348     M»  F.  Zollner  on  the  Origin  of  the  Earth's  Magnetism, 
ence  in  the  velocity  of  the  layers  is  observed,  but  in  the  inverse 
direction  of  that  on  the  sun's  surface. 
3. 
We  have  seen  that  the  two  causes,  which  alone  are  sufficient 
to  produce  the  phenomena  of  motion  observed  on  the  sun's  sur- 
face, are  a  continuous  flow  of  heat  into  space  (that  is,  a  perma- 
nent loss  of  heat  of  the  surface)  and  the  rotation  of  the  globe. 
The  sun  in  its  present  condition  loses  its  heat  directly  by  radia- 
tion of  the  glowing  liquid  mass.  In  a  later  state,  however, 
when  the  whole  surface  will  be  covered  by  a  solid  crust,  the  loss 
of  heat  of  the  liquid  mass  will  take  place  by  conduction.  The 
inner  parts  will  give  up  their  heat  to  the  cooler  crust,  which  in 
its  turn  loses  it  again  by  radiation.  Hence  it  is  to  be  seen  that 
immediately  under  the  solid  crust  the  essential  conditions  re- 
quired for  the  circulation  are  preserved.  The  liquid  masses 
cooled  down  by  contact  and  conduction  will  sink  down  as  before 
at  the  poles,  giving  rise  to  a  polar  undercurrent  which  is  ac- 
companied by  an  equatorial  upper-current.  This  equatorial  cur- 
rent will  touch  continually  the  solid  crust,  and  suffer  a  westerly 
deflection  analogous  to  that  of  the  upper  trade-winds.  For  the 
inhabitants  of  such  a  globe  the  direction  of  the  curents  in  the 
liquid  nucleus  below  their  feet  will  result  from  an  equatorial 
component  and  a  component  in  the  direction  of  rotation.  If, 
for  instance,  our  earth  were  a  globe  of  such  properties,  these  cur- 
rents would  have  a  south-westerly  direction  in  the  northern  he- 
misphere, whilst  they  would  have  on  the  southern  hemisphere  a 
north-westerly  direction. 
4. 
We  must  now  inquire  whether  the  motion  in  the  liquid  nu- 
cleus could  make  itself  perceptible  in  any  way  on  the  surface. 
As  long  as  the  crust  is  sufficiently  thin  to  be  affected  by  the 
changes  necessarily  accompanying  such  a  motion  (as  the  ice  co- 
vering a  river  is  affected  by  the  water  flowing  underneath),  such 
mechanical  influences  could  be  observed  and  measured  by  instru- 
ments comparing  the  direction  of  gravity  with  that  of  a  solid 
body  rigidly  connected  with  the  crust. 
But  as  the  thickness,  and  hence  also  the  relative  rigidity  of 
the  crust,  increases  in  such  a  manner  that  these  mechanical  in- 
fluences could  no  longer  be  observed,  except  perhaps  by  very 
delicate  instruments  indeed,  changes  extending  over  a  longer 
period  and  affecting  the  intensity  of  the  whole  streaming  mass 
could  still  make  themselves  perceptible  by  thermic  changes  in 
the  solid  crust.  For  whatever  may  cause  such  an  increase  or 
decrease  in  the  intensity  of  the  whole  streaming  process  in  the 
