Aug. 1 8 , 1923 
Gossypol in Cottonseed 
289 
be dispensed with by using the general average, as the increased accuracy 
obtained by considering variations in nitrogen content is within the 
limits of error of the method. 
The use of petroleum ether as a medium for precipitation facilitated 
the formation of a more filterable mass, and expedited the separation of 
the aniline compound. It also accelerated the rate of filtration. During 
the ether extraction some material other than fat or gossypol was dis¬ 
solved. This was removed from the petroleum-ether mixture before the 
aniline was added. 
When the crude ether extracts of cottonseed kernels or the mother 
liquors obtained in the process of recrystallization of gossypol were 
treated with aniline, the precipitate formed had a dull red color. When 
this precipitate was recrystallized from chloroform, a few crystals which 
differed from aniline-gossypol and which could be separated mechanically 
were usually obtained. When purified gossypol was converted into the 
aniline compound, crystals of this second type were not obtained. 
Although the second substance was not studied in detail, it seems probable 
that it is the aniline compound of the “D-gossypol” of Carruth. The 
aniline method precipitates not only gossypol, but also gossypollike 
substances, the quantity of which, however, was relatively small. 
Carruth states that the error of his method is less than 10 per cent 
when 0.5 gm. of gossypol is dissolved in 50 cc. of oil. The results (Table 
I) by his method, however, are somewhat lower than those subsequently 
obtained with the same seed by the authors’ modification of his pro¬ 
cedure. The results of the control analyses for the estimation of known 
quantities of free gossypol are given in Table II. 
Table II .—Estimation of known quantities of gossypol 
Weight of free 
gossypol taken. 
Dissolved in 25 cc. of — 
Total 
weight of 
precipitates.^ 
Aniline 
contained in 
precipitates. * 
Gossypol recovered. 
Gm • 
a O. 7002 
° O. 6642 
b O. 6902 
b O. 6902 
b 0. 6902 
b O. 6902 
Peanut oil. 
.do. 
Cottonseed oil c . 
.do. 
.do. 
Gm . 
O. 8736 
. 8424 
. 8296 
•8534 
.8594 
. 8562 
Per cent . 
26. 07 
25- 77 
26. 50 
27.10 
26. 67 
26. 67 
Gm . 
O. 6459 
•6253 
. 6097 
. 6221 
. 6302 
. 6279 
Per cent . 
92. 24 
94. 14 
88. 34 
90- 13 
9 1 * 37 
90.97 
« Dried in desiccator over calcium chlorid. 
& On moisture-free basis, 0.7000 gm. actually taken. 
c Refined in Oil, Fat, and Wax Laboratory of the Bureau of Chemistry, 
d Including weight of all subsequent small precipitates. 
« Calculated on basis of nitrogen analysis. 
The error would appear to be on the average about io per cent of 0.7 
gm. of free gossypol dissolved in 25 cc. of cottonseed oil, equivalent to 
about 3 mgm. per cubic centimeter of evaporated ether extract. The 
seeds showing a low gossypol content are not equally poor in oil. There 
is, therefore, a greater percentage of error in determining the gossypol 
in the samples of seed running low in gossypol. 
A gram sample of gossypol “acetate” gave no weighable ash. Kjel- 
dahl analysis of two preparations of “free” gossypol gave 0.0363 and 
0.033 P er cent of nitrogen. This small quantity is not significant and 
undoubtedly represents an impurity in the preparations. 
