DUBLIN NATUEAL HISTOEY SOCIETY. 87 



trie and parietal. It is true that it sometimes looks as if it were not so ; 

 but I am disposed to think that this appearance is only when it is up- 

 permost, and consequently towards the observer ; while, on the other 

 hand, as is frequently the case, it may appear absent when it is on the 

 side turned away from the observer. In Closterium it is true that a 

 rounded body occupies the middle of the clear space (possibly nuclear) ; 

 but it is, at all events, not parietal, but placed in the very centre of the 

 contents, and it looks, indeed, scarcely diiferent in nature from the 

 scattered or longitudinal series of amylaceous large granules frequently 

 present, of which it seems but to form one. Again, in Closterium there 

 is always a clear space at each extremity of the endochrome, in which a 

 greater or less number of opaque, very minute, but sometimes variably 

 sized, granules exert a remarkable constant active movement (as is well 

 known), which apparently is maintained during the whole life of the 

 organism. In some species there appears a special vacuole in which these 

 occur, in others they merely occupy a vacant space immediately beyond 

 the rest of the endochrome, leaving the rest of the frond clear. In An- 

 kistrodesmus there are no such graniiles. It is true, that Mr. Berkeley 

 alludes to a circulation of fluid contents in his Closterium Grijftthn (seen 

 only under a power ranging from 1000 to 1500 diameters); but this 

 circumstance is of no generic or specific importance. It is common, in- 

 deed, in Desmidiacese. Again, in Closterium the self-division is trans- 

 verse, taking place at the middle of the frond ; and, when completed, 

 the new segment is merely a rounded or somewhat triangular protube- 

 rance, and the frond, consequently unequal (Figs. 56 and 62 afford an ex- 

 ample), presently the new, short, rounded segment elongates into acomplete 

 counterpart of the older segment, assuming whatever may be the mature 

 characteristic specific form. In Ankistrodesmus the self-division com- 

 mences in a slightly oblique manner, presently rendered more and more 

 oblique by the younger portions growing alongside one another longitu- 

 dinally, the process being again and again repeated, until an aggregated 

 fasciculus of cells, greater or less in number, is produced. J^ow, the 

 aggregated fasciculated character of the cells in this genus has been, I 

 apprehend, looked upon as one of primary importance. I imagine it is 

 only of secondary, and in its place as primary, I should think, the ob- 

 liquely dividing, slender, attenuated cells should come, and the circum- 

 stance of those cells being aggregated into fasciculi (in A.falcatus form- 

 ing dense fagot- like bundles) be regarded rather as an accidental or se- 

 condary, but very far from unimportant character. Free cells are fre- 

 quently met with of even A. falcatus, in which, on the other hand, the 

 fagot -like bundles are often very large, many ceUs (thirty-two at most, 

 Nag.) being combined together. It is only the most minute species of 

 Closterium that are comparable in dimensions with any Ankistrodesmus. 

 Now, my plant has not a transverse band at the middle (in recent 

 living specimens), but only a rounded, clear spot at one side of the cell; 

 it has no clear space near or at the end of the cell containing moving 

 granules; its self- division is oblique, and the cells frequently remain 

 combined in twos or fours : therefore, I believe it is an Ankistrodesmus. 



