726  Estimation  of  Methyl  Alcohol  \^'^oc^^,^^i9To. 
standard  prepared  in  an  exactly  similar  way  from  a  sample  of  ethyl 
alcohol  containing  a  known  amount  of  methyl  alcohol.  When  a 
colorimeter  of  Schmidt  and  Haensch  was  used,  differences  could  be 
detected  between  samples  of  ethyl-methyl  alcohol  mixtures  in 
which  the  amounts  of  methyl  alcohol  differed  by  only  0.2  per  cent. 
Possibly,  with  a  more  modern  form  of  colorimeter,  more  accurate 
estimations  would  be  possible.  The  researches  on  the  determination 
of  the  amounts  of  methyl  alcohol  in  ethyl  alcohol  have,  up  to  the 
present,  been  confined  to  mixtures  containing  between  0.5  and  5 
per  cent,  of  the  former. 
The  sample  of  ethyl  alcohol  gave  a  faint  formaldehyde  reaction 
when  oxidized,  and  the  oxidation  products  were  tested  for  methyl 
alcohol  in  the  manner  described  above.  Whether  this  is  due  to  the 
fact  that  it  contains  small  amounts  of  the  lower  alcohol,  or  to  the 
fact  that  formaldehyde  is  formed  in  small  amounts  by  the  oxidation 
of  ethyl  alcohol,  it  is  impossible  at  the  present  to  say. 
C.  Estimation  01^  Me;thyl  AIvCOhol  in  Acetonk. — It  was  found 
that  a  similar  method  could  be  applied  to  the  estimation  of  methyl 
alcohol  in  acetone  to  that  employed  for  estimating  this  alcohol  in  water, 
as  the  acetone  is  only  oxidized  to  a  slight  extent  by  persulphate.  In 
the  former  case,  however,  there  were  certain  difficulties  in  carrying 
out  the  estimation,  which  have  necessitated,  when  the  percentage 
of  methyl  alcohol  present  is  small,  certain  minor  modifications  in 
the  process.  It  was  found  when  the  acetone  contained  4  per  cent, 
or  more  of  methyl  alcohol  that,  as  a  certain  concentration  of  per- 
sulphate was  exceeded,  there  was  a  sudden  transition  from  the 
bright  pink  color  yielded  by  the  formaldehyde  test  to  a  light  pink, 
and  this  latter  color  persisted  even  when  the  above-mentioned  con- 
centration of  persulphate  was  considerably  above  this  transition 
point.  This  phenomenon  is  probably  due  to  the  formation  of  a 
condensation  product  of  formaldehyde  with  acetone,  which  is  very 
resistant  to  oxidation  by  persulphate.  When  the  amount  of  methyl 
alcohol  in  acetone  falls  below  4  per  cent.,  the  transition  point  is  not 
very  marked.  At  4  per  cent,  and  over,  when  higher  concentrations 
of  persulphate  must  be  used,  the  point  is  unmistakable;  for  each 
given  concentration  of  methyl  alcohol  in  acetone,  under  the  condi- 
tions of  experiment  described  below,  there  corresponds  a  given 
concentration  of  persulphate,  a  small  addition  to  which  causes  a 
sudden  transition  from  a  bright  pink  to  a  light  pink,  when  the  for- 
maldehyde test  is  applied.    By  this  method  it  is  easy  to  determine 
