1 6 SAKYO KAXDA. 



Lyon then adopted a method "by which the density of small 

 organisms could be found without any injury or change in them." 

 He "made use of the haematocrit attachment of the centrifuge 

 by which very high velocities" (from 10,000 revolutions to 

 12,000 per minute) "of rotation and consequent high centrifugal 

 force are secured. At the bottom of a small 1 tube was placed a 

 little 1 gum-arabic solution of known density; above this a few 1 

 drops of water containing paramecia. The tube was then 

 centrifuged about one minute." He thus found that "an 

 average density of paramecia is about 1,048 or 1,049" (17, p. 



i. Experiments on Paramecium caudatum. 



In the results of Jensen, Platt, and Lyon, however, the differ- 

 ence of density is so great that the present writer thought it 

 well to repeat Dr. Lyon's experiments, under his direction, with 

 some modifications. The writer prepared a gum-arabic solution 

 using distilled water. The acid of the solution was carefully 

 neutralized with sodium carbonate. But he regrets that he 

 forgot to dialyze it. He made a series of 40 solutions of different 

 densities from 1.02 to 1.05 in Naples jars. The specific gravity 

 of each was carefully determined by means of a finely graded 

 hydrometer. After some trials the solutions above and below 

 the densities given in the following table were found unnecessary. 

 Then he prepared many small tubes about 12.5 cm. long and 

 0.4 cm. in diameter, one end of which was sealed. In each trial, 

 two of those tubes were filled about 1 1 cm. high with gum-arabic 

 solutions of known specific gravity, e. g., the one 1.02 and the 

 other i. 02 1. This pair was centrifuged to drive the air out. 

 Then on the top of each solution in the two tubes, one drop of 

 the culture containing dense paramecia was added by a very 

 small pointed pipette. The tubes so prepared were again centri- 

 fuged for one minute with a speed of about 1,300 revolutions 

 per minute. The distribution of the animals in both tubes was 

 then carefully compared. All 40 different solutions having been 

 thus tested, the nearest possible density of the animals was 

 obtained : 



1 Italics not in the original. 



