jSji'O  . . Mr.  de  Luc  on 
leaft  of  all  the  fubftances  which  are  not  affected  by  hu- 
midity. I would  always  ufe  flint-glafs,  that  its  changes 
in  this  refpetd  might  be  more  uniform  in  all  the  Areo- 
meters. 
Its  ball  fhould  be  one  inch  andean  half  in  dia- 
. * ' ■ L J ■ ■ *■  •«  - • 
meter,  and  there  fhould  be  at  the  bottom  of  it  a little 
hollow  cylinder,  which  fhoukl  communicate  with  it,  and 
contain  the  ballad,  in  order  that  it  might  be  able  to  keep 
upright,  at  the  other  end,  a branch  fo  much  ,the  longer; 
which  will  be  eafiiy  underftopd.  This. ball  fhould  not  be 
very  thick,  any  more  than  the  fuperior  branch  on  which 
its  divifions  fhould  be  marked.  The  different  thickneffes 
of  this  branch,  that  is,  its  different  external  diameters, 
will  produce  the  different  fenfibilities  of  the  inftrument. 
The  lefs  the  diameter  will  be,  the  more  will  the  Areome- 
ter 
Hers;  du  moms  entre  les  matieres  que l’humidite ' n’afFefte  pas.  Ce  verre  feroit 
toujours  le  flint-glafs,  afin  que  fes  changemens  a cet  egard  fuflent  plus  uniformes 
dans  tous  les  Areometres. 
Je  donnerois  a fa  boule  un  pouce  et  demi  de  diametre:  et  elle  auroitafon' 
fond  un  petit  cylindre  creux  qui  communiqueroit  avec  elle,  et  renferraeroat  le 
left;  afin  de  pouvoir  maintenir  de  bout,  au  cote  oppole,  une  branche  d’autant 
plus  longue  ; ce  qu’on  fentira  aifement.  Je  ferois  cette  boule  peu  epaifle,  ainft 
que  le  tube  fuperieur,  ou  la  branche  fur  laquelle  les  divifions  devroient  etre  mar- 
quees. Les  difterentes  epaifteurs  de  cette  branche,  e’eft  a dire  fes  differens  dia- 
metres  exterieurs,  feront  les  difterentes  fenflbilites  de  l’inftjrmnent.:  plus  fon  dia*» 
meter 
