658  '         Chemistry  of  Gossypol.  {AmsJpT%9ih8arm' 
which  separates  out  almost  quantitatively  on  cooling  the  solvent.  It 
is  possible  that  instead  of  being  a  simple  aniline  salt  of  gossypol  it 
may  be  a  condensation  product  of  carbonyl  groups  in  gossypol  with 
the  amino  group  of  aniline.  The  substance  dissolves  in  alcoholic 
potash  with  a  blood-red  color  which  seems  to  persist  for  a  period 
longer  than  one  would  expect  if  the  alkali  immediately  replaced  the 
aniline  at  the  acid  groups.  Both  "  B  "  gossypol  and  "  D  "  gossypol, 
described  later,  give  similar  orange-yellow  precipitates. 
The  insolubility  of  the  aniline  compound  of  gossypol  furnishes 
a  method  of  determining  the  amount  of  gossypol  in  cottonseed 
products,  and  therefore  a  chemical  method  for  determining  the  rela- 
tive toxicity  of  a  cottonseed  product.  These  compounds  do  not  dis- 
solve readily  in  the  usual  organic  solvents  but  are  least  soluble  in 
petroleum  solvents.  Warm  aniline  dissolves  them  readily  and  they 
separate  out  in  crystalline  form  on  cooling.  Alcoholic  potash  dis- 
solves them  with  a  beautiful  red  color  which,  however,  disappears 
after  the  compound  is  decomposed  by  the  reagent  into  gossypol  and 
aniline. 
The  aniline  compound  of  gossypol  was  first  made  by  Dr.  J.  T. 
Dobbins  in  this  laboratory.  It  was  desired  to  see  if  "  B  "  gossypol, 
which  at  that  time  was  thought  to  be  the  product  in  cottonseed  meal 
formed  from  gossypol  in  the  cooking  process,  could  be  made  with- 
out the  formation  of  the  black  material  which  accompanies  it  when 
gossypol  is  heated  to  180-1900.  It  was  thought  by  Dr.  Dobbins 
that  by  heating  the  gossypol  in  aniline  this  could  be  avoided.  It 
was  found,  however,  that  even  in  boiling  aniline  the  gossypol  was 
not  decomposed  as  in  dry  heating.  On  cooling  the  solvent  it  was 
noticed  that  an  orange-yellow  crystalline  substance  was  deposited. 
This  interesting  compound  is  not  appreciably  toxic,  probably  on 
the  most  plausible  explanation  of  the  reaction  since  gossypol  behaves  as  a 
fairly  strong  dibasic  acid.  Later  this  analysis  was  checked  by  another  chem- 
ist, who  reported  4.35  and  4.38  per  cent,  nitrogen  on  the  unheated  material. 
This  appeared  to  indicate  that  aniline  was  disengaged  on  heating.  I  have 
more  recently  made  analyses  of  this  same  preparation  and  have  found  4.61, 
4.64,  4.84,  4  70  per  cent,  nitrogen  in  the  unheated  substance.  There  is  no 
perceptible  odor  of  aniline  to  this  substance.  The  average  of  4.65  per  cent, 
for  the  three  most  closely  agreeing  results  agrees  better  for  a  compound  con- 
taining 2  molecules  of  gossypol  to  5  molecules  of  aniline,  2C3oH2s09-5C6H5NH2, 
which  requires  4.58  per  cent,  nitrogen.  My  analyses  of  the  aniline  compounds 
"  B  "  gossypol  and  "  D  "  gossypol  show,  respectively,  4.84  per  cent,  and  4.15 
per  cent,  nitrogen. 
