METOHDS OF EXAMINING MILK 253 
Rothenfusser it reached 41.5. The addition of a small 
quantity of water to milk lowers the refraction number, 
but on account of the wide range in the refraction number 
of normal market milk it is necessary, in order to detect 
slight additions of water, to have the herd under suspicion 
milked under supervision and to take a sample of the 
milk and compare the refraction number of this sample 
with the refraction number of the sample of market milk. 
The daily variation in the refraction number of the milk 
of a herd will not exceed 0.1 to 0.55. Changes of feed 
have only a slight influence and incomplete milking has 
no effect. When the refraction number is as low as 36.5 
or below the milk may be unhesitatingly pronounced 
watered. 
The refraction number is not affected by the per 
cent. of fat. A sample of milk taken from the top of a 
can and one taken from the bottom of the same can will 
both have exactly the same refraction number, although 
there will be a great difference in the fat per cent. This is 
an important advantage, because when a charge that milk 
has been watered is based on the specific gravity and fat 
per cent. the claim is often made that the sample was 
taken from the bottom of the can or after the top milk 
had been sold. 
The presence of boric acid and borax in the propor- 
tion usually used to preserve milk (0.1 to 0.2 per cent.) 
will increase the refraction of the milk serum. Salicylic 
acid also increases the refraction when present in as 
small amount as 0.05 per cent. which is the quantity used 
as a preservative, but formaldehyde causes no increase 
when present in the amount ordinarily used as a preserva- 
tive. Sodium bicarbonate does not have any effect when 
added up to 0.1 per cent. 
