112 ItEPORT OF THE DEPARTMENT OF BACTERIOLOGY OF THE 



It should also be stated that several samples which showed only 

 a few bacteria according to the Breed method showed no bacteria 

 at all by the Rosam method. The chief difficulty with this technique 

 is clue to evaporation at the edge of the cover glass, which causes 

 diffusion currents. These currents oftentimes move so rapidly as 

 to sweep the bacteria completely out of the microscopic field while 

 they are being counted. The bacteria in many preparations could 

 not be counted for this reason. Another difficulty came from 

 heating the mixed milk and stain to steaming. This caused the 

 formation of a film on the surface of the milk, which had to be 

 removed. It was impossible to determine how many bacteria were 

 removed at the same time. Another objection to the method 

 which cannot be overloookcd is that there is a wide variation in 

 the weight of the loopfuls taken. The technique calls for a platinum 

 loop of such a size that it will transfer an average of 0.004 gram of 

 milk and stain mixture. Even for rough work the percentage 

 variation between loopfuls ought not to be great, but when series 

 of weighings were made a very wide variation was found even when 

 all possible precautions were taken to make the weights uniform. 

 The following were the weights secured from different loopfuls of 

 ■milk measured by the same loop. 



First Trial — Milk Stained with 

 Loeffler's Methylene Blue. 



0.0070 grams. 



0.0044 



0.0052 



0.0064 



0.0068 



0.0043 



0.0055 



0.0053 



0.0061 



0.0053 



0.0065 



0.0052 



Second Trial — Milk Stained with 

 Methylene Blue and Pyridin. 



0.0026 grams. 



0.0032 " 



0.0024 " 



0.0046 " 



0.0050 " 



0.0045 " 



0.0023 " 



0.0025 " 



0.0029 " 



0.0027 " 



0.0039 " 



It is difficult to say how much these variations in weight were due 

 to evaporation which occurred during the* process of weighing. The 

 rate of evaporation is obviously high in thin films of warm milk. 

 An attempt was made to keep the interval of time of each weighing 

 the same, and as small as possible, in order to keep the error from 

 this source constant. In spite of this there is a percentage variation 



