New York Agricultural Experiment Station. 109 



tains very few bacteria and the precise count makes little difference. 

 If there are many bacteria present in the smear it is often impossible 

 to count the whole field of the microscope, owing to confusion caused 

 by lack of guide lines. An eye-piece micrometer, ruled in squares, 

 overcomes this difficulty. Here, too, it is not necessary to count a 

 large number of fields, for it is soon seen that the milk is of poor 

 quality. 



Uneven distribution. — A more real source of error than the agree- 

 ment or lack of agreement between duplicate counts is the uneven 

 distribution of bacteria in the smear. It is evident that for accu- 



Benton Hand Tally Register. 



rate quantitative work several fields must be counted to overcome 

 this error. This is a laborious operation if large numbers of bac- 

 teria are present. Fortunately practical experience has shown that 

 accurate counts are not necessary in order to form a satisfactory 

 idea of the amount of bacterial growth which has taken place in a 

 given sample of milk. Counting is much more easily done if a hand 

 tally register is used. This can be procured from hardware dealers. 

 Counting of the whole field. — Another error which becomes 

 important in some cases lies in attempting to count the whole 

 microscopic field. The margin of the microscopic field is invariably 

 so hazy as to obscure the bacteria lying in this region. This error 

 is not serious, however, where few organisms are present, but becomes 



