350  Processes  and  Laboratory  Notes.       { Am jJu^r \m&rm' 
conclusions  based  upon  the  other  seven  samples ;  it  will  be  seen 
that  there  is  a  decrease  of  from  0  21  to  0-71  per  cent,  between  the 
insoluble  ash  and  the  ash  of  the  dregs,  in  the  case  of  those  samples 
yielding  an  impure  crude  morphine,  whilst  No.  10,  yielding  a  pure 
morphine,  showed  no  decrease. 
While  these  experiments  were  going  on,  I  also  tried  in  various 
ways  and  with  different  indicators  to  determine  the  acidity  of  the 
opium  or  opium  infusion  directly,  but  these  efforts  were  fruitless. 
The  loss  in  the  reprecipitation  of  the  morphine  varies  from  0033 
to  O  0635  gramme.,  and  while  all  of  this  may  not  be  morphine,  owing 
to  the  influence  of  the  lime-water  upon  the  impurities  in  the  crude 
morphine,  it  opens  up  the  question  of  the  morphine  left  in  the  mother- 
liquor  in  the  assays  proper.  When  it  is  remembered  that  this  opera- 
tion was  carried  out  so  as  to  minimize  the  loss,  that  the  use  of  alco- 
hol and  of  larger  quantities  of  ether  in  the  assays  will  necessarily 
cause  greater  loss,  and  that  the  morphine  actually  weighed  must  be 
subjected  to  a  correction  which  itself  is  arbitrary,  one  can  realize 
that  much  work  will  yet  have  to  be  done  before  a  satisfactory  or 
accurate  opium  assay  process  is  arrived  at.  Of  the  two  problems 
to  be  solved,  the  one  disclosing  accurately  the  quantity  of  morphine 
in  mother-liquors  is  considered  the  more  difficult ;  in  fact,  the  solving 
of  this  will  practically  also  solve  the  purity  of  any  isolated  morphine. 
ANALYTICAL  PROCESSES  AND  LABORATORY  NOTES. 
By  Charles  H.  La  Wall. 
ASSAY  PROCESSES  FOR  KOLA,  GUARANA  AND  COFFEE. 
A  method  for  the  estimation  of  caffeine  in  kola,  guarana  and 
coffee,  which  obviates  the  use  of  the  Soxhlet  extraction  apparatus, 
has  been  used  with  success  during  the  past  year.  The  results  agree 
closely  with  those  obtained  by  the  long  and  tedious  processes  for- 
merly employed,  and  can  be  obtained  within  a  much  shorter  space 
of  time.  It  resembles  closely  the  process  recently  suggested  by  Dr. 
Keller  for  the  determination  of  caffeine  in  tea.  The  directions  are 
as  follows :  Into  a  separatory  funnel  of  convenient  size,  place  5 
grammes  of  the  drug  and  5  c.c.  10  per  cent,  ammonia  water.  Allow 
the  mixture  to  stand  for  thirty  minutes,  then  shake  out  the  alkaloid 
with  chloroform,  using  three  portions  of  20  c.c.  each. 
