^S&St""}     Report  of  Research  Committee  B.  3^7 
2-5  cc.  of  SnG2 :  reaction  begins  in  15  minutes  at  air  tempera- 
ture, in  less  than  5  minutes  when  heated. 
3  0  cc.  of  SnCl2:  reaction  slightly  less  intense  than  with  2-5  cc, 
both  hot  and  cold. 
3  5  cc.  of  reagent:  a  still  further  slight  decrease  in  intensity. 
No  difference  could  be  observed  in  the  reaction  of  As2Os  and 
As205;  the  sodium  metarsenite  appeared  to  be  very  slightly  more 
colored  than  As2Os. 
(b)  With  1  cc.  of  U.  S.  P.  solution  and  a  small  piece  of  metallic 
tin,  at  a  temperature  of  about  8o°  C,  1  cc.  of  the  solutions  contain- 
ing 0  05  mgm.  of  As  began  showing  a  brownish  color  at  13  minutes. 
At  35  minutes  the  reaction  was  quite  sharp  and  continued  to  grow 
slowly  in  intensity  to  the  end  of  the  hour. 
No  difference  was  perceptible  between  As203  and  As205. 
(c)  3  cc.  of  the  solution  of  the  "  Arzneibuch."  mixed  with  I  cc. 
of  the  solutions  containing  0-05  mgm.  of  As  began  showing  a  feeble 
reaction  at  35  minutes,  and  at  the  end  of  the  hour  showed  a  plain 
reaction,  as  compared  with  the  unchanged  solution,  but  did  not 
equal  in  intensity  the  reaction  produced  by  the  addition  of  metallic 
tin. 
On  repeating  the  experiments  with  solutions  containing  0-03 
mgm.  of  As  in  1  cc.  a  feeble  reaction  resulted  from  the  use  of  stan- 
nous chlorid  and  metallic  tin  at  the  end  of  an  hour,  but  even  after 
standing  over  night  the  other  methods  gave  such  a  faint  reaction 
that  it  required  close  inspection  in  the  colorimeter  to  perceive  the 
change.  A  solution  containing  002  mgm.  of  arsenic  failed  to  show 
any  reaction.  So  I  think  that,  for  practical  purposes,  the  utmost 
limit  of  reaction  is  reached  at  o  03  mgm.  of  As  in  1  cc,  and  that 
the  method  employing  metallic  tin  and  heat  is  to  be  preferred, 
where  the  presence  of  antimony  or  bismuth  does  not  forbid  its  use. 
Even  the  addition  of  pure  concentrated  sulphuric  acid,  which  was 
recommended  when  the  test  was  first  introduced,  has  not  yielded  to 
me  any  substantial  advantage  in  detecting  the  presence  of  arsenic  in 
the  greater  dilutions. 
2.  THE  BEST  PROPORTION  OF  REAGENT  TO  SPECIMEN. 
In  the  foregoing  experiments  and  a  number  of  others  conducted 
for  that  purpose,  it  was  observed  that  both  with  the  U.  S.  P.  solution 
(without  the  use  of  metallic  tin)  and  with  that  of  the  German 
