AmMJa0rch■,Sf.rm•}        The  Detection  of  Methyl  Alcohol.  109 
Copper  Wire  Spiral. — Copper  wire  of  size  18  is  taken  of  the 
length  of  1  meter.  It  is  wound  in  a  close  spiral  around  a  smooth 
rod  7  mm.  thick  to  make  a  coil  about  3  cm.  long.  A  handle  is 
made  by  twisting  together  the  two  free  ends  of  the  wire  in  the 
spiral,  one  of  the  ends  having  been  overlapped  in  winding,  begin- 
ning to  wind  at  about  30  cm.  from  the  end  of  the  covered  strand. 
The  handle  is  left  of  sufficient  length  and  is  bent  at  right  angles  6 
cm.  from  the  extremity,  the  horizontal  part  being  wound  with 
twine. 
Phloroglacinol  Test  Solution.-— -Take  of  phloroglucinol  5  deci- 
grammes; soda  (white),  10  grammes;  distilled  water  to  make  50 
c.c.  Dissolve  the  phloroglucinol  in  about  40  c.c.  of  the  distilled 
water  with  a  little  of  the  soda,  then  add  the  remainder  of  the  soda, 
and  enough  distilled  water  to  make  the  solution  measure  50  c.c.  A 
slight  color  in  the  fresh  solution  usually  disappears  on  brief  stand- 
ing and  may  be  disregarded.  On  long  standing  the  solution  dark- 
ens in  color  and  should  be  rejected. 
In  Haigh's  article  the  suggestion  is  also  made  that — 
"  The  Rimini  test  may  be  used  instead  of  the  phloroglucinol  test 
for  the  detection  of  formaldehyde  after  the  oxidation  of  the  alco- 
hols. After  the  removal  of  the  acetaldehyde  by  boiling,  I  c.c.  of  a 
dilute  solution  of  phenylhydrazine  hydrochloride  is  added,  then  a 
few  drops  of  a  fresh  solution  of  sodium  nitroprusside,  and  finally 
I  c.c.  of  a  50  per  cent,  solution  of  sodium  hydroxide.  If  formalde- 
hyde is  present  a  light  blue  or  green  color  will  result,  depending 
upon  the  amount  of  methyl  alcohol  in  the  original  spirit,  and  to 
some  extent  also  on  the  care  with  which  the  boiling  operation  is 
conducted.  The  boiling  of  the  liquid  should  not  be  carried  too  far 
and  should  be  conducted  as  slowly  as  possible.  In  case  the  original 
spirit  contained  no  methyl  alcohol  the  resulting  color  of  the  solu- 
tion will  be  a  greenish  yellow.  Satisfactory  results  are  obtained 
with  both  of  these  tests  for  quantities  of  methyl  alcohol  in  spirits 
as  small  as  one  part  to  twenty  parts  of  ethyl  alcohol. 
The  Sub-committee  of  the  U.  S.  Revision  Committee  then  under- 
took a  careful  testing  of  this  modified  Mulliken  and  Scudder  method 
and  tested  the  several  color  reactions  suggested  to  distinguish  the 
formaldehyde.  They  found  that  the  resorcinol  test  seemed  more 
delicate  than  the  phloroglucinol  test  and  could  be  made  to  distin- 
guish a  smaller  admixture  of  methyl  alcohol  than  the  other.  In 
