426     Prof.  Challis  on  the  Hydro  dynamical  Theory  of  Magnetism. 
night- time;  and  this  anticipation  is  confirmed  by  observation. 
But  besides  the  magnetic  variations  due  to  changes  of  the  tem- 
perature and  density  of  the  atmosphere,  we  may  suppose  that 
effects  are  produced  by  solar  gyrations  analogous  to  those  attri- 
buted above  to  lunar  gyrations.  This  cause  might  modify  the 
epoch  of  day-maximum  (which,  as  observation  shows,  does  not 
occur  at  the  hottest  time  of  the  day),  and  might  also  account  for 
what  has  been  called  the  nocturnal  episode.  As  to  the  reality  of 
the  solar  gyrations,  I  consider  that  we  have  ocular  evidence  in 
the  phenomenon  of  the  zodiacal  light ;  for,  according  to  the  ar- 
gument maintained  in  arts.  13  and  14,  the  rays  which  render 
the  zodiacal  light  visible  might  originate  in  a  collision  between 
the  gyratory  motion  and  vibrations  of  different  orders  propagated 
from  the  sun,  without  its  being  necessary  to  suppose  that  matter 
other  than  the  aather  exists  at  points  from  which  the  rays  pro- 
ceed. It  has  been  established  by  observation  that  the  zodiacal 
light  extends  beyond  the  radius  of  the  earths  orbit  (see  an 
article  in  the  Phil.  Mag.  for  February  1863). 
45.  The  foregoing  considerations  enable  me  now  to  state  the 
cause  alluded  to  in  art.  6,  to  which  I  conceive  that  minute  an- 
nual inequalities  of  Dip,  Total  Force,  and  Declination  might  be 
attributed.  Let  it  be  admitted  that  the  solar  gyrations  produce 
a  sensible  magnetic  effect  at  the  earth's  distance  from  the  sun. 
Then  on  the  supposition  that  the  gyrations  are  in  circles  sym- 
metrically disposed  relatively  to  the  plane  of  the  ecliptic  (which 
must  approximately  be  the  case),  it  will  follow  that  the  magnetic 
effect  varies  with  the  earth's  varying  distance  from  the  sun,  being 
less  as  the  distance  is  greater.  Inequalities  thus  produced  would 
be  the  same  for  the  northern  as  for  the  southern  hemisphere, 
agreeing  in  that  respect  with  the  small  annual  inequalities  which, 
as  stated  in  art.  6,  General  Sir  E.  Sabine  has  deduced  from  ob- 
servation. 
46.  On  the  hydrodynamical  principles  which  have  been  ap- 
plied in  the  foregoing  explanations,  I  make  the  following  sug- 
gestion relative  to  the  cause  of  the  secular  variations  of  ter- 
restrial magnetism.  From  what  has  been  already  argued  (art. 
42),  the  motions  immediately  impressed  on  the  aether  by  the 
earth's  constituent  atoms  in  consequence  of  its  rotatory  and  or- 
bital motions,  result  in  circulating  motions  which,  if  the  earth 
moved  uniformly  in  a  rectilinear  course,  would  be  steady  motions 
having  always  the  same  geometrical  relations  to  the  position  of  the 
earth's  centre.  But  the  earth  moves  with  varying  velocity  in  a 
curved  path.  Assuming,  therefore,  that  the  system  of  circulating 
motions  always  tends  to  be  steady  and  to  maintain  a  rectilinear 
course,  it  is  conceivable  that  there  may  be  continual  minute  shift- 
ings  of  the  axis  of  the  system  in  directions  transverse  to  the 
