IMPROVED  METHOD  OF  ULTIMATE  ANALYSIS. 
395 
From  the  convenience  and  success  attending  the  use  of  such 
rolls  in  this  manner  for  this  purpose,  it  was  but  a  step  in  the 
same  direction  to  adopt  them,  when  oxidized,  as  substitutes  for 
the  oxide  of  copper,  in  the  tube  to  furnish  the  oxygen  necessary 
for  combustion.  A  very  compact  roll,  6  inches  long,  was  made 
so  as  to  fit  the  combustion  tube  tightly,  and  was  forced  into  its 
place  from  the  largest  end  of  the  tube  leaving  about  7  inches  of 
the  tube  behind,  and  4  inches  before  it,  vacant.  The  tube  thus 
prepared  was  then  placed  upon  the  furnace  as  for  a  combustion, 
and  the  copper  surfaces  of  the  roll  thoroughly  oxidized  at  a  low 
red  heat  in  currents  of  dry  air  and  oxygen.  The  tube  was  then 
in  condition  for  immediate  use  ;  and  each  successive  combustion 
afterward  left  it  in  precisely  this  condition,  ready  for  the  next 
succeeding  one. 
The  combustions  were  made,  as  is  now  the  common  practice, 
in  a  thin  straight  tube  open  at  both  ends,  protected  and  supported 
in  a  shallow  trough  of  wire  gauze.  The  fuel  used  was  common 
illuminating  gas,  supplied  through  several  stop  cocks,  and  burned 
above  a  wire  gauze  after  due  admixture  with  air  in  the  common 
way.  The  prepared  substance  was  weighed,  introduced  and 
burned  in  a  boat-shaped  piece  of  platina  foil,  fitting  the  tube 
well,  but  easily.  A  roll  of  reduced  copper  was  placed  behind 
the  boat,  and  another  in  the  anterior  end  of  the  tube.  The  tube  was 
heated  gradually  first  from  the  anterior  end  toward  the  boat,  then 
from  the  posterior  end  toward  the  boat,  and  finally  the  boat  itself 
was  heated.  The  heating  and  the  first  part  of  the  combustions 
were  made  in  a  very  slow  current  of  air  admitted  through  the 
dessicators,  (concentrated  sulphuric  acid,  potassa,  and  chloride 
of  calcium)  in  bubbles  about  2  seconds  apart.  After  half  an 
hour,  or  when  the  flow  of  the  gasses  became  slow,  the  slow  current 
of  air  was  replaced  by  a  more  rapid  current  of  oxygen,  and 
this  continued  till  the  anterior  copper  roll  became  fully  oxidized. 
The  heating  was  then  discontinued,  and  after  a  few  moments  a 
slow  current  of  air  resumed  for  a  short  time.  The  results  were 
very  satisfactory. 
Some  of  the  advantages  of  the  suggested  method  areas  follows. 
The  copper  surfaces  only  being  oxidized,  the  metallic  con- 
tinuity of  the  wire  cloth  serves  to  distribute  the  heat  equally  in 
all  directions,  whilst  the  part  of  the  tube  least  accessible  to  the 
