96

POLITICAL  ECONOMY

the  data  from  which  he  calculates  at  what
output  maximum  gains  can  be  reaped  ;
but,  according  to  the  exposition  above,  the
comparisons  which  determine  his  action  are
comparisons  of  aggregates—of  total  gains  at
different  outputs—not  of  margins,  or,  in  other
words,  of  the  differences  made  to  gains  by
small  variations  of  output.  This,  however,
is  not  an  altogether  correct  account  of  the
matter.  The  monopolist  aims  at  an  aggregate,
a  maximised  revenue,  but  in  order  to  attain
it  he  can  be  guided,  and  is  not  unlikely  to
be  guided,  by  the  marginal  effects  of  variations ­
  of  his  output  on  his  net  gains.  He
need  not  make  his  supply  leap  about
bewilderingly  in  the  hope  of  fortuitously
hitting  upon  the  one  most  favourable  to  his
interests.  On  the  contrary,  if  he  is  wise,
he  will  decide  on  some  output  which  seems
to  offer  reasonable  prospects  of  yielding  a
high  monopoly  revenue  and  then  proceed
scientifically  by  making  small  variations  in  it
and  watching  their  effect  on  his  profits.  When
an  effect  is  favourable  he  will  naturally
make  another  variation  in  the  same  direction
and  so  on,  but  when  it  is  otherwise  he
will  retrace  his  steps.  When  monopoly
revenue  is  the  greatest  possible,  marginal
costs  equal  what  might  be  termed  differential