MICROSCOPIC ESTIMATE OF BACTERIA. 151 



and the method of making: the determination. Wrong con- 

 clusions will be reached, if emphasis is laid upon mere num- 

 bers without a consideration of the other factors. 



A numerical determination taken by itself without full 

 consideration of all the factors involved has no value. Per- 

 sons not familiar with the circumstances are apt to attempt 

 to make comparisons when such are inadmissible. The results 

 of the work of two laboratories cannot be compared, unless 

 there has been uniformity in every detail that would affect 

 results. Every bacteriologist doing much milk work has had 

 embarrassing experiences along this line. The man who has 

 had the longest experience with numerical determinations of 

 bacteria in milk in connection with milk commission work 

 believes that no good end is served by stating results publicly 

 in numerical terms, and acts in accordance therewith. In any 

 event, the publication of numerical results should be restricted 

 to a small circle of persons w r ho appreciate their significance. 



The bacterial count is undoubtedly the instrument by which 

 the effectiveness of good dairy methods is measured, and has 

 been a prominent factor in the development of those methods. 

 It reveals facts regarding the operation of high grade dairies 

 that no inspection could disclose. A count of certified milk 

 is better evidence of the observance of certain features of 

 cleanliness in handling the milk than a visit to the dairy. 



The value of the bacterial count in locating contamination 

 is recognized by careful dairymen. Some large distributing 

 firms use the method to good advantage in controlling the 

 producing dairies. Postal card reports sent from time to time 

 convince the dairyman that his product is under scrutiny and 

 stimulate him to greater efforts. 



Microscopic estimate of bacteria. Slack has observed that, in 

 connection with his method for the study of milk sediments, 

 the number of bacteria may also be roughly estimated. By a 

 long series of comparative tests, it has been found possible to 

 tell from the number of bacteria in the smear whether or not 

 plate cultures would show above or below 500,000 colonies per 

 cc. An error of less than \ ( /< was made in this regard. Slack 

 uses the test to eliminate those samples of milk that are well 



