Am.  Jour.  Pharm. 
Feb.,  1887. 
Amount  of  Caffeine,  in  Coffee. 
95 
still  very  impure,  and  the  recrystallization,  after  pressing  between 
bibulous  paper,  so  readily  gives  rise  to  loss  that  this  method  cannot 
be  relied  upon  to  give  accurate  results.  It  is,  moreover,  extremely 
tedious  and  troublesome.  Dr.  James  Bell1  recommends  as  a  better 
method  the  extraction  of  coffee  beans  mixed  with  magnesia  by  boiling 
with  strong  alcohol,  evaporating  off  the  spirit  and  treating  the  residue 
with  water  to  dissolve  the  caffeine.  In  this  way  a  considerable  portion 
of  the  coloring  material  is  separated  by  the  magnesia,  and  a  further 
quantity  is  got  rid  of  by  evaporating  the  aqueous  solution  to  dryness 
with  a  further  quantity  of  magnesia,  and  then  dissolving  the  caffeine 
in  hot  benzol.  We  have  not,  however,  found  this  method  to  give 
more  satisfactory  results  than  the  previous  one,  and  one  circumstance 
that  greatly  interferes  with  its  application  to  coffee  is  the  considerable 
amount  of  fat  present  in  the  bean.  After  several  attempts  to  modify 
these  methods  of  treatment  for  the  determination  of  caffeine  in  coffee 
beans  we  had  recourse  to  the  use  of  lime  as  a  means  of  separating  the 
tannic  acid  while  dissolving  out  the  caffeine  by  boiling  with  alcohol. 
For  this  purpose  it  is  advisable  to  mix  the  finely  powdered  coffee 
with  moist  lime  and  then  to  extract  the  mixture  in  a  continuous  per- 
colator of  the  kind  described  by  Wait!  in  this  Journal  (vol.  xiv.,  p. 
376).  The  alcohol  is  then  evaporated  off  and  the  dry  residue  is  mixed 
with  some  water  and  a  few  drops  of  dilute  sulphuric  acid,  the  addition 
of  which  has  the  effect  of  separating  the  fat  and  clarifying  the  solution 
by  converting  a  small  quantity  of  soluble  lime  salt  into  calcium  sul- 
phate. After  filtering  ^  the  cooled  liquid  it  is  quite  free  from  fat, 
and  may  then  be  evaporated  to  obtain  the  caffeine  in  a  crystalline  state. 
A  better  plan,  however,  is  to  extract  the  caffeine  from  the  solution  by 
shaking  it  with  chloroform,  in  which  it  is  freely  enough  soluble  to  be 
readily  taken  up,  and  on  evaporating  off  the  chloroform  caffeine  will 
be  obtained  in  a  condition  fit  for  weighing.  The  principal  points  to 
be  observed  in  carrying  out  this  operation  are  the  acidification  of  the 
wrater  solution  from  the  spirit  extract  and  the  shaking  with  chloroform, 
but  with  proper  care  very  uniform  results  can  thus  be  obtained.  We 
have  hitherto  been  in  the  habit  of  operating  upon  50  grams  of  coffee 
beans  in  'each  experiment,  but  after  confidence  in  the  method  has  been 
obtained  and  experience  in  operating,  a  much  smaller  quantity  might 
be  taken. 
On  applying  this  method  to  the  determination  of  caffeine  in  various 
11  Analysis  and  Adulteration  of  Foods.'   Part  I,  p.  16. 
