300 
Phillips. — On the Development of 
Polysiphonia violacea (Roth), Grev. 
Figs. 5 , 6 . 
P. violacea differs from those already described in the 
possession of four siphons, the smallest number known in 
the genus. The siphons of P. fastigiata reach as many as 
twenty in parts, and those of P. nigrescens a larger number 
still, but both are free from the so-called cortex of Rhodomela 
subfusca. Rosenvinge (16) in his paper on the morphology 
of the genus regards P. nigrescens , fastigiata , and violacea 
as types of as many groups within the genus, and to these 
species he largely devotes his attention. It is therefore 
interesting to find that in all essential respects the structure of 
the cystocarp in Polysiphonia violacea is cell for cell com- 
parable with that of the three species already enumerated. 
The only point of difference arises when the history of the 
sporogenous derivative of the auxiliary cell is traced sub- 
sequent to spore-formation. Under P. nigrescens it has been 
stated that the sporogenous cell seems to absorb the cell 
from which it sprang, and even the sterile derivatives of that 
cell. In P. violacea the process of fusion goes even farther, 
to the inclusion, that is to say, of the central cell itself. 
Here again I cannot say whether the cells of the sterile 
branches, more especially the peripheral cells, are absorbed 
or atrophy ; they cannot, however, be traced in the maturer 
cystocarps, where a large multinucleate mass confluent with 
the central cell may be observed. There can be no doubt, 
however, that the earlier carpospores are derived directly 
from the sporogenous derivative of the auxiliary cell, and 
we may fairly infer that when fusion with other cells occurs 
later, it is daughter-nuclei of this cell, or perhaps daughter- 
nuclei of the auxiliary cell, which furnish the nuclei of the 
carpospores. 
