Nov. 24, 1923 
Studies on Rancidity 
333 
ing a red condensation product with phloroglucin. The possibility of 
some relationship between the presence of active oxygen and a positive 
Kreis test is also suggested. 
RELATIONSHIP BETWEEN THE KREIS TEST TOR RANCIDITY AND PEROXID 
oxygen 
The literature contains many references to the presence of peroxid 
oxygen in rancid fats, and Vintilesco and Popesco (j 9) in their test for 
rancidity use the presence of active oxygen as a criterion of the rancid 
condition. Kerr and Sorber ( 21 ) in a recent study have shown a wide 
parallelism between the results of the Kreis test and those obtained by 
the test employed by these authors. To avoid risk of error in the prem¬ 
ises, however, a number of samples of rancid fats and of rancid oleic acid, 
all giving positive Kreis tests, were examined for the presence of peroxid 
oxygen in the following manner: 
To a small amount of a 10 per cent solution of potassium iodid (free 
from iodates and from free iodin) an equal quantity of glacial acetic 
acid (used for solvent purposes) and a like quantity of the fat to be 
examined were added. The mixture was shaken vigorously for about a 
minute and diluted with 5 to 10 volumes of water. The production of a 
blue color on the addition of starch paste was taken as an indication of 
the presence of peroxid oxygen in the original fat. A blank test on the 
reagents was performed simultaneously. 
Without exception all samples of rancid fats and of rancid oleic acid 
that were examined gave a positive peroxid test accompanied by a 
positive Kreis test, while fresh fats gave neither test. 
These observations suggested all the more strongly the possibility 
that the active oxygen of rancid fats might participate in the production 
of the red color in the Kreis test, or that the Kreis test might even be 
a general reaction for peroxids of unsaturated compounds. Such a 
compound is known to be formed when turpentine is exposed to the air; 
and a positive response to the Kreis test was in fact obtained from an old 
sample of turpentine that also responded strongly to the test for peroxid 
oxygen. Subsequently, however, the sample was found to give weak 
but positive tests for aldehydes with Schiff’s reagent and with ammoniacal 
silver oxid, so that no legitimate conclusion could be drawn regarding the 
possible participation of the peroxid group in the reaction with phloro- 
glucin-hydrochloric acid. 
vSeveral attempts were therefore made to discover what effect the 
destruction of the peroxid group in rancid fats might have on their 
behavior in the Kreis test. 
A sample of rancid oleic acid, prepared by aeration of the commercial 
product, was found to react intensely in the peroxid and in the Kreis 
tests. About 25 cc. of the sample was then shaken with potassium 
iodid and acetic acid as in the peroxid test, and the treatment continued 
for two days, the liberated iodin being titrated from time to time with 
sodium thiosulphate solution. After addition of water and ether, the 
ethereal solution of the oleic acid was separated, washed with water, and 
tested for peroxids and for its behavior in the Kreis test. A weak test 
for peroxids was still obtained, however, together with a somewhat 
weakened Kreis test. 
In a second experiment an attempt was made to destroy the oleic acid 
peroxid by hydrolysis. Twenty grams of the rancid sample were boiled 
