276  Estimation  of  Cocaine  by  Mayer's  Reagent.  {Amju0nue^Sarm- 
natant  fluid.  The  quantity  of  reagent  required,  no  doubt,  will  be 
different  if  this  latter  method  is  followed.  The  filtrate  runs  through 
rapidly  and  perfectly  bright.  The  proportion  of  acid  used,  of  course, 
should  be  always  the  same,  although  in  this  titration  the  result  is  not 
greatly  influenced  by  considerable  variations  in  the  proportion  of  acid. 
If  no  acid  is  added,  the  fluid  will  not  filter  clear.  It  is  some- 
times difficult  to  tell  just  when  the  precipitation  should  be  called  com- 
plete. My  practice  is  to  continue  addition  of  the  reagent  until  two 
drops  fail  to  produce  at  once  a  distinct  permanent  turbidity.  Occa- 
sionally, however,  it  will  be  found  that  such  a  turbidity  continues  to 
be  produced  for  some  time,  without  the  formation  of  a  weighable 
amount  of  precipitate.  In  such  cases  one  must  use  judgment,  an 
indispensable  factor  in  all  analytical  work. 
If  possible,  the  solution  for  titration  should  have  a  strength  of 
about  1  part  of  alkaloid  in  400,  since  the  results  of  titration  seem  to 
be  more  uniform  than  in  solutions  either  stronger  or  weaker  than 
this.    The  results  vary  most  in  fluids  containing  more  than  1  in  300. 
Some  experiments  made  with  a  modified  Mayer's  reagent,  contain- 
ing five  equivalents  instead  of  three  of  potassium  iodide,  promise 
greater  uniformity  than  those  obtained  with  the  ordinary  solution. 
A  considerably  smaller  quantity  of  reagent  is  required,  and  the  influ- 
ence of  dilution  is  not  so  great.  I  am  not  sure  but  that  in  nearly  all 
cases  the  modified  reagent  may  be  substituted  for  "Mayer"  with 
advantage,  but  the  quantity  of  alkaloid  precipitated  by  it  in  general 
is  greater.  The  addition  of  a  fixed  quantity  of  potassium  iodide  to 
the  solution  before  titration  has  a  similar  influence.  Thus  in  a  solu- 
tion 1  :  200,  requiring  of  the  ordinary  reagent  about  6  c.c.  to  precipi- 
tate 0*100  of  cocaine,  it  was  found  that  after  the  addition  of  2  c.c.  of 
a  10  per  cent,  solution  of  potassium  iodide,  5'04  c.c.  of  the  same 
reagent  sufficed.  Of  the  modified  reagent,  only  4.8  c.c.  were 
required. 
The  presence  of  alcohol  is  of  course  to  be  avoided  in  all  titrations 
with  Mayer's  reagent.  In  a  single  experiment,  however,  made  to 
ascertain  its  actual  influence  with  this  particular  alkaloid,  I  found 
that  in  a  solution  1  :  400,  20  per  cent,  had  very  little  effect.  With- 
out the  alcohol,  7*17  c.c.  of  reagent  were  required.  With  it,  7*28  c.c. 
A  single  experiment  with  glycerin  seemed  to  show  a  greater  influence, 
the  quantity  of  reagent  required  being  only  6'8  c.c. 
Confessedly  the  results  to  be  expected  in  the  use  of  Mayer's  reagent 
