FEEDING PARAMECIUM KNOWN BACTERIA 147 



lated with a portion of a head of timothy hay every day and 

 used when twenty-four hours old, thus securing fresh cultures of 

 bacteria on the hay. It was believed that in this way a chance 

 mixture of bacteria would be secured which would be more com- 

 parable to that of the fresh infusions in which Paramecia thrive 

 best when grown in mass culture. It turned out in actual prac- 

 tice that this assumption was probably incorrect. Infusions 

 seeded in this way were excellent for the growth of Paramecium 

 for four months. After this the division rate sharply declined 

 and the death rate increased until the lines in this mixture died, 

 and it was impossible to replace them. The possibility is sug- 

 gested that the actual change in the bacterial content of the 

 chance mixture was greater than if dependence had been placed 

 upon inoculation with bacteria carried over with the Paramecia. 

 This failure of the chance mixture could not have been foretold 

 during the course of the experiments, since its behavior up to 

 the actual time of death of the animals w^as not unlike that of 

 certain of the cultures of known bacterial content. Because of 

 the failure of this chance mixture during the latter part of the 

 w^ork, the culture C was used as the control, the chance mixture 

 serving as an additional index during the time it was furnishing 

 adequate food. This line C was chosen as the control because 

 of its uniform behavior throughout the course of the observa- 

 tions, and because it was a component of all infusions of known 

 bacterial content used during this time, save in those cases where 

 it was desired to make a complete change of food. 



From five to ten lines of Paramecium were grown in each cul- 

 ture of known bacterial content. An equal number of animals 

 w^ere grown in the chance mixture. Daily transfers were made 

 to fresh fluid except in a few cases during the winter when the 

 temperature of the laboratory was so low that bacteria would 

 not grow rapidly enough to furnish a good medium. But a single 

 animal was carried over in each transfer, so that a new series of 

 cultures was started each day. Contamination from the air 

 during these transfers was avoided by raising the cover of the 

 Petri dish only enough to admit the passage of a carefully flamed 

 capillary pipette. Since it was not deemed necessary to wash 



