300         Atropine  Per  iodides  and  I odomer  curates.    { Am  jS'if^1'111, 
ing  a  dark  red  color,  which  can  easily  be  seen  by  stoppering  the 
flask,  reversing  it  and  looking  through  its  neck.  The  addition  of 
atropine  is  then  stopped,  the  flask  shaken  till  the  supernatant  liquid 
is  perfectly  transparent,  and  then  it  is  diluted  to  100  c.c.  Of  this 
liquid  an  aliquot  portion  is  filtered  off  and  the  excess  of  iodine  in  it 
determined  by  a  decinormal  solution  of  sodium  thiosulphate.  Every 
part  by  weight  of  iodine  consumed  corresponds  to  0-2849  part  of 
alkaloid  or  every  cubic  centimeter  of  the  decinormal  iodine  solution 
consumed  corresponds  to  0  0036048  gramme  of  atropine  alkaloid. 
These  factors  are  obtained  on  the  ground  that  for  the  formation 
of  the  atropine  enneaiodide  only  eight  atoms  of  iodine  are  supplied 
by  the  free  iodine  used  to  make  the  decinormal  iodine  solution,  the 
ninth  atom  coming  from  the  potassium  iodide  of  that  solution.  The 
success  of  this  method  depends  solely  upon  the  atropine  solution 
being  sufficiently  diluted,  as  otherwise  a  resinous  mass  is  liable  to  be 
formed  instead  of  a  granular  precipitate,  and  added  in  small  portions 
at  a  time  to  the  iodine  solution,  care  being  taken  to  shake  the  mix- 
ture after  each  addition  and  to  stop  the  operation  as  soon  as  the 
supernatant  liquid  commences  to  become  transparent,  being  still 
red-colored.  It  is  easy  to  be  guided  by  the  very  dark  color,  as  an 
excess  of  atropine  makes  the  liquid  almost  colorless.  With  a  little 
care  the  method  gives  very  good  results,  as  can  be  seen  from  the 
following  analyses. 
Whether  this  method  of  assay  is  applicable  to  the  crude  drugs 
containing  atropine  and  their  galenical  preparations,  we  shall  try  to 
determine  by  experiments  to  be  continued. 
In  conclusion,  we  wish  to  say  that  the  method  of  using  a  solution 
of  iodine  and  potassium  iodide  for  the  quantitative  estimation  of 
alkaloids  has  been  long  since  proposed,1  and  recently  Kippenberger2 
has  given  considerable  prominence  to  this  method  in  his  toxicologic 
cal  and  other  researches.  But  he  rests  upon  the  assumption  of  the 
formation  of  the  atropine  iodide,  and  this  may  be  the  leading  pro- 
duct under  the  particular  conditions  he  prescribes;  but  his  method 
as  a  whole,  when  tried  in  this  laboratory,  has  not  given  satisfactory 
results.  The  estimation  of  caffein  as  a  periodide  recently  published 
by  Gomberg3  proves  highly  satisfactory. 
1  Bouchard  :  Compt.  rend.,  9,475  :  Wagner.  Ding,  poly.  J.,  igi,  40.  Ztschr. 
anal.  Chem.,  1,  102. 
2  Ztschr.  anal.  Chem.,  35,  10  ;  34,  317,  etc. 
•Jour.  Amer.  Chem.  Soc,  is,  331. 
