A™cS;r9i3™'}        Estimation  of  Formaldehyde.  457 
Methyl  salicylate  was  produced  with  salicylic  acid  and  sulphuric 
acid,  but  only  a  trace  of  iodoform  could  be  produced.  The  quantity 
of  methyl  alcohol,  96  per  cent.,  was  then  estimated  with  a  re- 
fractometer  and  by  the  method  suggested  by  C.  Simmonds  in  his 
notes  on  the  determination  of  small  quantities  of  methyl  alcohol, 
which  is  here  given. 
Note  on  the  Determination  of  Small  Quantities  of  Methyl 
Alcohol. 
By  C.  Simmonds:  Analyst.  27,  16  (1912). 
Small  proportions  of  methyl  alcohol  have  hitherto  been  some- 
what difficult  to  determine  readily  and  accurately.  Fairly  good 
approximate  results  can  be  obtained  by  comparative  experiments 
with  the  well  known  method  of  Riche  and  Bardy  (Compt.  rend.,  80, 
1076,  1875),  or  with  Wolff's  modification  of  Trillat's  process 
(Ann.  Inst.  Pasteu,  1902,  8),  but  these  methods  are  lengthy  and 
rather  troublesome.  The  process  described  by  Thorpe  and  Holmes 
(J.  Chem.  Soc,  85,  1 ;  1904)  gives  good  results  when  the  quantity 
of  methyl  alcohol  is  not  too  small.  It  is  not  well  adapted,  however, 
for  use  when  the  proportion  of  methyl  alcohol  is  less  than  about 
2  per  cent,  of  the  ethyl  alcohol,  since  the  necessary  subtractive 
correction  (loc.  cit.,  pp.  2,  3)  may  in  such  cases  be  equal  to  or  may 
exceed  the  quantity  it  is  desired  to  estimate.  For  determining  very 
small  portions  of  methyl  alcohol  the  method  is  quite  inapplicable. 
In  such  cases  satisfactory  determinations  can  be  made  by  applying 
the  principle  of  colorimetric  comparison  by  Deniges'  process  for 
detection  of  methyl  alcohol  (Compt.  rend.,  150,  332,  1910). 
The  possibility  of  thus  using  the  process  is  indicated  by  Deniges' 
(loc.  cit.,  p.  833).  The  object  of  the  present  note  is  to  give  the 
procedure  which  the  writer  finds  most  suitable  for  utilizing  the 
reaction  quantitatively  in  general  analytical  work  as,  for  example, 
in  examining  spirituous  beverages,  medical  tinctures,  flavoring  es- 
sences, and  so  forth. 
The  alcoholic  mixture  is  best  purified,  when  necessary,  either 
by  the  method  of  Thorpe  and  Holmes  (J.  Chem.  Soc,  83,  314, 
1903),  or  by  other  suitable  means.  It  is  then  diluted  with  water 
or  mixed  with  ethyl  alcohol,  as  the  case  may  require,  until  it  con- 
tains 10  per  cent,  of  total  alcohol  by  volume. 
To  5  c.c.  of  this  prepared  liquid  contained  in  a  wide  test  tube 
