Some  Properties  of  the  a  Rays  emitted  by  Radium.     723 
slit  to  plate,  1*94  cm.;  total  distance  a  +  b  =  4*085  cm.). 
The  magnetic  field  did  not  vary  by  more  than  0*3  to  0*5  per 
cent,  in  the  course  of  an  experiment. 
Since  the  time  when  I  carried  out  these  experiments, 
Rutherford  has  published  similar  ones  *,  and  has  besides 
proved  the  important  fact  that  the  separation  of  the  bands  is 
greater  in  air  than  in  vacuo. 
These  results  might  be  explained  by  supposing  that  the 
greater  deflectibility  of  the  pencil  corresponds  to  an  increased 
curvature  ol  the  trajectory — such  increase  being  due  to  a 
decrease  in  the  velocity  of  the  particles  carrying  positive 
electric  charges,  which  constitute  the  a,  rays.  This  decrease 
of  velocity  should,  moreover,  take  place  progressively  along 
the  air  path. 
I  proceeded  to  study  with  greater  precision  than  heretofore 
the  air  trajectory  of  the  a  rays  when  deflected  by  a  magnetic 
field,  employing  a  photographic  method  which  I  had  already 
used  for  a  number  of  years. 
The  method  consists  in  receiving  a  pencil,  defined  by  a 
linear  source  and  a  slit  parallel  to  it  and  at  a  distance  a.  from 
it,  on  an  inclined  photographic  plate  resting  against  the 
screen  containing  the  slit  and  normal  to  the  plane  of  the 
undeflected  pencil.  The  trace  of  this  pencil  will  be  a  straight 
line  ;  but  if  it  is  deviated  in  a  direction  parallel  to  the  plate, 
first  one  way  and  then  the  other,  the  trace  consists  of  two 
divergent  curves  which  intersect  each  other  at  the  point  of 
contact  of  the  slit  with  the  plate.  The  distance  of  a  point  in 
the  plate  from  its  lower  edge  is  proportional  to  the  height  y 
of  this  point  above  the  horizontal  plane  against  which  the 
plate  rests,  and  if  the  inclination  of  the  plate  be  known,  this 
height  may  be  determined.  Further,  by  measuring  the 
distance  apart  2x  of  two  points  in  the  two  curves  having  the 
same  y,  the  required  trajectory  may  be  constructed  point  by 
point. 
The  greater  part  of  the  new  tests  were  carried  out  by  using 
as  source  a  platinum  wire  0*1  to  0*2  mm.  in  diameter, 
uniformly  excited  by  the  radium  emanation.  The  method  of 
procedure  generally  adopted  was  as  follows  : — Owing  to  the 
rapid  decrease  in  the  intensity  of  the  source,  the  magnetic 
field  was  reversed  every  five  minutes  in  order  to  equalize  the 
impressions  of  the  two  deflected  pencils  ;  the  source  was 
first  screened  by  an  aluminium  sheet  0*015  mm.  thick,  and 
then,  at  the  end  of  about  half  an  hour,  the  aluminium  sheet 
was  removed,  and  the  photographic  plate  was  displaced 
laterally  so  as  to  obtain,  side  by  side  with  the  first  curves  and 
*  Phil.  Mao-.  January  1900,  p.  160. 
