^'"  si°p""x87t'"'-}    Development  of  the  Chemical  Arts.  421 
which  agrees  with  the  former,  and  which  scarcely  exceeds  that  of  coal- 
gas.  We  may  regard  this  process  as  the  final  and  successful  solution 
of  the  problem  as  to  the  economical  and  rational  production  of  oxygen. 
We  have  still  to  review  a  group  of  projects  which,  without  any 
chemical  agents,  aim  at  extracting  oxygen  from  the  atmosphere  by  a 
purely  mechanical  procedure.  They  are  based  upon  two  physical  prin- 
ciples, diffusion  or  absorption. 
Th.  Graham  who,  in  his  classical  researches,  investigated  the  laws  of 
the  escape  of  gases  through  narrow  apertures,  made  known  in  1866  f 
that  air  which  is  drawn  through  a  fine  chink  in  a  plate  of  caoutchouc 
passes  in  the  constant  proportion  of  41*6  per  cent,  of  oxygen  to  58*4 
per  cent,  of  nitrogen,  the  half  of  the  atmospheric  nitrogen  being  held 
back.  This  mixture  causes  glowing  chips  of  wood  to  burst  into  flame. 
Deville  j  tested  the  industrial  value  of  this  process,  and  found  that  the 
time  required  was  too  long. 
Absorption  has  been  utilized  in  two  distinct  forms.  Montmagnon 
and  De  Laire,  in  1868,  took  out  a  French  patent,  ||  based  upon  the 
observation  of  Angus  Smith,  §  that  charcoal  absorbs  from  the  air  more 
oxygen  than  nitrogen.  According  to  them,  100  litres  of  wood-char- 
coal absorb  925  litres  of  oxygen  and  only  750  litres  of  nitrogen.  If 
moistened  with  water,  they  give  off*  350  litres  of  oxygen  and  650  litres 
of  nitrogen,  so  that  575  litres  of  oxygen  and  55  (100  ?)  litres  of  nitrogen 
remain  and  can  be  extracted  with  the  air-pump.  By  repeating  this  pro- 
cess with  the  same  gaseous  mixture,  they  succeeded  in  bringing  the 
oxygen  almost  in  a  state  of  purity.  Whether  this  process  has  ever 
been  carried  out  on  the  large  scale  is  not  known.  An  attempt  has, 
however,  been  made  with  Mallet's  method,^!  based  on  the  property  of 
water  to  absorb  oxygen  rather  than  nitrogen. 
The  coefficients  of  absorption  of  the  two  gases  are  0*025  N",  and 
0*064  for  O.  If  multiplied  by  the  proportion  of  their  bulk  in  the  at- 
mosphere, 0*79  for  N,  and  o-2i  for  O,  these  numbers  give  the  vol- 
ume-proportion of  both  gases  in  water  =^0-0197  N,  and  0*0097  ^  ' 
*  or,  the  air  absorbed  in  water  contains  in  one  volume,  0*67  N,  and  0*33 
O.    If  the  unabsorbed  nitrogen  is' allowed  to  escape,  and  the  absorbed 
t  Graham,  "  Comptes  Rendus,"  Ixiii,  471. 
J  Deville,  Wagner,  "  Jahresberichte,"  1867,  216. 
II  "  Bull,  de  la  Soc.  Chim."  [2],  xl,  261. 
^  Angus  Smith,  "  Proc.  Roy.  Soc,"  xii,  424, 
^[  Mallet,     Dingler's  Polyt.  Journ.,"  cic,  112. 
