518 GROWTH 



give the volume of some species fairly accurately. This is true for only 

 a few solids, such as the cube. The method has an error of about 33 

 percent when used with spherical organisms. Chalkley (1929) measured 

 the volume of Amoeba by gently drawing it into a capillary tube of 

 known diameter and calculating the volume from the length of tube 

 filled plus the two hemispherical ends. 



Populations of Protozoa are usually measured by counting a sample 

 of the population in a Sedgewick-Rafter cell with a Whipple disc in the 

 eyepiece of the microscope, or with a hemocytometer (cf. Woodruff, 

 1912; Hall et al., 1935). The chief source of error of this method de- 

 pends on how closely the sample represents the population. Care must 

 be used that none of the animals are lost by sticking to the transfer 

 pipette and to make sure that all are counted once only. Berkson et al. 

 (1935) has given a quantitative treatment of the errors of counting 

 red blood cells with a hemocytometer, and their evaluation might be 

 applied to estimates of protozoan populations. 



Tippett (1932) has suggested that the mean number may be esti- 

 mated by counting the squares containing 0, 1, 2, and so forth animals 

 and using the tables prepared for the Poisson distribution. With Proto- 

 zoa, greater precision may be obtained by killing the animals before 

 making the count. Many killing fluids are hypertonic, and animals may 

 be lost from the osmotic efl^ects of the killing fluid. Jennings (1908) 

 and others have found that Worcester's fluid causes little change with 

 paramecia when a sufficient amount is used to overwhelm the animals. 

 Hardy's (1938) method of estimating numbers by comparing with 

 standards containing a known number of dots might be used when high 

 precision is not required. 



Protozoa may be centrifuged into a tube with a calibrated capillary 

 bottom, similar to that used by Carlson (1913) for yeast. Elliott (1939) 

 obtains greater precision and convenience by fusing a hematocrit tube 

 to a 10 ml. centrifuge tube. When the animals are killed before centrifug- 

 ing, it is necessary that the volume of the animals not be changed by an 

 anisotonic killing fluid. Commercially made tubes should be carefully 

 calibrated, as errors as great as 12 percent have been reported for some 

 makes of Hopkins vaccine tubes. Solid packing may not be possible with 

 the usual laboratory centrifuges, but for given conditions constant pack- 

 ing may be obtained in equal time intervals. If the values are to be used 



