38  Presence  of  Arsenic  in  Chemicals.  {^'J^^m^' 
the  standard  and  the  other  for  the  production  of  octahedral  crystals 
of  arsenous  oxide.  This  precaution  is  essential  to  eliminate  the 
possible  presence  of  such  disturbing  impurities  as  antimony,  mer- 
cury and  selenium. 
When  comparatively  large  deposits  of  antimony  are  heated,  the 
resulting  sublimate  may  contain  octahedral  crystals.1 
Selenium  and  tellurium  have  been  discussed  by  A.  E.  Berry2  and 
O.  Rosenheim.3  Tellurium  probably  has  no  influence  on  the  Marsh- 
Berzelius  test,  but  there  seems  to  be  much  uncertainty  at  present 
about  the  influence  exerted  by  the  selenium.  The  disturbing  influ- 
ence of  this  element  can  be  eliminated,  like  the  sulphur  compounds, 
by  placing  a  wad  of  cotton,  moistened  with  lead  acetate,  in  the  fore- 
part of  the  calcium-chloride  tube. 
The  arsenical  deposit  occurs  from  I  to  2  centimetres  back  of  the 
flame,  which  is  so  placed  that  the  greater  portion  of  the  mirror  is 
in  the  constriction  of  the  tube.  When  the  deposit  is  small,  a  fine 
brownish  mirror  results,  but  large  deposits  are  arranged  in  three 
rings.  The  ring  toward  the  flame  is  brown  and  semi-transparent. 
The  middle  portion  is  dense  and  almost  black,  while  the  outer  por- 
tion is  grayish  and  diffusing. 
The  character  and  formation  of  the  mirrors  are  influenced  by  cer- 
tain elements  like  antimony,  mercury  and  selenium. 
As  above  stated,  the  hydrogen  was  allowed  to  evolve  slowly  and 
pass  through  the  apparatus  for  thirty  minutes  before  the  substance 
to  be  tested  was  introduced.  The  flame  during  this  time  was  placed 
just  before  the  outer  constriction;  and  if  any  arsenic  could  possibly 
have  been  present  in  any  portion  of  the  zinc,  sulphuric  acid,  etc.,  it 
would  have  been  revealed  by  this  precaution.  The  test,  however, 
was  invariably  negative.  The  flame  was  then  placed  just  before  the 
inner  constriction  and  the  substance,  or  solution  of  the  substance  to 
be  tested,  introduced  into  the  flask,  little  by  little,  through  the  sepa- 
ratory  funnel.  It  is  generally  necessary  to  add  the  substance  to  be 
tested  slowly,  because  the  presence  of  arsenic  seems  to  augment  the 
reaction  very  materially.  In  most  cases  where  arsenic  was  present* 
it  was  revealed  in  from  five  to  eight  minutes,  and  all  seemed  to  have 
been  evolved  in  from  fifteen  to  twenty  minutes.    The  test,  however, 
1  Wormley,  1877,  Am.  Jour.  Med.  Science,  74,  399. 
2 1901,  Jour.  Soc.  Chem.  hid.,  20,  322. 
a  1901,  Chem.  News,  83,  280. 
