456 STATE BOARD OF AGRICULTURE. 



high as the average, and are therefoi*e not considered. An error must 

 have crept in somewhere, as the comparison of all data shows; prob- 

 ably the final number of cells is recorded one decimal place too low. 

 Excluding these three data, the variations are, 



The variation is quite large with Strain 2, the maximum is 5 times 

 as large as the minimum. The averages show that Strain 2 was quite 

 different from the other thi'ee lactic types. 



Another experiment showing a very wide variation in parallel experi- 

 ments is that given in Table IV. This experiment was made with a 

 dairy startei', carried on in pasteurized milk, and apparently a pure 

 culture, as far as could be judged from the gelatin plates. Dilutions 

 in decimal order were transferred into flasks with 100 cc. of sterile 

 milk. The acidities and numbers of bacteria were determined after 

 12 and 24 hours' growth at 21°. The highest fermenting capacity is 

 more than three times as great as the lowest. 



Smaller variations were obtained with Sti*ain II which has been used 

 in the majority of the following experiments. This strain has been 

 transplanted in the laboratory for about two jears at least everv^ second 

 day from milk to milk. It is a rapid acid producer and is used as 

 starter in several creameries. In Table V are compiled all data obtained 

 with 24-hour cultures of this strain. Some of the experiments have 

 been carried on more than a year apart, and the only identical factors 

 are the well-bred culture and the temperature of about 21-24°C. The 

 very last experiment shows both the highest and the lowest fermenting 

 capacity, 13.7 and 33.8X10-'° mg. The next extremes are 16.6 and 30.5. The 

 proportion of lowest to highest number is 1 :2.5 and 1 :1.6 respectively, 

 while in the experiments of Marshall and Farrand, with freshly isolated 

 cultures, the corresponding proportions of the four strains were 1:3.5, 

 1 :4.8, 1 :3.9 and 1 :1.9. Table IV with the dairy starter gives the pro- 

 portion of the two extremes 1:3.3. This shows that the greatest varia- 

 tion of the well-bred Strain II is smaller than that of freshly isolated 

 pure cultures. 



