90 ARCHER AND DIXON, ON DESMIDIACEjB. 



parasitic growth, possibly at the expense of the original cell- 

 contents, though it is questionable how the germ could find 

 an entrance into the apparently uninjured cell. Such I also 

 thought the flask-shaped bodies above-mentioned, until I 

 met with Professor Henfrey's remarks on the phenomenon 

 in Chlorosphsera, nor does it quite appear that his conjecture 

 is altogether proven. 



Before attempting, in a measure, to account for the above 

 described curious external aberrations from the normal form, 

 it will, I think, be well briefly to draw attention to the mode 

 of division in this family of Desmidiaceae. So far as I can 

 make out, there can be little doubt that in these plants the 

 first step in the process of division is the formation of a sep- 

 tum at the central space or isthmus, whereupon the segments 

 become gradually removed more distant from each other by 

 the growth of a new cell-wall being interposed between them, 

 eventually forming two new segments. At first the new 

 growth is simple in outline, and pale in colour, but afterwards 

 assumes the characteristic, more or less complex form and 

 degree of tint of the species, and becomes filled with endo- 

 chrome, exactly similar to the older segments; in the free 

 species separation taking place, each older segment bearing 

 with it a new one, to replace that from which it has been 

 separated by the above-mentioned process of growth. Ac- 

 cording to Hofmeister {he. cit.), " the new halves are at first 

 lined only by the protruded portions of the pellicle of the con- 

 tents belonging to the older half cells," and ' ' it is the margin 

 of the half shells which constitute the rings evident in many 

 species, e.g. Closterium, Docidium/' &c. I believe that the 

 portion of each new segment first formed to be the end-lobe, 

 beneath which, at each side, are then gradually evolved the 

 lateral lobes. Nor can it be that certain specimens in 

 Micrasterias and Euastrum, occasionally met with, are more 

 than an apparent exception to this, in which the end-lobe is 

 not only seemingly absent, but in which, in its place, a more 

 or less deep .shins exists. For I should think the phenomenon 

 referred to is due to the arrest of growth of the end-lobe, and 

 which, not keeping pace with the expansion of the lateral 

 lobes, is left behind, thus producing the sinus. 



I would here like to remark, parenthetically, that I do not 

 find that Hofmeister, in his paper alluded to, makes any refe- 

 rence to the circumstance of there being, occasionally at Least, 

 cast oil', immediately after division, from each of the new 

 segments a loose transparent coat, sometimes looking almost 

 like two empty segments, with their ends towards each other, 

 or hack to back. What I refer to is well shown by Mrs. 



