488 EXPERIMENT STATION RECORD. 



average daily yield of milk by the Holstein in the second period, when she was 

 insufficiently nourished, was 20.30 lbs., being about 10 lbs. less than in the first and 

 nearly 7 lbs. less than in the third period, when she was abundantly fed. 



"It is therefore manifest that the ration of straw, oats, and turnips diminished 

 considerably the yield of milk, and that the ration of hay, oats, peas, bran, oil cake, 

 and turnips considerably increased it. 



■• In the fourth aud fifth periods each cow yielded practically the same quantities 

 of milk; and the addition of peas and oil cake to the Ayrshire's ration of grass did 

 not increase her yield over that of the Holstein. . . . 



"The variations in the percentages of the constituents of the same cow's milk 

 during the several periods are so little that they come in nearly every instance 

 within the limit of experimental error. The food, therefore, has had, apparently, 

 no effect on the composition of the milk." 



Bacterial contamination of milk, F. C. Harrison (Ontario Agr. 

 Col. and Expt. Farm Bpt. 1896, pp. 105-114, pis, 4). — Studies were made 

 on the germ content of milk, and contamination from the foremilk, from 

 the animal and milker, from the air, and from the use of improperly- 

 cleaned dairy utensils. In each instance the number of germs from a 

 cubic centimeter of the milk, the number of species, and a description of 

 the species is given. The germ content of the milk of the college dairy 

 ranged from 8,000 to 18,000 per cubic centimeter, while that of the milk 

 sold in Guelph was found to range from 121,000 to 1,200,000. 



" Notice is especially directed to the great difference between the number of genus 

 in the mixed milk of a herd kept under cleanly conditions, immediately after milk- 

 ing, and the milk supplied to consumers in the city of Guelph." 



The germ content of the foremilk, or first part of the milking, was 

 compared with that of the rest of the milking. The first few streams 

 from each teat were drawn into a sterilized flask. This contained from 

 18,000 to 54,000 germs per cubic centimeter of milk, while the balance 

 of the milk contained 890 to 4,800 germs. 



"The results clearly demonstrate the fact that by milking the first few streams 

 from each teat onto the ground, or into a separate pail, contamination from the fore- 

 milk is in a large measure prevented.' - 



Cows were milked under clean, rather dirty, aud dirty conditions, and 

 also after the flanks and udder of the cow had been moistened with a 

 wet cloth. In each case the number of germs falling into a milk pail 

 12 in. in diameter during 1 minute was determined, and culture dishes 

 were exposed beside the milk pail for 1 minute. The results are given, 

 together with plates showing the cultures in the dishes. Under ordi- 

 nary conditions there were about 8,000 to 17,000 germs, while when the 

 Hanks and udder of the cow were moistened with a wet cloth before 

 milking from 640 to 2,300 germs fell into the pail. 



Determinations were made by means of culture dishes of the germ 

 content of stable air during bedding, feeding hay, cleaning up, etc., and 

 1 hour later. In the former case the number of germs which would fall 

 into a 12 in. milk pail in 1 minute ranged from 12,000 to 42,000, while 1 

 hour thereafter the number ranged from 483 to 2,300. 



