PERCENT CASEIN. 
Official. Van Slyke-Bosworth. | - Hart. Formol Titration. 
2.98 3.05 2.95 | 2.99 
2.96 3.05 2.90 2.98 
2.45 2.45 2.40 2,00 
2.40 2.40 PaaS fs) 2.48 
1.79 (d) 1.80 1.80 1.85 
ETc (oh) iP rAs | 1.85 1.83 
3.28 | Shs) | 3.18 3.18 
3.29 3.20 ! 3.15 3.20 
2.46 2.49 2.40 | 2.46 
PTL 3.80 3.65 | 3.70 
2.90 2.90 2.80 2.96 
2.47 | 2.50 2.45 2.48 
3.71 Sieh) 3. nO 3.74 
2.85 | 2.85 2 ee | 3.01 
2.80 2.74 2.00 2276 
2.89 2.85 | 2.90 2.91 
Note.—The two samples marked (d) were diluted milk. 
Samples were taken on different days from the same source. 
The above table shows the relative accuracy of the different methods. 
For the estimation of casein in milk the choice of the methods mentioned 
depends on the purpose for which the analysis is made. If total proteids 
are to be estimated, the Van Slyke-Bosworth and Hart methods must be 
excluded, unless an assumption is made as to the average amount of albumen 
in milk. This could be done on the same basis as that for the formol method 
and which would introduce only a slight error for normal milk and from a 
mixed herd. 
In reviewing these methods and considering speed, and ease of carrying 
out the work, the formol titration method is to be preferred. In all three 
volumetric methods it is very essential that the water used for dilution 
should be free from carbon dioxide. Very little distilled water found in 
laboratories is free from carbon dioxide. This factor alone may introduce 
errors to vitiate the results. Titration after the addition of the formaldehyde 
should be carried to a sharp pink color and remain so for at least five minutes. 
GEORGE SPITZER. 
Purdue University. 
