252 Dr. R. von Lendenfeld on Guard- Polyps 
not belong to the surface-cells, but are decidedly subepithelial. 
In isolation-preparations it is easy to detect the muscular cor- 
puscles in the isolated fibrille ; they le between the centri- 
petal ends of the surface-cells, and consequently may be 
called interstitial cells. 
The muscular fibres form a cylindrical tube by lying in a 
parallel position upon the supporting lamella. When the 
guard-polyp is contracted they he in a straight line and lon- 
gitudinally. At the extremity of the endodermal axis the 
muscular cylinder also terminates, and there are here in the 
subepithelial layer from two to four fine large multipolar 
ganglion-cells. In the superficial layer of the whole terminal 
portion of the machopolyp we find large elongated urticating 
capsules. Each urticating capsule is enclosed by a thin 
plasma-envelope; the centrifugal end alone seems to be free. 
In this plasma-envelope there is a flat nucleus. The enidocil 
is short and forms an angle of 60° with the surface; it is 
always situated opposite to the nucleus. Sometimes it is 
slightly bent, and then the concave side is always turned 
towards the urticating capsule. 
The centripetal end of the plasma-envelope is drawn out 
into a process which lies radially and always appears to be 
straight. All these processes unite at the centrifugal ex- 
tremity of the endodermal axis; therefore at the same place 
where the ganglion-cells are situated. Further, there are 
also in the terminal knob of the guard-animal radial mus- 
cular fibres, which are sometimes closely applied to the 
pedicles of the enidoblasts. Supporting cells with broad 
centrifugal terminal lamella also occur. Some of the super- 
ficial cells are remarkable for their slenderness, and may be 
interpreted as sense-cells. I have not succeeded, however, 
in detecting any connexion of these cells with the subepithelial 
ganglion-cells, so that [ am not sure whether or not these 
elements perform a tactile function. 
The guard-polyps are capable of extraordinary extension 
in length, which, however, takes place slowly, like the exten- 
sion of the tentacles of Hydra, and never shows that precision 
and rapidity of movement which characterize the solid ten- 
tacles. ‘The retraction of the guard-animal takes place much 
more quickly; it may be effected in half a second. The 
movements of the extended guard-animal are inconsiderable 
and really amoeboid. 
Both the extension and contraction and the movements 
which are performed in the extended state may be easily 
recognized as the effects of two antagonists. Just as in the 
tentacles, we have here to do with an elastic rod on all sides 
