278  Amy  I  Nitrite  and  Nitrous  Ether.  {^I^mt^ 
to  determine  if  it  would  be  justifiable  to  omit  the  silver  nitrate 
until  the  reaction  between  nitrite  and  chlorate  was  ended,  In  two 
estimations  the  silver  nitrate  was  added  at  the  beginning,  and  the 
mixture  excluded  from  light  as  much  as  possible ;  in  two  others  no 
silver  nitrate  was  added  and  the  solutions  placed  into  the  light. 
All  of  them  were  otherwise  treated  in  the  same  manner,  as  nearly 
as  possible,  and  titrated  after  standing  the  same  length  of  time. 
The  results  were  4-55  and  4-52  per  cent,  for  the  first  pair  and  4-65 
per  cent,  for  each  of  the  second.  It  is  seen  that  the  figures  are 
somewhat  higher  when  the  silver  is  not  added  until  reaction  is  com- 
plete, which  has  been  confirmed  by  subsequent  estimations,  but  it 
is  probable  that  they  are  more  nearly  exact. 
It  was  also  thought  expedient  to  substitute  potassium  or  ammo- 
nium sulphocyanate  for  sodium  chloride  in  residual  titration,  so  as 
to  permit  the  use  of  a  color-indicator,  which  much  facilitates  the 
work. 
In  the  following  experiments,  made  to  determine  the  time 
required  for  an  estimation,  25  c.c.  of  water,  5  c.c.  of  a  saturated 
solution  of  potassium  chlorate,  5  c.c.  of  spirit  of  nitrous  ether,  and 
5  c.c.  of  10  per  cent,  nitric  acid  were  successively  placed  into  a  100 
c.c.  flask,  which  was  at  once  securely  stoppered,  and  allowed  to 
stand  a  definite  length  of  time,  with  occasional  shaking,  and  then 
titrated.    The  spirit  used  assayed  4-28  per  cent,  by  nitrometer. 
Time  allowed  for  reaction.  Per  Cent. 
5  minutes   375  ;  — 
10     "   3-85;  — 
15     "        ■   •  •   3'9i;  4-n 
30    "    4-29 ;  4-34 
60     "    4*55  ;  4'6o 
Over  night   4-65  ;  4*65 
Under  the  above  conditions,  then,  at  least  an  hour  is  re- 
quired to  obtain  results  approaching  the  maximum  per  cent, 
obtainable.  The  highest  figures  also  show  a  decided  increase  over 
the  per  cent,  obtained  by  nitrometer.  An  increase  had  been 
expected,  as  the  gasometric  method  has  always  been  supposed  to 
yield  somewhat  low  results,  but  not  such  a  great  one.  A  similar 
difference  was  found  in  all  samples  of  spirit  of  nitrous  ether  exam- 
ined and  a  still  greater  one  in  amyl  nitrite,  but,  as  will  be  shown 
later,  the  results  by  the  new  method  are  entitled  to  more  credence. 
In  further  attempts  to  shorten  the  method,  it  was  found  that 
