724 
Estimation  of  Methyl  Alcohol. 
Am.  Jour.  Pharm. 
October,  1920. 
Thus,  if  the  end-point  is  taken  to  be  that  concentration  which 
just  gives  the  pink  formaldehyde  reaction  after  heating  for  ten 
minutes  with  the  persulphate  solution,  it  will  be  found  that  5  per 
cent,  of  the  latter  just  oxidizes  5  Cc.  of  a  solution  containing 
X  I  per  cent.  =0.29  per  cent,  methyl  alcohol.    The  percentages 
of  methyl  alcohol  corresponding  to  the  various  concentrations  of 
persulphate  are  given  in  the  final  column  of  Table  I.  It  is  obvious 
that  the  concentration  of  a  persulphate  solution  just  necessary  to 
oxidize  a  given  concentration  of  methyl  alcohol  is  a  linear  function 
of  that  concentration,  and  if  c  be  the  concentration  (per  cent.) 
of  the  persulphate,  x  that  of  the  methyl  alcohol,  and  k  a  constant, 
then  X  =  kc.  If  this  be  calculated  from  the  above  table  (neglecting 
the  data  for  percentages  of  persulphate  below  0.25  per  cent.)  k 
will  be  found  to  be  =  0.059.  These  data  were  obtained  with  a 
methyl  alcohol  sample  which  had  the  correct  boiling  point,  after 
careful  fractionation  over  lime,  and  with  a  sample  of  persulphate 
in  which  the  amount  of  ammonium  persulphate  was  estimated  by 
warming  with  a  definite  amount  of  standardized  ferrous  sulphate 
solution  and  determining  the  amount  of  the  latter  oxidized  by 
titration  against  permanganate  solution;  this  sample  was  found 
to  contain  98.6  per  cent,  ammonium  persulphate.  In  a  series 
of  experiments  carried  out  with  another  sample  of  persulphate 
which  was  less  pure,  the  constant  k  was  found  to  be  0.054. 
carrying  out  the  estimation  of  methyl  alcohol  by  the  method  above 
described,  with  any  given  sample  of  persulphate  it  is  best  to  stand- 
ardize 5,  4  or  3  per  cent,  solutions  against  methyl  alcohol  and  to 
determine  directly  the  constant  k.  Only  a  few  experiments  will  be 
necessary,  as  the  results  will  only  differ  slightly  (by  fractions  of  a 
Cc.)  from  the  numbers  given  in  Table  I,  column  3. 
B.  Estimation  of  Methyi,  AivCOhol  in  KthyIv  Alcohol. — ^An 
attempt  was  made  to  estimate  methyl  alcohol  in  ethyl  alcohol  by  the 
same  method  as  that  employed  for  estimating  the  methyl  alcohol  in 
water.  This  was  found,  however,  to  be  impracticable,  and  there  is  no 
preferential  oxidation  of  the  lower  alcohol  by  persulphate.  It  was 
found,  for  example,  that  the  amounts  of  the  latter  necessary  to 
oxidize  to  the  end-point  (disappearance  of  formaldehyde)  did  not 
differ,  outside  the  limits  of  experimental  error,  with  ethyl  alcohol 
mixtures  containing  4  and  5  per  cent,  of  methyl  alcohol.  Another 
