AmAp?u;S£rm"}    Nitrates  and  Ammonia  in  Water.  161 
the  later  distillates,  thus  avoiding  a  possible  precipitate,  and  saving 
the  analysis.  Total  ammonia  may  be  determined  in  several  succes- 
sive samples  of  water  if  the  quantity  of  alkaline  permanganate  is 
increased  (10  c.c.  additional  for  every  sample  of  water);  if  500  ex. 
of  any  sample  (as  is  generally  recommended)  are  to  be  used  for  a 
single  determination  a  corresponding  quantity  of  permanganate 
must  be  used. 
After  determining  the  total  ammoniat  the  retort  is  rinsed  with 
several  portions  of  water  and  again  charged  with  100  c.c.  of  the 
sample,  to  determine  the  free  ammonia;  instead  of  adding  a  solu- 
tion of  sodium  carbonate,  I  have  always  fused  a  little  of  the  dried 
salt  on  platinum  foil,  and  put  the  foil,  with  the  fused  salt,  into  the 
retort.  The  distillate  is  collected  in  four  or  five  portions  of  10  c.c. 
each  and  nesslerized  as  previously  stated.  The  difference  between 
the  two  determinations  gives  the  albuminoid  ammonia. 
In  matching  the  colors  produced  in  the  distillates  with  the  aid 
of  a  dilute  ammonium  chloride  solution,  it  was  noticed  on  several 
occasions  that  the  color  produced  by  the  ammonium  chloride  solu- 
tion was  not  as  intense  as  experience  led  me  to  expect ;  on  making 
a  fresh  solution,  and  testing  this  with  the  reagents  used  in  the  for- 
mer test,  a  much  more  intense  coloration  was  produced,  showing 
that  the  dilute  ammonium  chloride  solution  will  not  keep  indefinitely ', 
but  must  be  occasionally  compared  with  or  replaced  by  a  fresh  solu- 
tion. One  of  these  deteriorated  ammonium  chloride  solutions  gave 
a  very  satisfactory  test  for  nitric  acid,  indicating  the  oxidation  of 
the  nitrogen  originally  present  as  ammonia. 
The  suggestion  has  been  made  in  processes  in  which  500  c.c. 
water  are  used  for  the  ammonia  determination  to  match  the  color 
by  using  a  piece  of  amber  glass,  the  value  of  which  has  been  ascer- 
tained; this  I  have  not  found  satisfactory  when  only  100  c.c.  of 
water  are  used. 
A  large  number  of  experiments  have  been  made  to  obtain  a  per- 
manent solution  with  which  the  colors  obtained  in  the  nesslerized 
distillates  can  be  imitated  and  the  ammonia  estimated ;  success  in 
this  will  mean  the  saving  of  much  time  in  the  determinations,  as, 
after  a  set  of  solutions,  representing  various  quantities  of  ammonia, 
is  once  prepared,  there  will  be  no  occasion  for  the  preparation  of 
ammonia-free  water. 
