COMMERCIAL  GLACIAL  PHOSPHORIC  ACID.  195 
When  the  glacial  phosphoric  acid,  dissolved  in  cold  water,  is 
Immediately  oversaturated  with  ammonia,  filtered  from  the 
precipitate,  and  mixed  with  a  solution  of  chloride  of  ammonium 
and  magnesium,  a  precipitate  of  ammonio-phosphate  of  magne- 
s'ia  occurs  only  of  that  portion  of  the  phosphoric  acid  which  has, 
by  the  action  of  a  moist  atmosphere  and  the  consequent  gradual 
absorption  of  water,  been  converted  into  the  terhydrate  or  the 
common  phosphoric  acid ;  an  excess  of  the  double  chloride  pro- 
duces in  the  filtrate  no  further  precipitate,  until  it  is  heated  to 
boiling.  By  setting  the  solution  aside,  the  meta-  or  pyrophos- 
phoric  acid  is  gradually  converted  into  the  common  tribasic  acid, 
but  this  change  is  effected  in  a  short  time  on  the  application  of 
heat.  To  determine  whether  all  the  acid  has  been  changed  in 
this  manner,  it  will  suffice  to  add  more  of  the  double  chloride  to 
the  ammoniacal  solution  and  boil  again;  this  must  be  continued 
until  a  precipitate  ceases  to  be  separated. 
But  inasmuch  as  meta  and  pyrophosphate  of  lime  and  mag- 
nesia are  kept  in  solution  by  the  corresponding  phosphate  of 
ammonia,  it  will  be  necessary  to  boil  the  original  solution  of  the 
acid  for  some  time,  allow  it  to  cool  and  supersaturate  it  with 
ammonia.  The  filtrate  must  then  be  repeatedly  boiled,  and  some 
ammonia  must  be  added  to  the  cooled  liquor  until  no  further 
precipitate  occurs.  We  may  then  be  sure  that  all  the  earthy 
phosphates  have  been  separated.  The  combined  precipitates  are 
now  exposed  to  heat  until  they  cease  to  lose  weight ;  their  weight 
is  then  deducted  from  the  original  weight  of  the  phosphoric  acid, 
and  the  remainder  may  be  looked  upon  as  the  weight  of  pure 
hydrated  phosphoric  acid,  provided  that  in  the  beginning  we 
have  satisfied  ourselves  of  the  absence  of  sulphuric  acid  and 
ammonia.  Though  many  works  on  chemistry  recommend  to 
test  for  nitric  and  hydrochloric  acids,  I  have  never,  either  on 
any  former  or  the  present  occasion  found  either  of  these  two;  but 
sulphuric  acid,  I  have  in  one  or  two  cases  met  with,  not  how- 
ever in  Merck's. 
The  ammoniacal  solution  of  phosphoric  acid  is  then  precipitated 
with  a  solution  of  chloride  of  ammonium  and  magnesium,  and  after 
standing  for  several  hours,  filtered.  The  precipitate  is  washed 
with  a  diluted  solution  of  ammonia,  until  the  washings,  when 
oversaturated  with  pure  nitric  acid,  produce  no  precipitate  in 
nitrate  of  silver. 
