394 DR. A. FARRE ON THE STRUCTURE OF THE CILIOBRACHIATE POLYPI. 
For the process of retraction two distinct sets of muscles are provided, the one act- 
ing upon the animal, and the other upon the flexible part of the cell. 
The muscles for the retraction of the animal are contained in the visceral cavity, 
and consist of two bundles of delicate thread-like chords (fig. 3. a 8 and 9.) ; the 
one set (a 8.) arising from the bottom of the cell to be inserted about the base of the 
stomach ; the other (a 9.) also arising from near the bottom of the cell, though ge- 
nerally at the opposite side from the former, and passing up free by the side of the 
pharynx to be inserted around the line of junction of this organ with the base of the 
tentacula. The muscles provided for the retraction of the operculum, or flexible 
portion of the cell, have their origin from the inner surface and near the top of the 
stiff part, and are inserted into the flexible portion, on which they act. (fig. 3. & 2 and 3.) 
They are most distinctly seen when the flexible operculum is completely drawn in, 
at which time the latter is folded up, so as to occupy the axis of the upper part of the 
cell, and to it the muscles are seen extending inwards from the opposite sides of the 
cell from which they have their origin. They consist of six flattened bundles of fibres 
having a tri radiate arrangement. The upper three sets ( b 2.) act upon the flexible 
part of the cell and are inserted into it. The lower three (b 3.) are smaller, and are 
for the purpose of retracting the bundle of setse by which it is crowned. 
It is at this point that the best opportunity is afforded for investigating the struc- 
ture of this form of muscle. It would appear as if muscular fibre were here reduced 
to its simplest condition. The filaments are totally disconnected, and are arranged 
the one above the other in a single series. They pass straight and parallel from their 
origin to their insertion, and have a uniform diameter throughout their course, ex- 
cept that each filament generally presents a small knot upon its centre, which is most 
apparent when it is in a state of contraction, at which time the whole filament also is 
obviously thicker than when relaxed. The filaments have a watery transparency and 
smooth surface, and under the highest powers of the microscope present neither an 
appearance of cross markings nor of a linear arrangement of globules. 
These muscles, though apparently attached to the inner walls of the cell, must yet 
have the membranous parietes of the body interposed between their insertions and 
these walls, if as I suppose the cell is completely lined by the integument. In the 
lower part the integument is only occasionally seen separate from the walls of the 
cell, but above it may be easily discerned in the expanded animal passing up to be 
inserted around the tentacular ring, and thus distinctly bounding this part of the 
body which is always free within the expanded operculum. It is probable, therefore, 
that the retractors of the operculum as well as those of the body are within the visceral 
cavity, and that the relation of the origins of both, with regard to the integument 
and cell is similar to that which exists in the attachment of the muscles, with refer- 
ence to the mantle and shell of bivalve Mollusca. This is easily understood if we 
suppose that the integument in this case, as I have ascertained it to be in another 
species, is attached to the operculum on a line with the base of the setse, which is the 
