ZYGOSPOR^E. 19 



MI NATION is, if not of rare occurrence, very rarely detected. 

 Dr. de Bary, of Germany, enjoyed a favorable opportunity ; he 

 describes the whole process substantially thus : Two cells are 

 drawn together by a conjugal influence which we have no means 

 of detecting ; the semi-cells separate Tlate XXVII, fig. 27), and 

 pour out their contents ; those of the one cell unite with those 

 of the other, at first without a distinct investment, but soon the 

 mass acquires a membranous envelope (figs. 28. 29), forms a 

 matrix, or sporangium in which are reproduced new plants pos- 

 sessed of invigorated vitality. The envelope of the matrix or 

 zyyospwe is primarily thin and smooth, but by degrees it acquires 

 increased thickness, and in the Cosmariums, usually a granular, 

 tuberculated (fig. 28), or more frequently a spinous surface (fig. 

 29) ; the spines being sometimes simple but commonly forked at 

 their extremities. The next step so difficult to be traced is the 

 opening of the wall of the zygospore ffig. 29), setting free small 

 spheres of sarcode ; as they issue they enlarge and acquire a 

 gelatinous, or thin membranous wall. The wall thickens and 

 the sphere enlarges (fig. 30), the contents constrict (fig. 31), 

 first in one direction and then transversely to the plane of the 

 first incision 'fig. 32); these parts develop and set free two or 

 four new plants in size and form like the mother-cell, except in 

 the cytioderm or membrane ; this is not granular, but smooth, 

 and so remains until after the multiplication by division takes 

 place. After the first division the new semi-cells assume the char- 

 acteristic granular surface ; the result of this first division is two 

 plants each composed of one granulated and one smooth semi 

 cell. The second division will make two perfect cells and two 

 which retain the one smooth semi-cell; the third division pro- 

 duces eight cells, all of which except the original two semi-cells 

 will be of typical form. 



Hofmeister's views are somewhat at variance with those of 

 De Bary in regard to numbers; he affirms that the contents of 

 the sporangiums of Cosmuriums are transformed by repeated 

 binary subdivisions into eight or sixteen cells, which assume the 

 orginal form, bnt not full size of the parent, before they are set 

 free by the rupture or diffluence of the wall of the sporangium. 



The sporangium* of Clofiterium, Staut-aatrnm and of other genera 

 are supposed to germinate in the same manner as Cosmarium. 

 Plate XXVII, fig. 33, is a cluster of Closteriums surrounded by 

 a membrane, found by Ralfs ; lie however questions whether this 

 is a germinating sporangium. 



Encysted clusters of Desmids are not necessarily evidence of 



