September 18, 1896.] 



SCIENCE. 



399 



ble light. Analogy, however, leads us to ex- 

 pect that to get polarization effects we must 

 use in the case of short waves, polarizers of a 

 much finer structure than would be neces- 

 sary for long ones. Thus a wire bird-cage 

 will polarize long electrical waves, but will 

 have no effect on visible light. Rubens and 

 Dubois made an instrument which would po- 

 larize the infra red rays by winding very fine 

 wires very close together on a framework ; 

 this arrangement, however, was too coarse 

 to polarize visible light. Thus, though the 

 structure of the tourmaline is fine enough 

 to polarize the visible rays, it vadij be much 

 too coarse to polarize the E-ontgen rays if 

 these have exceedingly small wave-lengths. 

 As far as our knowledge of these rays ex- 

 tends, I think we may say that though 

 there is no direct evidence that they are a 

 kind of light, there are no properties of the 

 rays which are not possessed by some vari- 

 ety of light. 



It is clear that if the Eontgen rays are 

 light rays, their wave-lengths are of an 

 entirely different order to those of visible 

 light. It is, perhaps, worth notice that on 

 the electro-magnetic theory of light we 

 might expect two different types of vibra- 

 tion if we suppose that the atoms in the 

 molecule of the vibrating substance carried 

 electrical charges. One set of vibrations 

 would be due to the oscillations of the 

 bodies carrying the charges, the other set to 

 the oscillation of the charges on these bodies. 

 The wave-length of the second set of vibra- 

 tions would be commensurate with molecu- 

 lar dimensions ; can these vibrations be the 

 Rontgen rays? If so, we should expect 

 them to be damped with such rapidity as 

 to resemble electrical impulses rather than 

 sustained vibrations. 



If we turn from the rays themselves to 

 the effects they produce, we find that the 

 rays alter the properties of the substances 

 through which they are passing. This 

 change is most apparent in the effects pro- 



duced on the electrical properties of the 

 substances. A gas, for example, while 

 transmitting these rays, is a conductor of 

 electricity. It retains its conducting prop- 

 erties for some little time after the rays 

 have ceased to pass through it, but Mr. 

 Rutherford and I have lately found that the 

 conductivity is destroyed if a current of 

 electricity is sent through the Rontgenized 

 gas. The gas in this state behaves in this 

 respect like a very dilute solution of an 

 electrolyte. Such a solution would cease to 

 conduct after enough electricity had been 

 sent through it to electrolyze all the mole- 

 cules of the electrolyte. When a current is 

 passing through a gas exposed to the rays, 

 the current destroys and the rays produce 

 the structure which gives conductivity to 

 the gas ; when things have reached a steady 

 state the rate of destruction by the current 

 must equal the rate of production by the 

 rays. The current can thus not exceed a 

 definite value, otherwise more of the con- 

 ducting gas would be destroyed than is 

 produced. 



This explains the very characteristic fea- 

 ture that in the passage of electricity 

 through gases exposed to Rontgen rays, the 

 current, though at first proportional to the 

 electro-motive force, soon reaches a value 

 where it is almost constant and independ- 

 ent of the electro- motive force, and we get 

 to a state when a tenfold increase in the 

 electro-motive force only increases the cur- 

 rent by a few per cent. The conductivity 

 under the R5ntgen rays varies greatly from 

 one gas to another, the halogens and their 

 gaseous compounds, the compounds of sul- 

 phur, and mercury vapor, are among the 

 best conductors. It is worthy of note that 

 those gases which are the best conductors 

 when exposed to the rays are either ele- 

 ments, or compounds of elements, which 

 have in comparison with their valency very 

 high refractive indices. 



The conductivity conferred by the rays 



