ZYGNEMA. 137 



separate, it exhibits the appearance of two curved knife- 

 blades slightly ajiproximating to each other at their apices, 

 near to which usually lies the divided spiral thread, and 

 strongly impressing the superficial observer, from the position 

 and asjjcct of these blades, with the idea that they are the 

 instruments which effect its separation, and reminding him of 

 the beautiful provision whereby the section of pollen granules 

 is accomplished. 



" On transmitting a short time since a specimen of the 

 Zygncma quadratum, in a state of reproduction, to the Rev. 

 M. J. Berkeley and ISIr. Ralfs, but unaccompanied by any 

 remarks in reference to the structure of the cells, both these 

 gentlemen noticed their peculiar conformation, and from the 

 former I received correct sketches of their appearances. 



'^ The structure of the joints in Zygnema was long since 

 noticed in one species of the genus by Mold, who thus de- 

 scribes it in his paper upon the multiplication of cells by 

 division, inserted in the ' Flora ' : — ^ In Z. elongatum Ag., 

 the dissepiments have a very peculiar structure, which I have 

 found in no other species. The terminal surface of each cell 

 is not even, but elongated into a blunt conical process. This 

 process can only be observed in its true state when two joints 

 are separated one from the other ; when, on the contraiy, the 

 threads are unbroken the j^rocess is generally introverted like 

 the finger of a glove, and exhibits the form represented at 

 PL I. fig. 8. a, by c. Tills is the common condition, and ia 

 most threads no joint is found otherwise constructed. But I 

 have now met with a single thread in which a part of the 

 articulations has the ordinary length, while another part has 

 joints only half as long. In these shorter articulations it 

 Avas normal that only the alternate dissepiments had the 

 structure i^ecullar to this species (so that by these dissepi- 

 ments the thread was divided into articulations of the ordinary 

 length), Avhile, on tlie contrary, the intermediate dissepiments 

 exhibited the form usual in ConfervcB.''^ 



" The observation, that ' this process can only be observed 

 in Its true state (that is, everted) when two joints are sepa- 

 rated, the one from the other,' is inaccurate, for the cells may 

 be separated and yet the processes inverted, the cvcrslon of 



