PROD UCTION OF MILK FREE FROM CONTA MINA TION 277 



for purposes of general sale. In the case of the experiments with 

 the Ulax filter, he obtained the follow ng figures : 



Commenting on the results, Savage says : ' . . . On the whole 

 the experiments showed no bacterial advantage from nitration 



Coolerj 



of the milk, and this although one of the supposed best strainers 

 was employed, the apparatus steamed before each experiment, 

 and the wool used for the filtration of only half a gallon of milk/ 



In the case of the other filter, he says : ' The strainer in this 

 experiment consisted of wire gauze and a layer of loose flannel- 

 like material. It rapidly strained the milk. It was fixed in a 

 shed near the cow-sheds, and the farmer had unbounded faith in 

 its filtration properties, rather to the neglect of ordinary clean- 

 liness in milking.' In these samples the number of bacteria in 

 the unstrained sample was 5980 and in the strained sample 43,300. 

 Savage says : ' In this experiment the straining added a large 

 number of bacteria to the milk, no doubt partly from the strainer 

 being dirty, causing a breaking down of the manure or other 

 particles, and a better distribution of the bacteria.' 



