34» 
Journal of Agricultural Research 
Vol. XXVI, No. 
The compound C 15 H 12 0 6 , on the other hand, might easily result from 
the condensation of epihydrin aldehyde and phloroglucin, through anhy- 
dridization of the primary condensation product: 
H 
4 S 
: 4 > 0+ ’ 
4=o H0 \/ 
H 
o 
/\ 
i 
H- 
H- 
H 
4 
XI-v 
I >° 
H— C' +H 2 0 
H | H 
O H—C O 
And according to von Euler such anhydrid formation frequently occurs 
in the condensation reactions of phloroglucin. 
Evidently, therefore, the material analyzed was composed principally 
of the first anhydrid of the phloroglucid originally formed by the con¬ 
densation of epihydrin aldehyde and phloroglucin, a compound which 
corresponds to the formula C 15 H 12 0 6 , and which contains the epihydrin 
aldehyde radical intact. Presumably the anhydrid formation occurs 
under the dehydrating influence of concentrated hydrochloric acid and 
is responsible for the change of the originally water-soluble phloroglucid 
into the extremely insoluble substance that served for our analytical 
material. 
The experiments described in this section have shown (a) that sub¬ 
stance K is not glyceric aldehyde, dihydroxyaceton, methyl glyoxal, 
acrolein peroxid, malonicdialdehyde, or oxyacrolein; ( b ) that the colored 
phloroglucid obtained in the Kreis test from epihydrin aldehyde is 
identical with the phloroglucids similarly obtained from substance K 
and from rancid fats; (c) that in its reaction with phloroglucin, epihydrin 
aldehyde enters integrally into the reaction product. It therefore fol¬ 
lows that epihydrinaldehyde is formed in the reaction between acrolein 
and hydrogen peroxid in presence of concentrated hydrochloric acid, and 
that lie color obtained from rancid fats in the Kreis test is due to the 
formation of a phloroglucid of epihydrin aldehyde. 
