504 
THE  ADmEALTY  ASTEONOmC.AL  EXPEEI:MENT 
low  altitudes,  I tried  it  on  the  sun  at  mid-day  in  the  same  manner ; when  an  immense 
increase  in  the  number  and  definition  of  the  black  lines  at  the  violet  end  took  place. 
With  the  reflected  ray,  the  two  bars  of  H could  but  just  be  discerned  as  faint  nebulous 
streaks,  B,  C,  and  D being  as  sharp  and  as  black  as  silver  wires  in  a telescope ; with 
the  direct  ray  the  individual  hues  composing  the  bars  of  H could  be  distinctly  separated, 
and  many  lines  appeared  indistinctly  in  the  space  beyond.  The  conclusion  thence  to 
be  derived  was,  unfortunately,  that  our  particular  speculum  did  not  reflect  the  violet 
end  of  the  spectrum ; and  our  observations  were  therefore  not  comparable  in  that  part, 
even  if  they  were  in  others,  with  standard  observations  elsewhere.  The  discovery  was 
made  too  late  to  enable  a direct  determination  to  be  made  for  Guajara,  but  the  method 
was  employed  for  Orotava  when  we  returned  there  from  the  mountain  a few  days  after. 
Red  End  of  the  Solar  Spectrum. 
In  the  accompanying  Table  (Plate  XXXV,)  are  arranged  the  whole  of  the  drawings 
of  the  red  end  of  the  spectrums  that  were  taken  at  the  three  stations,  on  the  same  scale 
as  Fraunhofee’s  justly  celebrated  Munich  engra'ving.  The  times  and  circumstances  of 
the  observations  being  appended,  will  enable  any  one  to  judge  of  the  realitj'  of  the  facts 
purporting  there  to  be  represented. 
Bearing  in  mind  that  great  accuracy  of  position  is  not  pretended,  and  that  though 
much  care  was  bestowed  on  the  general  appearance,  thickness,  and  definition  of  the  lines, 
the  shortness  of  the  time  available  was  entirely  inadequate  to  procmmig  a good  dmwing. 
and  that  several  of  the  diagrams  should  be  employed  together  in  deducing  a result, — 
we  may  proceed  to  the  examination.  Comparing  the  eleven  Tenerifie  spectrums  of  the 
sun  with  Fraunhofer’s,  we  can  only  assume  identity  in  place  of  the  lines  A,  a.  B.  C. 
D,  E,  and  h : everything  else  appears  differently.  Of  this  difference  the  prevailing 
feature  appears  to  be,  that  whereas  Fraunhofer’s  spectrum  stretches  to  beyond  A. 
which  is  seen  by  him  as  a clear  and  distinct  hne, — he  gives  none  of  the  numerous  broad 
bands  and  groups  of  lines  between  A and  «,  and  a and  B,  that  were  Gsible  on  the 
mountain  and  below,  whenever  A was  quite  or  nearly  Gsible. 
From  the  fact  of  Fraunhofer’s  spectrum  including  A,  we  might  be  entitled  to  expect 
from  the  Teneriffe  observations,  that  it  represents  the  sunset  appearance  (see  Xo.  10)  ; 
but  then  how  can  the  omission  of  the  broad  bands  of  lines  between  A and  B,  and  C 
and  D,  and  especially  those  beyond  D,  be  explained  % If,  on  the  contrary,  the  absence  of 
those  marked  bands  is  to  be  regarded  as  a proof  that  a high  spectrum  was  intended. 
Nos.  1,  2,  and  3 indicate  that  A should  not  be  seen  in  such  a position.  Again,  while 
the  spaces  between  A a and  aB  are  blank  in  Fraunhofer,  and  well  filled  in  Tene- 
riffe,  he  has  some  lines  between  B and  C,  a compartment  always  remarkably  empty 
with  us. 
Comparing  the  Teneriffe  observations  inter  se,  we  may  assume  2 and  3 as  giving  a 
near  approach  to  a zenith  solar  spectrum,  and  4,  5,  6,  7,  8,  0,  10  as  gmng  the  effects 
superinduced  by  the  sun  being  seen  through  a greater  thickness  of  the  atmosphere  by 
