78 Behaviour of Becqiierel mid Rontgen Rays in a Magnetic Field. 



shadow obtained on development was mucli shorter, the time of ex- 

 posure being, of course, in each case the same. 



The numerical estimate of the curvature of the rays was obtained 

 from an experiment of the latter kind. 



Fig. 4. 



e 



^|\^ 



c I 7 ~~ 



Let us suppose that ec (fig. 4) represents in section the front surface 

 of the radiating substance, cf the surface of the photographic plate. 



Let / be the place furthest from c at which the darkening on the 

 photographic plate was perceptible. Now the rays which reach 

 furthest are those which proceed from <?, the highest point of the 

 radiating surface, as may easily be seen from geometrical considera- 

 tions. The rays which reach / must consequently proceed from e. 

 Eays proceeding from any lower point of the surface will either be 

 bent up so as never to reach the plate at all, or else they will strike it 

 short of /. The ray which reaches / from e will clearly just graze the 

 surface of the plate at /. 



If r be the radius of curvature, h the distance cc, and / the distance 

 <:j\ then 



If then we measure the height of the highest part of the radiating 

 surface above the plate, and I the greatest distance to which the 

 darkening of the plate extends, we have data for determining r. 



It must be admitted that the measurement of / involves great un- 

 certainty. The image gradually tails off, and any estimate of its 

 length must to a great extent be arbitrary. 



The value of r deduced is more uncertain still, since P is involved 

 in calculating it. But, in spite of these objections, the method may, I 

 think, be relied on to give the order of magnitude of r, and that is all 

 that is required, so far as the conclusions which it is here sought to 

 draw are concerned. 



In one experiment, the length / was estimated at 2 cm. ; h was 

 0-8 cm. Thus r = 3 cm. approximately. 



The strength of the magnetic field, measured as before, was 

 1 680 C.G.S. Thus the field required to produce a curvature of radius 

 1 cm. is about 5 x 10^. 



In another experiment, / was 1-8 cm., b was 0-8 cm., and the field 

 2140. This gives practicallj?^ the same result as the preceding. 



