50  M.  F.  Zollner  on  the  Spectroscopic  Observation 
longer  covered  the  line  when  the  light  of  portions  of  the  other 
limb  fell  on  the  slit.  In  order  to  avoid  flexures,  the  precaution 
was  taken  of  fixing  the  telescope  and  letting  the  image  of  the 
sun  pass  over  the  slit  by  the  diurnal  motion.  "With  direction  to 
the  vicinity  of  the  north  or  the  south  pole  of  the  sun  (where  no 
displacement  was  to  be  expected),  the  coincidence  of  point  and 
spectrum-line  remained  perfectly  unaltered;  and  this  was  a 
guarantee  that  any  alterations  that,  with  a  little  movement  of 
the  refractor,  may  happen  to  take  place  in  the  spectrum-appa- 
ratus are  so  small  as  to  be  without  influence  on  the  observations; 
"June  11.  The  observations  were  made  in  the  same  manned 
as  on  the  previous  day.  We  endeavoured,  by  numerous  estima- 
tions, to  fix  the  magnitude  of  the  displacement  of  the  lines 
towards  the  point  in  the  focus  of  the  telescope,  taking  as  unit 
the  distance  between  two  lines  near  together  in  the  spectrum; 
Our  accounts  varied  between  0010  and  0*015  millionth  of  a 
millimetre — which  would  give,  for  the  motion  of  a  point  in  the 
solar  equator,  a  velocity  of  0*42  mile  in  a  second. 
"June  15.  The  observations  were  carried  out  as  before;  only, 
instead  of  the  point  in  the  observing-telescope,  I  had  very  fine 
cross-threads  introduced.  A  twenty-four  times  magnifying- 
power  could  be  used  with  advantage ;  with  it  Fraunhofer's  lines 
appeared  extremely  sharp.  In  the  vicinity  of  the  line  F  and  the 
group  b  I  made  preliminary  estimations,  which  gave  0*008  mil- 
lionth of  a  millimetre  for  the  amount  of  the  displacement,  from 
which  0*35  mile  would  result  as  the  velocity  of  an  equatorial 
point.  It  is  a  striking  fact  that  the  observations  always  give  a 
greater  velocity  than  that  calculated  from  the  known  period  of 
revolution  of  the  sun ;  yet  it  would  be  hazardous  to  deduce  any 
conclusion  whatever  therefrom,  since,  on  the  one  hand,  the  esti* 
mates  are  very  uncertain,  and,  on  the  other,  the  wave-lengths 
of  the  lines  in  the  solar  spectrum  are  not  so  accurately  deter- 
mined that  the  uncertainty,  in  comparison  with  the  amount  to 
be  determined,  is  a  vanishing  quantity.  Thus  much  only  results 
from  all  the  observations — that  a  displacement  of  the  lines  by 
the  rotation  of  the  sun  may  be  regarded  as  certainly  proved." 
As  an  appendix  to  these  results  (which  evidently  promise  for 
the  reversion-spectroscope  a  very  extended  application  in  future 
for  quantitative  determinations  in  spectrum- analytical  investi- 
gations), I  take  leave  to  communicate  the  construction  of  a  new 
and  considerably  simplified  reversion-spectroscope.  I  have 
already  alluded  to  it,  in  the  description  of  the  one  before  men- 
tioned * ;  but  since  then  I  have  been  so  convinced  of  its  practical 
utility,  that  I  am  sure  the  principle  can  be  applied  with  great 
*  Berichte  derK.  Sachs.  Gesellsch.  d.  Wiss,  math.-phys.  Classe,  Feb.  &, 
1869,  p.  73. 
