Corona. 
the  Eclipse  of  December  l\th  and  Uth,  1871.  193 
the  atmosphere,  and  apparently  not  met  with  elsewhere  in  the 
earth,  forms  at  all  events  one  of  the  component  parts  of  the 
corona.  But  though  we  learn,  from  the  fact  of  the  line  being 
bright,  that  this  gas  is  incandescent,  and  from  its  proximity  to 
the  sun  we  should  scarcely  expect  any  thing  else,  yet  this  fact 
in  no  way  renders  it  impossible  that  much  of  the  light  we  re- 
ceive from  the  corona  should  be  reflected  or  scattered  by  minute 
particles  of  perhaps  denser  matter,  probably  incipient  cloud, 
suspended  within  it,  as  such  particles  are  supposed  to  exist  in 
the  earth's  atmosphere  in  order  to  account  for  the  polarization 
and  blue  colour  of  the  sky. 
It  is  well  known  now  that  the  polarizing-angle  of  such  mi- 
nute particles  is  45°;  and  it  is  evident 
that  a  much  greater  proportion  of  light 
reflected  at  this  angle  would  be  received 
from  the  portions  of  the  corona  at  a  dis-A 
tance  from  the  sun  than  from  those  close 
to  its  limb.  The  annexed  woodcut  will 
illustrate  my  meaning.  It  is  evident  that  ! 
only  in  case  of  the  ray  A,  and  rays  par- 
allel to  it,  will  the  angle  of  incidence  be 
45°.  The  light  coming  from  portions  of 
the  corona  nearer  to  the  limb  will  evi- 
dently consist  more  and  more  of  light 
incident  at  other  angles  than  45°;  and 
consequently  the  proportion  of  polarized 
light  will  be  less  and  less  as  we  approach 
the  sun. 
With  regard  to  the  photographs  of  the 
corona,  the  boundary  seems  in  all  to  be  well  defined ;  and  the 
streamers,  strange  to  say,  do  not  appear  in  any  of  the  photo- 
graphs I  have  seen.  Whatever,  therefore,  they  may  be,  the 
light  emitted  from  them  does  not  appear  to  be  nearly  so  actinic 
as  that  received  from  the  corona  proper.  I  have  not  tested 
them  for  polarization,  and  can  therefore  say  little  about  them. 
It  seems  to  me,  however,  impossible  that  a  gaseous  envelope 
round  the  sun  could  possibly  take  the  very  irregular  form  their 
appearance  would  indicate. 
The  glass  plates  used  in  the  polarimeter  were  the  thin  plates 
used  for  mountiug  microscopic  objects.  I  have  not  yet  deter- 
mined their  refractive  index  ;  but,  supposing  them  to  be  of 
crown-glass  and  that  p=l*54,  then,  from  the  Table  given  in 
Professor  W.  G.  Adams's  paper  in  the  last  March  Number  of 
this  Magazine,  we  get  the  following  figures  for  the  proportion 
of  polarized  light,  as  shown  by  my  measurements  : — 
Phil.  Mag.  S.  4.  Vol.  43.  No.  285.  March  1872.         0 
