362 DEDUCTIONS FROM ANALYSIS. 



will give results very nearly correct; this formula applies a 

 correction automatically for the dilution with excess of fat above 

 3, or for concentration of the solids not fat due to a deficiency 

 below 3. The figure 96 may be used if 4 per cent, is taken as a 

 reasonable upper limit. 



Another formula in which the percentage of water is calculated 

 from the aldehyde figure (A) is 



Water = 100 - ~ X 100. 



As these formulae fail with abnormal milks, the author has 

 proposed the two following formulae, which hold good, not only 

 with normal milks, but with abnormal milks in addition : 



100 

 (S L) IQQ _ y exceeds 4. 



G -j- F 4L exceeds 16. 

 S = solids not fat, L = milk-sugar, F = fat, and G = lactometer degrees. 



The formulae depending on solids not fat do not actually 

 give the amount of added water, but really the dilution ; 

 should the fat be greatly higher than that found in normal 

 milk a portion of the dilution will be due to fat; in 



100 



such a case it is advisable to multiply the fat by ^-^ ^ T 



1(X) W 



(W being the dilution), and if the figure is greatly above normal 



100 

 to subtract F x 100 _ w 4 from the dilution, and call the 



remainder added water. This involves the assumption that 

 4 per cent, is a reasonable upper limit for fat in milk. 



Calculation of Fat Abstracted. The detection of adultera- 

 tion by removal of cream can be effected only with certainty by 

 the estimation of fat ; if this falls below 3-0 per cent., a per- 

 sumption is raised that cream has been abstracted. From the 

 table on p. 306 it is seen that the mean percentage of fat varies 

 at different times of the year ; a limit of 3-25 per cent, could be 

 used from October to January with as much justification as a 

 limit of 3-0 for the other months (cf. p. 356). 



The percentage of cream abstracted is calculated by the 

 formula 



F 

 Cream abstracted = 100 y X 100, 



where F = percentage of fat. 



This formula gives a minimum percentage of fat abstracted. 

 The figure thus calculated is almost always seriously below the 

 truth ; the probable amount can be calculated by substituting 

 3-75 for 3, or, better still, the monthly average figure given in 



