NEW TO BRITAIN IN 1880. 205 



In Cosmarium the cell is more or less constricted in the centre, 

 without sinuses or lobes at the margin, and each serai-cell as broad 

 or broader than long (with a few exceptions). There is no notch at 

 the ends, as in Euastrum, and the side view is a double circle or a 

 double ellipse, whilst the end view is oval or elliptical, sometimes 

 warted, but without spines. 



In Xanthidium the general form, and the side and end views, are as 

 in Cosmarium^ except that the segments are spiny. It is, in fact, a 

 Cosmarium with spines. 



Artlirodesnius is a simple form of Xanthidium^ each segment hav- 

 ing but a single spine on each side. 



We may now compare these three genera which have a simple con- 

 stricted cell without lobes, and an oval end view, thus — 



Cosmarium — spineless. 



Xanthidium — many spined. 



Arthrodesmus — with a spine on each side of each semi-cell. 



We have now left to us the one genus Staurastrum, in which the 

 forms are exceedingly variable, but uniformly consisting of a deeply 

 constricted cell, with or without spines or projections, but the end view 

 is angular, often triangular, sometimes quadrangular, or, if circular, 

 then with a lobed margin. Hence it will be evident that the end 

 view is of the greatest importance in determining this genus and its 

 numerous species. -In a front view some species might easily be 

 mistaken for Cosmarium, others for Xanthidium^ and others for Ar- 

 throdesmus, but the end view settles the question at once by its 

 angular form. 



Staurastrum includes more species than any other genus, and 

 these again exhibit a greater variety of contour. In fact, so variable 

 are they, that most elaborate modes of sub-grouping have been 

 adopted, and although some method is absolutely essential, no one 

 yet proposed is wholly perfect. To my mind it seems preferable to 

 collect closely allied species around definite types, rather than to 

 subject them to the hard and fast lines of an elaborate technical 

 analysis. At any rate such a method will be all-sufficient for the 

 jDurpose of this evening. 



I have already said that the front view of some species closely 

 resembles Cosmarium, such as the large s^iecies Staurastrum tumidum, 

 and the smaller ones, St. orbiculare, and St. brevispina, notwith- 

 standing the small projecting nipple. 



Allied to these I have to record from Capel Curig the large Stau- 



