CANNING SARDINES 191 
PARTIALLY DRYING THE FISH 
APPARATUS 
The apparatus used in drying is pictured in Figure 22, page 120. The plan of 
this drier is the same in principle as that shown in Figure 25, page 145, except 
that heat was furnished by steam coils; there was no cooler and no mechanical 
method for handling flakes. Air velocity was controlled by the operation of a 
damper leading to the blower. Other dampers made it possible to recirculate 
all or any part of the air that had gone through the drier. The temperature of 
the air was controlled by regulation of the quantity and pressure of the steam 
that entered the coils, and the humidity was raised, when desired, by recirculating 
part of the air, either with or without the addition of steam, on the intake side 
of the blower. The drier was first designed to handle a tier of four flakes and 
was so operated at first. Later, to get higher velocities, it was necessary to 
send all the air over and under just one flake. 
A standardized Short and Mason anemometer was used for obtaining the air 
velocities, and wet and dry bulb thermometers were used for indicating the 
temperature of the air and its relative humidity. These were placed always 
in the same place—where they encountered the full velocity of the air—and 
in the case of the thermometers, at some distance back from the steam coils 
and just in front of the fish. The air temperatures, as obtained, are quite 
accurate, but the readings of the wet-bulb thermometer are undoubtedly high— 
probably 3 to 6 per cent. Carrier, in the American Heating and Ventilating 
Engineers’ Guide (see footnote, p. 120), shows that such errors exist and that 
they are more pronounced at lower air velocities. The air velocities, too, being 
taken at the most favorable place, are a little higher than the average for the 
cross section used. The variations in velocity during any one series of runs, 
while studying one variable, are due to experimental errors. In taking these 
air velocities the anemometer was placed in position right after the fish were 
put in and was removed immediately after they were taken out. Due to this 
procedure, the time intervals were never exactly the same. On the whole, 
during any one run the temperature and humidity remained quite constant 
although there were frequent small variations, which were quickly corrected. 
These errors, however, are not important here, since the experiments are, for 
the most part, comparative ones, in which case most of the errors tend to cancel 
each other. Then, too, the experimental results are only used for the purpose 
of showing general tendencies and not for the detection of fine differences in 
behavior. 

PROCEDURES 
In studying the effect of varying any one drying factor care was taken to use, 
wherever possible, like-sized fish, all of which had the same preliminary treat- 
ment. The supply to be used in a given experiment, consisting possibly of 
several runs of the drier, was placed in a large bucket with holes in the bottom, 
from which the fish were removed as needed, weighed, and arranged on the flake, 
always in the same general manner. ‘The fish were weighed again immediately 
after being removed from the drier. They were then returned or discarded, 
as conditions demanded. Although precautions might be taken which would 
assure quite comparable drying conditions in any one experiment, the changes 
that take place in fish that have stood for several hours might seriously interfere 
with the experiment. These changes can be ignored if they make little difference 
in the amount of water removed from the fish by similar drying conditions during 
the time the experiment is in progress. Several series of runs of the drier were 
made in order to get information on this point. The results, given in Table 34, 
indicate no great difference in the amount of water removed from fish that have 
stood varying lengths of time. The variation shown is not pronounced nor is 
it constant; in fact, the experimental error itself might well be greater. In 
view of these facts, this variation will not be considered in the experiments 
unless the differences in the amount of water removed in the experiments under 
ccnsideration are quite small, 
