Am  jour.  Pharm. ) 
Dec,  1876.  ( 
Analysis  of  Tea. 
559 
solution  carefully  decanted  ;  the  residue  boiled  again  with  the  same 
quantity  of  water  and  the  clear  solution  decanted  ;  the  residue  then 
dried  on  the  water-bath  and  weighed,  the  loss  being  taken  as  the 
quantity  sought. 
The  determination  of  the  tannin  in  the  decoction  was  tried  by  four  of 
the  numerous  methods  in  use,  without  obtaining  very  nearly  concur- 
rent results  from  triplicate  operations  with  each  method.  The  method 
recommended  by  A.  H.  Allen,1 — precipitation  by  standard  solution  of 
lead  acetate,  in  presence  of  potassium  ferricyanide  as  an  indicator,  was 
used  for  No.  1, — 19  per  cent,  being  the  mean  of  three  estimations, 
with  a  variation  of  nearly  5  per  cent.  The  diffculty  consisted  in  per- 
ceiving the  pink  color  made  with  the  indicator,  masked  as  it  is  by  the 
color  of  the  tea  decoction.  The  result  given  for  No.  2  is  the  average 
of  three  varying  determinations  by  Wagner's  method  with  a  standard 
solution  of  cinchonia.2 
The  variation  was  due  to  the  difficulty  in  recognizing  the  red  color 
of  the  indicator  in  the  solution,  distinct  from  the  colored  precipitate,  in 
presence  of  the  tea  color.  No.  3  is  given  in  the  table  as  the  average 
of  three  trials  of  the  method  of  Hager,3  by  the  diminution  of  density 
from  removal  of  the  tannin  by  copper  oxide.  The  results  were  widely 
divergent — showing  that  this  method  which  was  proposed  for  oak-bark 
is  not  applicable  to  tea.  No.  6  was  worked  volumetrically  by  solution 
of  gelatin  (MUller's  method)  ;  three  determinations  giving  consider- 
able differences. 
A  volumetric  method  for  tannin  was  then  tried,  with  use  of  standard 
solution  of  cinchonia  sulphate  in  excess,  titrating  back  in  the  filtrate  with 
standard  solution  of  potassium  mercuric  iodide.  Triplicate  operations 
were  tried  on  No.  4  and  No.  5 — the  results  for  each  number  differing 
but  slightly.  The  remaining  numbers  were  worked  by  the  same 
method:  Nos.  10  and  12  being  confirmed  by  closely  agreeing  dupli- 
cate results.  The  solution  of  cinchonia  sulphate  is  Wagner's  standard 
(before  quoted) :  4*523  grams  cinchonia  sulphate,  with  0*5  gram  sul- 
phuric acid,  in  water  to  make  one  litre.  The  potassium  mercuric 
iodide  solution  is  Mayer's  standard  :  13*546  grams  mercuric  chloride 
1 "  Chem.  News,"  xxix.,  169,  189. 
2"Zeitsch.  f.  anal.  Chem.,"  v.,  4,  8  \  vii.,  139. 
3  "  Untersuchungen,"  II.,  115. 
