Ani sept"1"'  i9i8rm'  )     Determination  of  Methyl  Alcohol.  637 
It  is  possible  from  such  a  compound  to  set  free  the  hydrogen  of 
the  hydroxyl  by  replacing  it  with  an  easily  oxidizable  metal  such  as 
sodium. 
Thirty-two  parts  by  weight  of  methyl  alcohol  treated  with 
metallic  sodium  yield  exactly  the  same  weight  or  volume  of  gas  as 
would  be  yielded  by  forty-six  parts  of  ethyl  alcohol  under  identical 
conditions.  However,  if  equal  weights  of  the  two  alcohols  are  sub- 
jected to  this  treatment,  very  different  weights  or  volumes  of  gas 
are  set  free  and  this  forms  a  positive  means  of  differentiating  be- 
tween the  two  alcohols.  The  fact  is  that  methyl  alcohol  yields 
nearly  twice  the  volume  of  gas  that  an  equal  amount  of  ethyl  alcohol 
gives. 
When  methyl  and  ethyl  alcohols  are  mixed  and  so  treated  the 
yield  of  gas  is  exactly  in  proportion  as  the  alcohols  are  mixed. 
The  alcohols  must  be  dehydrated  which  is  accomplished  by  treat- 
ing a  small  portion  of  alcohol  with  a  very  large  excess  of  fused 
potassium  carbonate. 
The  above  statements  are  true  of  chemically  pure  alcohols  but 
commercial  varieties  have  higher  alcohols  associated  with  them 
which  also  react  with  sodium  and  yield  gas  in  a  different  ratio  as 
their  molecular  weight  varies.  This  fact  operates  against  this 
method  as  a  precise  laboratory  operation,  but  does  not  affect  its  ap- 
plication in  the  work  for  which  it  is  designed,  namely,  to  protect  the 
government  formula  by  indicating  a  minimum.  The  higher  alcohols 
likely  to  be  present  in  such  mixtures  have  but  one  hydroxyl  also  but 
their  molecular  weights  are  still  higher  than  ethyl  alcohol.  This 
would  indicate  as  a  deficiency  of  wood  alcohol  in  any  test  and  re- 
quire further  additions  possibly,  to  make  the  tested  alcohol  satis- 
factory. Acetone  which  is  also  present  in  the  denatured  alcohol 
does  not  react  with  sodium,  as  there  is  no  hydroxyl  in  its  composi- 
tion, its  presence  however  indicates  against  the  test  and  in  favor  of 
the  government  formula.  While  such  impurities  must  be  mentioned 
to  completely  cover  the  subject,  their  presence  from  a  practical 
point  of  view  is  almost  negligible  because  of  the  small  percentage 
present. 
Apparatus. — Having  grounded  our  theory  in  substantial  fact, 
apparatus  for  carrying  it  into  effect  becomes  our  next  consideration. 
From  the  preceding  remarks  it  will  be  learned,  first,  that  we  hope  to 
set  free  hydrogen  gas  from  the  mixed  alcohols ;  second,  to  collect 
this  freed  gas ;  third,  to  measure  the  gas ;  fourth,  to  carry  out  these 
