TEMPERATURE CHANGES, ETC., DURING CANNING OPERATIONS. 21 
from the distilled-water curves are noted, however, in the case of 
those curves for cans processed at 116° and 121° C., in which an 
equilibrium is not attained throughout the entire period of the tests. 
Furthermore, these differences are increased with the higher process- 
ing temperatures, the final portion of the curves for cans processed 
at 121° C. being steeper than those for cans processed at 116° 
C. That this has no correlation with the rate of heat penetration is 
shown by the fact that the increase in pressure is continued long after 
the equilibrium of temperature has been reached. Experiments show 
that a temperature equilibrium is reached in string beans in about 
15 minutes in the No. 2 and in 18 to 22 minutes in the No. 3 can. In 
one case during the present tests the pressure continued to increase 
for four hours after the processing was begun, at which time the 
experiment was terminated. 
This seems to indicate that at the higher temperature there is a 
decomposition of the material in the can, with a consequent lbera- 
tion of gases. It was thought that the setting free of hydrogen by 
the action of the food material on the metal of the container, as 
pointed out by Bigelow (1) might explain this result, but compara- 
tive tests, using both plain and enameled cans under these conditions, 
failed to give evidence in support of this view. 
These curves also illustrate how strains are greatly lessened by 
sealing at as high temperatures as practicable, especially when proc- 
essing is to be done at the higher retort temperatures. It will be 
seen that even when cans are sealed at temperatures as high as 
80° C. the pressures obtained are well above the theoretical values. 
The processing period in these tests, except in the single instance 
noted, was 1 hour. This may not conform to regular practice, but 
the maximum pressures for any period less than this will be shown on 
these curves, for when the steam in the retort is cut off the pressure 
in the can begins to fall almost immediately. 
The actual strain on the can at any time during the process may be 
found by subtracting the retort pressure from the pressure indicated 
for that time in the curve. When the retort pressure is released, 
however, the strain on the can is greatly increased. If a complete 
release of pressure in the retort should be effected instantaneously 
the full pressure indicated would be felt, but the gradual decline in 
retort pressure in practice favors the can somewhat. The greatest 
_ strain occurs when the pressure in the retort reaches zero, but at this 
point the strain is, however, only 3 to 5 pounds, or (in some cases 
more than this) less than the maximum pressure indicated in ‘the 
curve. ‘The extent of this variation from indicated pressures is de- 
pendent upon the temperature of sealing, the processing temperature, 
the nature of the material, the amount of bulging of the can, and the 
rate of release of pressure in the retort. When processing is done at 
