Intelligence  and  Miscellaneous  Articles,  319 
ment  the  dispersion  of  the  luminous  pencil  at  its  issue  from  the  liquid, 
the  bands  spread  over  a  larger  surface,  and  lose  their  intensity ;  so 
that  we  gain  nothing  for  the  observation  of  the  absorption-spectrum 
of  a  liquid  by  using  a  spectroscope  with  several  prisms.  It  is  not 
the  same  when  we  have  to  do  with  vapours,  almost  all  the  lines  of 
which  are  very  fine,  and  appear  more  distinctly  the  more  the  lumi- 
nous pencil  is  spread  out  and  the  source  whence  it  emanates  is  more 
intense. 
This  remark  led  me  to  recognize  the  existence  of  the  absorption- 
spectra  of  chlorine  and  chloride  of  iodine.  In  the  examination  of 
that  of  chlorine,  I  had  to  take  particular  care  to  operate  on  pure  gas. 
In  fact,  the  phenomenon  being  insensible  for  a  thickness  of  some 
decimetres,  as  the  fruitless  endeavours  of  some  physicists  have  shown, 
it  was  necessary  to  employ  a  gaseous  column  of  greater  dimensions; 
and  as  very  small  quantities  of  chlorous  and  hypochloric  acids  were 
sufficient  to  exhibit  the  absorption-lines  of  these  substances,  the  im- 
purities of  the  gas  prepared  with  hydrochloric  acid  and  binoxide  of 
manganese  might,  in  the  conditions  of  the  experiment,  have  caused 
the  appearance  of  the  absorption-lines  of  the  chlorinated  compounds. 
To  destroy  these  products,  which  do  not  resist  the  action  of  heat,  I 
passed  the  dry  chlorine  through  a  glass  tube  heated  to  redness. 
On  causing  the  light  to  pass  along  the  axis  of  a  tube  of  15  decims. 
length  filled,  under  the  atmospheric  pressure,  with  chlorine  thus 
purified,  I  could  distinguish  very  clearly  the  absorption-lines  of  this 
gas  ;  but  the  phenomenon  is  more  striking  in  an  apparatus  of  greater 
length.  The  one  I  made  use  of  was  composed  of  three  glass  collars 
6  centimetres  in  diameter,  adjusted  end  to  end  so  as  to  form  a  tube 
of  468  decims.  length;  this  was  closed  at  its  two  extremities  with 
parallel  disks  of  plate-glass.  Having  placed  it  erect  in  order  to  fill 
it  with  chlorine  by  bringing  the  gas  through  a  tube  to  its  lower  part 
while  the  air  escaped  through  an  aperture  made  at  its  upper  extre- 
mity, I  placed  it  horizontal  and  caused  to  pass  along  its  axis  a  pencil 
of  the  Drummond  light.  On  emerging  from  the  gas,  the  rays  fell  on 
the  slit  of  a  spectroscope  with  two  prisms,  and  gave  a  spectrum  ex- 
tending into  the  violet  and  marked  with  very  distinct  lines.  In  the 
least-refrangible  region  and  as  far  as  the  place  of  the  line  D  the 
spectrum  is  continuous  ;  but  a  little  beyond  that  a  system  of  lines 
commences  which  presents  on  analogy  with  the  fine,  almost  equi- 
distant lines  observed  in  the  vapours  of  bromine  and  iodine.  Their 
aspect  and  intensity  vary  with  the  region  of  the  spectrum  examined ; 
and  they  extend  nearly  to  the  violet,  which  is  entirely  absorbed  in 
the  case  of  the  luminous  source  in  question. 
The  protochloride  of  iodine  is  much  more  favourable  than  chlorine 
for  the  observation  of  the  lines  of  absorption.  At  the  temperature 
of  40°  C,  a  thickness  of  30  centims.  of  this  substance  furnishes  suf- 
ficient vapour  to  produce  an  absorption-spectrum  composed  of  a  score 
of  fine  lines,  sensibly  equal  in  intensity,  their  distance  diminishing 
very  little  from  the  extreme  red,  where  they  commence,  to  a  little 
beyond  the  line  D,  where  they  end ;  two  other  lines,  rather  intense, 
