Am>Je°b^-I^arm  }  Detection  and  Determination  of  Alcohol.  89 
when  examined  in  a  long  test  tube  over  a  sheet  of  white  paper  or  a 
porcelain  slab. 
I  then  commenced  a  series  of  experiments,  having  for  their  object 
the  detection  of  minute  quantities  of  alcohol  by  converting  it  into 
aldehyd,  and  treating  the  resulting  solution  wijh  caustic  alkali.  After 
making  a  great  number  of  attempts  I  succeeded  in  devising  a  method 
which  is  not  only  reliable  as  indicating  with  certainty  the  presence  of 
alcohol,  but  which  within  certain  limits  gives  fairly  approximate  quan- 
titative results. 
Distillation  with  sulphuric  acid  and  permanganate  of  potash  was  first 
tried  and  it  was  found  that  with  a  given  proportion  of  permanganate  a 
•1  per  cent,  solution  of  alcohol  could  readily  be  detected,  but  with  even 
a  slight  excess  of  permanganate  the  results  were  entirely  negative. 
Other  oxidizing  agents  were  then  tried,  and  bichromate  of  potash 
found  to  be  the  most  reliable,  since  an  excess,  unless  large,  did  not 
materially  decrease  the  yield  of  aldehyd.  Still  the  quantitative  results 
obtained  from  the  same  solution  varied  very  considerably,  and  for  some 
time  the  cause  was  apparently  inexplicable.  At  length,  noticing  that 
when  the  determinations  were  very  low  considerable  bumping  took 
place  during  the  distillation,  several  substances  w'ere  added  to  obviate 
this,  and  it  was  found  that  when  a  few  small  pieces  of  pumice  were 
placed  in  the  flask  the  distillation  proceeded  evenly  and  the  results  were 
remarkably  uniform.  Fresh  pumice  must  be  used  for  each  distillation. 
For  an  aqueous  solution  containing  from  "04  to  '4  per  cent,  of  alcohol 
the  following  method  may  be  relied  upon  for  quantitative  determina- 
tions, whilst  it  will  detect  with  certainty  *oi  per  cent.,  or  1  in  10,000. 
When  the  quantity  of  alcohol  present  differs  much  from  the  proportion 
here  given  the  determination  is  too  low,  becoming  more  and  more 
unreliable  as  the*  percentage  rises  above  or  falls  below  these  limits. 
The  requisites  are  a  saturated  solution  of  bichromate  of  potash,  a 
•dilute  sulphuric  acid  (B.  P.  acid  and  water,  equal  quantities),  a  syrupy 
solution  of  caustic  soda,  methylated  spirit  free  from  aldehyd,  a  200 
cc.  flask  with  good  condensing  arrangement  attached,  and  a  long  narrow 
test  tube  graduated  to  3  and  23  cc. 
One  hundred  cc.  of  the  dilute  alcohol  are  placed  in  the  flask,  2  cc. 
of  bichromate  solution,  8  cc.  of  the  dilute  acid,  and  a  few  pieces  of 
pumice  are  added  and  20  cc.  distilled  (not  too  rapidly),  and  the  distillate 
conveyed  by  a  long  tube  to  the  bottom  of  the  test  tube  in  which  has  been 
