842 THE VARIABILITV OF COWs' MILK, i., 



The ranges of variation of these mean values of tlie composition 

 and properties of the milk of groups of cows are, of course, less 

 than those of the individual samples. The extents to which the 

 individual variations are masked by this process of averaging are 

 widely different in the different quantities measured. The relative 

 values of this masking-effect are shown by the figures in the column 

 headed Range 2, Kange 1. These figures show that, in the case of 

 the values of the weight of milk obtained at a milking, the mean 

 results have a range of variation 041 of the range of the indi- 

 vidual results. The masking of the individual variation, produced 

 by taking the mean figures, is less in the present case than in the 

 case of the values of any of the other cjuantities. 



The mean values of the percentages of solids not fat, on the 

 other hand, show a range of variation which is only Oil of that 

 of the individual results. The process of taking the mean values 

 of the percentages of these substances has, therefore, a very pro- 

 nounced effect of obscuring the range of variation which actually 

 occurs, even in the case of small groups of samples, like the 

 present. 



The difference between the mean composition and the com- 

 position of the mixed milk is more pronounced in the case of these 

 smaller numbers of samples than it was in the case of the larger 

 number. The difference does not in any case exceed 5% of the 

 value in question, however. 



These results raise the question, whether the fact, that the 

 samples of milk examined for commercial or legal jDurposes are 

 almost invariably mixed samples, may not have given rise to an 

 exaggerated notion of the constancy of the percentage of solids 

 not fat. The existence of such a notion would, perhaps, lead to 

 the establishment of a standard for the percentages of these sub- 

 stances, which would be too close to the mean value to make allow- 

 ance for the variation met with in individual samples of milk. 



The standard for the percentage of solids not fat is 8-5 in all 

 the States of Australia, and in England. It is mteresting to 

 observe that, even among the above mean and "mixed" values of 

 the percentage of solids not fat, one is at the standard, and one is 

 below it. Of the individual values, nearly one-half are at or below 



