62         ON  THE  ESTIMATION  OF  CREAM  OF  TARTAR,  ETC. 
which  had  begun  to  change,  and  had  undergone  several  changes 
(half  a  gramme  per  litre). 
In  no  case  was  the  proportion  of  cream  of  tartar  greater 
than  that  corresponding  to  a  saturated  liquid. 
Between  the  acidity  of  a  wine  and  the  amount  of  cream  of  tartar 
it  contains  there  is  no  relation.  In  fact,  in  two  wines  of  the  same 
acid  and  alcoholic  standards,  such  as  Formichons  of  1859 — 62, 
the  cream  of  tartar  varied  from  simple  to  double.  The  largest 
amount  corresponds  to  a  saturated  solution  of  cream  of  tartar 
and  to  the  newest  wine.  This  fact  is  essential,  for  it  shows 
that  the  variations  do  not  seem  due  to  a  decomposing  action  ex- 
ercised on  the  cream  of  tartar  by  the  free  acids  contained  in 
the  wines  examined. 
We  must  add  that  Formichon  of  1857,  one  portion  bottled 
and  another  portion  kept  in  a  balloon — sealed,  after  inducing  a 
vacuum,  for  the  last  three  years — contained  in  each  case  exact- 
ly the  same  quantity  of  cream  of  tartar. 
III.  Supposing  that  the  above  method  was  applicable  to  the 
approximative  estimation  of  the  total  amount  of  tartaric  acid, 
and  even  of  the  whole  quantity  of  potash  contained  in  wines, 
we  made  the  following  experiments  : — 
1.  Divide  a  weak  solution  of  tartaric  acid  into  two  equal 
portions,  neutralise  one  of  the  portions  exactly  by  potash,  and 
mix  the  two  liquids.  Add  some  of  the  ethero-alcoholic  mixture 
before  described,  which  will  precipitate  all  the  tartaric  acid  in 
the  form  of  cream  of  tartar  (except  the  trace  of  soluble  cream 
of  tartar  in  the  mixture).  Small  quantities  of  organic  acids 
may  be  added  to  the  liquid  without  materially  affecting  the  re- 
sults. 
2.  According  to  these  facts  it  suffices,  to  decide  whether  m 
wine  contains  free  tartaric  acid  independently  of  cream  of  tar- 
tar, to  take  fifty  cubic  centimetres  of  the  wine,  to  saturate  ten 
cubic  centimetres  of  it  by  potash,  to  add  to  them  the  forty 
others,  to  take  one-fifth  of  the  mixture  and  add  to  it  fifty  cubic 
centimetres  of  ethero-alcoholic  mixture.  If  the  wine  contain 
free  tartaric  acid,  a  more  abundant  precipitate  is  produced  than 
with  the  primitive  liquid.  The  excess  of  acid  in  the  precipitate  cor- 
responds nearly  to  half  the  weight  of  the  free  tartaric  acid  of  the 
wine.    The  process  has  always  been  applicable  to  the  wines  we 
