554 WALDO SHUMWAY 



b. On the gastric vacuoles. The gastric 'vacuoles, those 'im- 

 provised stomachs' in which the processes of digestion take place, 

 have been carefully studied by many investigators. Metalnikow 

 ('12) has shown that in any given line of Paramaecia, the rate 

 of formation of these vacuoles is constant under normal condi- 

 tions, subject to an half-hourly rhythm and not influenced by the 

 quantity of food present. The quality of the food however ex- 

 ercised a striking effect on the rate of vacuole formation. A more 

 easily assimilable food produced a decrease in the time intervals 

 between the formation of the vacuoles and an increase in the 

 length of their cyclosis. Experiments with changed environ- 

 mental conditions showed that this effect was also produced by 

 slight doses of arsenic and alcohol and by raising the tempera- 

 ture. As these agents have been reported as increasing the 

 division rate of Paramaecium, it seemed advisable to test the 

 effect of thyroid feeding with this index. 



The technique employed is a modification of that of Metal- 

 nikow. The Paramaecia left over in the watch glasses after the 

 daily isolations were allowed to multiply for not over three days 

 in small test tubes containing 4 cc. of hay infusion or thyroid 

 suspension. A large number of Paramaecia was removed from 

 each test tube and allowed to stand half an hour in ten drops of 

 freshly prepared hay infusion or thyroid suspension. Twenty or 

 more individuals from each culture were isolated on two slides 

 and a drop of carmine solution added. After five minutes they 

 were killed and fixed by adding on the underside of a coverslip 

 the following fluid. 



Glycerin 10 cc. 



Corrosive sublimate to sat. 



Glacial acetic acid 1 cc. 



Camphor water 10 cc. 



This fluid kills rapidly with a minimum of distortion and 

 clears the protoplasm so that the carmine filled vacuoles are re- 

 markably distinct and easy to count. The number of these 

 • vacuoles is counted for twenty individuals on each slide and an 

 average obtained. These slides may then be laid aside until 

 opportunity permits of verifying the count. 



