Secondary  Rontgen  Radiation.  821 
due  to  changes  in  the  primary  radiation,  though  the  absorption 
of  some  of  these  beams  was  greater  than  that  of  some  secon- 
dary radiations  whose  character  appeared  very  constant  when 
that  of  the  primary  was  varied. 
In  general  the  substances  from  silver  to  iodine  (see  fig.  1) 
emitted  secondary  rays  showing  remarkable  variations  in 
absorbability  in  different  experiments.  The  behaviour  of 
these  substances  and  of  those  in  the  middle  of  the  first  long- 
chemical  period  form  a  striking  contrast. 
This  points  to  the  conclusion,  though  the  evidence  is 
not  decisive,  that  at  least  some  substances  emit  a  radiation 
differing  considerably  in  absorbability  from  the  primary,  and 
exhibiting  considerable  variation  in  character  as  the  primary 
is  varied. 
[Note,  May  14th,  1906. — More  recent  experiments  with 
thin  metal  leaves  have  shown  this  conclusively.] 
Polarization. 
Further  information  regarding  the  nature  and  cause  of 
the  radiation  was  obtained  by  using  a  partially  polarized  beam 
of  X-radiation  for  the  primary,  and  measuring  the  intensity 
of  radiation  proceeding  in  different  directions  from  the  sub- 
stances investigated.  It  was  shown  in  the  paper  on  "  Polarized 
Rontgen  Radiation,"  that  from  those  substances,  in  which, 
during  the  passage  of  a  Rontgen  pulse,  the  corpuscles  or 
electrons  are  accelerated  in  the  direction  of  electric  displace- 
ment, the  secondary  radiation  differs  in  intensity  in  different 
directions,  giving  evidence  of  the  polarization  in  the  primary 
beam.  These  substances  are  also  the  origin  of  polarized 
secondary  radiation.  When  the  acceleration  of  electrons 
ceases  to  be  in  the  same  direction  as  the  electric  displacement 
in  the  primary  pulse,  evidence  of  polarization  of  the  primary 
disappears,  and  the  secondary  radiation  ceases  to  be  polarized. 
Conversely,  the  disappearance  of  evidence  of  polarization,^,  e. 
of  the  variation  of  intensity  of  secondary  radiation  in  different 
directions  at  right  angles  to  the  primary,  when  polarization 
has  been  demonstrated,  shows  that  the  secondary  radiation 
ceases  to  be  polarized  and  that  the  electrons  cannot  now  be 
accelerated  in  the  direction  of  electric  displacement  during 
the  passage  of  Rontgen  pulses. 
To  test  the  nature  of  the  secondary  pulses  from  this  point 
of  view,  experiments  were  made  using  different  substances  as 
radiators,  and  it  was  found  that  approximately  equal  evidence 
of  polarization  was  given  by  all  those  substances  of  low 
atomic  weight  which  emitted  a  radiation  whose   character 
