356 
Journal of Agricultural Research 
Vol. XXVI, No. S 
DISCUSSION OF RESULTS 
The experiments reported in this paper have shown, on the basis of 
spectrophotometric evidence, that epihydrin aldehyde diethylacetal and 
the Compounds that have been termed substance K and substance R 
yield identical condensation products with phloroglucin in the Kreis test. 
They have also identified substance K, formed by interaction of acrolein 
and hydrogen peroxid in strongly acid solution, as epihydrin aldehyde. 
The chief point of interest is the chemical nature of substance R. Most 
of the saturated acids and aldehydes up to C 9 compounds, together with 
hydroxystearic acid, dihydroxystearic acid, diketostearic acid, azelaic 
acid, acrolein, crotonic aldehyde, dihydroxyacetone, methylglyoxal, and 
acrylic acid have been definitely excluded as possibilities; while ket- 
oxystearic acid, azelaic half aldehyde, glyceric aldehyde and malonic dial¬ 
dehyde have been excluded either on theoretical grounds or because their 
described reactions do not harmonize with those of substance R. Even 
epihydrin aldehyde has been fairly definitely excluded in consequence of 
its probably differing from substance R in solubility, volatility, and 
stability. However, substance R undoubtedly yields epihydrin aldehyde 
on treatment with hydrochloric acid, as proved spectroscopically, and is 
derived from oleic acid; and presumably it contains an ethylene oxid 
group which is derived from the peroxid group of oleic acid peroxid. 
A revision of the hypothesis mentioned at the beginning of this paper 
regarding the mechanism of the atmospheric oxidation of oleic acid 
would seem to be necessary. In particular the decomposition of the oleic 
acid peroxid, which is formed as the initial step in the oxidation, must 
proceed, in part at least, by a route different from that which leads to the 
formation of the C 9 aldehydes and acids. At all events this was true of 
oleic acid ozonid, where hydrolysis with water, resulting in the formation 
of the C 9 compounds, materially interfered with the subsequent formation 
of substance R under the influence of concentrated hydrochloric acid. 
It might be argued that oleic acid peroxid itself is substance R, and 
that on treatment with the concentrated hydrochloric acid used in the 
Kreis test it is broken down directly to form epihydrin aldehyde, which 
in turn yields the red condensation product with phloroglucin. There is 
no direct means of verifying or disproving this hypothesis; but if it be 
well founded, then the peroxid, unexpectedly, must react differently 
from the ozonid, for the ozonid, unlike rancid fats containing substance 
R, gave a practically negative Kreis test until it had been treated with 
warm hydrochloric acid for an appreciable length of time. It seems 
more likely, therefore, that substance R is formed from oleic acid peroxid 
at a later stage in the atmospheric oxidation of oleic acid. 
What, then, are the steps leading to the formation of substance K 
from oleic acid peroxid, and at what point of the process is substance R 
formed? It seems unreasonable to suppose that the C 18 chain should 
rupture between two saturated carbon atoms while the weaker peroxid 
or ozonid grouping is left intact. To account for the formation of 
heptylic aldehyde, and of substance R and eventually of epihydrin 
aldehyde with its ethylene oxid group, it would seem necessary to 
suppose that the molecule is first weakened at the point of eventual 
cleavage by the introduction of new double bonds. 
The introduction of new double bonds into the oleic acid chain under 
the influence of atmospheric oxygen is not exactly an occurrence that 
would be expected. The substitution of a double bond for a single bond 
is ordinarily attended with considerable difficulty and, in the laboratory, 
