Am ' $^^mvm'} Assaying  Galenical  Preparations  of  Opium.         41 1 
washings  of  the  residue  till  the  contents  of  the  flask  weigh  not  less 
than  1-6  nor  more  than  2-2  of  the  weight  of  the  opium.  Add  half 
its  weight  of  alcohol,  shake,  add  ether  (1-5  of  the  opium  weight), 
shake,  add  ammonia  (0-4  of  the  opium  weight),  and  shake  vigorously 
till  morphine  begins  to  crystallize.  After  12  hours  pour  off  the 
ethereal  layer,  wash  the  morphine  with  ether  and  dry. 
Stillwell  follows,  in  the  main,  Fluckiger-Squibb,  washing,  how- 
ever, with  morphine  saturated  alcohol  and  similar  water.  He  uses 
hot  alcohol  to  purify  the  morphine. 
The  sampling  of  ophim,  as  now  done  by  different  chemists,  varies 
considerably.  Since*  uniformity  in  this  operation  would  result  in 
preventing  much  annoyance,  I  propose  the  following  plan  which  I 
have  followed  for  some  time  and  which  yields  good  results :  Take, 
with  a  knife,  a  small  piece  from  the  inside  of  every  lump  in  the  lot, 
usually  one  hundred,  mix  these  pieces  together,  take  10  gm.  for  the 
moisture  determination,  dry  the  remainder,  pulverize,  and  then  take 
from  this  homogeneous  mixture  5  gm.  for  another  moisture  determi- 
nation (of  the  powder),  and  10  gm.  for  the  determination  of  morphine* 
Calculate  the  percentage  of  morphine  in  the  crude  opium  taking 
into  consideration  the  amount  of  moisture  contained  therein  less 
the  moisture  contained  in  the  powder.  For  example  :  crude  opium 
contains  23-58  per  cent.,  and  the  powder  3  per  cent,  of  moisture. 
5  gm.  of  the  powder  yielded  0-577  gm.  morphine.  Hence  5  gm. 
powder  =  6-025  gm.  crude  opium,  and  6025  :  0-577  :  :'  100  :  9-5 
per  cent,  of  morphine.    (Experiment  4). 
Assay  of  galenical  preparations  of  opium. — The  percentage  of 
active  principle  is  calculated  by  weight.  But  liquid  medicines  are, 
in  the  United  States,  usually  dispensed  by  measure,  and  always  are, 
and  will  be  so  administered.  Hence,  I  prefer  measuring  galenical 
preparations  to  weighing  them,  for  analytical  purposes. 
In  the  subjoined  table  I  give  the  average  percentage  of  morphine 
as  obtained  from  the  preparation  and  compared  with  that  found  in 
the  opium  used.  In  such  assays  differences  are  sometimes  found, 
either  an  excess  or  deficiency.  Such  a  difference  cannot  actually 
exist,  if  the  opium  used  for  the  preparation  had  been  thoroughly 
exhausted  and  the  dregs  were  free  from  morphine ;  but  it  must  be 
remembered  that  no  sample  taken  from  opium  fully  represents  the 
whole.  However,  I  think  that  my  method  of  sampling  means  a 
step  nearer  to  the  truth.  When  working  on  a  large  scale,  it  would 
be  considered  a  waste  of  labor,  time  and  money,  to  subject  a  large 
