COOLING HOT-BOTTLED PASTEURIZED MILK. 13 
only about 5° F. (2.8° C). Therefore, if the crates containing pint 
bottles were stacked on top of those containing the quarts, the tem- 
perature throughout the stack could be equalized to a certain extent. 
This would require, however, extra care and trouble in stacking. 
The curves in figure 8 are plotted from temperatures taken in quart 
bottles in the first, third, fourth, and fifth tiers of crates, the bottles 
selected having been immediately above one another and in line. 
On account of not having a thermocouple available, the temperature 
in the second tier was not taken. Curves 3, 4, 5, and 6 show the 
difference in temperature of bottles in respect to one another when 
placed one above the other. It will be noted that the difference in 
temperature of adjacent bottles decreases rapidly as the distance from 
the bottom increases and also that there is a wide difference in the 
temperature of the top and bottom quarts. 
The curves 3, 4, 5, and 6, in figure 9, showing the drop in tempera- 
ture of pints and quarts, are nearer one another than those in figures 
6, 7, and 8. This is owing to the initial temperature of the pint bot- 
tles being about 13° F. (7.2° C.) lower than that of the quarts. It is 
of course impracticable to adjust the initial temperatures of different- 
sized bottles so that the cooling will be more nearly uniform through- 
out the stack, and these curves are simply included as a matter of 
general information. 
When the cold air is forced in at the bottom of the stack and up 
through the crates, the variation in temperature in different-sized 4 
bottles and the same-sized bottles at different positions in the stack 
is too great for satisfactory operation, and we do not believe this 
method would be commercially practicable. Consequently in our 
next experiments the direction of the air through the stack was 
reversed periodically, the cold air being blown through from the 
bottom for 15 minutes and then through from the top for the same 
period. 
BY FORCING THE AIR ALTERNATELY FROM BOTTOM AND TOP. 
The series of curves shown in figure 10 shows the effect of reversing 
the direction of air through the stacks. The air was forced through 
the stack from the bottom for 15 minutes, then reversed and forced 
through from the top for, the same length of time. This arrange- 
ment was tried with the view of bringing the temperature of the top 
and bottom bottles nearer together. It may be noticed that the 
numbers on the curves in figure 10 do not correspond to those 
in the previous figures. As before stated, these numbers refer to 
the thermometers used in the work, and in the present case ther- 
mometers 1 and 2 recorded the temperature of the incoming air — 
one for the upward flow, the other for the downward flow. Nos. 
8 and 9 recorded the outgoing air in a similar way* 
