Nov. 24, 1923 
Studies on Rancidity 
337 
that they form with phloroglucin, were out of the question, because of the 
impracticability of isolating the reactive substance from rancid fats, or 
of preparing its condensation product in a sufficiently pure condition. 
Spectroscopic comparison of the colored condensation products was there¬ 
fore resorted to. The validity of the method was established by extend¬ 
ing the spectroscopic comparisons to the corresponding condensation 
products of vanillin, eugenol, cinnamic aldehyde and aged turpentine. 
Twenty-five cubic centimeters of a rancid sample of commercial oleic 
acid were shaken in a separatory funnel with 25 cc. of concentrated 
hydrochloric acid, after which the mixture was shaken with 25 cc. of 
a 1 per cent ethereal solution of phloroglucin. A portion of the red- 
colored hydrochloric acid layer was drawn off into a glass cell for spectro¬ 
scopic comparison. 
Corresponding solutions were prepared from samples of rancid lard, 
rancid cottonseed oil, rancid oleo oil, vanillin, eugenol, aged turpentine, 
and cinnamic aldehyde, as well as from an acrolein solution to which 
hydrogen peroxid had been added. Care was used in each case to 
employ such quantity of the given material as to produce in the hydro¬ 
chloric acid layer a color of about the same intensity as that of the solu¬ 
tion obtained from rancid oleic acid. Minor differences in intensity were 
adjusted by suitable dilution of the colored solutions with concentrated 
hydrochloric acid, or by varying the thickness of the layer of solution 
examined. 
The red solutions thus obtained were viewed by transmitted light, 
two at a time, through a spectroscope designed for such comparisons. 
The spectra obtained from the solutions prepared from the several rancid 
fats, from rancid oleic acid, and from the acrolein-hydrogen peroxid 
mixture appeared to be identical. In each of these cases the spectrum 
exhibited a fairly narrow, well-defined, apparently symmetrical absorp¬ 
tion band, centered at 6.0 on the arbitrary scale of the instrument, and 
located in the yellow-green region to the right of the D line. The position 
of the band was identical in each case, at least to within the limits of 
error of the instrument. 
The spectra of the solutions prepared from vanillin, turpentine, 
eugenol, and cinnamic aldehyde, while showing slight differences among 
themselves, were similar in type and showed a general absorption begin¬ 
ning at a point in the yellow-green or the blue-green region and extending 
to the violet end of the spectrum. In no case was a localized band 
observed similar to that which had been observed in the spectra of the 
solutions prepared from rancid fats and from acrolein-hydrogen peroxid 
mixtures. The coloring matters were extracted with amyl alcohol and 
the solutions thus obtained were reexamined, but no change in the type 
of the spectrum could be noted in any case, and no significant changes 
in the length of the darkened areas. 
A sample of nonrancid cottonseed oil which gave a red color in the 
Kreis test was similarly treated and examined. The spectrum obtained 
resembled those previously obtained in the work with the aromatic 
substances, and showed no similarity whatever to those obtained in the 
experiments with rancid fats, rancid oleic acid, or mixtures of acrolein 
and hydrogen peroxid. 
From the above-described experiment it follows that the spectroscopic 
examination affords a valid means of distinguishing the colored substance 
formed in the Kreis test on rancid fats from other red condensation 
