ON TECHNIQUE 51 



of cells in a number of representative fields on the slide, and comput- 

 ing the area of each field, and thus obtaining the number in the 

 total preparation. This method presents certain difficulties. The 

 films do not dry uniformly; that is, they are much thinner at the 

 edge than at the center, and the cells are correspondingly more 

 numerous in the center than at the periphery. It is necessary, there- 

 fore, to count a number of fields chosen at regular intervals along 

 a line extending from one edge through the center to the opposite 

 edge of the slide. In general, twenty fields were chosen, ten along 

 each of two lines at right angles to each other, the spacing of the 

 fields being easily accomplished by means of a mechanical stage. 

 There is also some difficulty in accurately discharging .01 c.c. pipettes 

 onto a slide, and leaving behind on the tip of the pipette the same 

 quantity each time. It is therefore very desirable to prepare a num- 

 ber of slides, in most of the work here reported — five. The total 

 number of cells counted in most of the cases, therefore, was the 

 number contained in one hundred microscopic fields; that is twenty 

 fields from each of five different preparations. This is tedious but 

 necessary to obtain uniform results. Where the number of cells is 

 small, as during the early stages of growth, frequently many more 

 fields had to be counted to obtain a count of accuracy comparable 

 to that during the later stages of growth. 



Since practically all of the cell counts were made by direct micro- 

 scopic methods, death of the cells is not recorded in the curves 

 until the dead cells have completely autolyzed. Wilson has pointed 

 out that some cells are dying at practically all stages of growth, 

 but during the growth cycle proper the number of these is small, 

 and the error is practically negligible. In the death phase, however, 

 the microscopic counts do not give a true measure of the rate of 

 death, since of course the curve does not begin to decline until some 

 time after the cells have actually begun to die; and further, as will 

 be shown in Chapter IX, because the rate of autolysis is not depen- 

 dent upon the death rate. The death phase therefore began much 

 earlier and the rate was probably much higher than is indicated in 

 the curves given, particularly with the cholera vibrio as described 

 in Chapter VIII. No particular attention has been paid to the death 

 phase, however, save in studies of the involution forms of the colon 



