Amju°ner;il,9h8arm'}     Amy  I  Nitrite  and  Nitrous  Ether.  279 
further  limitation  of  the  bulk  of  the  solution  and  persistent  shaking 
materially  hasten  the  reaction.  With  10  c.c.  of  water,  instead  of 
25  c.c,  and  frequent  shaking,  4-58  and  4*6  per  cent,  was  obtained 
after  fifteen  minutes.  Without  any  addition  of  water  4-55  and  4-58 
per  cent,  was  obtained  in  the  same  length  of  time.  In  the  latter 
case  the  solution  became  yellow,  showing  secondary  reactions,  and 
this  degree  of  concentration  is  therefore  not  advisable,  although  the 
figures  obtained  are  normal.  Frequent  shaking  is  required  to  bring 
into  contact  with  the  solution  the  vapors  of  ethereal  nitrite  diffused 
through  the  air  above  the  liquid,  which  would  otherwise  remain 
unchanged  for  a  considerable  time. 
The  high  results  of  this  method,  as  compared  with  Allen's  gaso- 
metric  method,  led  to  the  suspicion  that  they  might  be  due  to  the 
presence  of  aldehyde,  but  a  few  experiments  showed  that  this  was 
not  the  case.  The  same  spirit  as  had  been  used  in  the  experiments 
given  above  was  assayed  with  the  addition  of  01  c.c.  and  0-2  c.c.  of 
99  per  cent,  paraldehyde,  respectively,  to  5  c.c.  of  the  sample,  a 
great  deal  more  of  aldehyde  than  could  possibly  be  present  in  any 
sample  tested.  The  results  were  practically  the  same  as  without 
aldehyde,  namely,  4-60  and  4-54  per  cent.  This  was  repeated  with 
0-5  c.c.  and  ro  c.c.  of  the  aldehyde  for  5  c.c.  of  spirit  and  the  mix- 
tures allowed  to  stand  over  night  before  tritrating.  The  results  were 
4- 3  and  4-12  per  cent.  The  presence  of  aldehyde,  therefore,  affects 
the  accuracy  only  when  a  very  large  quantity  of  it  is  present  and 
in  such  a  case  does  not  increase,  but  decreases  the  result.  No 
interference  need  be  feared  from  the  amount  of  aldehyde  naturally 
present. 
To  obtain  further  data  about  the  reliability  of  the  method,  an 
attempt  was  made  with  pure  silver  nitrite.  If  the  reaction  pro- 
ceeded according  to  theory,  exactly  one-third  of  the  silver  in  the 
salt  would  be  converted  into  chloride,  and  the  remainder  stay  in 
solution  in  form  of  nitrate,  according  to  the  following  equation : 
3AgN02  +  KCIO3  =  AgCl  +  2AgN03  +  KNO3. 
A  very  rapid  blackening  of  the  precipitate,  as  soon  as  exposed  to 
light,  however,  the  cause  of  which  is  not  known,  made  the  estima- 
tion impracticable.  The  silver  nitrite  used  had  a  good  appearance, 
and  gave  702  and  69-95  per  cent,  of  silver,  70-1  per  cent,  being 
required  by  theory. 
