268  Determination  of  Hydrocyanic  Acid.  {Am'jJx°^\l^rm- 
We  ascertained  the  accuracy  of  our  method  by  diluting  i  Cc.  of 
a  standard  solution  of  potassium  cyanide  to  25  Cc.  and  evaporating 
under  the  diminished  pressure  to  1  Cc.  When  the  residue  was 
tested  as  described  above  we  obtained  the  same  quantity  of  Prussian 
blue  as  that  formed  by  applying  the  test  to  1  Cc.  of  the  undiluted 
standard  solution. 
The  quantities  of  reagents  mentioned  are  suitable  for  1  to  2  mg. 
of  potassium  cyanide.  If  less  than  1  mg.  of  potassium  cyanide  is 
present,  the  quantities  of  the  reagents  should  be  reduced  accordingly. 
A  large  excess  of  reagents  must  be  avoided  in  order  to  obtain  the 
maximum  density  of  color.  When  the  analysis  is  carried  out  as 
described  the  maximum  error  should  not  exceed  more  than  1  part 
in  20.  Thus  in  a  plant  giving  20  mg.  of  potassium  cyanide  per 
100  g.  of  plant  the  results  might  vary  by  0.001  per  cent,  if  100  g. 
of  material  are  used  for  analysis. 
Application  as  a  Qualitative  Test. 
This  method  also  proved  to  be  a  very  delicate  qualitative  test  for 
the  presence  of  a  cyanide.  Owing  to  the  possibility  of  obtaining 
a  pure  blue  color  we  were  able  to  detect  with  certainty  0.00002  g. 
of  potassium  cyanide,  which  represents  less  than  0.00001  g.  of  hydro- 
cyanic acid. 
Application  to  Microchemical  Analysis. 
The  test  for  hydrocyanic  acid  was  applied  microchemically  to 
sections  of  cherry  bark  and  bitter  almond  with  marked  success. 
Sections  or  small  parts  of  the  material  were  treated  with  the  reagents 
in  a  test-tube  according  to  our  method.11 
Conclusions. 
1.  Dilute  alkaline  solutions  of  a  cyanide  can  be  concentrated 
under  diminished  pressure  without  appreciable  loss  of  cyanide. 
2.  The  maximum  quantity  of  Prussian  blue  can  be  obtained  from 
a  cyanide  only  when  the  volume  of  the  solution  to  be  tested  is 
sufficiently  small,  as  has  been  indicated  by  Berl  and  Delpy  and  by 
Lander  and  Walden. 
3.  In  the  test  for  a  cyanide  it  is  better  to  acidify  with  nitric 
11  Further  work  is  in  progress  on  the  application  of  this  method  to  micro- 
chemical  analysis. 
