AmAugU,r,x87h7!rm' }      Methods  of  Estimating  Tannins.  4 1 5 
actually  combine  with  copper  in  the  same  proportions.  In  fact,  as  we 
shall  see  later  on,  the  different  tannins  differ  notably  in  their  properties 
and  reactions,  only  agreeing  in  their  power  of  precipitating  gelatin,  and 
I  fear  tanners  will  have  to  give  up  all  hope  of  measuring  them  by  one 
common  standard.  Indeed,  to  a  chemist,  to  do  so  seems  about  as 
reasonable  as  to  compare  the  values  of  nitric  and  sulphuric  acids  by  a 
standard  solution  of  hydrochloric  acid.  Probably  the  differences  be- 
tween gallotannic,  quercitannic  and  catechutannic  acids  are  quite  as 
great  as  those  of  the  mineral  acids  I  have  named.  Chemists  may 
fairly  undertake  to  compare  sumach  with  sumach,  or  bark  with  bark, 
but  the  relative  values  of  the  tannins  of  bark  and  sumach  are  commer- 
cial matters  which  no  analysis  can  decide,  though  it  doubtless  might  be 
done  by  carefully  conducted  technical  experiments. 
The  disadvantages  of  the  copper  method  are  that  it  is  slow,  trouble- 
some and  difficult,  and  that  the  washing  and  drying  must  be  rapidly  and 
carefully  done,  as  the  precipitate  is  easily  decomposed.  This  difficulty 
might  be  overcome  by  igniting  and  weighing  the  CuO,  but  this  can 
only  be  done  easily  in  oxygen,  as  otherwise  the  copper  is  so  much 
reduced  that  it  is  apt  to  deflagrate  with  nitric  acid  or  ammonic  nitrate. 
I  think  the  best  way  is  to  filter  on  a  vacuum  filter,  and  dry  in  an  air- 
bath  (with  a  thermostat)  at  ioo°. 
The  mean  error  of  such  result  in  a  series  of  eight  analyses  of  com- 
mercially "pure"  tannin,  containing  apparently  about  85  per  cent, 
digallic  acid  by  the  total  employed,  was  only  +  4*2  per  cent.,  a  much 
better  approximation  than  any  of  the  foregoing.  It  is  not  likely  that 
the  results  with  tanning  materials  would  be  quite  so  good.  Analyses 
of  bark  showed  considerable  divergence,  and  combustion  of  the  precip- 
itate proved  that  it  was  somewhat  inconstant  in  composition,  the  Cu 
varying  from  21*6  to  25*4  per  cent.  I  fancy,  too,  that  for  oak  bark 
tannin  two-thirds  of  the  weight  of  precipitate  is  decidedly  too  high  an 
estimate.  It  must  also  be  borne  in  mind  that  if  lime  be  present,  as  is 
often  the  case  with  tanyard  liquors,  it  will  be  precipitated  as  carbonate. 
This  might  be  prevented  by  filtering  off  the  precipitate  before  washing 
with  ammonic  carbonate  ;  but  the  method  is  troublesome  enough  with- 
out this,  besides  being  of  rather  questionable  accuracy. 
Another  process  which  has  been  much  recommended  is  Mr.  A.  H. 
Allen's  volumetric  one,  with  a  standard  solution  of  acetate  of  lead, 
using  as  an  indicator  a  mixture  of  ammonia  and  potassic  ferricyanide. 
