16 
Mr.  N.  R.  Campbell  on  the 
Table  I. 
f=6. 
a=1. 
a=8. 
a  =  9. 
I 
a=10.  U=ll.  U=12. 
r/  =  13.  a  =  14. 
x=    1. 
•4839 
•4308 
•3881 
•3532 
•3239 
•2992 
•2779 
•2596 
•2434 
*=  2. 
•7559 
•6932 
•6394 
•5922 
•5513 
•5154 
•4839 
•4558 
•4308 
#=  3. 
•9034 
•8520 
•8023 
•7559 
•7133 
•6744 
•6392 
•6071    -5778 
x=  4. 
•9728 
•9408 
•9034 
•8648 
•8268 
•7903 
•7559 
•7235)  -6932 
a?=  5. 
•9962 
•9831 
•9609 
•9334 
•9034 
•8725 
•8384 
•8119  j  -7836 
a-=  6. 
1-000 
•9980 
•9889 
•9728 
•9522 
•9286 
•9034 
•8776    -8520 
x=  7. 
1-000 
•9987 
•9923 
•9804 
•9645 
•9457 
•9249 
•9034 
a?=  8. 
1-000 
•9990 
•9943 
•9854 
•9728 
•9576 
•9408 
:t=  9. 
...      1-000 
•9993 
•9957 
•9889 
•9788 
•9664 
#=10. 
1000 
•9995 
•9962 
•9912 
•9831 
a?=ll. 
1-000 
•9996 
■9974 
•9930 
■   a?=12. 
... 
1-000 
•9996 
•9980 
a?=13. 
1-000 
•9997 
re =14. 
...      1-000 
1 
It  should  be  noted  that  in  all  cases  except  that  of  platinum, 
where  the  secondary  radiation  is  so  far  larger  than  the  in- 
trinsic, the  same  value,  within  half  a  unit,  would  have  been 
assigned  to  a,  if  it  had  been  calculated  Prom  curve  (A) 
instead  of  (E).  Hence  an  error  in  the  estimation  of  the 
proportion  in  which  the  radiation  was  cut  down  by  the 
screen  would  only  affect  the  absolute  values  of  the  intensities 
of  the  radiations,  and  would  leave  the  penetration  constants 
unchanged. 
§  10,  fi  was  determined  by  subsidiary  experiments.  After 
the  form  of  the  screened  curve  had  been  determined  for  any 
metal,  the  sliding  side  was  pushed  back  as  far  as  possible 
from  the  opposite  end.  Two  pieces  of  the  same  wire  gauze 
as  formed  the  cage  were  then  inserted  in  the  cage  parallel  to 
the  sliding  side,  each  being  about  10  cms.  from  the  nearer 
end.  The  ionization  in  the  portion  enclosed  entirely  by  wire 
gauze  was  then  measured ;  the  difference  between  this  ioni- 
zation and  the  value  of  vabx+fisc,  as  found  from  the  curve 
for  the  value  of  x  equal  to  the  distance  between  the  two 
sheets  of  wire  gauze,  must  be  due  to  the  rays  from  these  two 
sheets ;  for  the  distance  from  the  gauze  to  the  sides  of  the 
box  is  too  great  to  allow  the  rays  from  the  latter  to  have  any 
effect  at  all.  Thus  we  know  the  ionization  given  by  a  known 
area  of  the  gauze  and  the  quantity  p  can  be  calculated 
immediately.  _  It  was  found  to  be  small,  the  average  value 
being  3x10"'.  fj,  as  found  in  this  manner  has  been  sub- 
tracted from  each  ordinate  before  it  was  plotted  in  the 
diagrams.  Similar  observations  were  taken  for  the  unscreened 
curves,  and  the  value  3*5  x  10~7  found  for  /jl. 
