AmDe°cU"im.arm'}  Ferric  Phosphate  and  Pyrophosphate.  587 
As  for  other  phosphates,  solution  of  ammonium  molybdate  in 
nitric  acid  forms  an  excellent  reagent  for  quickly  demonstrating  the 
presence  of  phosphoric  acid  in  such  a  scale  salt.  The  amount  of 
citrate  in  the  samples  tested  did  not  interfere  with  the  reaction. 
The  yellow  precipitate  must  appear  in  a  few  moments  if  phosphate 
is  present.  By  taking  the  same  amount  of  the  solutions  of  definite 
strength,  when  applying  the  test,  a  practised  eye  can  see  by  the 
quantity  of  precipitate  after  a  few  minutes  whether  a  true  phos- 
phate is  present  or  phosphoric  acid  only  as  a  slight  contamination 
of  a  pyrophosphate.  In  the  few  minutes  the  test  requires  a  solu- 
tion of  pure  pyrophosphate  of  sodium  acidified  with  nitric  acid  gave 
no  precipitate  with  the  molybdanic  solution.  This  test  of  the  salts 
under  consideration  is  shorter  than  the  U.  S.  P.  test  and  is  better 
because  citric  acid  does  not  give  any  reaction  with  the  reagent 
used.  But  it  suffers  from  the  same  defect  inasmuch  as  the  presence 
of  pyrophosphate  is  also  here  demonstrated  only  by  the  negative 
result,  viz  :  The  absence  of  any  considerable  quantity  of  phos- 
phates. 
By  careful  addition  of  acid  to  the  aqueous  solution  of  these  scale 
salts,  real  iron  phosphate  and  pyrophosphate  are  precipitated. 
This  reaction  could  serve  to  remove  the  citric  acid  before  making 
the  silver  test.  It  was  found,  however,  that  the  precipitates  are 
very  difficult  to  filter  and  wash.  Even  after  thorough  washing,  an 
examination  showed  that  they  still  held  back  citrates,  ether  chemi- 
cally or  mechanically.    So  this  reaction  proved  unavailable. 
The  separation  from  the  citric  acid  by  means  of  a  calcium  salt 
was  tried  briefly,  but  proved  too  unpromising  for  me  to  proceed 
further  in  this  direction. 
The  test  involving  the  use  of  excess  of  magnesium  sulphate  for 
separating  the  acids  proved  satisfactory ;  I  obtained  good  results 
by  adding  to  the  alkaline  solution  of  the  acids,  called  "  A  "  below, 
as  obtained  from  the  scale  salts,  an  excess  of  ammonium  chloride, 
and  pouring  the  fluid  into  quite  a  large  excess  of  concentrated 
magnesium  sulphate  solution  (33  per  cent.)  The  magnesium  am- 
monium phosphate,  precipitated  in  the  cold,  was  identified  further, 
as  below,  by  being  converted  into  silver  phosphate.  In  case  iron 
pyrophosphate  was  present,  even  with  some  iron  phosphate,  the 
filtrate  from  the  precipitate  of  Mg(NH4)P04  yielded  magnesium 
pyrophosphate  by  boiling.    The  latter  was  identified  by  its  silver 
