BY J. BROWNLTE HENDERSON, F.I.C., AND L. A. MESTON. 169 



selling as watered milk, genuine milk which simply failed 

 to reach the legal standard. On the freezing point test, 

 we have been able to warn three vendors in the last three 

 years that the milk they were vending was genuine but 

 below the legal standard, and that they had better take 

 •steps to improve their herds by feeding or *' culling '^ in 

 order to meet the standard. 



It has been pointed out by one or two critics of the 

 freezing j)oint method that the results can easily be masked 

 by the addition of substances \^'hich depress the freezing 

 point. 



Any of the substances likely to be used in this direction 

 <}an be readily detected by an analyst. Many tests at the 

 disposal of the analyst are liable to be masked, and it is 

 part of his duty to look out for this masking, as for example 

 the masking of the heat test of explosives by the addition 

 of mercuric chloride, the masking of Becchi & Halphen 

 reactions for cotton seed oil by boiling the oil and other 

 treatment, and the masking of Hehner's reaction for 

 formaldehyde by adding sodium nitrite. 



The freezing point is such a sensitive test however 

 that if the ordinary dairyman did start tampering with 

 the milk he would either add too little to cover the addition 

 of water or add too much and make the freezing point 

 abnormal. 



For the last four years in the Government Chemical 

 Laboratory, Brisbane, every legal sample of milk for prose- 

 cution purposes has been checked by this method, and most 

 of the results are shown in Tables A. B., and C. 



Table A. shows results obtained on legal samples of 

 milk taken during the last three years, wherein an increased 

 acidity had developed. When the milk becomes acid the 

 iarger molecules are decomposed into a number of smaller 

 ones. The osmotic pressure increases and the freezing 

 point is further depressed. Without this knowledge the 

 results in Table A. would, make it appear that some of 

 the samples were abnormal before watering, while others 

 were only slightly above the legal standard of 8-5 S.N.F. 

 The added water, however (calculated on the minimum 



