Secondary  Bontgen  Radiation.  819 
Diation  and  which  gave  no  evidence  of  polarization  o£  the 
primary  beam,  was  extremely  little  affected  by  consider- 
able changes  in  the  character  o£  the  primary,  though  in 
all  cases  it  appeared  slightly  more  penetrating  when  a  pene- 
trating primary  beam  was  used.  The  change  was  remarkably 
small  in  some  cases. 
The  slight  change  in  the  character  of  the  secondary  ra- 
diation from  copper  when  the  primary  was  changed  is  shown 
by  the  following  results. 
The  absorptiuii  of  the  secondary  beam  from  copper  by  a 
plate  of  aluminium  "01  cm.  in  thickness,  when  the  primary 
beam  came  direct  from  an  X-ray  tube,  was  found  to  be 
72*75  per  cent.  When  only  the  penetrating  portion  of  the 
primary  which  had  got  through  an  aluminium  plate  '079  cm. 
thick  was  used  as  the  primary  beam,  the  absorption  of  the 
secondary  was  found  to  be  71*8  per  cent.  Now  the  former 
primary  beam  used  was  found  to  be  absorbed  to  the  extent  of 
about  35  per  cent,  by  the  same  plate  of  aluminium,  whereas 
the  second  primary  beam  was  absorbed  by  only  10  or  12 
per  cent. 
But  by  placing  the  aluminium  plate  '079  cm.  thick  in  the 
primary  beam  before  it  fell  on  the  radiator  of  copper,  the 
ionization  produced  by  the  primary  was  reduced  to  14  per 
cent.,  and  that  by  the  secondary  to  19  per  cent,  of  the  original 
ionizations. 
We  thus  see  that  though  81  per  cent,  of  the  secondary 
ionization  was  produced  by  the  secondary  rays  set  up  by  the 
absorbable  portion  of  the  primary  beam,  yet  when  this  was 
cut  off  the  enormously  more  penetrating  portion  of  the  primary 
set  up  secondary  rays  differing  in  absorbability  by  something 
of  the  order  of  1  per  cent. 
When  iron  was  used  as  the  radiator,  the  change  in  character 
was  still  less,  the  absorption  of  the  secondary  set  up  by  the 
primary  beam  direct  from  the  bulb  being  90'b*  per  cent., 
90*9  per  cent.,  90*9  per  cent.,  and  90*5  percent,  in  successive 
experiments ;  while  it  was  90*2  per  cent,  and  90  per  cent, 
when  the  thick  aluminium  plate  was  placed  in  the  primary 
beam. 
The  absorption  of  the  radiation  from  lead  appeared  much 
more  variable.  From  several  experiments  the  absorption  of 
the  secondary  beams  set  up  in  the  same  way  in  lead  as  in  the 
two  experiments  on  copper,  were  found  to  be  about  39  per  cent, 
and  35  per  cent.,  indicating  a  greater  dependence  on  the 
character  of  the  primary. 
As  a  general  rule,  it  was  also  found  that  experiments 
on  those  substances    which   emitted  a  very  easily  absorbed 
