Am.  Jour.  Pharm.  ) 
February,  1914.  J 
Colloids  and  Crystals. 
73 
water  boils  at  about  ioo°,  freezes  at  about  o°,  and  exhibits  the  be- 
havior of  a  colloidal  solution  in  general.  On  the  contrary,  a  soap 
solution  in  alcohol  shows  the  normal  change  in  freezing  and  boiling 
points  corresponding  to  the  molecular  weight,  and  conducts  itself  in 
all  respects  like  an  ordinary  molecular  dispersion. 
II. 
Every  one  is  familiar  with  the  distinctions  between  solutions  and 
suspensions.  Suspensions  are  turbid  in  aspect,  and  the  solid  can  be 
removed  by  letting  it  settle,  or  by  nitration.  Solutions  are  clear,  dis- 
solved matter  does  not  subside  and  is  unaffected  by  filtering.  Col- 
loidal solutions  occupy  an  intermediate  position. 
Consider  for  a  moment  the  effect  of  increasing  subdivision  on 
a  suspension  of  finely-divided  gold  in  water.  So  long  as  the  diameter 
of  the  particles  is  much  greater  than  a  thousandth  of  a  millimetre,1 
the  system  will  be  turbid  and  the  gold  will  settle  rapidly.  But  the 
wave-length  of  visible  light  ranges  between  0.4  \x  and  0.7  /x,  and 
when  the  particles  become  smaller  than  this  they  can  no  longer  reflect 
light  and  the  liquid  will  appear  clear.  At  the  same  time  there  will 
be  a  rapid  falling  off  in  the  speed  of  settling.  Stokes  has  derived  a 
formula  for  the  velocity  of  subsidence,  V ,  of  small  spheres  of  radius 
R  and  density  ^  falling  in  a  liquid  of  density  S'  and  internal  friction  / 
under  the  force  of  gravity  g: 
Substituting  the  proper  values  for  gold  and  water  and  assuming 
a  radius  of  /x  for  the  particles,  the  value  for  V  is  about  14  centi- 
metres per  hour.  This  means,  of  course,  that  the  system  would  be 
a  coarse  suspension  and  would  clear  up  at  once.  But  when 
R  =  10  fx  fx,  V  is  only  about  a  centimetre  a  month.  This  begins 
already  to  be  fairly  permanent.  It  must  be  remembered  that  the 
high  density  of  gold  (19.5)  increases  the  rapidity  of  subsidence. 
If  we  make  the  calculation  for  S  =  3,  which  is  about  the  density 
of  arsenious  sulphide,  V  comes  out  only  about  a  millimetre  a  month. 
1  It  is  usual  to  employ  the  symbol  v>  (the  Greek  letter  mu)  for  the  thou- 
sandth of  a  millimetre.  In  the  same  way  ix  fx  indicates  the  millionth  of  a 
millimetre. 
