MICKO-ORGANISMS IN MILK 309 



slightly less in machine-drawn milk. The parts were kept in 

 lime-water during the intervals between milkings. 



Harding, Wilson, and Smith made extensive studies of the 

 germ content of machine-drawn and hand-drawn milk. The 

 teat cups remained in 10 per cent, salt solution after the machine 

 had been carefully cleaned. Comparative counts of milk obtained 

 after the teat cup had been kept in brine with those of milk 

 not so treated gave the following results: Average of 11 samples 

 after immersion of the cups in brine, 17,086 bacteria per cubic 

 centimeter, and 9 samples with the cups not immersed in brine 

 gave an average of 188,580 bacteria per cubic centimeter. 



The authors call attention to the importance of removing the 

 air from the tubes of the machine while they are immersed in brine. 

 Tests showed that when air was present the average of 11 samples 

 was 4740 bacteria per cubic centimeter, while without air in the 

 tubes the average of 12 samples was 1990 bacteria per cubic 

 centimeter. 



These authors confirmed Stocking and Mason's work on the 

 efficiency of relief filters, and determined that absorbent cotton 

 is more efficient than common cotton. An improvement in the 

 construction of the teat cups made possible a further reduction 

 of germ content by allowing a larger filter to be inserted. Milk 

 drawn by a machine of the old type contained 8340 bacteria per 

 cubic centimeter in an average of 24 samples, while milk drawn 

 with the new machine had 3210 bacteria per cubic centimeter in 

 an average of 24 samples. 



Changing of teat cups before milking increased the germ con- 

 tent from 4150 to 6410 bacteria per cubic centimeter. The im- 

 portance of proper handling of the machine is further emphasized 

 by tests made of milk that had been milked in a pail the cover 

 of which had been removed and replaced before milking. The 

 bacterial content was 5850 7 against 4220 in a pail the cover from 

 which had not been removed. 



The influence of machine milking on milk secretion has been 

 studied by Lane, Woll and Humphrey, Beach, Price, Alexander, 

 Smith and Harding, and others. The difference in quantity pro- 

 duced by machine and hand milking, if there is any, is too small 

 to be measured. 



The investigations recorded show that milking by machine is 

 not only feasible, but has some advantages. The expense of run- 

 ning a machine is small, amounting to 4 cents per hour, 6 cents per 

 milking, and $3.60 per month (Woll and Humphrey). The ex- 

 pense for repairs is relatively small and the saving in help con- 

 siderable. 



In order to obtain favorable results, however, intelligent care 



