Tests op Cream Separators. 



703 



trials at factories, thus placing each machine tested on an equality 

 and giving equal weight to each series of tests whether it was com- 

 posed of few or many trials. In making the average in this way a 

 single trial of one machine has as much value in determining the 

 average as fifty trials of another machine although the latter prob- 

 ably indicates more accurately the true efficiency of the machine. 



The other method is to average the individual tests having no 

 regard to the number of machines used. In this case the influence 

 of each machine upon determining the average is in proportion to 

 the number of times it was used. If then a poor machine is tested 

 many times and a good one but once or twice, or vice versa, the 

 result may be misleading. In Table XII the average computed in 

 both ways is given together with the maximum and minimum 

 amounts of fat found in the skimmed milk in each group of 

 machines in any single trial. This brings out more forcibly what 

 has been said, that some machines of each style of manufacture do 



efficient work. 



TABLE Xn. 



KIND OF MACHINE. 



Accumulator 



Alexandra Jumbo 



Columbia 



Danish Weston . 



DeLaval 



Sharpies 



United States ... 

 Victoria 



Maximum. 



.20 

 .33 

 .U 

 .25 

 .50 

 .65 

 .60 

 .38 



SUMMARY. 



The results of all these trials show that it is possible to separate 

 the cream from milk with a loss of not more than one-tenth of one 

 per cent, of fat in the skimmed milk. 



That in all probability there is nearly as much difference in effi- 

 ciency of separation between different machines of the same make 

 as there is between the different makes themselves. 



HENRY H. WING. 



