356     M.  F.  Zollner  on  the  Origin  of  the  Earth's  Magnetism, 
"  The  cause  that  the  magnetic  poles  of  the  earth  are  shifted 
relatively  to  the  geographical  poles,  lies  in  the  different  action  of 
this  backward  tendency ;  and  if  the  statement  which  has  only 
been  made  as  an  assertion  is  true,  we  ought  to  be  able  to  cal- 
culate the  situation  of  the  magnetic  poles  from  the  configuration 
of  land  and  water." 
It  appears  to  me,  from  the  indefinite  expression  "this  back- 
ward tendency/5  that  the  author  was  not  acquainted  with  the 
relations  shown  above  to  exist  between  electrical  currents  and 
currents  of  liquids.  At  any  rate,  the  fact  remains  qualitatively 
correct ;  as  to  the  sufficiency  of  the  quantitative  data  for  the  ex- 
planation of  the  facts  in  question,  I  do  not  take  upon  myself  to 
judge. 
The  result  of  the  numerical  calculation  is  certainly  remarkable 
for  the  very  exact  coincidence  with  Hans  teen's  observations  *. 
At  the  end  of  his  paper  the  author  makes  the  supposition 
that,  by  a  gradual  rising  or  falling  of  the  continents,  a  change  in 
the  position  of  the  magnetic  poles,  and  hence  secular  variations 
of  the  magnetic  constants,  may  be  effected.  According  to  my 
theory,  the  distribution  of  heat  and  cold  in  the  earth's  crust,  if 
with  decided  differences,  ought  to  be  of  great  influence.  A  great 
lowering  of  temperature  on  the  surface  will  cause  a  more  rapid 
cooling  at  the  particular  spot  in  the  crust  and  thus  cause  in  the 
liquid  streams  a  tendency  to  descend.  To  such  a  tendency  the 
poles  of  cold  owed  at  an  early  phase  their  formation.  I  have 
shown  in  my  paper  on  the  Law  of  the  Sun's  Rotation  that  the 
distribution  of  heat  on  the  sun's  surface,  as  observed  by  Secchi, 
may  be  very  simply  deduced  from  the  thermic  reactions  which 
the  atmospheric  currents  exert  on  the  glowing  surface.  As  this 
distribution  of  temperature  is  always  such  that  by  itself  it  would 
produce  the  currents  by  which  it  is  generated,  any  irregulari- 
ties which  at  the  earliest  state  of  development  had  influenced 
these  currents  would  increase  with  increasing  lowering  of  tem- 
perature, and  so  fix  permanently  the  position  of  the  poles  of  cold. 
It  is  to  be  seen  from  this  that  generally  these  poles  will  not 
coincide  with  the  poles' of  rotation,  on  the  sun  as  well  as  on  the 
earth.  It  follows  that  in  these  points  the  descending  currents 
will  predominate,  and  poles  of  currents  will  be  formed  which 
must  coincide  with  the  magnetic  poles. 
It  is  therefore  evident  that,  also  looked  at  from  this  point  of 
view,  permanent  differences  in  the  temperature  of  the  earth's 
surface  must,  if  great  enough,  have  an  influence  on  the  inner 
currents,  and  hence  also  on  the  earth's  magnetism. 
*  Hansteen  finds,  from  his  observations,  long,  of  the  magnetic  pole 
east  of  Ferro  290°  21 ',  lat.  69°  30';  the  calculations  of  Menzzer  give  long. 
289°  37'  2'8"-5,  lat.  69°  11'  53"'8. 
