CONFERVOID FILAMENTS OF MOSSES. 575 



branches. This I have drawn on PI. LVII. fig. 12. The precise point from which the 

 roots (true) are given off I have not observed. In PL LVII. fig. 13, the filament 

 from which this ascending axis sprang is tinged with brown. This is not a necessary 

 point ; I have seen it arise from ceUs of the most distinctly marked Confervoid type. 

 This mode of development is frequently to be met with, and seems to hold an analogous 

 position to the buds on the creeping stems of Phanerogamous plants. 



The next plan to be described, by which these filaments assist in reproduction, is that 

 which has been described by Kutzing as the " ■peculiar fruit " of the Protonemea?. They 

 are, however, by no means confined to the filaments, but can be found on any part of 

 the plant except the spore-frond (sporange and peduncle). They are found at the end 

 of a branch of a filament whose cells, instead of elongating, become broader, the whole 

 assuming the form of a club, containing 5 to 8 cells. The cell-waUs become thickened, 

 and the green granular contents closely approximated. After some time the cell-walls 

 assume a reddish-brown colour, which becomes darker by age. Segmentation sometimes 

 takes place in a direction parallel with the axis of the filament, or even irregularly 

 (PL LVII. fig. 4ffl). When the cell- wall has assumed a full brown colour, the green 

 colour of the contents becomes fainter, and at last they are frequently transformed into 

 oil-granules, after the manner of the formation of the oily contents of the hypnospores 

 of Volvox and other Algse. When these gemmae have become quite brown, they easily 

 separate from the filament and become free. They occur at aU times of the year, but 

 are most observable during drought. When they are excited to grow, they begin to 

 shoot out into a confervoid filament, generally from the extremities, although sometimes 

 each cell may push out a branch laterally. At PL LVII. figs. 4 a, 5, 15 a, PL LVIII. 

 fig. 23 b, are shown these " gemmae." No doubt these are "resting gemmce" one of the 

 means by which the life of the plant is preserved during severe trials of drought and 

 cold *. 



They have been called gemmae, and considered the homologues of the gemmae (bulbils, 

 &c.) of the higher plants ; but although they are subservient to the reproduction of the 

 parent, yet the ascending axis does not immediately spring out of them, but, as has been 

 already remarked, they sprout out into confervoid filaments first, from any one of the com- 

 ponent cells, which then pass on to the formation of the ascending axis, as above noticed. 

 They differ, therefore, from the bulbils in not possessing a true " stirps " within them- 

 selves. It seems that we should rather consider each as a resting coiupound gonicUum, 

 — the reason for which will be more apparent as we proceed to the consideration of the 

 single-celled gonidium itself. 



But this is not the only method by which the confervoid filaments are reproduced : any 

 one of the cells, detached from the other, is capable of continuing the growth of the fila- 

 ment, in the same manner as each is capable of doing it whilst forming part of the filament, 

 by branching and division, as I have before noticed. And there is a great tendency for 

 these cells to separate from each other, more particularly in the older filaments (see 

 PL LVII. figs. 4 b, 10, 11, 12, 14, 15) ; but whether old or young, they may bulge out 



* Some of these points have been already shown by Schimper, but they are here repeated in order that a clearer 

 idea may be formed of the single gonidium. 



VOL. XXIII. 4 H 



