vol. i.i Bitter-Davis. — Enter opneusta. 191 



makes the former seem more probable, for rotation should, on 

 the assumption of the latter, correspond in rate to that of the 

 ciliary wave, which, however, is more rapid than the rotation. 



As lias already been indicated, vertically upward swimming 

 is rare. Of the large number of individuals examined such 

 movements occurred in a few instances, but only for very short 

 distances. Usually a spiral course was taken; aud likewise when 

 sinking in perfectly quiet water the animal moved slowly in a 

 spiral path until it reached the bottom of the aquarium. A 

 detailed study of the rate and manner of sinking was made to 

 supplement the observations on rising. Sinking rates were 

 determined in two ways. In one, the animals were put in a 

 long test-tube filled with water and corked tightly. Two marks 

 200 mm. apart were scratched on the tube some distance from the 

 ends so as to give the animal space for orientation, aud also to 

 prevent counter influences of small air hubbies which were nec- 

 essarily, to some extent, present. Trials were made by carefully 

 reversing the tube from one vertical position to another, and the 

 time of falling through the distance of 200 mm. noted by means 

 of a stop watch. In the second method a large battery jar was 

 at first used, but later, in order to get a longer column of water, 

 a tall specimen jar was substituted. A fine line some distance 

 below the surface of the water was marked on the side of the jar 

 for the starting point. Animals were carefully placed below the 

 surface film by means of a large pipette or floated from a watch- 

 glass, care being taken to prevent water currents. By the time 

 the starting line was reached the animals had oriented themselves 

 and were moving normally. The tube method was abandoned 

 because it was thought that the occasional friction against the 

 sides of the tube might be a source of error. The results, how- 

 ever, were practically the same for both methods. Trials were 

 made with individuals of different sizes, with all other factors 

 constant except specific gravity, to determine variation in sinking 

 rate due to difference in size, and to correlate this with variations 

 in specific gravity. It was found that the smallest individuals 

 fell through the distance of 300 mm. 20-30 seconds more quickly 

 than the largest ones. Other trials were made in which speci- 

 mens of different sizes were started from the same point at the 



