520 GROWTH 



filter, an alundum crucible, or an asbestos mat; washing them rapidly 

 to remove the culture fluid, but not to burst the cells, and drying them 

 to a constant weight. A vacuum desiccator containing sulphuric acid or 

 phosphorous pentoxide at room temperature may give better results 

 than a drying oven. This is a difficult method to control, so as to get 

 consistent results. 



Bacteriological methods of diluting and plating are not often ap- 

 plicable to Protozoa, but may be useful for testing the culture medium 

 to make certain that it is bacteria-free. Standard texts should be con- 

 sulted for methods. The ordinary nutrient agar is not a certain medium 

 for estimating the bacteria found in water, and special media must be 

 used. The number of colonies on an incubated plate may be less than 

 the number of bacteria unless care has been used to prevent clumping of 

 the bacteria. The errors of plate counts have been evaluated by Mattick 

 et d. (1935) and by Ziegler and Halvorson (1935). Gordon (1938) 

 has questioned these probability tables. 



Other methods which might be useful to protozoologists are the 

 measurement of the suspension in terms of viscosity (Shapiro, 1937) 

 and the determination of the velocity of sedimentation (Nielsen, 1933) • 

 These would be used with killed or nonmotile animals. 



The Growth of Individual Protozoa 



Simpson (1902) measured the length and breadth of Paramecium 

 caudatum with an ocular micrometer at a few and at many hours after 

 fission. Jennings (1908) supplied the first detailed measurements, and 

 the data from his summary table are plotted in Figures 130 and 131. 

 At division the animal decreases in breadth and increases in length. 

 After the separation the increase in both dimensions is increasingly rapid, 

 then proceeds at a nearly constant relative rate, and finally slows until 

 the cycle is repeated. 



The graphs of the growth are made on arithlog, or semilogarithmic 

 paper, to facilitate analysis. This equivalent to plotting the logarithm 

 of the size against a linear time axis. The slope of the growth curve at 

 any point is the relative rate of growth [dy/ydt). Two growth curves, 

 parallel to each other, are changing at the same relative rates. When no 

 change in form occurs, the curves for area and volume will be corre- 

 spondingly above and have slopes two and three times as great as a 

 linear dimension. 



