326  Martins  Butter  Color  Method. 
A  MODIFICATION  OF  MARTIN'S  BUTTER  COLOR 
METHOD,  AS  APPLIED  TO  THE  DETECTION  OF 
ADDED  COLORS  IN  SPICES. 
By  Charges  H.  UWa^. 
Several  authorities  on  food  adulteration  state  that  oil  soluble 
coal-tar  dyes  and  certain  added  vegetable  colors  may  be  detected 
in  such  spices  as  capsicum  and  mustard  by  an  adaptation  of  Mar. 
tin's  butter  color  method,  shaking  the  sample  with  a  mixture  of 
two  volumes  carbon  disulphide  and  fifteen  volumes  alcohol.  It  is 
further  stated  that  the  carbon  disulphide  dissolves  the  oil  and  natural 
color  of  the  spice  while  the  overlying  alcohol  layer  holds  in  solution 
many  of  the  artificial  coloring  matters  that  may  be  employed. 
When  the  test  is  applied  as  above  described,  no  separation  into 
layers  takes  place  because  there  is  so  much  cellular  tissue  present 
in  proportion  to  the  amount  of  fixed  oil  of  the  spice  that  a  homo- 
geneous pasty  mass  results  if  five  grammes  of  the  spice  are  used, 
while  with  a  smaller  quantity,  say  two  grammes,  the  spice  simply 
remains  in  suspension  without  separation  into  layers. 
The  principle  of  Martin's  test  as  applied  to  butter  depends  upon 
the  fact  that  the  five  grammes  of  butter  which  are  used  in  the  test 
combine  with  and  separate  out  with  the  carbon  disulphide,  which 
under  ordinary  circumstances  is  perfectly  soluble  in  alcohol  and 
remains  in  solution  until  mixed  with  at  least  an  equal  proportion 
(to  the  carbon  disulphide  used)  of  some  fatty  material  which  is 
insoluble  in  alcohol. 
The  failure  of  the  added  spice  to  properly  separate  in  the  test  as 
applied  led  to  the  thought  that  the  addition  of  some  fixed  oily  sub- 
stance insoluble  in  alcohol  might  effect  the  necessary  separation 
and  accordingly  the  addition  of  liquid  petrolatum  and  of  lard  was 
tried  with  very  satisfactory  results  in  both  cases.  The  separation, 
as  effected  by  either  of  these  fats,  both  of  which  are  colorless  and 
thus  unobjectionable  in  their  application  to  this  particular  test,  was 
both  prompt  and  effective,  and  the  overlying  alcohol  layer  could  be 
decanted  easily  and  subjected  to  the  usual  dyeing  tests  for  confir- 
mation of  the  suspicion  afforded  by  the  unusual  depth  of  color  of 
the  alcoholic  solution. 
The  test  as  modified  is  as  follows :  Mix  1 5  cc.  alcohol  with  2  cc. 
carbon  disulphide  and  add  2  grammes  of  the  suspected  spice  ;  shake 
