NANOMIA CARA. 



203 



Fig. 336. 



the shape cf the swimming-bell is rectangular. There can be but little 

 doubt that these swimming-bells, as I have called them, are genuine 

 Medusa) ; they have all the characters of Medusa), and when they be- 

 come detached, move like them, the only difference being the absence 

 of a proboscis to admit food. This, however, they do not need as long 

 as they remain connected with the main axis, 

 the cavity of which opens directly into the 

 chymiferous tubes, and thus circulates in 

 them whatever food is taken in at the feed- 

 ing mouths, and from them passed into the 

 cavity of the main axis. I have not been 

 able to detect any opening leading directly 

 into the system of chymiferous tubes. These 

 Medusa? are the locomotive organs of the 

 community ; they force the water in and out 

 of their cavity, and thus propel the w r hole community by a sort of 

 alternating motion, resembling that of sculling a boat ; the bells on one 

 side of 'the axis are filling with water, while those of the other side are 

 forcing the water out violently ; the motion begins at the bottom bell, 

 passes on to the top one of the same side, then begins at the bottom of 

 the other row, and so on, throwing the whole of the upper part of the 

 community violently from one side to the other, while the remainder 

 is dragging lazily after it. I have not found any specimens with more 

 than eight swimming-bells fully developed ; the younger bells are 

 added between the first-formed pair and the float, where we find a 

 cluster of swimming bells in different stages of development. These 

 young bells are formed, as the Medusa? buds of the Tubularians, by 

 folds of the outer wall, which gradually grow larger and larger, and 

 circumscribe parts of the main cavity to form chymiferous tubes. 



In their younger stages, the swim- 

 ming-bells resemble still more the 

 Medusa) of Hydroids, when they 

 have not yet assumed an irregular 

 outline, and while their chymiferous 

 tubes are still straight. In the cluster 

 of young bells here given (Fig. 337), 

 we find a few of the different stages 

 through w T hich one of these bells 

 passes, from the time it appears as 

 a mere bud, till it has gone through 



Fig. 33G. The same bell as Fig. 333, seen from the other side, to show the course of the bent 

 tubes, and the mode of connection of the tubes leading into the main axis ; I, bent tube. 



Fig. 337. Group of swimming bells, in different stages of development, o, the chymiferous 

 tubes are simple sacs ; ?>, the tubes, having united, make a circuit ; c, first signs of bending of the 

 tubes, t, of the preceding figures. 



Fig. 337. 



