30 
OX  A  NEW  METHOD  OF  ALKALIMETRY. 
by  its  presence  on  a  solution  of  litmus,  but  owing  to  the  diminished 
sensibility  of  litmus  thus  tinted. 
In  the  hope  of  avoiding  those  sources  of  error  to  which  I  have 
alluded,  and  of  facilitating  the  determination  of  the  per  centage 
of  alkali,  I  adopted  a  method  of  alkalimetry  which  in  substance 
is  the  following : — To  the  alkali  under  examination  is  added  a 
known  excess  of  a  normal  solution  of  oxalic  acid  ;  and  after  the 
expulsion  of  carbonic  acid  from  the  solution  by  boiling,  the  excess 
of  oxalic  acid  remaining  is  determined  by  means  of  a  standard 
solution  of  ammonia. 
Some  objection  might  be  made  to  the  employment  of  a  solution 
of  ammonia ;  but  it  will  be  found  that  a  dilute  solution  of 
ammonia  ;  if  kept  in  properly  constructed  apparatus,  will  remain 
more  constant  than  might  be  expected.  The  standard  solutions 
of  ammonia  and  of  oxalic  acid  are  most  easily  prepared  by  means 
of  a  standard  solution  of  sulphuric  acid,  care  being  taken  that 
perfectly  pure  acid  be  employed,  and  that  the  amount  of  real 
acid  be  carefully  determined. 
Having  prepared  the  standard  solutions  of  a  desired  strength, 
the  determination  of  an  alkali  or  of  an  alkaline  carbonate  may 
be  thus  effected  :— 10  grs.  of  an  alkaline  carbonate,  carbonate 
of  soda  for  example,  after  having  been  placed  in  a  flask,  a  solu- 
tion of  oxalic  acid  corresponding  to  10  grs.  of  pure  carbonate  of 
soda  is  added ;  the  solution  is  then  boiled  until  the  expulsion  of 
carbonic  acid  be  effected,  when  the  solution  is  diluted  with  dis- 
tilled water ;  and  after  the  addition  of  a  few  drops  of  a  solution 
of  litmus,  the  excess  of  oxalic  acid  is  determined  by  a  standard 
solution  of  ammonia.  The  excess  of  oxalic  acid  remaining  will 
of  course  indicate  the  impurities  present,  or  the  absence  of  alkali, 
which,  by  deduction  from  the  quantity  originally  taken,  will  give 
the  amount  of  available  alkali. 
Care  must  be  taken  that  the  solution  be  only  tinted  with  litmus, 
and  not  too  deeply  colored,  as  the  more  feeble  the  coloration 
within  certain  limits,  the  more  easily  detected  is  the  change  of 
tint  produced  by  an  excess  of  alkali  or  acid. 
I  have  found  it  necessary  to  employ  distilled  water,  failing,  as 
I  have  done,  to  obtain  accurate  results  with  other  water,  owing 
to  the  presence  of  carbonic  acid.  The  preparation  of  standard 
solutions  is  so  well  understood,  that  it  is  unnecessary  to  enter 
into  further  details. 
