6  lEyperimental  Xeotslatton
memory  the  favourable  or  unfavourable  results  of
every  kind  of  action.  Charles  Babbage  proposed
to  make  an  automaton  chess-player,  which  should
register  mechanically  the  numbers  of  games  lost
and  gained  in  consequence  of  every  possible  kind
of  move.  Thus,  the  longer  the  automaton  went
on  playing  games,  the  more  experienced  it  would
become  by  the  accumulation  of  experimental  results.
Such  a  machine  precisely  represents  the  acquirement
of  experience  by  our  nervous  organisation.
But  Erasmus  Darwin  doubtless  meant  by
experiment  something  more  than  this  unintentional ­
  heaping-up  of  experience.  The  part  of
wisdom  is  to  learn  to  foresee  the  results  of  our
actions,  by  making  slight  and  harmless  trials
before  we  commit  ourselves  to  an  irrevocable  line
of  conduct.  We  ought  to  feel  our  way,  and  try
the  ice  before  we  venture  on  it  to  a  dangerous
extent.  To  make  an  experiment,  in  this  more
special  sense,  is  to  arrange  certain  known  conditions, ­
  or,  in  other  words,  to  put  together  certain
causal  agents,  in  order  to  ascertain  their  outcome
or  aggregate  of  effects.  The  experiment  has
knowledge  alone  for  its  immediate  purpose  ;  but
he  is  truly  happy,  as  the  Latin  poet  said,  who  can
discern  the  causes  of  things,  for,  these  being  known,
we  can  proceed  at  once  to  safe  and  profitable
applications.
It  need  hardly  be  said  that  it  is  to  frequent  and
carefully-planned  appeals  to  experiment  in  the