^ZrtilTo]  Assay  of  Iron  Filings,  245 
is  not  so  important  as  in  ferrum  redactum,  which  depends  for  its 
medicinal  properties  on  the  metallic  iron  it  contains.  It  is  necessary 
however,  that  it  be  nearly  pure,  as  the  presence  of  much  oxide  would 
seriously  alter  the  iron  content  of  the  preparation  if  allowance  were 
not  made.  An  estimation  of  its  iron  content,  both  free  and  combined, 
is  therefore  necessary. 
Iron  filings  are  official  in  the  German  Pharmacopoeia  as  ferrum 
pulveratum,  but  the  assay  process  therein  given  estimates  the  total 
iron  only,  whether  free  or  in  any  state  of  oxidation.  The  proportion 
of  free  metallic  iron  may  be  estimated  by  the  official  B.  P.  method 
for  ferrum  redactum,  by  allowing  the  filings  to  react  with  copper  sul- 
phate solution  and  titrating  the  ferrous  sulphate  thus  formed  with 
potassium  permanganate.  In  using  this  process  for  ferrum  redactum, 
it  has  been  pointed  out  by  Peck  {Y ear-Book  of  Pharmacy,  1898,  p. 
399),  that  it  is  difficult  to  secure  accurate  or  even  concordant  results 
unless  certain  experimental  details  be  adhered  to.  This  lack  of 
concordance  is  due  chiefly  to  the  incompleteness  of  the  reaction, 
in  varying  degree,  due  to  the  formation  of  a  protective  coating  of 
metallic  copper  on  the  iron  particles,  and  to  the  massing  together  of 
the  particles,  preventing  the  reaction  proceeding  with  sufficient 
rapidity  to  be  completed  in  the  specified  time.  These  objections 
apply  to  filings  also,  with  greater  force  if  they  are  somewhat  coarse, 
and  if  care  is  not  taken  results  may  be  obtained  which  are  below  the 
truth  or  above  for  a  reason  to  be  discussed.  Peck  directs  that  the 
copper  sulphate  solution  be  added  cold  to  the  iron,  the  solution  heated 
in  a  flask  with  Bunsen  valve  for  half  an  hour,  then  rapidly  filtered  and 
titrated.  In  assaying  a  number  of  samples  of  iron  filings  by  the 
B.  P.  process  another  complication  made  itself  evident,  and  the  results 
obtained  were  not  only  inconcordant,  but  in  some  cases  impossibly 
high,  being  several  figures  over  100  per  cent.  Using  Peck's  modifi- 
cation, similar  results  were  obtained.  This  was  not  explainable  for 
some  time,  till  on  one  occasion  a  smell  of  phosphine  was  noticed  in 
the  hot,  unacidified  solution.  The  odor  was  like  that  of  acetylene 
from  calcium  carbide.  This  latter  smell  is  really  due  to  traces  of 
phosphine,  pure  acetylene  being  odorless.  The  presence  of  phosphine 
suggests  the  presence  or  formation  of  hydrogen,  either  as  occluded 
hydrogen  or  formed  by  galvanic  action  between  the  copper  and  the 
iron,  and  combing  while  nascent  with  the  phosphorus  in  the  iron. 
The  latter  explanation  is  the  more  probable.  As  phosphine,  PH3, 
reacts  with  potassium  permanganate,  KMn04,  according  to  the 
