soil vegetation i). 

 1. The seagrass (and algse) formation. 



B. The halophyte vegetation. 



I. The muddy soil vegetation. 



1. The mangrove formation. 



2. The Salicornia formation. 



3. The Conocarpus formation. 

 II. The sand strand vegetation. 



1. The Fes caprce formation. 



2. The Toiirnefortia formation. 



3. The Cocco/o6a-Manchineel formation. 

 III. The rocky coast vegetation. 



With regard to the associations included under the formations, 

 1 may refer to the more special sections. 



Regarding the section III, the rocky coast vegetation, one can 

 naturally also separate it into several formations e. g. a lichen 

 (and algae) formation, a Chasmophyte formation etc. I have not 

 made sufficient observations however, to permit such divisions. 



Before proceeding to the description of the formations 1 wish 

 to give some few remarks on the external conditions affecting the 

 halophilous vegetation. 



As to the soil the shores of the Danish West Indian islands 

 differ very much. In some places we have sandy or rocky shores, 

 open to the whole power of the sea, in others, the coast is shel- 

 tered by off-lying coral-reefs, allowing the deposit of mud and clay ; 

 and such altered conditions often differ at a very short distance, 

 and thus lead to a very different development of the vegetation, 

 in connection with the other, more or less, altered conditions of life. 



To understand the following it must be emphasized, that the 

 tide is scarcely perceptible, or in any case, of very slight signifi- 

 cance in the Danish West Indies and the else, somewhat arranging 

 and dividing influence of the tide on the vegetation is thus ab- 

 sent here, and produces the effect that the boundaries, between 



The lithophyte vegetation I do not deal with here. 



