130 
MURIEL ROBERTSON. 
double rhizoplast (figs. 1-5, 21). One is therefore here, as in 
Clathrina, dealing with an extra-nuclear structure. Both 
in the earlier paper published with Prof. Minchin and in the 
present account the work has been done with the view of 
elucidating the blepharoplast-centrosome question, and not 
from the standpoint of the morphology of the sponge. 
The literature of the subject was fully discussed in the earlier 
paper, and I propose only to consider here a few additional 
publications which have particular bearing on the questions 
under investigation. These will be treated under the General 
considerations” at the end of the paper. 
Material. 
The material was obtained from the Marine Laboratory at 
Plymouth, and was fixed in various ways. Corrosive sublimate 
and acetic acid, Flemming’s fluid and Bouiu’s fluid were all 
used ; of: these Bouin’s fluid gave unquestionably the best 
result. The stains used were Heidenhain’s iron-hasmatoxyliu • 
iron haematein, according to Dr. Seidlin’s method (‘ Parasi- 
tology/ vol. iv, p. 94), and Twort’s combination of neutral 
red and Lichtgrim ; these all gave good results, and were used 
with good effect to supplement and control each other. 
The collar-cells of these sponges form, as is well known, a 
single layer of epithelium lining the radial tubes or flagellated 
chambers. The shape of the cell is subject to considerable 
variation, according to the exact condition of the sponge, and, 
indeed of the different parts of the sponge, at the moment of 
fixation ; they are, however, usually either flask-shaped (fig. 1), 
or shaped like chimneys *(fig. 2). The collar is a delicate 
tubular structure which projects from the free end of the 
cells, and varies somewhat in length according as the cell is 
fully extended or not. These structures are of extreme 
delicacy, and are easily destroyed by unsuitable fixation. 
The cytoplasm of the cell is granular and vacuolated, and 
may show large inclusions. The nucleus lies a little below the 
