42  Mr.  L.  R.  Ingersoll  on  the  Faraday  and 
the  visible  spectrum  may  be  considered  as  accurately  known 
for  many  typical  substances,  measurements  in  the  ultra- 
violet and  infra-red  have  been  scanty,  and,  especially  in  the 
latter,  the  broader  and  more  important  field,  far  from 
satisfactory. 
As  a  reason  for  this  lack  of  attention  to  the  infra-red,  it 
may  be  urged  that  the  problem  of  measuring  rotations  in  this 
part  of  the  spectrum  is  one  of  rather  peculiar  difficulty, 
necessitating  as  it  does  magnetic  and  electrical  complications 
in  spectrobolometric  work  which  requires  the  highest  order 
of  sensibility.  But  even  if  an  accuracy  equal  to  that  of 
visual  measurements  cannot  be  expected,  the  extension  of  the 
work  into  this  field  is  nevertheless  a  matter  to  which  con- 
siderable interest  attaches  ;  for  it  is  there  that  evidence  must 
be  sought  in  support  of  rotatory  dispersion  formulae,  which 
cannot  as  yet  be  considered  as  generally  established,  because 
of  the  narrow  range  of  spectrum  only  over  which  they  have 
been  tested.  The  effect  of  absorption-bands  on  rotation  must 
also  be  considered  in  this  region,  and  whether  or  not  their 
action  may  be  accounted  for  by  a  type  of  dispersion  formula 
analogous  to  the  Ketteler-Helmholtz.  It  is  in  the  infra-red, 
too;  that  one  must  seek  for  an  explanation  of  the  anomalous 
character  of  the  rotatory  dispersion  shown  by  magnetic 
metals,  and  question  whether  the  rotation  continues  to  increase 
with  increasing  wave-length  throughout  the  infra-red  spec- 
trum, and  therefore  perhaps  be  very  considerable  for  electro- 
magnetic waves  metres  long. 
It  was  in  the  hope  of  throwing  a  little  more  light  on  these 
and  similar  questions,  that  the  present  work  was  attempted. 
The  two  distinct  cases  which  arise  will  be  treated  separately. 
They  are  (1)  Faraday  rotation,  or  rotation  suffered  by  the 
plane  of  polarization  of  light  transmitted  through  some  sub- 
stance in  a  magnetic  field;  and  (2)  the  Kerr  effect,  or  rotation 
due  to  reflexion  at  the  polished  pole-face  of  a  magnet,  or  at 
a  paramagnetic  mirror  in  a  magnetic  field. 
Part  I. 
Faraday  Rotation. 
That  radiant  heat  as  well  as  light  may  show  Faraday 
rotation  was  first  proved  by  Wartmann  *,  and  measurements 
of  the  rotation  for  a  number  of  different  substances  were 
later    made    by    De   la    Provostaye    and    Desains  f   and  by 
*  Compt.  Rend.  xxii.  p.  745  (1846). 
f  Ann.  Chim.  Phys.  [3]  xxvii.  p.  232  (1840). 
