FOR NEW HAMPSHIRE FARMERS 



155 



of cream is lighter iu weight than the same vol- 

 ume of milk, also because the uecessary mixing 

 of the cream is apt to cause a mixture of air with 

 the liquid, increasing its bulk, and because more 

 or less cream is sure to adhere to the pipette, 

 rendering any form of measurement inaccurate. 

 The figures in the following table give the aver- 

 aoe result of a larse number of tests and show 

 the approximate discrepancy when cream is sam- 

 pled, by the use of an 18 cc. pipette, and illus- 

 trates the injustice resulting from its use : 



Sampled with pipette (18 cc). 



Sample No. 1 22.00 per cent. fat. 



Sample No. 2 24.00 per cent. fat. 



Sample No. 3 29.50 per cent. fat. 



Sample No. 4 31.00 per cent. fat. 



Sample No. 5 38.50 per cent. fat. 



Sampled by weight 18 grams. 



Sample No. 1 22.00 per cent. fat. 



Sample No. 2 24.25 per cent. fat. 



Sample No. 3 30.35 per cent. fat. 



Sample No. 4 32.00 per cent. fat. 



Sample No. 5 40.00 per cent. fat. 



- It 



=r- «» 





Fig. 15. 



It will be seen that in sample No. 5 there is a difference of 

 one and one half per cent, between the pipette and the scale, 

 or one and one half pounds of butter fat for every hundred 

 pounds of cream. Dealers in cream should use the scale in 

 sampling cream for testing, and farmers selling cream should 

 insist on this method. If necessary a state law should compel 

 its use. 



SKIM MILK — BUTTER MILK 



Skim milk and butter 

 milk are sampled and test- 

 ed same as whole milk. 

 Samples of skim milk 

 should not be taken direct 

 from the separator as it 

 may^ then contain small 

 air bubbles. A special 

 test tube (Fig. 17) has 

 Yig, 16. been devised for this work. 



