Am.  Jour.  Pharm. ) 
July,  1885.  i 
Commercial  Tannin . 
345 
position  during  the  process  of  manipulation.  These  disadvantages  are, 
if  not  entirely  obviated,  at  all  events  lessened  in  the  method  by  Lowen- 
thal,  and  as  this,  in  tlie  present  state  of  our  knowledge,  may  be  regarded 
as  the  most  satisfactory,  I  have  employed  it  in  the  estimation  of  the 
tannic  acid. 
The  Avorking  details  of  the  process  are  given  very  fully  by  Sutton, 
and  for  these  it  is  only  necessary  for  me  to  refer  to  his  standard  work. 
That  there  are  difficulties  in  the  actual  carrying  out  of  the  process  is 
undeniable,  the  most  formidable  of  these  being  the  uncertainty  of  the 
end-reaction  and  the  doubt  as  to  the  exact  amount  of  correction  required 
for  the  gelatin.  Sutton  directs  the  permanganate  to  be  delivered  in 
very  slowly  with  constant  stirring  until  a  faint  rose  color  appears 
round  the  edges  of  the  liquid,'^  while  DragendorfP  says  the  solution  is 
to  be  run  in  till  the  blue  color  changes  to  green."  It  will  be  found 
in  actual  practice  that  the  rose  color  is  considerably  distant  from  the 
green,  and  as  everything  depends  on  knowing  exactly  when  to  stop  it 
must  be  admitted  that  in  this  res})ect  the  process  is  not  yet  |)erfect.  I 
found  that  it  was  necessary,  in  order  to  get  good  comparative  results, 
to  run  in  permanganate  solution  till  the  color,  a  faint  yellow,  was  as 
nearly  as  possible  identical  with  that  of  a  standard  solution  kept  side 
by  side  with  the  liquid  beiug  titrated.  Auth.orities  also  differ  as  to  the 
correction  required  for  the  gelatin.  Proctor,  quoted  by  Sutton,  says 
that  probably  the  nearest  approach  to  the  truth  may  be  obtained  by 
deducting  half  the  correction  f<  .r  gelatin  from  the  oxidizable  matter 
other  than  tannin,"  Avhereas,  according  to  DragendorfP,  the  addition 
of  gelatin  as  described  by  Lowenthal  introduces  only  a  slight  source 
of  error  which  may  be  generally  neglected."  As  to  the  latter  state- 
ment, it  is  obvious  that  if  only  comparative  results  are  required,  the 
error  may  be  neglected,  provided  always  that  a  carefully  standardized 
gelatin  solution  be  employed,  but  if  even  approximately  correct  figures 
are  wanted,  it  is  absolutely  necessary  that  the  correction  for  gelatin  be 
made.  lam  also  decidedly  of  opinion  that  Proctor  is  within  the  mark 
when  he  deducts  only  half  the  gelatin  correction,  and  prefer  rather  to 
go  to  the  other  extreme  by  deducting  the  whole.  This  I  have  done 
in  these  estimations,  and  consequently  it  is  just  possible  that  my  results 
as  regards  the  percentage  of  tannic  acid  may  be  rather  high. 
Having  ascertained  the  quantity  of  permanganate  required  for 
oxidation,  the  calculation  of  the  results  is  readily  accomplished.  For 
this  purpose  I  have  used  the  coefficients  of  Neubauer  and  Oser,  as  given 
