BRITISH DESMIDIE^. 181 



and the others are not affected. In all the species the free margin of 

 the outer cells is bipartite, a character which I consider important, and 

 in fact a modification of the form observed in Cosmarium and other 

 genera with constricted fronds. The cell in Pediastrum^ however, is 

 different from theirs, since its division into two segments exists only 

 on one side, and in regard to the inner cells is at most a slight con- 

 cavity of the external margin. 



The flat, star-like fronds of Pediastrum are so characteristic, that 

 there is no risk of mistaking it for any other genus, Crucigenia per- 

 haps excepted ; that however differs by having entire, quadrangular 

 cells. 



It is far more difficult to distinguish its species. Ehrenberg relies 

 chiefly on the number of the circles, but this character, as Meneghini 

 and Professor Bailey have observed, cannot always be depended on : 

 the latter says, "^ There appears to me to be much confusion in the 

 specific characters, arising from the circumstance that the number 

 of corpuscles in the different rows has been made a character of spe- 

 cific importance. From what I have seen of the species, I am satis- 

 fied that the number of corpuscles in a star is liable to great variation 

 in the same species." 



I have myself noticed that, in one species at least, the number of 

 circles varies from one to three ; and the same observation has been 

 made by Mr. Jenner. He has also remarked that the number of cells 

 in the inner circle, on which I was inclined to place greater reliance, 

 is subject to variation. Some species have one cell in the centre and 

 others two ; but the central cells are said to be sometimes deficient, 

 a variation which I have not observed. 



All the above characters, it must be allowed, are more or less un- 

 certain ; still upon them we must for the most part depend in discri- 

 minating between nearly allied species. 



Meneghini adopts as a specific character the number and position 

 of hyaline vesicles in each cell ; but I am afraid that both of these 

 are also variable. 



The synonyms in this genus are so confused that I quote them with 

 much hesitation. Ehrenberg, having relied almost entirely on the 

 number of the circles and the cells in each, has neglected the form of 

 the cells, and consequently his species are intermixed, states of some 

 being referred to others. Mr. Hassall indeed has constituted several 

 new species ; but as he has attempted this from an examination merely 



N 2 



