g6  Estimation  of  Sugar  and  Tannin  in  Wines.  { Am'jebT,'S&aTVa' 
wards  an  increase  in  the  weight  of  the  copper  precipitate,  and  unless 
the  filtration  has  been  very  carefully  effected,  the  sugar  may  come 
out  from  0-2  to  2-2  per  cent,  too  high,  the  error  increasing  with  the 
amount  of  lead  solution  used.  For  wines  rich  in  tannin  and  color- 
ing matters,  the  charcoal  process  is  the  best.  If,  for  this  purpose, 
powdered  purified  animal  charcoal  is  used,  25  grams  of  the  charcoal 
will  be  found  sufficient  for  200  cc.  of  wine.  The  time  the  charcoal 
is  allowed  to  act  varies  from  15  to  60  minutes,  according  to  the 
amount  of  coloring  matters  present.  A  little  sugar  is  also  absorbed 
by  the  charcoal,  but  the  author,  who  has  thoroughly  investigated 
the  matter,  finds  the  amount  never  to  exceed  0  3  per  cent,  under 
the  most  unfavorable  conditions.  In  some  cases,  it  is  advisable  to 
first  dilute  the  sample  with  a  known  volume  of  water  before  attempt- 
ing to  decolorize  with  the  charcoal.  The  author  next  investigated 
the  two  chief  processes  used  for  the  estimation  of  the  tannin.  It 
must  be  borne  in  mind  that  this  acid,  as  it  occurs  in  wines,  is  not  at 
all  a  definite  chemical  compound,  and  is  not  identical  with  gallo- 
tannic  acid.  The  percentage  found  by  analysis  is  therefore  not  the 
true  one,  but  only  expresses  its  equivalent  in  gallo-tannic  acid.  The 
process  which  was  found  to  answer  best  was  that  of  Lowenthal. 
According  to  this  method,  the  solution  containing  the  tannin  is 
largely  diluted  with  water,  mixed  with  a  definite  quantity  of  solution 
of  indigo-carmine,  containing  sulphuric  acid,  and  then  titrated  with 
a  weak  solution  of  potassium  permanganate.  Operating  on  a  known 
quantity  of  tannin,  and  deducting  the  permanganate  necessary  for 
the  oxidation  of  the  indigo  alone,  the  exact  strength  of  the  perman- 
ganate expressed  in  tannin  is,  of  course,  obtained. 
In  applying  the  process  to  wine,  which,  of  course,  contains  many 
other  organic  matters,  also  oxidizable  by  permanganate,  the  author 
proceeds  as  follows :  20  cc.  of  the  sample  is  mixed  with  2  litres  of 
rainwater,  20  cc.  of  solution  of  indigo-carmine  (30  grams  per  litre), 
and  10  cc.  of  sulphuric  acid.  A  solution  of  potassium  permanganate 
(1  :  1000)  which  has  been  carefully  standardized  is  now  run  in  until 
the  liquid  just  loses  its  green,  and  changes  to  a  bright  yellow  color. 
To  obtain  the  amount  of  permanganate  absorbed  by  the  organic 
matter,  50  cc.  of  the  sample  is  mixed  with  100  cc.  of  a  solution  of 
gelatin  (1  :  1000),  and  60  cc.  (  =  20  cc.  of  original  sample)  is  filtered 
off,  and  again  titrated.  The  difference  between  the  two  titrations 
represents  the  true  amount  of  tannin.    The  permanganate  absorbed 
