188 
Determination  of  Ethyl  Nitrite. 
f  A.m.  Jour.  Pharni. 
t       April,  1885. 
it  will  afford  the  pharmacist  or  medical  man  a  valuable  means  of 
assaying  his  sweet  spirit  of  nitre,  requiring  as  it  does  no  out  of  the  way 
reagents,  or  even  a  standard  solution.^ 
In  the  course  of  the  experiments  made  in  testing  the  foregoing  pro- 
cess, in  many  instances  I  observed  the  amount  of  nitric  oxide  set  free 
simply  by  agitation  with  iodide  of  potassium,  without  any  acid  being 
added.  It  might  be  supposed  that  the  gas  given  off  at  this  stage  of 
the  process  was  that  derived  from  the  free  nitrous  acid  of  the  sample ; 
but  I  do  not  think  this  is  strictly  the  case,  as  in  experiments  in  which 
acetic  acid  was  substituted  for  dilute  sulphuric  acid,  nearly,  if  not 
quite,  the  whole  procurable  gas  was  evolved.  Hence  a  sample  which 
by  decomposition  contained  free  acetic  acid  would  evolve  nitric  oxide, 
though  no  nitrous  compound  other  than  ethyl  nitrite  were  present. 
With  regard  to  Dott's  titration  process,  if  the  treatment  with  potas- 
sium iodide  be  effected  in  an  open  basin  in  presence  of  air,  the  amount 
ot  nitrite  found  is  liable  to  be  seriously  in  error;  but  if  air  be 
excluded  Dott's  method  gives  fair  approximate  results,  somewhat  in 
excess  of  the  truth.  The  process  can  be  advantageously  employed  on 
the  solution  wliich  has  already  been  decomposed  with  potassium  iodide 
in  the  nitrometer.  The  nitric  oxide  is  allowed  to  escape  into  the  air, 
and  the  brown  liquid  is  washed  into  a  basin,  and  at  once  titrated  with 
decinormal  thiosulphate.  1  cc.  of  this  solution  (containing  24'S  grams 
of  crystallized  Na2S203  per  litre)  will  react  with  the  iodine  liberated 
by  '0075  gram  of  ethyl  nitrite. 
The  results  obtained  in  this  manner  show  a  constant  difference  of 
about  O'OOo  gram  of  nitric  oxide  above  that  corresponding  to  the 
volume  of  gas  liberated  in  the  nitrometer,  the  true  amount  doubtless 
lying  between  the  two.  The  difference  is  most  probably  due  to  a 
small  amount  of  nitric  oxide  remaining  dissolved  in  the  aqueous 
liquid,  which  causes  the  volume  of  gas  to  be  slightly  low,  and  becoming 
oxydized  to  nitrous  acid  during  the  subsequent  titration  liberates  a 
small  additional  amount  of  iodine.    This  source  of  error  becomes  very 
1  The  use  of  brine,  in  the  nitrometer,  instead  of  mercury,  renders  tlie 
manipulation  more  simple,  witliout  appreciably  affecting  the  accuracy  of 
the  results.  The  nitrometer  itself  may  be  extemporized  from  an  inverted 
Mohr's  burette,  to  the  nose  of  which  a  short,  wide,  glass  tube  has  been 
fitted  by  means  of  a  perforated  cork.  The  open  end  of  the  burette  is  also 
fitted  with  a  cork,  through  wliich  passes  a  short  length  of  glass  tubing 
which  is  connected  by  narrow  India  rubber  tubing  with  a  wide  glass  tube 
or  funnel  to  serve  as  a  reservoir. 
