308  Estimation  of  Morphine  in  Pills,  Etc.  {Am- jJ^yr'Jurm- 
and  the  extraction  and  decantation  continued  until  complete  ex- 
traction was  obtained.  An  aliquot  part  of  the  extract  was  then 
taken  for  the  determination. 
The  method  gives  very  accurate  results  in  the  hands  of  a  careful 
worker.  The  yellow  color  of  the  oxime  does  not  interfere  with  the 
titration,  as  by  proper  dilution  it  colors  the  liquid  only  slightly. 
Tablets  of  phenolphthalein  frequently  contain  milk  sugar  or 
cane  sugar,  but  as  cane  sugar  does  not  give  an  oxime  with  hydrox- 
ylamine,  and  as  milk  sugar  is  practically  insoluble  in  absolute 
alcohol,  they  do  not  interfere  with  the  reaction.  In  case,  however, 
that  the  method  should  give  too  high  a  result,  it  is  better  to  make  a 
volumetric  determination  of  sugar  in  order  to  be  more  sure. 
THE  ESTIMATION  OF  MORPHINE  IN  PILLS, 
TABLETS,  ETC.* 
By  J.  B.  Williams. 
In  a  paper  read  before  the  last  meeting  of  the  Pharmaceutical 
section  of  the  American  Chemical  Society  Mr.  A.  D.  Thorburn  sug- 
gests the  estimation  of  morphine  in  pills,  tablets,  etc.,  by  making 
the  aqueous  solution  alkaline  and  extracting  with  a  mixture  of 
phenyl-ethyl  alcohol  and  benzene,  partially  evaporating  the  alka- 
loidal  solution,  extracting  the  residue  with  n/io  acid  and  titrating 
back  with  n/io  or  N/50  alkali,  using  hematoxylin  as  indicator. 
This  method  has  recently  been  tried  out  with  the  following  results. 
Duplicate  assays  of  a  2  per  cent,  solution  of  morphine  sulphate, 
using  10  c.c.  (  =  0.2  Gm.  morphine  sulphate  or  0.1506  Gm.  anhy- 
drous morphine  alkaloid). 
^=0.1827  Gm.  morphine  sulphate  =  91.35  per  cent. 
5  =  0.1919  Gm.  morphine  sulphate  —  95.95  per  cent. 
The  phenyl-ethyl  alcohol  mixture  did  not  separate  sharply  even 
after  standing  two  hours.  The  morphine  is  apparently  not  dissolved 
in  the  mixture,  but  appears  to  be  held  in  suspension  by  it.  N/50 
alkali  was  used  in  titration,  but  the  end  point  with  hematoxylin  as 
indicator  is  not  sharp  or  satisfactory. 
*  Presented  at  the  meeting  of  die  American  Chemical  Society  at  Wash- 
ington, D.  C,  December,  191 1. 
