THE RADIO-ACTIVITY OF MATTER. 203 



Kaufmanii has attempted to elucidate the matter. Tie comhincd. at 

 rioht ano-les, the mai>"netic and the electric actions; uiifoi'tuiiately, the 

 expei'iment. which is very difficult to perform, did not ui\'e him one 

 plate tit to measur(\ For the values of T/p comprised between l.SOO 



and 4.<iO(), he found that the i-ejation ' xaried from 1.;! : lo', to 



().() : lU', and the speed r from '2.'.) : 1(>"' to L\.S : lo'". 



The ])ro()f of a reu'ular \ariatioii in the calculated relation ' is of 



/;/ 



considerable theoretical importance; if this relation was constant, as it 

 seemed to l)e as ih' result of a lar^'e number of measurements, we 

 might conclude that (lie slig-htly deviable rays, for which ///' is more 

 than 5,000, haA'e speeds consid(M'a])ly greater than that of lieht. 



On the other hand, theoretical considerations have oiv(Mi the idea 

 that the spec^l could not surpass that of the pi'opagation of eU^ctro- 

 magnetic disturl)ances — that is to sa}', the speed of light — and we have 

 been led to consider the mol)ile masses in a magn<»tic tield as endowed 

 with a particular inertia, which is a function of the speed. Tiider 

 those conditions the calculated mass ought to be appai'ent, or at least 

 partly so, and it should increase indetinitely as the actual speed 

 approaches that of light. The figures ])ublisii(Ml by M. Kaut'mann 

 bear out this hypothesis. 



Another consequence of this manner of looking at the question 

 would l)e that there should l)e continuity f)etween the deviable rays 

 and those which are not, as the radius of curvature of the ti-ajectories 

 l^ecomes intinite at the same time as the apparent mass. 



The photographic print alread}' mentioned (tig. 0). as well as one of 

 the following ones (tig. 18), bowed, on the contrary, a \ery distinct 

 discontiiuiity, although in the second one the exposure was sufficiently 

 prolonged for the impression of the least active rays, snch as the 

 penetrating nondeviable ones to be distinctly visible. 



This proof was obtained in (be following maimer: In the uniform 

 magnetic field of a permanent magnet I placed, normally to the tield, a 

 photographic plate; then on this latter I arranged screens of h^id fixed 

 on a sheet of glass, as shown in iig, 12. These screens are pierced 

 with openings normal to the plate, and destined to limit (he width of 

 the beam; in the path of these beamslai'ianged other screens, such as 

 aluminium oiu>s. l)elow the plate o^jposite a narrow slit in a strip of 

 lead a small bloc;k of lead is placed ha\inga deep cavity normal to the 

 plate, and in which the radiant ])ody is ])laced. ^V'e have thus a nar- 

 row lineai' source normal to the plate and sev(M-al millimetei's in length. 

 The cavity is covered with a thin sheet of aluminium to s(o]) the light 

 rays. 



The tigure represents a section made iiorinaliy so the tield of the 

 beam, of which a part is deviated. Each beam corresponding to a 

 determined speed gives an impression which is noticeably curved, as if 



