576 DR. J. B, HICKS ON THE GONIDIA AND 



on any side and form a branch, which, segmenting, becomes a true filament (PL LVII. 

 fig. 4 c). Not nnfrequently these cells retain their linear form, especially after the cell- 

 wall has become dense by age, though sometimes whilst the linear growth is very 

 active. In the latter case, the contents of these actively growing cells occasionally be- 

 come more or less homogeneous, with a distinct central nucleus (PL LVII. fig. 15 c), 

 and much resemble a single cell of Falmoglcea. As a rule, when the separated cell 

 retains its linear form, and is in active growth, it does not in general branch laterally 

 for some time, but continues in the linear direction by binary division for a considerable 

 period. But frequently, under certain conditions, the terminal cells, instead of possessing 

 an elongated form, become more or less globular, which shape is not confined to the last 

 one, although it may be more marked in it, but is frequently seen in the four or five (or 

 even more) nearest the end of the branch (PL LVII. figs. 10 b, 13, 15). Of these the terminal 

 one possesses the greatest tendency to become free, although all of them do so very readily. 

 These free cells, then, appear quite globular, with green contents, which are sometimes 

 granular, though more frequently more or less homogeneous, with or without a central 

 nucleus, as is observed in the gonidium of the Lichens (PL LVII. fig. 10 c). Where 

 there is this tendency to the globular form of the cells near the extremity of the filaments, 

 there is also a disposition in the cell-contents to become more /leeply green. This is 

 partly owing to the deeper tint the contents assume, and partly to the closer crowding of 

 the chlorophyll-granules of which they are composed. 



These terminal globular cells have, under certain conditions, a tendency to become 

 quiescent ; the cell-wall then becomes thicker, and, as in the cells mentioned before, it 

 frequently assumes a reddish-brown colour, whereby the whole becomes very dark — 

 indeed, so as to become almost a black ball (PL LVII. fig. 15 b, PL LVIII. fig. 16). These 

 cells also segment in a less regular manner than do those of the same filament below it ; 

 there is an imperfect tendency to branch. The cells thus formed at the side tend to the 

 globular form (PL LVIII. fig. 16 b, c). The cells, thus altered in appearance, generaUy 

 detach themselves in groups of two or three cells to each, and can be carried about by the 

 elements. They, however, occasionally do not become globular, but more or less oval, 

 springing from the end of one cell in groups of three or four, as shown at PL LVII. 

 fig. 15 b, each cell of which, becoming free, appears like the cell of a Palmoylcea, especially 

 as in some the contents become homogeneous, with a central nucleus. As this mode is 

 observable on the branch of the same filament which possesses the other forms at the 

 same time, this peculiar appearance is thus traced to its true origin, which would 

 otherwise scarcely be guessed at. The cells are more active than the globular forms ; 

 and some, as soon as they separate, begin to segment linearly, and to form a filament. 

 These conditions Avere looked upon as Confervse by the older algologists, under the 

 name of G. multicapsularis (PL LVIII. fig. 16; see Dillwyn's ' Confervoe ') and C. 

 umbrosa (PL LVII. fig. 15 6). I have already pointed out Kiitzing's arrangement of 

 these and other forms. 



These free dark cells with crowded contents, when influenced by warmth and moisture, 

 form branches in the manner already described, thus giving rise to new filaments. 

 They, however, may become mother-cells, as will appear below. In PL LVIII. fig. 16 a 



