THE ACTIVITIES OF CORYMORPHA 325 



and the hydranth is reached. Here commonly a slight block is 

 encountered, due probably in part to the contraction of the 

 sphincter in the circular muscle of the stalk and in part, per- 

 haps, to the fenestrated membrane (Torrey, '07, p. 283). By 

 continued slight pressure the ink will suddenly spurt through 

 this neck and quickly fill the cavity of the hydranth where 

 again it may meet a block if the mouth is closed. This obstruc- 

 tion, which seems also due to a sphincter, may likewise be over- 

 come, whereupon the ink will flow in a fine stream throug 1 the 

 mouth into the outer water. I have never been able to dis- 

 cover any evidence in favor of the view expressed by some of 

 the older workers that extensions of the gastro vascular cavity 

 make their way from the canals into the regions between the 

 vacuolated cells of the stalk (Agassiz, '62, p. 278) or from the 

 cavity in the hydranth to the axes of the tentacles (Allman, 

 71-72, pi. 19, fig. 7). 



The gathering of food into the gastrovascular cavity is one 

 of the best organized and most characteristic sets of responses 

 in Corymorpha. In an aquarium through which a gentle cur- 

 rent of water has been sweeping and in which the specimens of 

 Corymorpha have become well acclimated, they will be found 

 well fixed in the silt, with stalks upright and with their hy- 

 dranths pointing in the direction in which the current flows. If 

 the current is now shut off, feeding movements will begin in 

 about three to four minutes. These have been briefly described 

 by Torrey ('04 a, p. 397) and consist of the following steps. 

 At the beginning of a feeding movement the expanded proximal 

 tentacles with two or three convulsive efforts are drawn together 

 around the proboscis and at the same time the stalk is short- 

 ened. The stalk then curves so that the hydranth is brought 

 close to the mud. The proximal tentacles are opened and the 

 distal tentacles and proboscis are applied to the mud. The 

 stalk then slowly straightens and becomes vertical, the proximal 

 tentacles attaining full expansion as the resting position from 

 which the response started is again assumed. The whole re- 

 sponse partakes of the nature of a deep bow and requires for its 

 completion about one minute. As a result of it many small 



