II. A FLORA WITHIN ANIMALS. 25 



and correspond to the description of the penultimate secondary cell above given, 

 while the third is like the ultimate one described. When a single secondary cell 

 exists in the last-mentioned species of Enterdtryus, it has usually the same appear- 

 ance as the terminal cell of those cases in which two arc found. 



In a half dozen instances only, among several thousand filaments of Entero- 

 bryus attenuatus, have I been able to detect the existence of a secondary cell. 

 In some cases it was relatively very short and cylindrical, a trifling degree nar- 

 rower than the principal cell, truncated at its free extremity, concave at the other, 

 and fitting upon the convex termination of the principal cell, and .entirely filled 

 with granular matter. In others, the secondary cell was a very little broader or 

 more bulging than the principal cell, longer than in the former instances, obtusely 

 rounded at the free extremity, and filled with a hyaline protoplasma, with a few 

 granules. 



The connection of the secondary cells with the principal cell arid with one an- 

 other is very variable in its appearance in Enterobryus, and indicates that the 

 former are derived from the latter by a process of division ; or, in other words, it 

 indicates at least one of the modes of reproduction of the plant to be by division. 



The first indication of the separation of a secondary cell from the principal cell 

 is a faint line corresponding to a plane passing transversely through the contents 

 of the latter in the vicinity of its distal extremity (PI. IV. 25, c). Examples are 

 met with in which such a plane exists with the merest trace of a circumferential 

 contraction of the primordial utricle at the edge of the plane without the slightest 

 mark of such a character existing in the permanent cell-wall. The cell-contents 

 above the plane of separation and for some distance below are always more or less 

 granular, frequently to the total exclusion of globules, although these are frequently 

 met with within the fully formed secondary cells. 



In well developed principal cells of Enterobryus without attached secondary 

 cells, the free extremity very generally is more or less filled or even distended with 

 the characteristic globules, and is clavato in form (PL III. 4, 14), but when second- 

 ary cells are observed in their commencement, they are filled with granular mat- 

 ter (PI. IV. 18, 21, 24), from which it appeafs the globules within the distal 

 extremity of the principal cell are converted into granules in the development of 

 the secondary cells. 



I never observed this transformation step by step, but only saw the conditions 

 which I have just described. In these cases, if the globules are really converted 

 into granules, it is most probably not by the former being broken up into corre- 

 sponding masses of the latter, but by a separation of granules more or less rapidly 

 from the circumference of the globules; for when secondary cells are in course of 

 development from the principal cell, the distal extremity of the latter always con- 

 tains more or less granular matter enveloping the globules, which are spherical, 

 and, relatively to those ordinarily found lower down, small in size. 



Following the plane of division of the principal cell-contents in the production 

 of a secondary cell, examples of Enterdbryus are met with in which the circumfe- 

 rential contraction before mentioned, existing at the edge of the plane in the 



