428 Mr. H. J. Carter on the Development of Sorastrum, 



bond of attachment which extends from centre to centre of the 

 proximate liat triangular sides of the divisions, while two pairs 

 miite to form the zygospore, which is echinated. On the other 

 hand, the individual of Sorastrum is spined onl^ at two ends, 

 the other corner of the cuneate cell being stipitate, while in its 

 normal condition it forms one of a group of eight, sixteen, or 

 thirty-two individuals. The latter, again, do not appear to 

 undergo binary division, but produce one or more baby groups 

 of Sorastrum, and, if we are right in our conjecture, a smooth 

 sporangium, formed probably from the impregnation of a ma- 

 crogonidium by microgonidia. 



Thus the former, by its zygospore, is essentially a Desmid, 

 and the latter, by its mode of generation, essentially allied to 

 Pediastrum (see A. Braun's figures &c. of the development of 

 Pediastrmn gramdatum, pi. 3, in ' Rejuvenescence of Nature,' 

 translated in Botanical Reports by Henfrey, Ray Soc. Pub. 

 1853) — a view at which Mr. Archer had also arrived by hav- 

 ing frequently witnessed the evolution of young groups from 

 Coelastrum and Scenedesmus. Hence, in his last letter to me, 

 this able authority states : — " At present, and so far as obser- 

 vation has yet gone, I could assume Sorastrum (as well as 

 Pediastrum, Coelastrum, and Scenedesmus) as not belonging at 

 all to the Desmidiese." Of course the observations which have 

 led to this conclusion have been made since the last edition of 

 Pritchard was published, in which these genera are all placed 

 by Mr. Archer, as heretofore, among the Desmidiege. Further, 

 Mr. Archer's present view is also corroborated by Rabenhorst, 

 who [oj). cit. 1868) has assigned all these genera to his family 

 of Protococcacese. 



(It is curious, too, as showing the gradual development of 

 our knowledge in these respects among people widely sepa- 

 rated and without intercommunication, although probably of 

 contemporaneous education previously, in the same kind of 

 seminaries, that, in the month of June 1861, I had myself 

 made drawings of Pediastrum, Scenedesmus, &c., to show at 

 some future period that these organisms belonged rather to the 

 Protococcacejfi than to the Desmidiese.) 



Although, however, Rabenhorst figures and places all these 

 genera under his family Protococcaceee as ^^Algce unicelhdares 

 sensu strictisstmo," still a group of eight, sixteen, or thirty- 

 two individuals linked together in the form of Sorastrum can 

 hardly be considered " unicellular," any more than the conca- 

 tenated cells of a filament of Spirogyra. But, be this under- 

 stood as it may^ these organisms, for reasons above stated, un- 

 doubtedly belong much more to the Protococcacea3 than to the 

 Desmidieas. 



