ANoVf.XiNPih9um' }  Methods  for  Determination  of  Calomel  513 
solid  mercurous  chloride  to  a  Gooch  crucible  which  has  been  pre- 
viously ignited.  The  crucible  and  its  contents  are  washed  thoroughly, 
dried  at  no°  C,  weighed,  ignited  in  a  Bunsen  burner  to  drive  off 
the  mercurous  chloride,  cooled,  and  weighed  again.  The  loss  in 
weight  is  assumed  to  be  mercurous  chloride.  This  method  is  shorter 
and  simpler  than  the  sulphide  method,  but  it  can  not  be  used  for 
the  analysis  of  certain  calomel  tablets.  It  was  found  in  several 
cases  that  fillers  consisted  of  some  ingredients  that  were  insoluble  in 
water  and  various  organic  solvents,  but  were  at  the  same  time 
volatile  or  combustible,  like  cellulose,  thus  giving  a  result  too  high. 
A  method  was  also  tried  for  the  volatilization  of  the  free  mercury 
after  the  calomel  had  been  reduced  with  formaldehyde.  Here  there 
is  no  means  of  telling  when  the  reduction  is  complete,  and  even  after 
considerable  effort  consistent  results  were  not  obtained. 
Merrill  also  suggested  a  method  for  the  reduction  of  mercury 
to  the  metallic  state  with  formaldehyde,  filtering  off  the  free  mercury, 
and  determining  the  chlorine  gravimetrically  as  silver  chloride.  This 
is  open  to  the  same  objections  as  the  preceding  method,  for  com- 
plete reduction  with  formaldehyde  is  somewhat  slow  and  difficult. 
An  iodine  titration  'method  3  is  suggested  by  various  text-books 
on  quantitative  analysis  which  is  rapid  and  gives  good  results  with 
most  samples  of  calomel,  but  very  poor  results  with  a  few  others, 
success  seeming  to  depend  on  the  character  of  the  filler  used.  The 
method  consists  in  treating  the  tablets,  which  have  been  disinte- 
grated with  water,  with  potassium  iodide  and  standard  iodine  solu- 
tion in  excess,  and  titrating  this  excess  with  sodium  thiosulphate 
solution.  A  control  is  run  under  similar  conditions,  and  the  differ- 
ence between  the  amount  of  thiosulphate  solution  used  in  the  control 
and  the  sample  represents  the  iodine  consumed  by  the  calomel.  The 
end  point  in  this  titration  is  not  decisive  and  changes  on  standing. 
It  was  found  in  other  cases  that  the  iodine  did  not  react  as  rapidly 
as  the  method  indicated,  but  that  after  the  iodine  had  been  added  it 
was  better  to*  allow  the  solution  to  stand  before  titration. 
Other  methods  might  be  cited,  but  the  above-mentioned  are  suffi- 
cient to  indicate  the  difficulty  of  devising  a  method  that  is  at  once 
rapid,  accurate,  and  applicable  to  all  cases.  The  need  for  such  a 
method  arose  in  this  laboratory  by  reason  of  the  number  of  samples 
3  Schimpf,  "  Manual  of  Volumetric  Analysis,"  p.  408;  Sutton,  "  Methods  of 
Volumetric»Analysis,"  p.  248. 
