334 Journal of Agricultural Research voi. xxvi. no. 8 
with about 200 cc. of water over night under a reflux condenser. The 
separated oleic acid caused no immediate liberation of iodin in the 
peroxid test, though a gradual liberation was soon observed and at the 
end of several hours the test had become strongly positive. The Kreis 
test was strongly positive from the first, as was also the aldehyde test 
with Schiff's reagent. 
While this experiment seemed to be unfavorable to the idea that 
active oxygen is a factor in the production of a positive Kreis test in 
rancid fats, the evidence which it furnished could not be regarded as 
conclusive, owing to some question as to whether the peroxid test had 
indicated a temporary absence of peroxids or the presence of a sluggishly 
reacting peroxid. 
In a third experiment, a sample of rancid fat was boiled with water for 
several hours, the water being replaced from time to time in order to 
effect the removal of any hydrogen peroxid that might conceivably retard 
the decomposition of the organic peroxids. Previous experience had 
shown that but little of the substance that causes the Kreis test is ex¬ 
tracted from rancid fats by hot water. At the end of the hydrolysis, the 
clear fat, separated from the water, gave an almost completely negative 
test for peroxids and a practically negative Kreis test, even on standing. 
Such experiments as those described above, of course, in no case could 
have proven that peroxid oxygen is a factor in the phloroglucin-hydro- 
chloric acid reaction of rancid fats. But by showing that this reaction 
persisted after the destruction of the peroxids, they might have eliminated 
peroxid oxygen from consideration in this connection. This they have 
failed to do, and the participation of peroxid oxygen in the phloroglucin- 
hydrochloric acid reaction of rancid fats must for the present be regarded 
as a possibility. 
In the light of this possibility, all the oxidation products of oleic acid 
that were originally examined as to odor and as to their response in the 
Kreis test were reexamined in such a manner that the Kreis test was per¬ 
formed after the substance had been treated with hydrogen peroxid. 
Also mixtures of these substances were similarly examined, both in pres¬ 
ence and in absence of hydrogen peroxid. Uniformly negative results 
were obtained all in cases. 
COLORED CONDENSATION PRODUCTS OF PHUOROGUUCIN 
Kobert (22), in a study of the phloroglucin-hydrochloric acid reaction, 
concluded that only compounds containing an allyl group (—CH 2 — 
CH: CH 2 ) or a substituted allyl group (—CH 2 — CH: CRR') are capable 
of forming red condensation products with phloroglucin in the Kreis 
test, and in support of this contention he mentions a whole series of 
compounds so constituted that react in this way. Yet the fact that 
the same reaction is given by vanillin and other compounds which contain 
no such grouping, seems to destroy the validity of his generalization. 
According to von Euler ( 12 ), who has exhaustively reviewed the 
literature of this subject, the condensation between aldehydes and phlor¬ 
oglucin generally occurs in several stages. As a rule, the primary 
condensation products formed from one molecule of aldehyde and two 
molecules of phloroglucin are said to be colorless, the formation of a 
colored compound being dependent upon subsequent anhydridization 
followed by oxidation. In this manner a complicated triple ring system 
results. 
