PETERS. — METABOLISM AND DIVISION IN PROTOZOA. 447 



tive experience with the animal's normal condition and its reaction to 

 various substances aids in forming a judgment as to the numbers to be 

 used. Before one knows the normal adjustments of the animal, larger 

 numbers and more repetition may be necessary than later in the course 

 of the work. The present experiments were often so planned either 

 that each involved the one preceding it, or else that the new experiment 

 was based upon the assumption of the validity of those preceding, and so 

 led to discrepancy if the assumption was false. The above discussion 

 applies to a certain class of experiments the validity of which is but 

 little increased by repetition, but such are not numerous in biology. 



The animals were segregated into groups, five to ten being placed in a 

 watch-glass and constituting a preparation. This was a convenient 

 number, for it was possible to count them under the dissecting micro- 

 scope even after multiplication had taken place. The condition after 

 twenty-four to forty-eight hours, usually the latter interval, was regarded 

 as the one most representative of the effects of the medium upon division. 

 Experiment shows that, in general, division keeps pace with metabolism, 

 the latter of course depending upon the supply of food. In experiments 

 with liquid media it is necessary to exclude every form of natural food, 

 if it be desired to study the effect of the pure medium consisting, e. g., of 

 a salt solution. Hence the Stentors in these experiments carried on 

 their metabolic processes at the expense of the reserve nutriment which 

 they contained when taken from their original culture. Consequently 

 the effects upon division have to be observed early in the period of their 

 subjection to a medium practically free from food. Some results observed 

 after long intervals show that Stentor can live a long time without other 

 food than physiological salts, but the process of multiplication always 

 had its maximum in about the initial twenty-four to forty-eight hours. 

 Sometimes it was necessary for a special purpose to take results after 

 still shorter intervals. 



The question whether division occurs in cycles has an important 

 bearing upon any method of investigation whose results are based upon 

 (Diluting the number of organisms after a period of division. If there is 

 such periodicity, it would be important to know whether the animals 

 taken for comparative experiments had been taken in the maximum or 

 in the mimiinum of the cycle. In the case of Stentor neither direct 

 observation nor the experiments made, furnish evidence of any inherent 

 periodicity of division. The present experiments show that, except when 

 some special modification of the medium exists (e. g. presence of potassium 

 chloride in excess), multiplication runs, in the main, parallel to meta- 



