Pyrometry  and  Areometry , 8c  c.  473 
be  only  accidental;  for  I do  not  believe  that  all  glafles 
have  an  equal  dilatability  by  heat.  Their  disability 
often  appears  different  when  they  are  foldered ; for  it  is 
no  doubt  owing  to  that,  that  the  parts  which  are  united 
when  they  are  melted,  often  feparate  when  they  grow 
cold,  which  does  not  happen  when  the  glafs  is  exactly 
the  fame.  It  is  pollible,  therefore,  that  this  apparently 
exaft  conformity  was  occafioned  by  fome  compenfation, 
rather  than  by  real  exaflnefs. 
I faid  before,  that  the  irregularities  I obferved,  when 
the  glafs  and  the  brafs  were  combined,  were  not  to 
be  attributed  to  the  glafs,  and  here  is  a proof  of  it. 
When  I had  adjufted  the  immoveable  wire  of  my  Mi- 
crofcope  to  a fharp  point  which  terminated  my  rod  of 
glafs,  the  water  being  at  the  temperature  of  1 o°  of  my 
Thermo- 
bien  n’etre  qu’accidentelle;  carje  ne  crois  pas  que  les  differens  verres  aysnt  tous 
une  egale  dilatabilite  par  la  chaleur:  on  ne  voit  que  trop  fouvent  quand  on  les 
foude,  que  leurs  dilatabilites  peuvent  etre  differentes;  car  c’efl  fans  doute  par  la, 
que  les  parties  reunies  quand  elles  font  fondues,  fe  feparent  quelquefois  en  fe 
refroidiffant;  ee  qui  n’arrive  pas  quand  c’eit  exaCtement  le  merne  verre.  II  fe 
pourroit  done  que  cette  exaCte  conformite  apparente  vint  de  queique  compenfa- 
tion  plutot  que  d’une  exactitude  reelle. 
J’ai  dit  ci-delfus  que  les  irregularites  que  je  reinarquois  lorlque  !e  verre  et  le 
leton  etoient  combines,  ne  devoient  pas  etre  attributes  au  verre  ; eten  voief 
la  preuve.  Lorfque  j’avois  ajufte  le  fil  immobile  de  men  Microfcope  fur  urie 
pointe  aigue  qui  terminoit  ma  branche  de  verre,  tandis  que  1’eau  etoit  a la  tern- 
Yol.  LX VIII.  L i I perature 
