43Q      A  PROCESS  FOR  ESTIMATING  TANNIC  ACID  IN  GALLS. 
tails  of  another  method,  which  will,  I  think,  be  found  equally 
simple,  more  accurate  and  much  more  convenient  than  can  be 
the  case  with  one  depending  on  hydrometrical  measurement. 
It  must  be  remarked,  however,  that  the  method  now  described 
is  not  applicable  to  all  substances  containing  tannic  acid,  as  its 
indications  with  regard  to  catechu  and  myrobalans  are  not  to 
be  relied  on,  owing,  probably,  to  the  presence  of  some  other 
acid  possessing  the  power  of  precipitating  salts  of  copper. 
Gallic  acid  has  not  this  power,  and  its  presence  is  therefore  no 
obstacle  to  a  satisfactory  experiment.  With  galls  this  process 
gives  accurate  results,  as  I  have  seen  proved  by  careful  com- 
parison with  the  centigrade  gelatine  process. 
The  solution  I  employ  is  one  of  ammonio-sulphate  of  copper,  con. 
taining  in  each  decern  634  grains  of  crystallized  sulphate  of  cop- 
per, and  just  sufficient  liquid  ammonia  to  re-dissolve  the  precipi- 
tate first  occasioned.  Two  septems  of  this  solution  indicate  one 
grain  of  tannic  acid.  It  occasions  an  olive-green  precipitate, 
and  at  the  same  time  deepens  the  color  of  the  liquid  operated 
on.  As  the  experiment  approaches  completion,  the  color  again 
becomes  fainter,  and  with  perfectly  pure  tannic  acid  would  pro- 
bably vanish  altogether. 
It  is  at  first  almost  impossible  to  filter  the  liquid,  but  when 
the  proper  quantity  of  the  copper  solution  has  been  added,  the 
precipitate  is  readily  separated.  No  great  excess  of  copper  can 
be  put  in  unobserved,  owing  to  the  intense  color  of  the  test 
solution.  I  find  it  most  convenient  to  exhaust  the  powdered 
galls  with  water  in  the  proportion  of  a  decern  to  100  grains, 
and,  in  that  case,  if  half  the  liquor  be  operated  on  at  once,  the 
number  of  septems  of  test  solution  employed  directly  indicates 
the  percentage  of  tannic  acid. 
This  method,  while  equally  accurate,  has  several  advantages 
over  the  gelatine  process.  Owing  to  the  color  of  the  precipitate, 
it  is  more  readily  employed  by  artificial  light.  A  smaller 
quantity  of  test  solution  is  required,  for  it  is  impossible  at 
ordinary  temperatures  to  employ  a  gelatine  solution  much 
stronger  than  so  as  to  require  ten  septems  to  precipitate  one 
grain  of  tannic  acid,  consequently  the  alkalimeter  may  require 
re-filling  several  times,  occasioning  loss  of  time  and  liability  to 
error.    Lastly,  the  copper  solution  may  be  kept  for  a  long  time 
