Pyrotnetry  and  Areometiy,  See.  477 
fcale,  in  order,  if  the  refults  fhould  happen  to  be  dif- 
ferent, to  take  a mean  of  them.  In  a word,  for  this  one 
meafure  one  might  ufe  all  the  precautions  that  are  not 
grudged  in  a fundamental  experiment,  though  one  is  apt 
to  neglect  them  in  common  ufe- 
The  expanfion  of  a certain  rod  of  glafs  might  there- 
fore be  thus  determined;  and  by  fixing  to  it  afterwards 
any  other  fubftance,  in  the  manner  which  I have  ex- 
plained, one  would  have,  bv  means  of  the  immoveable 
wire  alone,  their  abfolute  expansibility,  free  from  any 
fenfible  error  occafioned  from  the  infirument. 
Notwithftanding  that  the  expanfions  of  the  glafs  were 
regular  in  my  experiments,  they  did  not  obferve  the  fame 
progrefs  as  my  Thermometer  in  their  degrees.  Thofe 
of  the  glafs  were  always  increafing,  or  its  condenfations 
decreafing, 
prendre  enfuite  le  milieu  entre  les  refultats,  s’ils  etoient  differens.  En  un  mot  on 
pourroit  prendre  dans  cette  mefure  unique,  toutes  les  precautions  qu’on  ne  re- 
grette  pas  dans  une  experience  fondamentale,  mais  qu’on  neglige  fi  aifement  dans 
l’ufage  ordinaire. 
On  determ  ineroit  done  ainfi  l’expanfion  d’une  certaine  branche  de  verre,, 
<Tapres  laquelle,  en  attachant  enfuite  a cette  branche  toute  autre  ma^iere  comme 
je  l’ai  explique,  on  auroit, . par  le  fil  immobile  feul,  leur  expanfibilite  abfolue. 
exempte  de  toute  erreur  fenfible  provenant  de  1’inflrument. 
Quoique  les  expanfions  du  verre  fe  trouvaffent  regulieres  dans  raes  expe-  * 
riences,  elles  ne  fuivirent  pas  la  marche  du  Thermometre  dans  leurs  degres : 
celles  du  verre  furent  toujours  croiffantes,  ou  fes  condenfations  decroiffantes, 
compa-r 
