Dr. G. C. Wallich on Desmidiacea from Lower Bengal. 191 



eventually the common cell-wall becomes either broken up or 

 absorbed, the sporangium is invested with its own proper tunics, 

 and, separating entirely from the parent filaments, it wanders 

 forth on its own peculiar reproductive mission. 



TTe find in this process a remarkable coincidence with what 

 is observed to take place in the conjugation of the Zygnemacese, 

 asj for instance, in Tyndaridea, Mesocarpus, and Staurocarpus ; 

 and another argument is thus afforded, if indeed any be necessary, 

 to prove the close affinity of the class of organisms under notice 

 with the Confervoid Algse in general. 



Length of joint -0006"; breadth -0014". 



Lower Bengal, 1855. 



Plate VII. fig. 1. A portion of a filament of D. Swartzii. Fig. 2. A 

 similar portion of var. /3, with rounded angles. Fig. 3. End view, showing 

 cuneiform processes. Fig. 4. Sporangia! filament in process of formation. 



3. Aptogonum, Ralfs. 



Filament triangular or quadrangular. Constriction entirely 

 absent. Margins plane or very slightly crenated. 



The intermediate spaces, in both the Bengal forms of this 

 genus, are subject to considerable variation in size. The pro- 

 minent character on which I ground the necessity for retaining 

 them in a distinct genus has no reference, however, to this cha- 

 racter, on which j\Ir. Ralfs lays such emphasis, but to the entire 

 character of the joints, and the total absence of the constriction 

 which produces it, and at the same time offers a marked distinc- 

 tive feature between this genus and the rest of the filamentous 

 plane or angular Desmidiacese, without exception*. Two forms 

 occur in Lower Bengal : one, the triangular typical species, A. 

 Baileyi of Mr. Ralfs and Prof. Bailey; the other a quadrangular 

 variety, which I have nevertheless referred to the same species, 

 as daily experience in this and the allied family of the Diato- 

 macese clearly proves how liable these humble organisms are to 

 variations of such a nature, when placed under varying condi- 

 tions of growth, or derived from different localities, which is 

 much the same thing. 



1. A. Baileyi. Filaments linear; joints, in front view, quadran- 

 gular, nearly equal in length and breadth. Uniting cushions 

 circular. End view triangular. 



It has already been mentioned that the margins are faintly 

 irregular, a large oblong portion of the free sides of each joint 

 being slightly raised above the sun-ounding surface, and thus 



* In empty joints, the junction of the segments is indicated merely by a 

 faint transverse line, as in Closterium. This line is omitted in the figures. 



