JAPANESE PALOLO, CEPvATOCEPHALE OSAWAT, N. SP. 31 



from the beginning, the swarm was thickest ; fhll-sized individuals 

 were now seen in abundance together with smaller ones. All 

 swam about rapidly, somewhat after the manner of eels, darting 

 in all directions. I ascertained that the swarm reached to a 

 depth of three or four feet from the surface of the water. Within 

 that extent, the worms in the height of the swarming, were so 

 plentiful that one could not dip his hand into the water without 

 touching some. About an hour and a half from the beginning, 

 the larger worms first began to gradually disappear and as the 

 end of the swarming approached, it was only the smaller specimens 

 that could be found swimming. Two hours after the beginning 

 of the swarming (at 9 p.m.), there were none to be found 

 swimming. 



The above account, except as it concerns the hours of the 

 day, may in general be considered to hold good for all observed 

 cases in which the swarming took place in large numbers. 



Possibly a part of the w r orms after the swarming sink to 

 the river-bottom in an exhausted condition. At the same time 

 it seems certain that a large number of them are carried down 

 stream and some distance out into the sea by the ebbing tide. 

 On more than one occasion I have followed the shifting swarm 

 to a distance of two or three miles from the river-mouth — to the 

 neighborhood of the old forts off Shinagawa, — but always to lose 

 sight of it gradually and altogether at the end. 



In the aquarium, in which I had kept a number of large 

 atocaB and in which the natural conditions were imitated as far 

 as possible, the worms changed, into the epitocous phase and 

 began to swim almost simultaneously with those at large. The 

 ebb and the flood of each day w T ere imitated in the aquarium at 

 proper hours, the former by gradually drawing off the water so 



