446 
ORGANIC  MATTER    OF  THE  AIR. 
inches  long  when  stretched  out,  and  about  4  in  diameter,  and  is 
compressible  into  the  thickness  of  about  2  inches.  The  sides 
are  made  of  thin  Mackintosh  cloth.  By  the  use  of  the  pump  the 
air  of  the  vessel  is  removed,  and  the  external  air  of  course  enters. 
A  few  strokes  of  the  pump  are  sufficient,  i.  e.  from  6  to  10. 
After  ten  strokes  I  perceive  no  change,  and  am  inclined  to  think 
that  it  is  an  unnecessary  number.  The  test  liquid  is  poured  into 
a  graduated  tube  or  burette,  containing  somewhat  more  than  will 
be  required.  A  portion  is  then  poured  into  the  tube  which 
passes  through  the  stopper,  and  the  stopcock  is  opened  to  allow 
it  to  pass.  Small  quantities  are  used ;  when  it  has  entered  the 
bottle,  the  liquid  is  made  to  spread  over  the  sides,  and  time  given 
it  to  be  exposed  to  the  action  of  the  air  ;  it  is  found  that  in  five 
or  six  minutes  a  decided  epoch  is  attained  from  which  to  date 
the  comparative  action. 
In  order  to  see  the  color  the  liquid  must  be  allowed  to  trickle 
down  the  sides  of  the  vessel,  and  collect  itself  at  one  point  of  the 
circumference  at  either  end  of  the  cylindrical  part  of  the  bottle. 
This  part  must  be  raised  up  to  the  level  of  the  eye,  so  that  the 
longest  axis  may  be  presented  to  the  sight,  and  thereby  the 
deepest  shade  of  color.  It  requires  some  time  to  accustom  one- 
self to  the  sight  of  such  a  small  amount  of  color;  but  when  it  is 
once  well  observed,  it  will  be  found  to  be  a  method  which  will 
admit  of  the  greatest  precision.  The  first  few  drops  which  are 
poured  in  will  probably  be  decolorized  at  once  :  a  few  drops  more 
must  then  be  added  ;  if  they  become  decolorized,  a  few  more 
must  be  used ;  and  so  on  until  there  is  a  perceptible  amount  of 
color  remaining.  When  this  occurs,  the  experiment  is  con- 
cluded. The  amount  of  the  reagent  used  is  then  read  ofi°  from 
the  graduated  measure.  If  the  liquid  be  of  the  proper  strength, 
and  the  bottle  the  required  size,  the  number  of  grains  gives  the 
comparative  quantity  at  once.  Sometimes  the  amount  of  organic 
matter  is  so  small  that  there  is  no  appreciable  action,  on  even 
the  smallest  amount  of  solution  by  one  vessel  of  air.  In  this 
case  it  is  necessary  to  fill  the  bottle  several  times.  The  mode 
of  doing  this  is  apparently  extremely  rude,  but  the  results  are 
such  as  not  to  demand  a  finer  method  at  present.  A  finer 
method,  of  course,  would  need  little  ingenuity  to  contrive.  At 
present  I  merely  remove  the  stopper  and  fill  again  with  air  as 
