32  Presence  of  Arsenic  in  Chemicals.    {A  j;nuarry.Pioo.3.m' 
The  method  next  important  to  Marsh's  is  undoubtedly  Reinsch's,1 
in  which  metallic  copper  is  employed  to  deposit  the  arsenic,  which 
it  is  claimed  forms  a  definite  chemical  compound  with  the  copper 
Cu5  As2.  Fleitman's2  method  is  practically  a  modification  of  Marsh's 
test,  in  which  the  hydrogen  is  generated  by  means  ot  an  alkaline 
solution,  acting  on  metallic  aluminum  or  zinc.  By  BettendorfT's3 
method,  the  arsenic  is  deposited  as  a  metal,  in  a  stannous  chloride 
solution,  upon  metallic  tin.  Gutzeit4  utilized  the  well-known  reac- 
tions of  arsine  on  paper  moistened  with  a  solution  of  mercuric  chlo- 
ride, or  acidulated  silver  nitrate. 
The  last  three  methods  are  the  ones  employed  by  the  U.S.P.  to 
detect  arsenic  in  the  compounds  recognized  by  it.  The  object  in 
employing  these  methods  was  primarily  because  they  were  compar- 
atively simple  and  at  the  same  time  gave  fairly  accurate  results- 
When  it  is  a  question  of  testing  for  the  presence  of  arsenic  in  any 
compound,  nothing  short  of  the  very  best  available  method  should 
be  applied  ;  inasmuch  as  it  so  frequently  happens  with  rapid  meth- 
ods, they  give  inaccurate  reactions  and  leave  the  worker  in  doubt. 
For  example  :  Gutzeit's  test  is  evidently  superior  in  point  of  delicacy 
to  even  Marsh's  or  Reinsch's,  but  depending  on  the  formation  of  a 
yellow  compound,  by  the  action  ot  arsine  on  mercuric  chloride  or 
acidulated  silver-nitrate  solution  the  results  are  ambiguous,  because 
of  the  fact  that  very  minute  traces  of  hydrogen  sulphide  or  phos- 
phine  produce  a  stain  similar  to  the  one  produced  by  arsenuretted 
hydrogen,  and  there  is  no  means  of  distinguishing  between  them. 
It  is  the  experience  of  chemists  in  general  that  the  Marsh  test 
modified  by  Berzelius,  and  the  Reinsch  method  are  the  most  reliable, 
some  being  in  favor  of  one  method  and  others  in  favor  of  the  other. 
After  considering  the  various  methods  in  detail  and  making  some 
experiments  with  the  same,  it  was  decided  to  compare  BettendorfT's 
methods  with  the  two  above. 
Before  attempting  to  apply  the  tests,  it  is  important  to  ascertain 
whether  the  chemicals  and  every  part  of  the  apparatus  to  be  em- 
ployed are  relatively  free  from  arsenic  or  other  interfering  agents- 
The  word  "  relatively "  is  used  because  chemists  realize  that  it  is 
1 1842, /cwr.  de  Pharm.,  2,  361. 
2  1851,  Ann.  (Liebig),  72,  126. 
3 1869,  Ztsch.  fur  Chemie,  12,  492  ;  Wittsteiri 's  Vierteljahrschr.,  1870,  430. 
*  1879,  Pharm.  Ztg.,  p.  263. 
