and  Composition  of  Uranium  Compounds.  185 
conclusions  drawn  from  my  preliminary  experiments,  viz.  : 
that  uranium  is  the  parent  not  only  of  radium,  but  also  o£  all 
the  other  active  substances  which  accompany  it.  Apparently 
the  experimental  error  of  the  present  work  is  about  +  1  per 
cent.  ;  but  it  may  be  somewhat  greater.  In  consequence 
there  may  be  comparatively  small  amounts  of  active  bodies 
( polonium,  actinium,  &c.)  which  are  not  derivatives  of 
uranium.  But  as  far  as  radium  and  its  known  products  are 
concerned,  the  case  is  clear  ;  for,  in  addition  to  the  evidence 
I  have  presented,  it  has  been  shown  by  Boltwood  *  that  the 
quantity  of  radium  emanation  obtained  from  various  uranium 
ores  is  strictly  proportional  to  the  weight  of  uranium  in  the 
-ample.  In  addition  Soddyf  na*  finally  obtained  radium 
emanation  from  a  quantity  of  pure  uranium  nitrate  which  two 
years  previous  had  been  entirely  freed  from  radium  and  its 
products.  It  is  true,  however,  that  the  amount  of  emanation 
so  obtained  was  much  smaller  than  one  should  expect  if 
radium  is  produced  directly  from  uranium  X{. 
It  is  possible  to  calculate,  from  data  here  presented, 
the  relative  activity  of  radium  and  uranium,  on  the  assump- 
tion that  all  of  the  excess  activity  of  uranium  ores  is  due  to 
radium,  together  with  its  products.  These  bodies,  collectively, 
are  4'15 —  1  or  3'15  times  as  active  as  that  mass  of  uranium 
with  which  they  are  associated.  Now  Rutherford  and 
Boltwood  §  have  shown  that  one  part  of  radium  is  in  equi- 
librium with  1*35  X  106  parts  of  uranium  in  ores.  Therefore 
radium  (with  its  products)  is  3*15  x  1*35  X  106  =  4*25  x  106 
times  as  active  as  uranium.  The  value  usually  accepted  is 
only  half  as  great  as  this.  This  ratio  may  also  be  calculated 
from  other  data.  Rutherford  and  McClung  ||  have  found  the 
maximum  saturation  current  due  to  1  sq.  cm.  of  a  thick  layer 
of  U308  to  be  about  4*5  x  10~13  ampere.  Therefore  the 
total  current  due  to  1  g.  of  uranium  is  791  X  4*5  x  10-13  — 
o'6  x  10-10  ampere  H.  The  total  current  due  to  1  g.  of 
radium  (plus  products)  is  1*2  x  10~3  ampere **. 
From  these  results  it  follows  that  radium  is  3  3  x  106 
times  as   active  as  uranium  ft-      This  estimate  may  be  too 
*  Loc.  cit. 
t  Phil.  Mag.  [6]  ix.  p.  768  (1905). 
t  Cf.  Boltwood,  Amer.  J.  Sci.  xx.  p.  239  (1905). 
§  Amer.  J.  Sci.  xx.  p.  55  (1905). 
||  Phil.  Trans.  A,  1901,  p.  25. 
H  McCoy,  J.  Amer.  Chem.  Soc.  xxvii.  p.  402  (1905). 
**  Eutherford,  '  Radioactivity,'  1904,  p.  156. 
ft  The  direct  comparison  of  the  activity  of  a  thick  film  of  pure  1.0 
with  that  of  a  minute  (known)  quantity  of  a  pure  radium  salt  will  doubt- 
less give  a  more  reliable  result. 
