Amju0ne!'ils6arm'}  Estimation  of  Cocaine  by  Mayer's  Reagent  275 
of  the  results  obtained  in  actual  titrations.  In  each  experiment  O'lOO 
gm.  of  alkaloid  was  used. 
Dilution,  1  :  200.  After  adding  4  c.c.  of  reagent,  the  filtered  fluid 
began  to  give  a  precipitate  with  cocaine  solution.  It  was  not  until 
5'85  c.c.  of  reagent  had  been  added,  however,  that  precipitation  was 
complete,  and  it  was  then  found  that  the  fluid  contained  mercury 
corresponding  with  0*55  c.c.  of  reagent. 
Dilution,  1  :  400.  After  adding  5  c.c.  of  Mayer's  reagent,  excess 
was  indicated  by  testing  with  cocaine  solution.  For  complete  pre- 
cipitation, however,  7*2  c.c.  of  reagent  were  required,  the  fluid  being 
found  to  contain  2 -3  c.c.  in  excess. 
Dilution  1  :  600.  For  complete  precipitation  8  c.c.  of  reagent  were 
required  ;  excess  of  reagent  not  estimated. 
In  these  experiments  the  solution  titrated  contained  in  each  case 
ten  minims  of  6  per  cent,  sulphuric  acid.  The  Mayer's  reagent 
actually  used  in  all  my  experiments  was  made  of  one-half  the  ordi- 
nary strength,  such  a  reagent  being  most  convenient  for  general  use. 
It  was  found  in  general  that  in  all  cases  there  was  required  for  com- 
plete precipitation  of  the  alkaloid  an  excess  of  reagent  amounting 
approximately  to  8 "5  per  cent,  of  the  solution,  and  if  the  precipitate 
were  washed  with  water  not  containing  Mayer's  reagent,  it  was  appre- 
ciably dissolved.  From  a  study  of  the  results  of  numerous  titrations 
I  deduced  the  following  rule  for  interpreting  the  results  of  an  ex- 
periment :  1 .  Correct  the  reading  from  the  burette  by  subtracting, 
for  each  c.c.  of  fluid  present  at  the  end  of  the  titration,  0'085  c.c.  (if 
the  half-strength  reagent  is  used).  The  remainder  multiplied  by  10 
will  give  the  quantity  of  alkaloid  indicated,  in  milligrams.  If  the 
Mayer's  reagent  of  full  strength  is  used,  a  correction  of  about  0*036 
c.c.  would  have  to  be  applied  for  each  c.c.  of  fluid  at  the  end  of  the 
titration,  and  the  multiple  would  be  20  instead  of  10. 
In  order  that  the  results  obtained  shall  be  at  all  trustworthy,  it 
is  necessary  that  the  reagent  be  added  always  in  the  same  way.  The 
first  addition  should  not  exceed  two-thirds  of  the  entire  amount  re- 
quired, and  the  successive  portions  should  consist  of  about  0*3  c.c. 
at  a  time,  until  the  end  of  the  titration  is  approached.  My  practice 
is  to  throw  the  fluid  at  once  on  a  filter,  just  large  enough  to  hold  it 
comfortably,  and  run  it  through  the  same  filter  once  or  twice  after 
each  addition  of  reagent.  This  is  a  much  more  expeditious  method 
than  that  of  allowing  the  precipitate  to  subside,  and  testing  the  super- 
