Am.  Jour,  pharm.  |  Mercuric  Chlorid  Tablets. 
January,  1914.  j 
3 
somewhat  different  in  details  from  either  of  the  two  modifications 
already  noted,  the  points  of  difference  and  the  reasons  therefor  will 
be  discussed. 
In  the  first  place,  while  it  is  true  that  the  use  of  large  amounts 
of  substance  and  reagents  tends  to  reduce  the  relative  magnitude  of 
experimental  errors,  there  is  a  point  beyond  which  any  slight  possible 
gain  in  accuracy  is  attained  only  at  the  sacrifice  of  an  unwarranted 
measure  of  simplicity,  convenience,  and  rapidity.  It  is  certainly  open 
to  question  whether  this  point  is  not  passed  by  Smith's  modification 
with  its  considerable  consumption  of  standardized  solutions,  espe- 
cially when  the  total  volume  of  liquid  worked  with  is  increased  by 
the  75  c.c.  of  water  here  found  necessary  to  prevent  interference  by 
hexamethylenamin  compounds.  The  writer  therefore  recommends 
the  employment  of  0.20  to  0.25  gram  of  mercuric  chlorid  for  each  test, 
and  the  addition  of  25  c.c.  of  tenth-normal  iodin. 
Experiment  1. — To  test  the  limits  within  which  the  results  of  parallel 
determinations  may  fall  if  the  above  proportions  are  used,  a  series  of  seven 
parallel  tests  were  made  on  a  solution  of  commercial  tablets,  using  the  equiva- 
lent of  one-half  tablet  for  each  test.  Aside  from  the  use  of  volumetric 
apparatus  which  had  passed  the  requirements  of  the  Bureau  of  Standards,  no 
precautions  not  employed  in  ordinary  quantitative  work  in  any  laboratory 
were  observed.  The  cubic  centimetres  of  iodin  solution  (tenth-normal  X 
0.978)  consumed  in  the  solution  of  mercury  were  as  follows:  (1)  17.72; 
(2)  17.73;  (3)  17.68;  (4)  17.69;  (5)  17.72;  (6)  17.73;  (7)  17-65;  an  extreme 
difference  of  0.08  c.c.  The  average  weight  of  mercuric  chlorid  per  tablet  was 
therefore  found  to  be  from  0.4677  to  0.4698,  a  difference  of  0.45  per  cent. 
The  range  of  variation  reported  by  Smith  in  a  series  of  five  tests  by  his 
method  on  pure  mercuric  chlorid  was  0.5  per  cent. 
Secondly,  the  proportion  of  potassium  iodid  should  be  consider- 
ably increased  over  that  employed  by  either  Rupp  or  Smith  in  order 
to  avoid  the  formation  of  mercurammonium  compounds  which  result 
when  caustic  alkali  in  excess  is  added  to  a  solution  containing  potas- 
sium mercuric  iodid  and  ammonium  salts,  as  in  the  well-known 
"  Nessler  test." 
Thirdly,  the  presence  of  ammonium  salts  in  one  way  or  another 
operates  to  retard  the  reduction  to  metallic  mercury  so  that  it  is  not 
complete  in  5  minutes  though  apparently  so  after  10  minutes. 
Lastly,  commercial  formaldehyde  solutions  often  contain  foreign 
substances,  some  of  which  conceivably  may  consume  iodin,  and  in 
fact  such  solutions  have  here  been  found  which  did  possess  this 
