14  Constitution  of  the  Alkaloids.        {A ja^uaryfigoo™' 
that  if  this  substance  abstracts  water  from  ammonium  oxalate  and 
yields  cyanogen  as  a  final  product,  why  should  it  not  yield  anilo- 
cyanogen,  as  he  calls  it,  from  aniline  oxalate. 
COONH4  .     CONH2  CN 
|  —  2H20=     |  —  2H20=     |  and 
COONH4  CONH2  CN 
Ammon.  Oxalate.  Urea.  Cyanogen. 
COOH.C6H7N  CO.NHC6H5  CNC6H4 
|  —  2H80=  |  -2H20=  | 
COOH.C6H7N  CO.NHC6H5  CNC6H4 
Aniline  Oxalate.  Oxanilide.  Anilocyanogen. 
He  obtained  all  these  products  except  the  last — anilocyanogen, 
but  never  succeeded  in  getting  it.  Why  this  peculiar  action  of  ani- 
line ?  asks  Hofmann.  Berzelius'  theory  stays  or  falls  with  it  ;  and, 
as  it  could  not  be  coaxed  into  existence,  Hofmann  looks  to  Liebig's 
theory  to  explain  the  trouble,  and  at  once  sees  that  the  last  stage 
is  impossible,  since  the  last  molecule  of  water  could  not  be  removed 
without  splitting  up  the  radical  phenyl,  and,  as  we  all  know,  phenyl 
is  a  pretty  stable  radical.  After  converting  himself  to  his  master 
Liebig's  theory,  Hofmann  starts  out  on  a  new  line  of  work.  If 
aniline  is  a  substituted  ammonia,  says  he,  why,  then,  if  I  treat  it 
with  ethyl  bromide,  I  ought  to  replace  one  of  the  hydrogen  atoms 
by  ethyl. 
Sure  enough,  by  the  action  of  dry  aniline  upon  ethyl  bromide,  he 
obtains  the  so-called  imidogen  compound  predicted  by  Liebig,  and 
the  proof  of  the  substitutability  of  ammonia  hydrogen  atoms  is 
made.  The  substance  obtained  was  the  hydrobromate  of  ethyl-* 
phenyl-amine.  Next  he  substitutes  the  rem  lining  hydrogen  atom 
of  ammonia  by  producing  di-ethyl-phenyl-amine.  Hofmann  then 
finally  succeeded  in  adding  to  this  tri-substituted  ammonia  a  mole- 
cule of  ethyl  iodide,  and  he  obtained  phenyl  triethylammonium 
iodide,  which  was  quite  different  from  the  di-  and  tri-substitution 
products.  It  was  metallic  in  nature  and  acted  like  ammonium 
iodide.  On  treating  it  with  moist  silver  oxide  he  obtained  phenyl- 
triethylammonium  hydroxide  in  crystals,  which  deliquesced  and 
gave  as  decomposition  products  H20,  C2H4  and  C6H5(C2H5)2N.  It 
•was  basic  and  caustic,  destroying  tissue  and  vegetable  colorsas  readily 
as  caustic  potash,  and  bitter  as  quinine.  Wurtz,  in  1849,  proved  the 
same  thing  that  Hofmann  did,  but  from  a  different  point  of  view. 
He  made  the  methyl  and  ethyl  esters  of  cyanic  and  cyanuric  acids 
and  heated  these  with  water,  which  yielded  him  the  alkyl-ureas,  i.  e.t 
