Electrically  Prepared  Colloidal  Solutions.  A27 
Clouds  o£  finely  divided  metal  scatter  from  the  cathode  during 
the  sparking  and  remain  suspended  in  the  water  for  a  time 
depending  on  the  nature  of  the  metal.  The  dish  (D)  was 
surrounded  by  running  tap-water  in  order  to  counteract  the 
heating  effect  of  the  arc.  The  sparking  was  continued  until 
the  water  became  so  clouded  as  to  be  quite  opaque.  The 
solutions,  which  usually  contained  a  little  sediment  from  the 
pulverization,  were  then  carefully  filtered.  Using  gold,  silver, 
and  platinum,  the  resulting  filtrates  retained  nearly  the  whole 
of  the  metal  sputtered  from  the  electrodes  and  remained  con- 
stant for  months.  When  bismuth,  lead,  and  iron  were  used, 
the  filtrate  still  retained  a  large  proportion  of  the  metal,  but 
these  solutions  were  by  no  means  as  stable  as  the  preceding 
ones.  After  sparking  with  electrodes  of  zinc  and  tin,  the 
clouded  solution  began  at  once  to  clear,  and  the  filtrate  failed 
to  retain  any  of  the  metal. 
The  water  used  in  the  experiments  was  first  distilled 
through  the  large  laboratory  still,  a  trace  of  acid  potassium 
sulphate  added,  and  then  redistilled  through  a  silver  spiral 
tube  ;  the  specific  conductivity  of  this  water  was  about 
3  X  10-6. 
III.  Properties  of  these  Colloidal  Solutions. 
The  properties  of  such  solutions  are  quite  analogous  to 
those  of  the  so-called  colloidal  solutions  of  arsenic  sulphide, 
gold,  silver,  ferric  hydroxide,  and  many  organic  substances, 
which  have  been  prepared  chemically  and  carefully  in- 
vestigated by  Carey  Lea*,  Linder  and  Pictonf,  Barus 
and  Schneider  {,  Zsigmondy§,  Hardy  ||  and  others.  They 
diffuse  in  pure  water  very  slowly,  and  may  be  separated 
by  dialysis  from  impurities  of  various  salts.  They  may  be 
readily  filtered  through  paper  filters.  When  viewed  with 
a  microscope  of  the  highest  powers,  no  particles  can  be 
detected  in  the  solution  ;  yet  when  a  beam  of  light  is  passed 
through  a  tube  containing  some  of  the  solution,  it  is  diffused 
and  polarized.  Bredig^f  points  out  that  these  two  latter 
properties  fix  the  upper  and  lower  limits  to  the  size  of  par- 
ticles which  must  be  in  the  solution.  The  microscope  can 
detect  directly  a  particle  of  diameter  as  small  as  14  X  10~5  cms.; 
while  the  fact  that  the  particles  diffuse  light  shows  that  their 
diameter  is,  at  least,  quite  comparable  with  the  wave-length 
*  Amer.  Jour,  of  Science,  xxxvii.  p.  476  (1889). 
f  Jour,  of  Chem.  Soc.  vol.  lxi.  p.  148,  vol.  lxvii.  p.  63,  vol.  lxxi.  p.  568. 
X  Z&its.  f.  Phi/s.  Chem.  vol.  viii.  p.  278. 
§  Liebi^s  Ann.  ccci.  p.  29  (1898). 
||  Proc.  Rov.  Soc.  vol.  lxvi.  p.  110. 
i[  Loc,  at.  and  Drude's  Ann.  xi.  p.  218  (1903). 
2F2 
