A^wUhr;iP909rm'}    Laboratory  Notes  on  Assay  Work.  123 
chloroform)  very  vigorously  and  it  will  be  found  that  an  almost 
complete  separation  results  in  a  few  seconds ;  allow  the  liquid  to 
stand  for  one  minute  or  so,  as  necessary,  and  collect  the  acid  solu- 
tion in  another  separator.  Shake  the  froth  and  ethereal  layer  with 
30-40  c.c.  more  of  1  per  cent.  H2S04,  let  the  liquids  separate,  which 
now  takes  only  one-half  minute  or  so,  and  add  the  acid  layer  to  the 
other  combined  acid  solutions  and  finish  the  assay  according  to  the 
U.  S.  P.  method. 
It  is  not  my  intention  to  criticize  any  U.  S.  P.  methods,  but  I 
would  say,  that  the  amounts  of  solvents  for  alkaloidal  assays  given 
in  this  valuable  book  are  in  most  cases  far  too  small  to  exhaust  the 
drugs  thoroughly;  furthermore,  in  using  small  amounts  of  solvents 
or  diluents  the  danger  of  getting  emulsions  increases.  Conium  fruit 
and  its  preparations  are  especially  liable  to  form  emulsions  and  on 
account  of  their  high  volatility,  it  requires  skilful  and  quick  work 
to  avoid  any  loss.  If  treated  as  outlined  under  aconite  there  will 
be  no  trouble  in  effecting  separation  of  the  liquids,  and  the  quick 
work  will  not  allow  an  appreciable  loss  of  coniine. 
To  make  a  reliable  assay  of  any  drug  is  not  so  very  easy,  as  most 
or  all  of  us  know.  We  need  not  only  practice  and  skilful  hands,  but 
also  brains,  and  by  no  means  least  of  all  common  sense.  Every  little 
detail  requires  different  treatment.  I  find  some  drug  analysts  who 
are  satisfied  if  the  results  obtained  with  duplicate  assays  fall  within 
fair  limits.  As  they  mostly  make  these  duplicates  at  the  same  time 
we  must  expect  that  they  get  the  same  results  under  the  same 
conditions. 
It  is  certainly  and  undoubtedly  a  good  check  to  have  duplicates 
made,  but  the  results  of  these  duplicates  if  identical  or  nearly  so  do  in 
no  way  give  a  guaranty  that  they  are  right.  I  had  to  make  up  some 
vears  ago  a  solution  of  a  mydriatic  alkaloid  of  known  strength  in  an 
alcoholic  indifferent  fluidextract  and  gave  portions  of  this  entirely 
uniform  preparation  to  eight  different  analysts  at  eight  different 
places  for  assay  with  the  request  to  hurry  these  assays,  to  avoid 
any  possible  deterioration.  In  about  five  or  eight  days  I  received 
all  their  results,  most  of  them  written  on  nicely  filled  out  blanks. 
From  three  men  whom  I  knew  to  be  reliable,  I  expected  close 
results  and  I  got  them  ;  the  five  others  were  all  too  low,  ranging  from 
80  per  cent,  down  to  22  per  cent,  of  alkaloids  present,  except  one 
sample  which  ran  160  per  cent.  As  every  one  of  these  analysts  said 
duplicates  were  made  by  them,  these  duplicates  did  certainly  not 
prove  the  correctness  of  their  assays. 
