July 30, i8c 



NATURE 



305 



primordial atoms did its share in stopping the Rontgen rays, 

 we should have that intimate connection between density and 

 opacity which is so marked a feature for these rays. 



I now pass from the consideration of the rays themselves to 

 some of the effects they produce on bodies through which they 

 pass. 



There seems considerable evidence that the energy associated 

 with these waves is small. I am not acquainted with any effects 

 produced by them which involve the expenditure of an amount 

 of energy comparable with that emitted in a second by a candle. 

 They do not produce any appreciable rise in temperature when 

 they fall on the thin metallic strips of a bolometer. Mr. Skinner 

 has found that they exert no appreciable effect on the combination 

 of hydrogen and chlorine, though this is a good test of the in- 

 tensity of very faint light ; and, what is more unfortunate, they do 

 not exert any of those deleterious effects on bacteria which are 

 fortunately associated with ultra-violet light. Some of the other 

 effects exerted by ultra-violet light seem to be associated with 

 these rays ; thus some observers who have had undue curiosity 

 about their bones, and have in consequence exposed their hands 

 frequently to these rays, have found that the hand so exposed 

 became sunburnt. There seems considerable evidence, too, that 

 these rays are not good for the eyes, though it is difficult to dis- 

 entangle any distinctly injurious effect due to the rays from the 

 had effect that may be produced by the straining of the eye in 

 the endeavour to see only a faintly luminous object. 



There is one property of substances which seems peculiarly 

 suitable for testing if these rays affect the substance through 

 which they pass : it is the property of transmitting electricity. 



When we investigate the effect of the Rontgen rays on this 

 property, we find the remarkable result that bodies which, 

 when shielded from these rays, insulate to all appearance, per- 

 fectly allow electricity to pass through them wlien exposed to 

 the action of these rays. I will, first of all, show an experiment 

 illustrating this property in the case of gases which in their 

 normal slate are of all substances the most perfect insulators. 

 The details of the experiment are shown in the diagram (Fig. 3). 

 The coil and bulb are placed in this box, lined inside with tin-plate, 

 and covered over the top with sheet-lead. A hole is cut in the 

 box just over the bulb, and this hole is covered with a plate of 

 aluminium, which is transparent to these rays. The air space 

 between the electrodes is placed over this hole. One electrode 

 is connected to one pair of quadrants of the electrometer, the 

 other electrode is connected to one terminal of a battery, the 

 other terminal of which is to earth ; the two pairs of quadrants 

 of the electrometer are connected together and with the earth, 

 and the connection between them broken. If there is no leak- 

 age acro.ss the air space, the needle of the electrometer will 

 remain at rest. Vou see it does so when the coil is not in 

 action. As soon, however, as the coil is turned on, the spot of 

 light moves rapidly across the scale, showing that electricity is 

 passing across the air space. The rapidity of movement of the 

 spot of light is a measure of the rate of leak. Now the electrical 

 leakage produced by these rays depends on the nature of the 



NO. 1396, VOL. 54] 



gas. The gas I have just used was air. I will now replace the 

 air by another gas — chlorine. Again you see the leak, but it is 

 now much faster than before. Mr. McClelland and I have in- 

 vestigated the rate of leak in different gases, and we find that 

 they can be arranged in the following order : hydrogen, coal 

 gas, ammonia gas, air, carbonic acid gas, sulphuretted hydrogen, 

 chlorine, mercury vapour. 



That the gas itself is put into a peculiar state by the passage 

 through it of these rays — a state which it attains for an appre- 

 ciable time — is .shown by the following experiment, which I 

 described some time ago in Nature. I have here an electrode 

 shielded from the direct action of these rays. I charge it to a 

 high potential, and even though the rays are on, it does not 

 leak. I now blow some of the air through which the rays have 

 passed on to the electrode, and you see at once we get a rapid 

 leak. The rate at which electricity passes through the gas 



depends upon the pressure ; the lower the pressure the slower 

 the leak. Mr. McClelland and I found that for an air space of 

 about I cm. the rate of leak over a considerable range of pressure 

 varies as the square root of the pressure. In .some experiments 

 recently made by Mr. Rutherford and myself, we found that 

 using a constant potential difference the rate of leak was 

 smaller across a very thin plate of air than across a thicker 

 one ; it thus appears that the process of conduction through 

 a gas is one that requires a considerable amount of room. 



Another very interesting point about the rate of leak is the 

 connection between the rate of leak and the electromotive force. 

 This can, perhaps, be most easily understood by means of a curve 

 (Fig. 4). The ordinate represents the rate of leak, the abscis.sa 

 the electromotive force. At first, when the E. M.F. is small, 

 the curve is a straight line, showing that the current is pro- 

 portional to the electromotive force ; in other words, that the 

 conduction of electricity through the gas, like the conduction 

 through metals and electrolytes, obeys Ohm's law. But it is 



only when the E.M.F. is small that the curve is straight. We 

 soon get to a stage where the current increases more rapidly than 

 the E. M. F. ; beyond this, again, we reach a part of the curve 

 where the current increases but slowly as the electromotive 

 force increases, and we finally reach a stage where the current 

 seems independent of the E. M. F. , and is, to borrow a term from 

 magnetism, " saturated." I have here a diagram (Fig. 5) of three 

 curves taken for the same gas, but at different distances from the 

 bulb. Vou see that the first ascent is much steeper near to the 

 bulb— that is, when the rays are strong than when it is far away 

 and the rays are weak, and practical saturation is attained sooner 

 when the rays are strong th.in when they are weak. These 

 curves bear a remarkable resemblance to those which represent 

 the relation between the magnetisation of a piece of iron and 

 the magnetic force acting upon it. When the rays are strong, 

 the curve is like that of soft iron ; when the rays are weak, it is 

 like steel. 



Gases are not the only substances that conduct when trans- 



