WATERING OF MILK.

Otago Daily Times, Issue 21165, 24 October 1930, Page 4

WATERING OF MILK.

DIFFICULTY OF DETECTION. (From Oi/r Own Correspondent.) . LONDON, September 10. Dr J. F, Tocher, of Aberdeen, in discussing the adulteration of milk by water at the British Association, said that the present law was unjust to the seller and had created an impression that the public analyst could always tell when milk had been watered. Even at .the present day it was not possible for an analyst to state, as the result of analysis, the percentage of water added in an adulterated sample. Nevertheless, the regulations imposed, and still impose, upon the seller of milk the necessity of proving his innocence in court when a sample of milk, sold by him, fell below 3 per cent, butter-fat or below 8.5 per cent, solids-not-fat. The seller of such milk came into court a guilty man, because it was presumed, until the contrary was proved, that the seller had been watering hie milk. It was, however, now an established fact that proportions ranging up to 14 per cent, of all genuine samples fell below 3 per cent, in butter-fat and proportions ranging from 22 per cent, to 30 per cent, of all genuine samples fell below 8.5 per cent, in solida-not-fat. A MATHEMATICAL PEOBLEM.

After discussing the advantages of a legal standard, the difficulties in connection with bulking and various tests of watering, including the freezing point method, he explained the basis on which he proposed to elaborate a criterion for watering. The problem was not analytical but mathematical in character. He explained that he found the proportions of the various constituents of milk to be quite different for low values of total solids, when compared with the proportions of these same constituents for high values of total solids in genuine milk. For low values of total solids, milk sugar contributed largely to any increase, while, for high values, casein contributed more than any other constituent. The first stage in the process was to obtain an equation which would determine with high probability the fact of watering, while the second stage was the development of a function, based on the actual variation of the constituents, the values of which would indicate the various degrees of watering. Unfortunately the second stage was a long one, involving more than 18 months’ calculations in order to find the values of 15 constants which would have to be used by analysts when they desired to calculate'the percentage of watering. He gave examples of preliminary tests of watering and showed that the crterion was likely to be of a practical character, giving values reasonably near the actual amount of watering in known mixtures of milk and water.