4 1 6  Methods  of  Estimating  Tannins.      { Am  '^;S^T' 
This  is  described  in  the  last  edition  of  Sutton,  but  in  its  original  form 
is  quite  inadmissible,  since  lead  precipitates  gallic  acid  as  well  as  tannin, 
and  both  react  equally  on  the  indicator.  In  combination  with  some  of 
the  differential  processes  in  which  the  tannin  is  removed  by  gelatin  or 
hide  raspings,  it  may  no  doubt  give  useful  results,  and  as  the  lead  com- 
pounds of  the  different  tannins  are  better  known  than  most  others,  pos- 
sibly factors  might  be  calculated  to  give  percentage  results.  I  cannot 
insist  too  strongly  that  any  calculation  of  percentages  by  comparison 
with  "pure"  tannin  is  utterly  fallacious,  both  because  the  various  tan- 
nins are  of  totally  different  constitution,  and  because  really  pure  tannin 
is  quite  unattainable.  That  met  with  in  commerce  only  contains  80 
to  90  per  cent,  of  really  pure  tannin,  and  is  very  variable. 
Mr.  Stoddart  uses  Mr.  Allen's  process  in  conjunction  with  absorp- 
tion of  tannin  with  hide  raspings,  when  of  course  the  loss  is  proportional 
to  the  tannin.  He  also  employs  Nelson's  gelatin  swollen  in  cold  water 
as  an  absorbent  in  the  same  manner.  Time  and  patience  are  necessary 
for  the  absorption  of  tannin  thus,  and  it  is  seldom  so  complete  that  the 
results  are  not  altered  by  prolonged  digestion.  In  my  experience  the 
end-reaction  of  Allen's  method  is  not  very  distinct,  and  it  is  necessary 
carefully  to  filter  the  drops  tested,  as  the  indicator  is  affected  by  the 
precipitate.    This  makes  the  process  somewhat  tedious. 
The  remaining  methods  which  I  shall  describe  are  all  based  on  the 
oxidation  of  tannin  by  various  agents,  and  all  involve  double  analyses 
after  absorption  of  the  tannin,  as  tannin  and  gallic  acid  are  almost 
identical  in  their  behavior  with  oxidizers. 
Mittenzwey,  and  afterwards  Terreil,  proposed  to  estimate  it  by  the 
direct  absorption  of  atmospheric  oxygen  in  alkaline  solution — a  difficult 
and  tedious  proceeding,  though  doubtless  capable  of  some  accuracy  in 
skillful  hands. 
Monnier  proposed  to  determine  with  permanganate  direct,  but  this 
proved  quite  impracticable,  since  the  oxidation  is  rapid  at  first,  and  then 
slow  and  with  no  definite  termination. 
To  Dr.  Lbwenthal  is  due  the  capital  improvement,  which,  with  his 
recent  additions,  constitutes  to  my  mind  the  most  practical  method  of 
tannin  analysis  yet  discovered.  He  adds  to  the  very  dilute  tannin  infu- 
sion a  considerable  quantity  of  indigo,  not  only  to  act  as  an  indicator, 
but  to  control  the  oxidation  of  the  tannin.  This  reaction  is  both  rapid 
and  accurate,  and,  combined  with  his  process  of  precipitation  by  gelatin, 
