292     CHLOROFORM  AND  THE  TESTS  FOR  ITS  PURITY  IN  P.  B. 
that  we  did  not  believe  any  commercial  chloroform  could  or  would 
stand  the  test.  It  was  thought  well,  however,  to  examine  the 
matter  a  little  more  closely. 
A  perfectly  pure  chloroform  was  very  carefully  prepared  from 
alcohol  by  the  process  given  in  the  B.  P,  Its  specific  gravity 
was  1-500  ;  it  gave  no  color  to  S  03 ;  left  no  residue  or  odor 
of  anything  after  evaporation,  but  it  yielded  gas  with  potassium. 
Specimens  of  chloroform  were  got  from  various  makers,  and 
others  were  prepared  from  acetone  and  methylated  spirit,  and 
one  and  all  tried  by  the  potassium  test  gave  off  gas.  At  first 
the  evolution  of  the  gas  was  rapid,  but  very  soon  became  slower, 
and  a  white  crust  was  formed  on  the  metal.  On  removing  that 
crust  the  gas  was  again  rapidly  produced.  Sodium  acted  like 
potassium.  The  collection  and  examination  of  the  gas  presents 
a  number  of  difficulties  not  very  easily  overcome  :  the  use  of  , 
water  is  out  of  the  question  ;  mercury  is  liable  to  explosive 
amalgamation  with  the  K  or  Na  ;  other  fluids  present  other  in- 
compatibilities with  the  necessities  of  the  case,  and  we  can  only 
manage  well  when  our  pneumatic  trough  is  filled  with  chloro- 
form ;  and  that,  being  rather  an  expensive  fluid  to  work  with, 
gives  off  its  own  vapor  with  the  gas,  and  so  complicates  the  re- 
sult. 
The  gas  got  by  the  action  of  Na  on  chloroform,  and  before 
washing  it  with  water,  burns  with  a  yellow  flame,  somewhat 
smoky,  and  sometimes  with  a  tint  of  green  at  the  edges  ;  and 
mixed  with  about  its  own  bulk  of  air,  and  a  light  applied,  it  ex- 
plodes with  almost  no  violence,  leaving  sharp  acrid  acid  fumes 
in  the  jar.  A  little  water  shaken  with  the  vapor  left  in  the  jar 
after  explosion  is  distinctly  acid  to  litmus,  and  gives  with  AgN05 
abundant  evidences  of  chlorine.  On  washing  the  gas  with 
water  some  absorption  takes  place,  and  it  then  burns  much  more 
like  Hydrogen,  explodes  more  violently  with  air,  and  leaves  less 
acid  and  chlorine  in  the  jar.  It  is  far  too  soon  yet  to  say  what 
the  gas  is  or  .is  not ;  so  far  as  we  have  seen,  however,  it  may 
be  Hydrogen  with  the  vapor  of  chloroform  diffused  through  it, 
and  due  to  water,  or  alcohol,  in  the  chloroform.  But  an  exam- 
ination of  the  crust  formed  on  the  Na  sets  that  aside,  for  it  al- 
most wholly  consists  of  NaCl,  which,  under  the  circumstances, 
can  only  derive  its  chlorine  from  the  decomposition  of  chloro- 
