TOXIC ACTION OF KCN ON PARAMECIUM AND DIDINIUM. 213 



vidual cells with a definitely known history. This necessitated 

 a great deal more labor, but was necessary in order to know 

 exactly the age and history of each cell that was used. 



EXPERIMENTAL. 



The first two experiments were preliminary and served to 

 orient the problem. 



Experiment I. Suspecting that starvation of Paramecia had 

 some relation to its power of survival in KCN solutions, a quan- 

 tity of Paramecia from a pure line hay infusion culture was 

 centrifuged and placed in one half native medium and one half 

 tap water. This was used as a control. Twenty c.c. of this 

 was added to a jar containing 600 c.c. tap water, on March 18; 

 and the same to another jar, on March 19. I made no counts, 

 but there were not over seven or eight hundred animals in each 

 jar, so that there was a large volume for each individual. Forty- 

 three of the control animals were tested immediately with the 

 KCN. About twenty of those added to the pure tap water were 

 tested as soon as added, and about the same number on each 

 successive day following. The animal to be tested was trans- 

 ferred from the tap water to about 2 c.c. N/$o KCN solution in 

 a watch glass by means of a capillary pipette. As small an 

 amount of water as possible was transferred with the animal. 

 It was then watched, a binocular microscope being used, until all 

 movement ceased, cessation of movement being considered the 

 death point. Soon after this, as a rule, the animal completely 

 disintegrated. The average time of survival is given in Table I. 



As can be seen by glancing at the table the survival time varies 

 greatly in both lots, and apparently chaotically. The standard 

 deviation 1 each day was large, varying from 0.63 .06, lot I., 

 for the sixth day to 2.73 .24 on the fourth day; and similarly 

 for lot II. This did not decrease as time went on. 



Before proceeding further, the question arose as to whether 

 the time of beginning cytolysis might be a more unvarying death 

 point than the time of cessation of movement. If the animals 

 are in the same condition as regards food and age then the death 



1 The standard deviation is a measure of the degree of variation. See C. B. 

 Davenport, "Handbook of Statistical Methods." 



