280 EXPERIMENT STATION RECORD. 



greater the speed of the machine the higher the percentage of fat in the cream, 

 and this effect was much more marked when the cream screw was set for 

 thiclj cream. Within ordinary limits the greater tlie speed the Jess the amount 

 of fat remaining in the sidm millv. This did not vary to any marlced extent 

 whether the separator was set for thin or thiclc cream. The rate of inflow was 

 uniformly Increased by the speed. The explanation given for this is not that 

 the greater the speed of the separator the more milk will run through it in a 

 given time, but that it is due to suction caused by the flow of air through the 

 machine. The capacity of the separators did not vary to any extent whether 

 set for thick or thin cream. 



In ascertaining the effect of the temperature of the milk tests were made at 

 70°, 80°, and 90° F., respectively. Contrary to the general assumption, the re- 

 sults indicate that the higher the temperature of the milk separated the thinner 

 the cream. This was found to be more marked when the separator was set 

 for thick cream than for thin cream. The results show that the colder the 

 milk separated the greater was the percentage of fat in the skim milk. The 

 capacity of the machines did not vary materially with milk of high or low 

 temperature. 



In testing for the effect of the richness of the milk separated, one lot was 

 separated under normal conditions, and the skim milk obtained was used in 

 reducing the second lot to about 3 per cent fat. The only marked variation 

 was the fjercentage of fat in the cream, which was found to vary practically 

 in direct proportion to the pex'centage of fat in the milk separated. Within 

 ordinary limits the richness of the milk had no appreciable effect upon the loss 

 of fat in the skim milk or upon the capacity of the machine. 



In the experiments on the effect of the rate of inflow, three runs were made 

 with each lot of milk: First, with the feed can full of milk; second, half fuh; 

 and third, with the milk 1 inch from the bottom. The rate of inflow was 

 changed in other trials by the faucet. In general a decrease in the inflow 

 resulted in (1) less cream in proportion to skim milk, (2) higher percentage of 

 fat in the cream, (3) no appreciable effect as to the loss of fat in the skim milk. 



Acidity had but little effect except that with sour milk there was a tendency 

 for the percentage of fat in the cream to be increased, especially when the 

 cream screw was set for thick cream. 



The effect of dirt in the tubes and of clogging of the bowl is also discussed. 



Milk powder starters in creameries, C. Larsen and W. White {SoutJi, 

 Dakota Sta. Bui. 123, pp. S-lJf). — In a preliminary investigation on the use of 

 milk powder solutions in the place of natural milk for starter making it was 

 found that the best degree of concentration to supply conditions for bacterial 

 growth was in the proixtrtion of 3 ounces of ix)wder to 1 quart of water. 

 The comparative development of acidity and bacteria in mother starters, pre- 

 pared from both milk powder and natural skim milk, is presented in tabular 

 form. The growth of both bacteria and the development of acid and flavor 

 was normal. Stained preparations of the bacteria fi'om milk powder solutions 

 appeared the same under the microscope as those grown in natural milk. 



A comparison of the specific gravities, amount of acid developed, and num- 

 ber of bacteria per cubic centimeter showed that the solution of 3 ounces of 

 powder in 1 quart of water furnishes the bacteria conditions resembling very 

 closely those formed in natural milk. Butter made from the milk powder starters 

 scored equally as high as when made of natural milk starters, both when fresh 

 and when 2, 4, and 6 weeks old. The scorched flavor in the milk powder was 

 not detected in any of the samples of butter made. The cost of the powder 

 starters is somewhat higher than that from those made of milk, it being esti- 

 mated at 0.3 to 1 cent per pound of butter made, yet they are recommended for 



