THE aIÉRIAL ALGÆ OK ICELAXD 369 



opinion of the nature ol" the smallest syslematic units. We must 

 liere consider three possibilities: 



1) We can assume a simple fluctualing variability. 



2) It can be supposed that the same genotyj)e mav give rise to 

 variants determined by environment and tinally, 



3) There can be a possibility of genotypically dilTering forms. 



Of these three kinds of forms only the latter two are of any 

 interest for ecology. Unable for the present to distinguish between 

 the three kinds of variants, ecology must take them all into account, 

 and we must hope that we shall subsequently obtain a better insight 

 into the significance of their occurrence in the various associations. 



In the present work 1 have atlempted to unravel the forms of 

 some of the most commonly occurring species. Owing to the pro- 

 blematic nature of these forms I have not, in some cases, formally 

 labelled them as varietates and formæ, but have simply given them 

 numbers. In other cases I have set up new varieties and forms, 

 and notably of species in which several such had already been 

 described and named. 



I have previously pointed out (1915, p. 298) that the terrestrial 

 Diatoms nearly all belong lo the Pennatæ and among these to the 

 Raphideæ, the Centricæ and Arraphideæ being only sparingly re- 

 presented. I connected this faet with their mode of living, thinking 

 that the presence of a raphe and the consequent molility might 

 probably be of great use to them, enabling them to penelrate be- 

 tween the particles of earth and in similar piaces in periods of 

 desiccation, and thus securing protection for them. Furthermore I 

 showed that most aerial Diatoms belong to small species, and often 

 to particularly small forms of the species which also occur in fresii 

 water. This, too, might be of importance to these organisms under 

 their special life conditions, since it seems to be a faet that small 

 organisms on the whole withstand desiccation better than large ones. 



These general observations were fuUy confirmed by my investiga- 

 tion of the Icelandic aerial Diatoms, though the facts do not perhaps 

 appcar so clearly in the present paper because I have included 

 localities of a considerably greater degree of moisture than Ihose 

 contained in my former investigation. On the whole the line of 

 demarcation between the aerial and the hydrophiious Dialoms is 

 not nearly so sharp in Iceland as in Denmark. This is doubtless 

 due to climatic dilTerences between the two countries. 



