AmFebu,ri8P77arm }        Wash-Blue  and  its  Analysis.  67 
reaction  AgCy  ,*  boiling  with  a  drop  of  amnionic  sulphide  and  addition 
•of  a  drop  of  ferric  chloride  Fe2Cl6 — red  color).  As  it  takes  a  very 
low  heat  to  decompose  Berlin  blue,  said  aqueous  extract  sometimes 
contains  undecomposed  potassic  ferrocyanide.  Test  with  ferric  chlo- 
ride. 
Pure  Berlin  blue,  ferric  ferrocyanide,  (FeCy6)Fe4-h  18 HzO,  loses,  at 
2120  F.,  7*22  per  cent,  of  water,  a  fact  the  manufacturers  are  well 
aware  of,  and  therefore  never  allow  the  temperature  of  the  "  blue  dry- 
ing room"  to  exceed  100  to  120°  F. 
On  ignition,  depending  on  the  intensity  of  heat,  varying  mixtures  of 
ferric  and  magnetic  oxide  remain  as  residue.  The  loss  on  ignition, 
therefore,  does  not  allow  any  conclusions  on  the  quantity  of  blue  pres- 
ent (for  instance,  in  a  mineral  color),  and  an  estimation  of  the  ferric 
oxide,  either  by  titration  or  weight-analysis  will  always  be  necessary. 
It  is  evident  that  this  method  would  become  incorrect  if  applied  in  the 
quantitative  analysis  of  a  blue  rendered  soluble  by  potassic  ferrocya- 
nide ;  it,  however,  can  be  successfully  followed  in  the  estimation  of 
oxalic  acid  blues."    Wash-blue  is  hardly  ever  adulterated. 
In  order  to  ascertain  the  quantitative  relation  between  the  ferric 
oxide  and  the  total  quantity  of  blue  I  prepared  a  pure  sample  by  the 
same  process  as  done  on  a  large  scale,  and  dried  it  slowly  at  8o°  F. 
It  corresponded  to  the  formula  2[(FeC6N6)34Fe+i8H20]==2368, 
which  will  yield  7Fe203=i  120.  100  blue  correspond  to  47*290  per 
cent,  oxide  of  iron. 
i*oo  gram  of  this  pure  ferrocyanide  of  iron  lost  on  ignition  0*6058 
=  60-58  per  cent,  of  its  weight.  Actual  residue  ==  0*3942  ==  39*42 
per  cent.  This  residue  then  was  dissolved,  the  Fe  reprecipitated  as 
Fe203,3H20,  and  finally  weighed  as  Fe203  =  0*4750  =  47*50  per 
cent.,  nearly  the  theoretical  amount. 
From  these  data  the  quantity  of  pure  blue  can  be  calculated.  I 
have,  however,  found  that  the  pure  "  commercial  blues  "  dried  at  a 
somewhat  higher  temperature,  contain  less  than  18  H20,  yielding 
49*75  to  50  per  cent,  of  Fe203,  and  consider  it  "practically  correct" 
to  multiply  the  quantity  of  Fe2Os  found,  by  two,  in  order  to  find  the 
quantity  of  "  commercial  blue  "  present. 
The  numerous  other  blue  pigments  of  this  series,  as,  for  instance, 
TurnbuWs  blue  (ferrous  ferricyanide,  (Fe2C12N12)Fe3),  chromate  blue,  so 
called  from  potassic  bichromate  and  sulphuric  acid  forming  the  oxidiz- 
