344 
Journal of Agricultural Research 
Vol. XXVI, No. S 
with the 3.35 cc. required in the preceding test. The titrated solution 
gave a strongly positive Kreis test. 
Repetition of Experiments b and c showed beyond doubt that when 
any neutral acrolein-hydrogen peroxid solution is treated successively 
with potassium iodid, excess hydrochloric acid and thiosulphate, in the 
order named, the resultant solution does not respond to the Kreis test. 
But when the order is changed so that a large excess of concentrated 
hydrochloric acid is added first and the other reagents afterwards, a posi¬ 
tive test is obtained. Evidently substance K is not destroyed by the 
removal of peroxid oxygen and is not formed except in presence of a con¬ 
siderable concentration of acid. It may be definitely stated, therefore, that 
substance K does not contain a peroxid group and is not acrolein peroxid. 
Epihydrin aldehyde. —A priori , it was perhaps to have been ex¬ 
pected that epihydrin aldehyde should be formed in the reaction between 
hydrogen peroxid and acrolein, more especially since other ethylene oxids 
are known to be formed in analogous reactions between peroxid oxygen 
and the ethylene linkage. Such a reaction, too, might occur in rancid 
fats, where also an ethylene oxid might result by degradation of a peroxid. 
Peroxid 
oxygen. 
i-o 
U 
v x 
/\ X\ 
V. 
+ A - i> 
+ AO 
But in spite of these relationships the absence of any description of 
epihydrin aldehyde in the literature consulted, led to the belief that this 
compound must be extremely unstable and that, even if formed in the 
reaction between acrolein and hydrogen peroxid, it immediately would 
be converted, through addition of water, into glyceric aldehyde. 
It now appears, however, that of the C 3 compounds originally sug¬ 
gested as being possibly identical with substance K, all but epihydrin 
aldehyde and mesoxaldialdehyde have been eliminated; and of these 
compounds, epihydrin aldehyde, being in effect a simple acrolein oxid, 
would seem to be the more closely related to acrolein. 
While epihydrin aldehyde itself seems never to have been prepared, two 
of its acetals have been synthesized, the diethylacetal by Wohl (43), and 
the dimethylacetal by Wohl and Momber (44). The latter authors also 
attempted to prepare the dimethylacetal by the direct action of various 
peroxids on the corresponding acrolein acetal, but were unsuccessful in 
their efforts. 
The preparation of epihydrin aldehyde diethylacetal by the process em¬ 
ployed by Wohl was eventually undertaken, the reactions involved being 
as follows: 
I CH,=CH—CHO+HCl+aCjHjOH 
Acrolein 
II CHjCl—CH 2 —CHfOCjHjlj+KOH 
III CH 2 :CH-CH(OC 2 H 5 ) J +HOa 
IV CHjOH—CHC1 - CH(OC 1 H 5 ) i +KOH 
->CH 2 C1 - CHj - CH(OC 2 H s ) 2 +h 2 o 
0 -Chlor-propionic aldehyde diethylacetal 
—>CH 2 : CH - CH(OC 2 H 5 )j+KC1+H 2 0 
Acrolein diethylacetal 
->CH 2 OH - CHC1 - CIEOCaH*), 
a-Chlor-£-oxy-propionic aldehyde diethylacetal 
r° 1 
—>CH 2 —CH—CHfOCjHjh+KCl+HjO 
Epihydrin atdehyde diethylacetal 
