184 
Determination  of  Ethyl  Nitrite. 
(  Am.  Jour,  Pharm, 
\      April,  1885. 
always  considerably  enhanced  by  the  sulphuric  acid,  and  strong  acid  is 
far  preferable  to  dilute. 
I  have  tried  the  experiment  of  shaking  a  sample  of  spirit  of 
nitrous  ether  with  neutral  sodium  carbonate,  filtering,  and  testing 
the  filtrate  with  a  solution  of  ferrous  sulphate  without  adding  acid. 
No  reaction  was  at  first  produced,  but  the  brown  color  gradually 
developed,  and  after  a  few  minutes  was  very  intense. 
The  iron  test  has  the  great  advantage  of  really  indicating  the  pre- 
sence of  nitrous  compounds  instead  of  leaving  their  existence  to  be 
inferred  from  some  such  fallacious  reaction  as  the  decoloration  of  per- 
manganate. No  doubt  the  permanganate  processes  of  assaying  spirit 
of  nitrous  ether  are  capable  of  giving  useful  comparative  results  in 
the  case  of  samples  of  good  quality,  but  when  applied  to  inferior 
samples,  or  to  specimens  containing  much  aldehyde  (such  as  the  Lon- 
don Pharmacopcjeia  preparation),  the  indications  are  utterly  worthless 
and  misleading,  and  the  results  should  on  no  account  be  expressed  in 
terms  of  nitrous  ether.^ 
The  well  known  process  of  Prof.  J.  F.  Eykman  for  assaying  spirit 
of  nitrous  ether  is  practically  a  quantitative  application  of  the  iron 
reaction,  but,  instead  of  relying  on  the  depth  of  the  brown  coloration, 
the  nitric  oxide  gas  evolved  is  collected  and  measured.  Eykman's 
process  has  given  excellent  results  in  the  hands  of  Messrs.  P.  MacEwan 
and  F.  S.  Dymond,  and  Mr.  D.  B.  Dott  has  confirmed  the  general 
accuracy  of  the  process,  and  has  found  the  results  very  constant,  even 
when  the  conditions  of  the  experiment  are  varied.  Their  reports  have 
been  confirmed  in  the  main  by  an  extensive  experience  of  its  capa- 
bilities in  my  own  laboratory,  where  the  accuracy  of  the  process  has 
been  fully  verified  when  a  known  quantity  of  pure  sodium  nitrite 
(made  from  silver  nitrite)  is  employed.  With  solutions  of  ethyl  nitrite 
several  sources  of  error  exist,  which  tend  to  cause  the  method  to  give 
results  somewhat  below  the  truth.  The  chief  cause  of  error  is  pro- 
bably incomplete  reaction,  but  the  solubility  of  nitric  oxide  in  aque- 
ous liquids  also  affects  the  results.  The  loss  from  the  latter  cause  is 
reduced  to  a  minimum  if  a  nitrometer  be  employed  instead  of  the  gas 
being  caused  to  bubble  up  through  a  solution  of  soda.    Probably  still 
1  A  sample  of  sweet  spirit  of  nitre,  which  gave  only  faint  indications  of 
nitrous  compounds  when  independently  examined  by  the  iron  test  by  four 
different  chemists,  was  reported  by  another  chemist  to  contain  1'3  per  cent, 
of  nitrous  ether,  as  the  result  of  an  anal^'sis  by  Muter's  modification  of  the 
permanganate  process. 
