574 DR. J. B. HICKS ON THE GONIDIA AND 



red or reddish-brown colour in the cellnlose of the cell-wall. It occurs most frequently 

 in those found on bark of trees and walls, and is more noticeable in summer and avitumn, 

 although it is to be met with at all seasons in the older cells (PL LVII. figs. 7, 11). 

 "WTien a new growth springs up, the young cell-walls possess no colour, showing that it 

 is the production of age. On one occasion, I saw this colour produced in aU the cells of 

 a large quantity of confervoid filaments which had grown in water in the shade. I had 

 placed the glass out of doors in the sun, and was surprised to find that all the older 

 cell-walls had become dyed of a reddish-brown hue, although previously colourless : 

 the contents within remained green. There is another change I have observed, ap- 

 parently connected with drought, and probably also with a dimunution of vital activity, 

 namely, the contents become brown or reddish brown. PI. LVII. fig. 12 b shows the 

 filament green, while at a it is changed to brown. That they were from the same source 

 was readily proved, by finding the green and brown cells on one filament ui varying 

 degrees of change. These brown cells separated very readily from one another, as 

 is the case in all the older cells of the filaments. The filament possessed somewhat 

 the appearance of the filamentous diatoms [Melosira, e. g.) ; and this was more striking 

 upon the separation of the ceUs ; for both the ends of each were marked with ra- 

 diating Knes, which, seen in profile, proved to be ridges (PI. LVII. fig. 12 c). At the 

 same time the cell-wall was very rigid, thin, transparent, and comparatively inde- 

 structible. It seems to me that tliis condition is to be considered as a "resting'" form, 

 equivalent to that state well known in the Algse. Their subsequent history I was not 

 able to trace. 



Having thus endeavoured to point out the more actively growing condition of the 

 confervoid filaments, I shaU attempt to show the various means by which they assist in 

 the reproduction of the parent Moss *. 



The first I shall mention is one variety of those modes in which in all cases the 

 ascending axis is formed out of the ceU of the filament. It is also one of the most 

 direct methods by which the filament attains that end. Eor other modes I must refer 

 the reader to Schimj)er's work above quoted. 



Prom any cell of the filament, except from those which are concerned in the processes 

 to be related hereafter, a branch is produced, the end cell of which divides into two 

 cells in the ordinary way above described, the terminal one increasing considerably in 

 size. Prom this, many (three or four) branches spring, in the mode of branching before 

 mentioned, in a row or verticel ; the cells of these branches are delicate and tapering, 

 and have the property of curving in towards the centre. There is also a similar row 

 of smaller branches springing from the same cell within the former row, surrounding as 

 it were an imaginary axis. From this springs, by gradual increment in the number of 

 the cells, the stem of the Moss. The first attempt at differentiation on the part of the 

 confervoid filament is thus shown to be in the cell producing the rows of curving-in 



* In the following remarks I have assumed that all these various phenomena tend, sooner or later, to the formation 

 of confervoid filaments, and through them to the other parts of the Moss. I have not possession of definite proof 

 of this ; but, from our knowledge of kindred conditions, we may, I think, fairly assume such to be the direction of 

 their growth. 



