138 VARIABLE MASS [CH. xiv. 



velocity, for each of the particles which have produced 

 the impression at that point. And inasmuch as the 

 particles of different speeds will be sorted out to 

 different parts of the spectrum, it may be possible 

 to select those points which correspond to the highest 

 speeds, and indeed to compare the ratio of the two 

 deflexions for various speeds, if by any means the 

 velocity corresponding to each point can be determined. 

 A little calculation is needed to bring out the details 

 of the theory, and that shall be given directly, but 

 first I will give an idea of the kind of apparatus used. 



Experimental Device used by W. Kaufmann. 



A minute quantity of radium salt in a little 

 brass box acts as source, and a pencil of its rays 

 penetrates a small hole, about half a millimetre 

 diameter, in a plate of platinum at a distance of 

 2 centimetres from the source ; on the way, they 

 pass between a pair of parallel and insulated plates 

 of brass which are separated by about 2 millimetres 

 from each other and connected to a high-tension 

 battery of from 2,000 to 5,000 volts. After then 

 travelling another 2 centimetres, they encounter 

 the photographic plate placed to receive them. The 

 apparatus is contained in a thoroughly exhausted 

 vessel, and the whole is placed between the poles of 

 a large electromagnet giving a nearly uniform field, so 



FIG. 19. Kaufmann's apparatus for measuring simultaneously the 

 electric and magnetic deflexions of particles possessing very high velocity. 

 The source of radiation is a minute quantity of radium placed in a box 

 at C. All except the highest-velocity rays are deflected out of action by 

 the magnet NS ; some of the highest-velocity rays pass upwards through 

 the aperture D, being deflected forward by the magnetic field, and side- 

 ways by an electric field, whose lines are coincident with the magnetic 

 lines, between the adjustable plates PiP 2 , which are kept as highly 

 electrified as possible through the electrodes R. These thus doubly- 

 deflected rays then fall upon the photographic plate E, where they are 

 spread out into an oblique sort of very minute spectrum, more or less in 

 accordance with diagram 18 ; on which spectrum micrometric measure- 

 ments are subsequently made. 



