PETERS. — METABOLISM AND DIVISION IN PROTOZOA. 445 



jectives of low power. It was finally rejected, perhaps unneeessarilj, 

 for the method uext described in consideration of the fact that conditions 

 in a drop might not be entirely normal for so large an animal as Stentor. 

 For almost all other Protozoa this objection would not hold. 



Solid watch-glasses were then adopted ; these were polished so as 

 to fit closely when piled on top of one another. They were shallow, 

 about 50 mm. in diameter and 10 mm. deep, and had concave internal 

 surfaces, thus presenting no angles in which the animals might lodge 

 and so escape observation and counting. A watch-glass was only par- 

 tially filled with the required medium, 2 to 4 cc. being placed in each. 

 No more animals were introduced than could be conveniently counted 

 under a dissecting microscope. When piled up the watch-glasses formed 

 a series of closed chambers with sufficient air space and with a much 

 larger proportion of liquid to each Stentor than in the hanging-drop 

 method. As nearly as could be imitated the conditions of a minute 

 natural pond existed here. This method was used for nearly all 

 experiments. 



An important point in any method is the transference of the organ- 

 isms from their native culture medium to the test-medium in the observa- 

 tion glass. To avoid contamination of the latter, or to reduce it to an 

 infinitesimal amount, the following practice was adopted. A consider- 

 able number of Stentors were taken from their usual location upon one 

 side of the culture jar with a medicine dropper and placed in a watch, 

 glass (No. 1). Soon these would gather together in one part of the 

 vessel on account of the presence there of zoogloea, the particular direc- 

 tion of the light, etc. With as small a quantity of liquid as possible, 

 they were then transferred to watch-glass No. 2, which contained a con- 

 siderable quantity of the medium to be used for the given experiment. 

 In the same way, except that pipettes of smaller bore and a dissecting 

 microscope were now used, they were repeatedly transferred to successive 

 " ponds " of the required medium. From the last watch-glass " pond " 

 a few at a time were transferred with a capillary pipette to the test watch- 

 glass containing the pure medium. This I cull the method of multiple 

 transference. 



How many such transfers are necessary ? Fortunately there is a 

 good test to determine this point. When the electrical conductivity of 

 the medium — using electrodes that can be dipped into the liquid of a 

 watch-glass and then removed — shows no immediate change upon the 

 introduction of Stentors, then there is no contamination. In the case 

 of multiple transference from culture liquid to distilled water, it was 



