^jun'^im'"'*}    Finely  Divided  Metals  and  Ferric  Salts.  293 
bile  equilibrium  be  simultaneously  transformed  into  an  insoluble  hy- 
drated  form.  According  to  this  explanation  it  would  follow  that  more 
iron  hydroxide  would  be  precipitated  the  higher  the  temperature  at 
which  reduction  takes  place,  for  the  dissociation  of  ferric  chloride  in- 
creases with  the  temperature.  And,  indeed,  experiment  proved  that 
about  twice  as  much  iron  is  precipitated  as  insoluble  hydroxide  when 
the  reduction  is  effected  at  100°  as  when  it  takes  place  at  ordinary  tem- 
peratures. But  according  to  this  explanation,  one  would  predict  the 
improbability  of  reduction  if  absolute  alcohol  were  substituted  for 
water  as  the  medium  of  the  change ;  whereas  experiment  shows  that 
even  under  these  conditions  reduction  readily  takes  place  with  great 
rise  of  temperature. 
4.  I  am  thus  driven  to  the  conclusion  that  the  zinc  acts  merely  as 
a  dechlorinating  agent,  much  as  stannous  chloride  acts — Fe2Clg  +  Zn 
=  ZnClg+Fcg  CI4,  and  that  the  precipitate  of  iron  hydroxides  which 
occurs  in  neutral  solutions  is  partly  due  to  the  zinc  oxide  which  is 
always  present  in  the  dust  to  the  extent  of  about  50  per  cent.,  partly 
to  the  zinc  hydroxide  formed  during  the  reduction  by  the  action  of 
water  on  the  finely  divided  zinc.  Zinc-dust  merely  effects  instantane- 
ously the  dechlorination  which  I  found  zinc-foil  required  several  hours 
to  effect. 
The  change  represented  above  is  an  exothermic  one ;  the  heat  of 
formation  in  aqueous  solution  of  ZnClg  [112,840]  is  greater  than  the 
negative  thermal  change  in  the  passage  from  the  system  FegClg,  Aq  to  the 
system  Fe2Cl4,Aq  [55,540]. 
5.  If  this  explanation  of  direct  dechlorination  be  valid,  it  seemed 
probable  that  all  those  metals  whose  chlorides  have  a  heat  of  forma- 
tion in  aqueous  solution  greater  than  55,540  gram-units  would,  in  the 
finely-divided  state,  reduce  ferric  solutions.  I  have  made  many  ex- 
periments in  this  direction,  and  I  find  that  the  following  metals  in  a 
finely-divided  state  reduce  ferric  solutions  with  varying  degrees  of 
rapidity : — iron,  mercury,  silver,  aluminium,  and  copper,  as  well  as 
zinc.  Sometimes  the  metals  were  employed  in  the  shape  of  foil  [alum- 
inium, copper  silver],  sometimes  in  the  state  of  fine  division  in  which 
they  are  precipitated  from  boiling  alkaline  solutions  of  their  formates, 
or  from  hot  solutions  of  any  of  their  salts  by  means  of  zinc-dust,  fol- 
lowed by  repeated  digestion  with  dilute  acids  suited  to  the  occasion. 
In  the  cases  of  aluminium  and  silver,  it  was  definitely  proved  that  no 
precipitation  of  a  salt  of  iron  occurred.    This,  without  doubt,  would 
