70  Mr.  L.  R„  Ingersoll  on  the  Faraday  and 
results,  both  in  visual  and  infra-red  observations,  and  while 
the  existing  arrangement  could  not  be  considered  as  very 
suitable  for  the  detection  of  minute  rotations,  still  it  may  be 
stated  with  reasonable  certainty  that  the  effect,  if  it  exists  at 
all,  is  less  than  a  tenth  of  the  rotation  for  steel. 
Interpretation  of  Residts. 
As  to  the  physical  interpretation  of  the  curves,  not  as 
certain  conclusions  can  be  drawn  now  as,  it  is  hoped,  may  be 
possible  after  the  work  has  been  extended  as  already  men- 
tioned, bnfc  certain  points  may  be  considered.  The  most 
evident  feature  is  the  striking  similarity  between  the  com- 
pleted curves  as  given  in  iig.  6  and  a  typical  dispersion-curve 
running  through  an  absorption-band.  JSTow  the  general 
theory  of  magnetic  rotation  calls  for  an  anomalous  character 
of  the  rotatory  dispersion -curve  in  the  neighbourhood  of  an 
absorption- band,  very  similar  in  fact  to  the  case  of  ordinary 
dispersion,  and  indeed  intimately  connected  with  it  in  theory. 
Whether  the  curve  should  have  oppositely  directed  branches 
on  the  two  sides  of  the  band,  or  not,  depends  on  whether  the 
hypothesis  of  molecular  currents,  or  of  the  Hall  effect,  be 
admitted  in  forming  the  equations.  (See  Drude's  '  Optics/) 
The  latter  condition  is  now  well  known  to  occur  in  the  case 
of  sodium  vapour,  while  curves  of  the  former  type  have  been 
obtained  by  Schmauss  *,  and  quite  recently  by  Wood  f,  who 
worked  with  a  saturated  solution  of  praseodymium  chloride. 
In  view  of  this,  the  most  natural  conclusion  to  be  drawn  from 
the  curves  above  given  is  that  the  magnetic  metals  present  a 
case  of  anomalous  rotatory  dispersion  similar  to  that  found  in 
certain  absorbing  solutions,  with  the  difference  that  the  region 
of  resonance-absorption,  if  such  it  be,  instead  of  being  very 
limited  in  extent,  covers  nearly  the  whole  visible  spectrum  as 
well  as  a  small  part  of  the  infra-red  ;  in  the  case  of  steel  it 
may  even  extend  into  the  ultra-violet.  This  might  be  con- 
strued as  evidence  pointing  to  the  existence  of  free  resonance 
periods  in  metals,  although  the  definite  vibration  time  charac- 
teristic of  common  cases  of  resonance  would  seem  to  be 
lacking.  This  general  view  is  somewhat  strengthened  by  the 
fact  that  the  magnetic  metals  show  an  increase  of  refractive 
index,  as  well  as  of  magnetic  rotation,  towards  the  red,  in  the 
visible  spectrum ;  and  while  the  dispersion-curves  as  given  by 
du  Bois  and   rhibens  J  show  a  regular  rise  without  maxima 
*  Ann.  der  Phys.  ii.  p.  281  (1900),  and  also  several  later  papers  in  this 
publication. 
t  Phil.  Mag.  Mav  1905. 
\  Wied.  Ann.  xl'i.  p.  522  (1890). 
