8 
BULLETIN" 944, U. S. DEPARTMENT OF AGRICULTURE. 
proportion of this milk is not too high no trouble will be experienced, 
but the proportion in some batches may be such as to cause consider- 
able trouble. 
It is interesting to know that raw milk with a low titratable 
acidity and which coagulates with alcohol and shows curdiness after 
evaporating and sterilizing can be made to stand sterilization with 
no resulting curdiness by mixing with it the proper amount of milk 
showing a high titratable acidity but no coagulation with alcohol. 
RELATIVE VALUE OF ACID AND ALCOHOL TESTS. 
The value of any practical test for determining the quality of milk 
for condensing must depend largely upon its ability to determine 
whether or not that milk, when evaporated to the required density, 
will withstand the necessary sterilizing temperature without becom- 
Alcohol Test 
SAMPLES* 
Aci'dity IE 1/ /O 9 a 7 6 5 4 3 
SAAfPL£S- 
2 3 4 5 6 7 3 9 10 II 12 
M No curd at £35°r: for 30 m In 
111 Curd at 235° F. for 30m In. 
Fig. 3. — The relation of titratable acidity and the alcohol test to coagulation of the 
evaporated milk. 
ing curdy. The flavor and odor are things which have to be deter- 
mined by the senses of taste and smell. The tendency of milk with 
objectionable flavor and odor to coagulate with alcohol indicates 
additional value of the test from this standpoint, especially where 
little attention is given to these factors. 
The preliminary work indicated clearly the unreliability of the 
acid test, as ordinarily used in the condensery, for determining the 
quality of milk for condensing. A comparison of the acid and 
alcohol tests in determining what milk when evaporated would 
stand sterilization without showing curdiness reveals the following: 
Of 90 samples of raw milk taken at the weigh room and repre- 
senting a wide range of condition, 46 samples of the finished evapo- 
rated milk showed curdiness after sterilization and shaking. Forty- 
four of the evaporated-milk samples withstood sterilization, show- 
ing no curdiness after shaking. Analysis of Table 2 and figure 3 
shows that the alcohol test would have rejected 43 of the 46 samples 
