492     M.  F.  Zollner  on  the  Origin  of  the  Earth's  Magnetism, 
end  of  this  wire  is  fixed  to  a  projection  20  millims.  long  at  the 
foot  of  a  vertical  brass  stand.  The  end  of  a  second  wire,  of  equal 
length,  is  attached  to  the  glass  rod  at  10  millims.  distance  from 
the  point  of  attachment  of  the  former,  and  is  fixed  to  a  projection 
at  the  upper  end  of  the  stand,  nearly  vertically  over  the  point  of 
attachment  of  the  first  wire. 
"  A  micrometer-screw  of  the  stand  enables  us  to  diminish  in- 
definitely the  momentum  of  direction  of  the  'horizontal  pen- 
dulum/ It  is  reduced  to  zero  if  the  two  points  of  attachment 
lie  in  the  same  normal,  supposing  we  neglect  the  torsion  of  the 
wires.  Having  regard  to  this  torsion,  however,  the  two  points 
must  lie  in  a  line  which  is  a  little  inclined  to  the  normal. 
"  The  described  horizontal  pendulum  was  put  up  in  the  cellar 
of  the  university  buildings,  which  is  12  feet  deep ;  the  tempera- 
ture in  it  is  very  constant.  A  mirror  attached  to  the  end  of  this 
pendulum  allows  us  to  read  off  the  changes  in  the  direction  ac- 
cording to  the  method  of  mirror-readings  on  a  scale  2500  mil- 
lims. distant  from  the  mirror.  It  is  easily  conceived  that  this 
apparatus  can  indicate  very  small  changes  of  the  normal  with 
respect  to  the  horizon.  When  the  time  of  vibration  was  brought 
to  52  seconds  by  means  of  the  tangent- screws,  a  deviation  of  10 
millims.  on  the  scale  corresponded  to  an  inclination  of  the  horizon 
amounting  to  0*035  second  of  an  arc ;  and  as  the  tenth  part  of 
a  division  may  be  estimated  with  surety,  000035  second  of  an 
arc  could  be  perceived.  One  can  easily  form  an  idea  of  the  sen- 
sitiveness of  the  apparatus  if  I  remark  that  the  difference  in  pres- 
sure which  was  caused  in  the  foundations  of  the  building  when 
a  lecture-room  on  the  second  floor  was  filled,  originated  a  devia- 
tion in  the  dot  of  light  of  twenty  divisions  of  the  scale.  When 
the  lecture-room  was  empty  again,  the  mirror  regained  its  former 
position.  This  instrument  is  therefore  at  the  same  time  an  ex- 
ceedingly sensitive  seismometer. 
"  The  apparatus  are,  of  course,  protected  by  suitable  covers 
from  the  currents  of  air  and  influences  of  radiant  heat." 
After  many  trials  and  much  trouble*  I  constructed  the  ap- 
*  I  discovered  in  the  above-mentioned  cellar  a  space  lying  about  15  feet 
deeper,  which  led  directly  to  the  bottom  of  the  foundations.  It  was,  how- 
ever, too  narrow  to  receive  the  telescope  and  scale  down ;  so  that  I  was 
obliged  to  content  myself  with  bringing  only  the  instrument  dowrn  and  re- 
flecting the  image  of  the  scale  into  a  telescope  directed  vertically  downwards. 
When  the  whole  instrument  was  protected  by  a  brass  cover,  which  had  only 
an  opening  closed  by  a  plane  glass  plate  opposite  the  mirror,  the  opening 
leading  to  this  subterranean  space  was  shut  up,  with  the  exception  of  the 
parts  necessary  for  the  reading-off.  The  instrument  was  thus  excellently 
secured  from  disturbing  changes  in  temperature.  But  the  above-mentioned 
elastic  compression  of  the  building,  or  rather  the  variation  of  level  in  the  fun- 
damental plane  of  the  building,  could  still  be  observed  ;  the  magnitude, 
however,  of  the  deviation  was  reduced  to  about  two  thirds  of  its  original  value. 
