292  Finely  Divided  Metals  and  Ferric  Salts.  {^Vune'iSs^'""'" 
When  this  filter  is  immersed  in  the  beaker,  the  clear  ferrous  solution 
rises  in  it  to  the  same  level  as  the  liquid  in  the  beaker,  and  may  then 
be  withdrawn  by  a  pipette.  The  object  of  the  muslin  is  to  prevent 
rupture  of  the  filter-paper  by  the  narrow  end  of  the  pipette.  A  titra- 
tion which  formerly  demanded  an  hour  or  more  can  be  executed  by 
this  method  in  three  or  four  minutes  with  an  accuracy  deducible  from 
a  comparison  of  the  following  numbers  selected  at  random  from  a 
series  of  titrations  in  which  no  especial  refinements  were  employed: 
25  cc.  of  an  acidulated  ferric  chloride  25  cc.  of  same  solution  re- 
solution reduced  by  iron-free  magnesium  re-  duced  as  above  required  15*2  cc. 
quired  15*3  cc.  of  a  decinormal  permangan-  of  same  permanganate  solu. 
ate  solution.  tion. 
If  one  is  to  judge  from  Aldendorff's  analysis  of  a  typical  specimen 
of  zinc-dust  (Fremy's  Encyclopedie),  this  substance  is  usually  free 
from  iron.  The  specimen  with  which  I  worked  contained  only  0'08 
per  cent,  of  iron,  or  rather  reducing  substance  equivalent  to  0*08  per 
cent.  iron. 
"3.  As  to  the  nature  of  the  reaction  between  zinc-dust  and  neutral 
ferric  chloride,  I  was  at  first  inclined  to  ascribe  the  efficacy  of  the  dust 
in  this  respect  to  the  occluded  hydrogen  which,  as  Greville  Williams 
has  shown,  is  always  contained  therein.  I  was  all  the  more  strength- 
ened in  this  view  by  experiments  which  showed  that  hydrogen  oc- 
cluded by  palladium  quickly  reduces  ferric  chloride  in  solution.  It 
is  interesting  in  this  connection,  coupled  with  the  experiments  of  Glad- 
stone and  Tribe,  to  note  that  hydrogen  occluded  by  platinum  has  no 
action  on  ferric  salts. 
To  put  this  hypothesis  to  the  proof,  I  heated  some  of  the  zinc-dust 
in  a  Sprengel  vacuum  till  all  the  hydrogen  was  given  off ;  I  then 
tried  the  action  of  the  volatilized  zinc  crystals  on  ferric  chloride. 
Their  reducing  power  was  unimpaired ;  hence  I  had  to  seek  for  a  new 
explanation. 
It  is  well  known  that  ferric  chloride  in  aqueous  solution  is  in  a  state 
of  partial  dissociation,  as  is  roughly  represented  in  the  equation — 
Fe^Clg  -f  (n  -1-  3)  H20±^(FeA^H20)  +  Fe^Cl  -f  6HC1. 
It  might  be  urged  that  on  adding  zinc-dust  to  such  a  system,  the 
hydrochloric  acid  would  be  removed  from  the  sphere  of  action,  with 
formation  of  zinc  chloride  and  hydrogen,  and  that  the  nascent  hydro- 
gen would  reduce  the  ferric  chloride  existing  as  such,  while  the  solu- 
ble hydrated  iron  oxide  might  in  virtue  of  this  upsetting  of  the  mo- 
