CONFERVOID FILAMENTS OF MOSSES. 577 



will be seen a number of filaments crowding each, other and running parallel, the 

 terminal cells of which, having assumed the dark colour and globular form, produce a 

 dark velvety covering to the surface upon which they grow — and thus presenting an 

 appearance easily mistaken for ConfervaB. 



I have hitherto in this communication shown the multiplication of the cells upon the 

 linear-binary form of growth ; we now, however, come to another stage, where the 

 segmentation proceeds upon the quaternary or its multiples ; and this leads us into still 

 more interesting ground. 



The first time I observed this was in some filaments which I placed in water ; and 

 these, after growing some time, produced cells at various points, chiefly at their ends, 

 which had not only segmented once upon the quaternary plan, but their subdivisions 

 had also repeated the process. This I have shown on PI. LVIIL fig. 17- In some the 

 process had extended still further, and in some less regularly, so as to produce irregular 

 masses of green cells, in varying degrees of the same form of segmentation (PI. LVIIL 

 fig. 17 b). However, in some the contents had divided into six or more portions round 

 a common centre (PI. LVIIL fig. 17 c), and the parent-cell wall, bursting, set free a group 

 of cells ready again to divide. 



Guided by these facts, I pursued my investigations in the same direction, and found 

 that the globular cells, which I have ah-eady described, separating from the ends of the 

 filaments, frequently underwent quaternary subdivision, and that in them the process 

 went forward to the infinite multiplication of these green cells, the result of which 

 repeated and rapid segmentation was to produce cells of exceeding minuteness. This, 

 as I have formerly shown in a similar condition in the gonidia of Lichens, is dependent 

 on the preponderance which the process of subdivision holds over individual cell-growth. 

 Wlien the former process is iu abeyance, then the latter regains the ascendency ; and 

 these little divisions, so small as scarcely to show any distinction between cell-wall 

 and contents, gradually increase in size so as at last to equal the original parent 

 cell *. 



At first the contents of the cells were somewhat granular, but after a generation or 

 two they became homogeneous and, in every respect, could not be distinguished from 

 the subdivisions of the Lichen-gonidium. 



And in another respect they much resembled these latter, namely, in the great 

 tendency for the process to keep on unvaryingly in the form in which it had begun. 

 This can readily be observed by any one who will take the trouble. Large areas may 

 thus be covered by the growth of these cells, which may continue for a long period of 

 time, certainly over a year, and probably, as far as I can make out, for many years. 



It may always be noticed on the face of any wall where Mosses grow, that underneath 



* In both the Lichen- and Moss-gonidium, this property of repeated and incessant subdivision is commenced so 

 early in the cell, that one subdivision is hardly fairly perceptible before the next can be recognized ; indeed, in some of 

 the cells, three and four generations are included in one parent. Upon this point Braun says, in ' Rejuvenescence in 

 Nature' (Ray Soc. 1853, p. 239): — "There are cells which never become old, but in their earliest stage, by 

 dividing, give up their existence again, or rather continue it in a new generation, till age finally is attained in a last 

 generation, which never undergoes division." 



4h2 



