Fig.  4. 
Electrically  Prepared  Colloidal  Solutions.  435 
will  enable  us  to  find  values  of  /  V '  -,  J  for  any  solid  and  liquid 
if  for  known  values  of  X  and  fi  we  can  observe  the  corre- 
sponding values  of  v.  The  experiments  described  in  the 
present  paper  will  supply  the  corresponding  values  of  r,  X,  fx 
for  a  number  of  cases. 
VI.    Velocities  of  the  Particles  of  Metal  Solutions. 
The  first  systematic  investigation  of  the  velocities  of  these 
particles  in  an  electric  field  is  that  afforded  by  the  interesting- 
experiments  recently  published  by  Whitney  and  Blake*. 
The  colloidal  solution  was  contained  in  tubes  such  as  shown 
in  fig.  4;  they  were  30-50  cms.  long  and  2'46-275  cms. 
diameter.  The  ends  were  closed  with 
goldbeaters'  skin  and  platinum  elec- 
trodes placed  in  water-cells  just  outside 
these  ends.  When  A  was  connected 
to  the  negative  electrode  and  B  to  the 
positive  electrode  of  a  set  of  storage- 
cells,  a  migration  of  the  particles  set 
in  down  the  tube.  A  reversal  of  the 
current  in  the  tube  caused  the  particles 
to  rise  again.  In  this  way  the  authors 
have  deduced  the  velocities  of  particles 
in  solution  of  platinum  and  gold 
to  be  respectively  34  x  10-5  and 
43  x  10-5cm.  per  sec.  per  volt  per  cm. 
Various  circumstances,  unavoidable  in 
the  use  of  a  tube  of  this  construction, 
render  the  determination  of  the  ab- 
~r  solute  values  of  the  velocities  of  the 
particles  under  unit  electric  force 
somewhat  unsatisfactory.  The  rate 
with  which  the  surface  of  the  particles 
in  the  solution  moves  down  or  up  the  tube  varies  in  a 
non-uniform  manner  with  the  time.  As  the  writers  found, 
the  specific  resistance  of  the  clear  liquid  above  the  surface 
of  colloidal  particles  may  be  continually  changing  ;  this  would 
naturally  affect  the  potential  gradient  along  the  tube.  Even 
when  this  is  corrected  for,  as  Whitney  and  Blake  took  pre- 
cautions to  do,  still  the  specific  resistance  of  the  upper  clear 
layer  of  liquid  remains  always  different  from  that  of  the 
solution  ;  a  fact  which  must  tend  to  affect  the  movement  of 
the  particles  at  the  surface  of  separation,  on  account  of  the 
*  Jour.  Amer.  Cheni.  Soc.  vol.  xxvi.  no.  10,  p.  1339  (1004). 
