Interzooecial Communication in Flustridds. By A. W. Waters. 281 
With the exception of Levinsen, few authors have noticed them, 
but in more than one paper he has supported my view as to 
their value in classification, and recently, in a most valuable work,* 
he has given particulars of the mode of communication in 34 species 
of Chilostomata. 
The mode of connection is, throughout the Bryozoa generally, a 
useful character, but in what we may call the calcareous forms the 
examination is not easy, though sections ground on a hone, or cut 
sections of decalcified stained material, will usually give the chief 
points. 
As Flustra seemed the most suitable genus for making compara- 
tive studies of these rosette plates, I have prepared sections, mostly 
from dried specimens,! of all the species which, were available, with the 
result that all show some distinguishing character. I must here 
thank Miss Jelly, Mr. Levinsen, and Mr. Kirkpatrick for specimens 
enabling me to make the list more complete. 
The characters of importance are the position, the number on both 
walls, and the number of pores on each plate through which com- 
munication takes place. First, there is a connection between the 
distal end of one zooecium and the proximal end of the next, and this, 
for simplicity, I speak of as the distal wall. In the distal wall there 
may be only one rosette plate, as in Flustra securifrons, papyrea,. 
nobilis , pisciformis, biseriata. 
Or two, as in F. serrulata. 
Or three rosette plates, as in F. membranaceo-truncata and 
Barleei . 
Or four, as in F. tenella and oblonga. 
Or six, as in F. papyracea, cribrif ormis, denticulata, dentigera , 
and foliacea. 
Or the distal plates may be numerous, as in F. carbasea, armata, 
ovoidea, militaris, reticulum, crassa ; and Euthyris bombycina, epi- 
scopalis, and obtecta. 
The plate may have only one perforation for the protoplasmic 
connection, as in F. tenella (plate VII. fig. 25), or there may be a large 
plate with many perforations, as in F. nobilis (plate VII. fig. 10). 
Turning to the lateral walls, there is a question how the plates 
should be counted. The entire lateral wall of one zooecium is in con- 
tact with the halves of the lateral walls of two zooecia. This may be 
seen in such species as Membranipora pilosa (plate VIII. fig. 1) and 
F. crassa (plate VIII. fig. 5), though where the zooecium is more 
hexagonal it might seem correct to speak of six walls. Levinsen gives 
the number of rosette plates in the upper half of the lateral wall, and 
* Levinsen, G. M. R., ‘Mosdyr,’ Zoologica Danica, 9 te Heft. See also his 
‘ Bryozoer fra Kara-Havet’ and his ‘ Polvzoa,’ from which many of the figures in his 
last work are copied. 
t It is often easier to make out these plates in dried than fresh specimens, and 
many of my sections have been cut without decalcifying. 
