288 
Journal of Agricultural Research 
Vol. XXV, No. 7 
petroleum ether. Rub off the material adhering to the precipitating 
flask with a rubber-tipped glass rod or dissolve it in ether, from which it 
may be precipitated by almost completely evaporating the ether, adding 
petroleum ether, and reevaporating to a small volume. Then pour it 
carefully into the Gooch crucible. Bring the Gooch crucible and its 
contents to constant weight at ioo° C. by hourly heatings. This precipi¬ 
tate is slightly hygroscopic. Carry on the heating no longer than is nec¬ 
essary. If filter paper has been used in the Gooch crucible, the precipi¬ 
tate can be removed without admixture of foreign substances. 
Preserve the filtrate containing the petroleum-ether-aniline mixture 
and transfer it to an Erlenmeyer flask. Stopper the flask and let it stand 
in the cold for from seven days to one month, to ascertain whether all the 
gossypol has come down. If gossypol appears, allow the petroleum ether 
to evaporate partly and let stand again. Corrections based on these 
subsequent precipitates may be made. 
APPLICATION OF MODIFIED ANILINE METHOD 
Table I shows the gossypol content of several varieties of cottonseed 
determined by the aniline method of Carruth ( 4 ) and also by the quali¬ 
tative “acetate” method (3), in which care was taken to make the yields 
as large as possible. These data show that the recoveries by the “ace¬ 
tate” method are from 16 to 33 per cent lower than those by the aniline 
method. The approximate agreement of the results obtained by the two 
procedures lends weight to the assumption that the aniline method gives 
data representing the gossypol content. 
Table I .—Gossypol obtained by the “ acetate ” and aniline methods 
Variety of cottonseed. 
Acetate 
method. 0 
Aniline 
method. 6 
Lone Star. 
Per cent. 
O. 27 
'IIA 
Per cent. 
O. 320 
. 426 
. 614 
nc? 
Acala..... 
Trice... 
• 0 £t + 
Durango. 
. 401 
C - 636 
C . 6lO 
I. Ol8 
Egyptian. 
Sea Island. 
° The acetic acid in gossypol “acetate” is approximately to per cent and has been deducted. 
b A Kjeldahl analysis was made and the nitrogen, calculated as aniline, was deducted from the weight 
of the precipitate. 
c Some was lost. 
The nitrogen content of the aniline precipitate was from 3.75 to 4.035 
per cent. The variation was due, to some extent at least, to the adsorp¬ 
tion of free aniline. A lower nitrogen content has been observed and a 
preparation free from the odor of aniline has been obtained upon re¬ 
crystallizing several times from chloroform. Usually more nitrogen was 
present in the more bulky precipitates from the Gooch crucible, owing 
probably to less favorable conditions for removal of aniline in the heating. 
The variability in aniline content of the aniline-gossypol compound was 
noted by Carruth also (3). The weight of gossypol was calculated by 
deducting the weight of the aniline (N X 6.64) from the weight of the 
precipitate. This correction of the deduction, however, could perhaps 
