274  Estimation  of  Cocaine  by  Mayer's  Reagent.  {Xm-ju0nl'^tTm- 
about  0*246  gm.,  but  accurate  weighings  are  not  practicable,  since  the 
precipitate  cannot  be  washed  with  water  without  loss.  The  observed 
variations,  however,  are  too  large  to  be  due  solely  to  this  cause,  and 
indicate  clearly  a  variable  composition  in  the  precipitate  itself.  The 
actual  weights  in  five  experiments  were  O230,  0*245,  0*248,  0*252  and 
0*271  gm.  The  lowest  result  was  obtained  when  the  reagent  was 
added  little  by  little,  the  highest  when  an  excess  was  added  at  once. 
Analysis  of  some  of  these  precipitates  showed  that  they  contained 
much  less  than  the  theoretical  proportion  of  mercury,  which  would  be 
0*066  (for  0*1  of  alkaloid).  At  the  same  time  the  quantity  was  too 
great  to  comform  with  the  formula  (RI)2HgI2,  or  0*033.  The  quan- 
tity found  was  about  *043  (in  a  precipitate  weighing  0*252),  corres- 
ponding with  the  formula  (RI)3(HgI2)2,  but  containing  either  an  ex- 
cess of  iodine  or  some  adhering  potassium  iodide  from  the  reagent. 
Variability  in  the  composition  of  the  precipitate  must  necessarily 
render  its  use  in  exact  quantitative  estimations  inadmissible.  At  the 
same  time,  if  under  similar  conditions  results  approximately  identical 
can  be  obtained,  it  may  be  possible  to  use  it  in  such  approximate  esti- 
mations as  the  pharmacist  has  occasion  to  make. 
The  use  of  Mayer's  reagent  offers,  indeed,  a  ready  means  of  ascertain- 
ing the  alkaloidal  strength  of  a  complex,  fluid,  of  which  the  pharmacist 
may  avail  himself,  when  time  and  skill  for  carrying  out  an  exact  quanti- 
tative estimation  are  wanting.  J  t  is  necessary,  of  course,  to  be  familiar 
with  the  precautions  to  be  observed  in  the  use  of  the  reagent,  and  to 
understand  the  sources  of  possible  error.  With  a  thorough  knowledge 
of  these,  it  may  be  possible  to  utilize  Mayer's  reagent*  in  the  practical 
examination  of  preparations  of  coca  leaves,  and  the  liability  to  error 
is  perhaps  no  greater  than  in  the  case  of  most  of  the  alkaloids  esti- 
mated by  its  use. 
When  Mayer's  reagent  is  added  drop  by  drop  to  an  acid  solution  con- 
taining cocaine  (1  :  200  to  1  :  600)  there  is  at  first  produced  a  heavy 
white  precipitate  which  collects  at  once  in  curdy  masses.  After 
adding  a  certain  quantity  of  the  reagent,  it  will  be  found  that  the 
filtered  fluid,  which  still  gives  a  heavy  precipitate  with  Mayer's  re- 
agent, produces  a  precipitate  also  in  a  fresh  solution  of  cocaine.  It 
is  thus  evident  that  precipitation  is  complete  only  when  a  certain  ex- 
cess of  reagent  is  present  in  the  fluid.  If  reagent  is  added  until  pre- 
cipitation is  just  complete,  it  will  be  found  that  the  fluid  does  in  fact 
contain  a  considerable  quantity  of  mercury.    The  following  are  some 
