396 JOHS. BOYE PETERSEN 



kind^V II mav secm remarkable llial I unlv have it liimi Iv leeland 

 and tlie Vestmannacyjar. Tliis. howcver, is presumably (lue lo accident, 

 most of my collections ol" alj^a' Ironi roeky walls and drijjpinf^ rocks 

 orit^inalin^ from tliese two paris of tlie coiinliy 



Tliiis lliere can liardly be any doiibl llial lliis species is a true 

 aérial alj^a whosc habitat is chiefly aniong mosses and on more or less 

 damp rocks. It seenis able lo withstand ralher strong desiccation. while 

 a|)parenlly il retiuires littlc nilroj^enous nourishmcnl (^Kolkwitz und 

 Marsson 19U8, p. 517, olijfosaijrobe,. 



Navicula perpusilla Grun. var Flotowii (Grun.' Boye P. n. conib. 

 N. Flotowii Cl. Syn. 1. j). 132. Van Ileurck Syn. PI. 14. l\<^. 41. Van Heurck 

 Types Xr. 148. 



E. Icel. 68, 85, L. 99, L. 128. 



If we compare the diaj^noscs of .V. Flotoivii and A'. pcrpasilla in 

 Clcve 1. c.) we are struck by the slij^ht dill'ercncc belween thcm. Also 

 the two figures which Grunow himseif has given of the two species in 

 Van Heurck's Synopsis show great resemblance to cach other. No 

 one has therefore been in doubt that thcy were very closely allied. An 

 examination of the two prcparalions in Van Heurcks Types Nos. 212 

 and 148 has, however, convinced nie that the resemblance is even more 

 pronouneed than could be inferred from the (iescri|)tions and illustra- 

 tions, hence I have thought it right to reduce N. Flotowii to a variety 

 of N. perpusilla, and we might jjcrhaps even witli justice go so far as 

 to include the two species within one. The chief dillerence stated by 

 Grunow and Van Heurck, (Syn. p. 106, Traité p.226; viz. that X.Floloivii 

 is said to be joined into long bands, whereas this it not the case with 

 X. perpusilla . cannol very well be given as a character as it has often 

 been seen that the same species has occurred in chains as well as se- 

 j)arately. In Van Heurck, Types No. 148 both X. perpusilla and X. Flo- 

 towii are found. which, strangely enough , is not pointed out in the 

 corresponding tcxt. In all the samples from Iccland in which I found 

 var. Flotowii. typical Xauicnla perpusilla were also present. 



Naniciilæ hacillares Cl. 



Navicula pseudobacillum Grun. Cl. Svn. I \). 137. V. Ileurck Svn. 

 PI. 13, fig. i). 



E. Icel. 14. 



Only a single s|)ccimcn observed. The species is hardly a true 

 aérial species. 



Navicula terrestris Hove P. Boye Petersen 1915, p. 288, fig. 7, 8. 

 E. Icel. 129 - N. Icel. L 138. L. 253. 



') Tinis it is evidciitly not by lucideiit tliat Van Heurck. in Types No. 148, 

 which contains the species in great (juanlity, gives as its hal)itat >entre les mous- 

 ses dun rocher«. 



