a  Particles  of  Uranium  and  Thorium. 
765 
The  results  are  in  rather  better  agreement  with  theory  than 
the  last  column  seems  at  first  sight  to  show  ;  for  the  first  and 
third  observations  fell  on  such  points  of  the  curve  that  errors 
of  measurement  were  magnified  in  calculating  the  result. 
The  second  measurement  has  the  greatest  chance  of  being- 
accurate,  and  is  very  nearly  the  mean  of  the  three.  An 
independent  experiment  showed  that  tinfoil  for  which  pd  = 
•000536  was  equivalent  to  1*88  cm.  of  air.  Hence  the  range 
of  the  a  particle  found  by  this  method  =  210  x  1-88/535  -  7*4. 
The  value  found  by  Kleeman  and  myself,  using  the  direct  and 
more  accurate  method,  was  7*1  nearlv. 
Thorium. 
The  material  was  used  in  the  form  of  thorium  oxide  which 
had  been  freed  as  far  as  possible  from  other  radioactive 
substances  by  means  of  the  processes  described  by  Rutherford 
and  Soddy.  The  treatment  employed,  which  included  heating 
to  a  bright  red  heat  as  the  final  stage,  was  judged  to  have 
been  successful  for  the  following  reasons.  In  the  first  place, 
the  recovery  of  activity  was  not  marked  by  an  initial  drop, 
so  far  as  could  be  observed  :  in  the  second,  it  rose  at  a  rate 
which  showed  that  it  would  be  halfway  to  the  final  value  in 
four  days,  the  final  value  being  about  four  times  the  initial. 
In  the  third  place,  no  emanation  came  off  the  material  when 
first  prepared ;  even  when  no  draught  was  employed  the 
readings  did  not  alter  in  15  minutes  :  and  in  the  fourth  place, 
the  observed  results  fitted  closely  to  the  calculated  curves, 
showing  only  a  slight  variation  as  in  the  case  of  the  radium. 
The  results  of  one  experiment  are  shown  in  the  folio  wing- 
table  : — 
I. 
II. 
III. 
IV. 
pd  of  A\  foil 
X  10°. 
i/L 
From  Curve  A. 
pd  of  full  range 
X  103. 
244 
•813 
•055 
444 
474 
•670 
•108 
439 
780 
•544 
•162 
480 
1061 
•412 
•227 
468 
1573 
•271 
•328 
480 
2073 
•173 
•417 
499 
2607 
•106 
•504 
517 
The  mean  of  the  figures  in  the  last  column  is  '00177. 
