THE FRESH-WATER CYANOPHYCEÆ OF ICELAND 255 



Ali the usual methods have certain bad points so that the ideal 

 must be said to be woik on living material, in faet quite to avoid 

 preservation. 1) The dried material offers the advantage that the 

 algae retain the colour quite well and in many cases it is possible 

 to soak them out (in ammonia or lactic acid) so much that the 

 natural forms can distinctly be observed; but the softer forms, for 

 instance species of Anabæna never, however, assume the original 

 forms even with the most careful soaking out. 2) When preserved 

 in alcohol the forms and cell contents as a rule are very perfect, 

 their colours, however, are entirelj^ lost. 3) Formalin often pre- 

 serves the forms well, but frequently Strange granulations come into 

 existence in the cells and sometimes Cyanophyceæ become almost 

 irrecognisable in this fluid. 



Therefore I shall recommend other investigators , wishing to 

 collect Cyanophyceæ in localities where an examination of the living 

 piants cannot be practiced to divide each sample into 2 parts, dry 

 the one and place the other in alcohol; it will then be possible 

 both to examine the colour of the plant as well as the form and 

 content of its cells and trichomes. 



THE SYSTEMATIC CHARACTERISTICS WITHIN THE 



CYANOPHYCEÆ. 



The determination of Cyanophyceæ frequently causes consider- 

 able difficulty, especially when dealing with undeveloped specimens, 

 and in not a few cases a determination has to be given up. The 

 difficuities are most frequent within the Coccogoneæ, but this is 

 especially due to the faet that this group is greatly in need of a 

 revision. In the course of time a considerable number of species 

 have been recorded, among w^hich many will certainly prove to be 

 identical, and a comparative summary of the species lacks almost 

 entirely. An attempt at comparison was first made by Hansgirg 

 (1892) later by Lemmermann (1910). W. B. Crow (1922) has 

 written about the principles regarding a natural sj'^stem within the 

 Coccogoneæ, but I shall not go into this treatise here. Even within 

 the Hormogoneæ, in spite of the excellent existing revisions of this 

 group (Bornet et Flahault, Gomont) a determination is often 

 difficult. 



The older systematists exclusively employed such characters 

 that simply could be observed either by macroscopic or microscopic 



