8 BULLETIN 1022, U. S. DEPARTMENT OF AGRICULTURE. 
conductivity of copper at about 1.0, of brass at about 0.25, and of 
glass at 0.002, the chances of error from heat conduction are hun- 
dreds of times greater with this thermocouple than with the ther- 
mometer. ite 
It must be borne in mind that in experimental work with canned 
’ foods small inaccuracies are of minor importance. Because of the. 
differences in the con- 
sistency of the material, 
of difficulty in prepar- 
ing uniform packs, of 
difficulty in adjusting 
closely the retort tem- 
peratures throughout 
the entire process, and 
of other factors, it is 
practically impossible 
to check time-tempera- 
ture curves closer than 
0.5° C., and it is doubt- 
ful whether a difference 
of 0.5° C., or perhaps 
even more with mate- 
rials of this kind, is of 
any practical signifi- 
cance. 
| 
| 
: 
®) 
= ( 
‘ss 
PRACTICAL VACUUM TESTS. 
In an attempt to de- © 
termine the optimum 
temperature for sealing, | 
cans of various mate- 
rials were prepared in 
lots which were sealed 
at the temperatures of 
20° 50° 2 602-707, 80. 
90°, and 100° C., respec- 
tively. In some cases, 
instead of this proce- 
dure, the cans were ex- 
hausted for two, four, or six minutes and then sealed. In each case 
the lots were divided and one half were tested for vacuum with a 
common Young’s vacuum gauge, and the remainder were subjected 
to concussion tests to learn to what extent the vacuum affects the 
denting and collapsing of cans in ordirary handling. 
Fie. 2.—Apparatus used in the oncussion tests. 
