and  the  Magnetic  Relations  of  the  Heavenly  Bodies.        457 
mentioned  above  from  a  great  number  of  observations  made  at 
New  Haven,  Boston,  and  Canada,  and  extending  over  a  range  of 
113  years.     The  comparison  yields  the  following  results  : — 
Frequency  of  Aurora  Boreales  as  dependent  on  the  time  of  the 
year, 
tit     ,1  Number  of 
Mouth.  ,        i 
aurorae  boreales. 
January  .....  173 
February      ....  210 
March 240 
April 267 
May 191 
June 179 
July 244 
August 238 
September    ....  293 
October 236 
November    .     .     .     .  215 
December     ....  159 
Here  also,  therefore,  the  maxima  are  in  the  same  two  months. 
"We  must,  however,  bear  in  mind  that,  in  consequence  of  the 
great  changes  of  the  radius  vector  of  the  earth  in  these  months, 
the  secondary  inductive  actions  between  sun  and  earth  must 
also  have  then  a  maximum,  so  that  only  by  this,  without  regard 
to  any  influence  of  the  position  of  the  sun's  axis,  the  magnetic 
induction  would  be  increased  in  these  two  months. 
So,  for  instance,  in  the  year  1870  the  greatest  velocity  with 
which  the  earth  was  going  away  from  the  sun  was  between  the 
24th  and  26th  of  March,  498-1  metres  a  second.  The  maximum 
value  for  its  approach,  between  the  1st  and  3rd  of  October,  was 
502*4  metres  a  second.  It  is  clear  that,  according  to  the  laws  of 
electrical  induction,  such  rapid  changes  of  the  distance  between 
the  earth  and  the  sun  must  cause  electrical  currents  in  these 
bodies,  even  if  only  one  of  them  exerted  a  magnetic  action 
upon  the  other.  How  great  the  intensity  of  these  currents 
would  have  to  be  in  order  to  produce  aurorse  boreales  and  mag- 
nctical  disturbances  on  the  earth's  surface  can  only  be  decided 
by  observation.  It  is  sufficient  for  us  to  have  shown  the  neces- 
sity of  the  existence  of  these  currents  without  regard  to  their  in- 
tensity. 
17. 
The  above  inquiries  have  shown  that  the  consequences  de- 
duced by  the  theory  are  confirmed  by  facts  proved  by  observa- 
tion. All  phenomena  on  the  solar  surface  which  are  connected 
with  changes  in  the  streams  necessarily  call  forth   analogous 
