45  6 'Mr.  de  luc  on 
mature  between  x o°  and  40°  of  my  Thermometer, the  mean 
proportions  of  the  dilatations  of  the  brafs  and  glafs  were 
as  2 1 to  1 o;  confequently,  I had  need  of  a length  of  brafs 
which  lbould  be  to  that  of  the  glafs  in  my  Hygrometer 
.as  10  to  21,  in  order  to  make  up  for  the  dilatability  of 
its  frame;  whereas  it  needed  to  have  been  from  10 
to  22  in  the  tranfition  from  water  in  ice  to  boilinp 
•j  a 
water. 
I do  not  give  this  latter  proportion  as  being  equally  to 
be  depended  upon  as  the  other;  but  it  is  fufficient 
that  there  be  really  a difference  between  the  two,  to 
ground  the  general  reflexions  I fhall  make,  after  having 
previoufly  noticed  another  difference  of  the  fame  kind 
which  . I have  found  by  means. of  this  Pyrometer. 
But  before  I mention  thefe  new  experiments,  I will 
-be flow  another  moment  upon  thofe  which  relate  to  the 
compa- 
rapport  moyen  cles  dilatations  du  Icton  et  du  verre  fut  comme  21  a 10 ; et  que 
par  confequent  j’avois  befoin  d’une  longueur  de  leton  qui  fut  a celledu  verre  dans 
mon  Hygrometre,  comme  10  a 21,  pour  compenfer  la  dilatabilite  de  fa  mon- 
ture;  au  lieu  qu’il  l’auroit  fallu  de  10  a 23  dans  le  paffage  de  l’eau  dans  la  glace 
a l’eau  bouillante. 
Je  nedonne  pas  ce  dernier  rapport  comme  aufli  fur  que  1’autre;  mais  il  fuffit 
qu’il  y ait  certainement  une  difference  entre  les  deux,  pour  fonder  les  reflexions 
generates  aux  quelles  je  viendrai,  apres  avoir  explique  une  autre  difference  du 
$neme  genre  que  j’ai  trouvee  au  moyen  de  ce  Pyrorrietre. 
Mais  avant  que  de  rapporter  ces  nouvelles  experiences,  je  m’arrgterai  encore 
im  moment  fur  cellesqui  regardent  les  expanfibilites  comparatives,  oudes  rapport 
entre 
