Nov. 24, 1923 
Studies on Rancidity 
357 
generally involves the use of vigorous reagents. Linolenic acid is not 
readily nor profitably prepared from oleic acid. Yet, in Nature the 
transition from the more saturated to the less saturated acids, and from 
the less saturated to the more saturated, must frequently be made (<?), 
and cleavage of even the saturated acids into smaller fragments, and 
eventual decomposition into carbon dioxid and water, must occur. 
The idea that the substitution of double bonds for single bonds in the 
oleic acid chain may be brought about by natural processes has, there¬ 
fore, nothing intrinsically unreasonable about it; and this is what is 
believed to occur, possibly only as a subordinate side reaction, at an 
early stage in the atmospheric oxidation of oleic acid. 
According to this conception, the first stage in the formation of sub¬ 
stance R would still be the formation of oleic acid peroxid. Some of the 
oxygen thus activated might then detach some of the hydrogen com¬ 
bined with saturated carbon atoms in the same or other molecules of the 
peroxid to form new unsaturated linkages, which in turn could react 
with molecular oxygen to form new peroxids. Among the compounds 
formed in this way one might expect to find the triple peroxid of linolenic 
acid, or an ethylene oxid compound derived therefrom by loss of an 
atom of oxygen from its central peroxid group. 
Salway (33), in studying the atmospheric oxidation of linseed oil, has 
detected acrolein among the products formed, but was unable to detect 
acrolein among the products formed in the atmospheric oxidation of 
oleic acid. He advances the hypothesis that in the oxidation of linolenic 
acid, molecular oxygen is first added at the first and third double bonds 
with the formation of a double peroxid, which in turn ruptures at the 
peroxid groups with the formation of heptylic aldehyde, the half alde¬ 
hyde of pimelenic acid and butendial. From the latter compound 
acrolein is formed with evolution of CO and C 0 2 . 
Our present idea would postulate a somewhat analogous process in 
the decomposition of the double peroxid of the oxid of linolenic acid, 
assuming that this compound be formed in the atmospheric oxidation of 
oleic acid, in which process the central ethylene oxid group would remain 
intact, while cleavage would occur at the two peroxid groups. In this 
way we would have formed heptylic aldehyde, the half aldehyde of 
pimelenic acid, the oxid of butendial, and possibly by further oxidation 
of the latter compound, the oxid of acrolein- 0 -carboxylic acid. By a 
process analogous to that indicated by Salway for the formation of 
acrolein from butendial, epihydrin aldehyde, i. e., substance K, might be 
formed from the oxid of butendial or from that of acrolein-/ 3 -carboxylic 
acid. Whether or not the latter reactions occur naturally or only under 
the influence of the concentrated hydrochloric acid employed in the Kreis 
test is a matter for conjecture; although some evidence has been obtained 
that would seem to indicate the possible presence of a simple compound 
of epihydrin aldehyde in rancid fats. 
According to this conception, substance R, the constituent of rancid 
fats that is responsible for their behavior in the Kreis test, would be the 
oxid of a double peroxid of 7-8, 9-10, 11-12 linolenic acid, the oxid of 
butendial, the oxid of acrolein -0- carboxylic acid, or a simple compound 
of epihydrin aldehyde. In any one of the latter three cases the Kreis 
test would have an added significance, inasmuch as the formation of 
substance R would then involve the simultaneous production of a chem¬ 
ically equivalent amount of heptylic aldehyde, the compound that Scala 
has held to be responsible for the rancid odor. And in this case a 
