494 
Mr.  de  Luc  on 
A practical  method  of  approximation  in  the  determination 
of  the  co-effecls  of  the  fame  caufes. 
It  has  been  thought  hitherto,  that  in  order  to  leflen 
the  effects  of  the  errors  which  are  unavoidable  in  obfer- 
vations  and  experiments,  one  ought  to  look  for  the  rela- 
tions of  the  co-effefts  at  the  greateft  poffible  diftances; 
becaufe  in  that  cafe  the  errors  being  divided  upon  a 
greater  fpace,  each  feparate  part  is  lefs  affedted  by  them. 
Thus,  in  order  to  find  the  relation  of  the  dilatations  of 
brafs  and  fteel  employed  in  the  Pendulum,  one  would 
willingly  expofe  thefe  metals  to  an  artificial  congelation 
and  to  the  heat  of  boiling  oil,  that,  by  meafuring  greater 
lengthenings, 
Moyen  pratique  d? approximation , dans  la  fixation  des  rapports  des  co-effets  des  tnemes 
caufes . 
On  a imagine  jufqu’ici,  que  pour  diminuer  Ies  effets  des  erreurs  inevitables 
dans  les  obfervations  et  Its  experiences,  il  falloit  chercher  les  rapports  des  co- 
effets  auxplus  grandes  diftances  polfibles;  parce  que  ces  erreurs  fe  divifant  fur  un 
plus  grand  intervalle,  deviennent  plus  infenfibles  fur  chacune  de  fes  parties. 
C’eft  ainfi  que  pour  trouver  le  rapport  des  dilatations  du  leton  et  de  l’acier  qu’on 
employe  au  Pendule,  on  expoferoit  volontiers  ces  metaux  a une  congelation  arti- 
ficielle  eta  la  chaleur  de  Fhuile  bouillante,  afin  que  pouvant  mefurer  de  plus 
grands 
