98 U. S. BUREAU OF FISHERIES 
It was not possible to determine with any degree of accuracy 
how rapidly the sardine-oil content of the frying oils increased dur- 
ing use by analyzing the samples of oil collected from time to time 
during the frying tests. The oils had been heated and exposed to 
the air for a long time, during which period much oxidation must 
have taken place. Such changes make the usual methods of analysis 
impracticable. 
A good idea concerning the change that must take place in the 
composition of oil used for frying sardines and the way in which 
it takes place can be obtained by using certain experimental data 
in connection with a series of calculations. These calculations are 
based partly on assumptions. They are reasonable ones, however, 
and the conclusions reached must be quite close to conditions actually 
attained in practice. The results of such calculations are very 
helpful in indicating ways in which frying can be improved. This 
work is summarized below.* 
When fat sardines are fried in cottonseed oil some oil cooks out 
of them and mixes with the oil in the bath. They leave more of 
this oil in the bath than they remove from it, and in this way the 
amount of sardine oil in the bath continues to increase. The rate 
of increase is rapid at first but lessens as frying continues. Never- 
theless, it is not long before the frying oil is almost all fish oil. The 
nature of this change in composition is shown in Figure 21. Other 
calculations show that the fish-oil content of a given quantity of 
frying oil increases less rapidly the larger the size of the individual 
units cooked at one time. An example will illustrate: If a ton of 
fish sufficiently fat to keep the oil content of the bath constant is 
fried 4 pounds at a time in 12 pounds of oil the fish-oil content of 
the frying oil will be less than if the same quantity of fish is cooked 
in units smaller than 4 pounds each. Still other calculations show 
that when the fish are fat enough to cause the oil in the fry bath 
to increase the percentage of fish oil increases less rapidly if the 
oil in the bath is allowed to increase than if the excess oil is removed 
as it collects. The application of these findings to improving frying 
procedures is discussed later. 
CHEMICAL AND PHYSICAL CHANGES 
Examination of the oil samples from the frying experiments showed 
that the oils gradually darkened with use, becoming red in color. 
At the end of both experiments the oil appeared almost black when 
viewed through thick layers. However, thin layers about an inch 
in thickness showed a deep red color. The viscosity of the oils 
increased greatly with use. The longer the oil was used the more 
unpleasant it became, acquiring a tallowy, paintlike odor and taste, 
both of which are characteristic and hard to describe. The free 
fatty-acid content of the oils increased but slightly, going from 
about 0.1 to 0.6 per cent in each lot of oil.*! The oils became only 
slightly rancid. Everything taken into consideration, it can not be 
said that they became insanitary—merely unpalatable. 
As far as the examination went, the oils remaining from the two 
runs (four lots) seemed to be quite comparable in quality; that is, 

3 Actual calculations are given on pp. 164 to 167, *! More complete data are given in Table 25, p. 168, 
fe a 
4 
PPE pind BBE OS A. | Stl a Ae SS eM al NA a 
ot Slee Saeed, 
a 
Pence, wri. 
