Am.  Jour.  Pharm.  \ 
January,  1900.  J 
Substitute  for  Ethyl  Alcohol. 
s 
Then  4  c.c.  of  91  per  cent,  alcohol  was  added  to  the  contents  of 
the  separator  and  shaken  in.  Then  20  c.c.  of  fresh  chloroform,  five 
minutes'  vigorous  shaking,  with  separation  and  drawing  off  the 
chloroform  stratum  into  the  beaker  as  before. 
This  addition  of  alcohol  and  chloroform,  shaking,  separating  and 
drawing  off  was  repeated  for  a  third  and  final  washing. 
The  chloroform  solution  was  evaporated  from  the  beaker  without 
boiling,  leaving  a  varnish-like  extract  that  weighed  015  gramme. 
This  was  dissolved  in  6  c.c.  of  decinormal  acid  by  water-bath 
heat  and  agitation,  and  was  titrated  back  with  decinormal  alkali  to 
the  neutral  point,  using  litmus  paper  indicator  in  the  way  described 
in  the  old  process. 
6-oo  c.c.  acid  used  less  3*65  c.c.  free  acid  =  2-35  c.c.  saturated  with  alkaloid, 
•28838  X  2-35  =  '67769  =  *  o*68oper  cent. 
Duplicate  assay  .  o-68o   "  " 
These  percentages  are  true  only  to  the  second  decimal  place,  and 
are,  therefore,  so  stated.  But  they  are  usually  trustworthy  to  0  005 
per  cent.  It  is  claimed,  however,  that  a  simple  and  easy  process 
that  will  give  a  result  true  within  this  first  range  of  error  is  more 
practically  useful  than  a  more  elaborate  process  with  half  that  range 
of  error. 
This  is  equivalent  to  stating  the  writer's  conviction,  long  held, 
that  for  pharmacopceial  purposes  a  process  that  five  pharmacists 
out  of  ten  can  apply  within  a  variable  error  of  1  per  cent,  of  result 
is  of  more  use  than  a  more  elaborate  process  that  not  more  than 
one  in  ten  can  apply  to  within  half  that  range  of  error  of  result. 
The  new  process  has  been  applied  to  the  percolates  as  well  as  to 
the  powder  with  the  same  degree  of  success,  10  c.c.  of  the  stronger 
percolates  being  evaporated  for  each  assay,  and  multiples  of  this 
volume  for  the  weaker  percolates. 
Throughout  the  many  applications  of  the  new  process  it  was 
found  that  unless  the  proportion  of  solid  extract  was  required  it 
was  better  not  to  carry  the  evaporation  lower  than  to  a  thick  liquid 
of  5  or  6  grammes,  because  it  is  much  easier  to  wash  it  clean  into 
the  separator  with  the  prescribed  volume  of  the  ammonia-alcohol 
mixture. 
The  one  great  difficulty  in  applying  the  "  shaking  out "  process 
in  alkaloidal  assaying  is  the  formation  of  emulsions.  This  difficulty 
is  entirely  avoided  in  this  process  by  the  use  of  alcohol :  First  to 
