March 5, 190b] 



NA TURE 



425 



analysis of the number of a-ray products present in a 

 substance. For example, suppose the amount of ionisa- 

 tion in the gas produced b\" a narrow pencil of a rays is 

 examined at varying distances from the radium. At a 

 distance of 7 cm. there is a sudden increase in the amount 

 of ionisation, for at this distance the a. particles from 

 radium C enter the testing vessel. There are again sudden 

 changes in the ionisation at distances of 4-8 cm., 4.3 cm., 

 and 3.5 cm. These are due to the rays from the radium A, 

 the emanation and radium itself respectively entering the 

 testing vessel. The o-ray analysis thus discloses four types 

 of a rays present in radium in equilibrium — a result in 

 tonformity with the more direct analysis. This method 

 allows us to settle at once whether more than one a-ray 

 product is present in a given radio-active material. For 

 example, an analysis by Hahn by this method of the radia- 

 tion from the active deposit of thorium has disclosed the 

 existence of two a-ray products instead of one as 

 previously supposed. We can consequently gain informa- 

 tion on the complexity of radio-active material, even 

 though no chemical methods have been found to separate 

 the products concerned. The range of the o particle from 

 each product is a definite constant which is characteristic 

 111 each product. 



The a particle decreases in velocity as it passes through 

 matter. This result is clearly brought out by photographs 

 showing the deflection of a homogeneous pencil of a rays 

 in a magnetic field before and after passing through an 

 absorbing screen. The greater divergence of the trace 

 of the a rays on the plate, after passing through the 

 screen, shows that their velocity is reduced, while the 



, sharpness of the band shows that the a particles still move 



I at an identical speed. 



In order to make an accurate determination of the con- 

 stants of the a particles, it is necessary to work with 

 homogeneous rays, and we consequently require to use 

 a thin layer of matter of one kind. For experiments of 

 this character, a wire coated with a thin film of radium C 

 by exposure to the radium emanation is very suitable. 

 The velocity of the a particle and the value elm, the ratio 

 of the charge carried by the a particle to its mass, can 

 be deduced by observing the deflections of a pencil of 

 a rays exposed in a magnetic and in an electric field of 

 known strengths. The deflection of a pencil of a rays in 



I an electric field is small under normal conditions, and 



[special care is needed to determine it with accuracy. 

 In this way I have calculated the velocity and value 

 •of c'm for a number of a-ray products. The velocity of 

 ■expulsion varies for different products, but is connected by 

 a simple relation with the range of the a. particle in air. 

 The value of eim has been determined for selected pro- 

 ducts of radium, thorium, and actinium, and in each case 

 the same value has been found. This shows that the 

 a particles expelled from radio-active substances in general 

 are identical in constitution. They have all the same 

 mass, but differ from one another in the initial velocity 

 ■of their projection, .\lthough we are sure that the 

 ■a particles, from whatever source, are identical atoms of 

 matter, we are still unable to settle definitely the true 

 nature of the a particle. The value of ehn found by 

 experiment is nearly 5x10'. Now the value of eim for 

 the hydrogen atom in the electrolysis of water is 10'. 

 If the charge carried by the o particle and the hydrogen 

 atom Is the same, the mass of the o particle is twice that 

 of the hydrogen atom, i.e. a mass equal to the hydrogen 

 molecule. But we are not certain that they do carry the 

 same charge. Here we are, unfortunately, confronted 

 by a number of possibilities, for the magnitude of »n for 

 the a particle is conditioned by the value assumed for e. 

 If the charge of the a particle is assumed to be twice the 

 value of the hydrogen atom, the mass comes out four 

 times the hydrogen atom — the value found for the heliurh 

 ato'ii. The weight of evidence still supports the view that 

 the o particle is in some way connected with the helium 

 atom. If the a particle is a helium atom with twice the 

 Ionic charge, we must regard the helium produced by 

 radio-active bodies as actuallv the collected a particles the 

 charges of which have been neutralised. This at once 

 offers n reasonable explanation of the production of helium 

 by actinium as well as bv radium. In addition, Strutt 

 has recently contributed strong evidence that helium is a 



NO. 2001, VOL. 77] 



product of thorium. Such results are only to be expected 

 on the above view, since the a particle is the only common 

 product of these elements. 



The determination of the true character of the a particle 

 is one of the most pressing unsolved problems in radio- 

 activity, for a number of important consecjuences follow 

 from its solution. Unfortunately, a direct experimental 

 proof of its true character appears to be very difficult 

 unless a new method of attack is found. We have seen 

 that if the charge carried by the a particle could be ex- 

 perimentally determined, the actual value of m could be 

 determined in terms of the hydrogen atom, since the value 

 of the charge carried by the latter is known. This could 

 be done if we could devise a method of detecting the 

 emission of a single a particle, and thus counting the 

 number of particles expelled from a known quantity of a 

 radio-active substance, for example, from radium. In 

 considering a possible method of attack of this question, 

 the remarkable property of the a particles of producing 

 scintillations in zinc sulphide at once suggests itself. 

 Apart from the difficulty of counting the scintillations, it 

 is very doubtful whether more than a small fraction of 

 the o particles which strike the screen produce the scintilla- 

 tions. Viewed from the electrical side, a simple calcula- 

 tion from the data at our disposal shows that the 

 ionisation produced in a gas by a single a. particle should 

 be detectable. The electrometer or electroscope used for 

 measurement would, however, require to be extremely 

 sensitive, and under such conditions it is known that small 

 electrical disturbances are very difficult to avoid. 



In order to obtain a reasonably large effect, we require 

 some method of magnifying the ionisation produced by 

 the a particle. In conjunction with Dr. Hans Geiger, I 

 have recentlv developed a method whereby the electrical 

 effect produced by the a particle can be magnified several 

 thousand times. From the work of Townsend it is known 

 that if a strong electric field acts on gas at low pressure, 

 anv ions generated in the gas by an external agency are 

 set in motion by the electric field, and under the proper 

 conditions produce fresh ions by collision with the gas 

 molecules. The negative ion is the most effective ioniser 

 in weak fields, but when the voltage is increased near the 

 point at which a discharge passes, the positive ion also 

 produces fresh ions by collision. In the experimental 

 arrangement the a particle from the active matter is fired 

 through a small opening about 2 mm. in diameter, covered 

 with a thin layer of mica, into a cylinder 60 cm. long 

 and 2-5 cm. in diameter, in which the gas pressure is 

 about 3 cm. of mercury. A thin insulated wire connected 

 to the electrometer is fixed centrally in the cylinder. If 

 the outside cylinder is charged negatively, for a difference 

 of potential of about 1000 volts any ionisation produced in 

 the cylinder is increased about 2000 times by collision. 

 This can be simply illustrated by using the 7 rays of 

 radium as a source of ionisation. When a dilTerence of 

 potential is applied to the cylinder, the ionisation produced 

 by the 7 ravs onlv causes a slight movement of the electro- 

 meter needle. By applying, however, a voltage nearly 

 equal to that required for a discharge through the gas 

 there is a verv rapid inovement of the needle. On re- 

 moving the radium there is no appreciable current through 

 the gas. On placing a source of o rays near the small 

 opening in the cvlinder so that some of the a particles 

 can be fired along the axis of the cylinder, the electro- 

 meter needle does not move uniformly, but with a 

 succession of rapid throws with a considerable interval 

 in between. Each of these throws is due to the discharge 

 produced by a single o particle entering the cylinder, in- 

 creased several thousand times by the intermediary of the 

 strong electric field. If a sheet of paper which stops the 

 o ravs is placed before the opening, the electrometer 

 needle at once comes to rest. The interval of time 

 between the throws is not uniform. This is exactly what 

 we should expect if the number of o particles entering 

 such a small opening is governed by the law of probability. 

 On the average, a certain number of a particles are fired 

 through the opening per minute, but in some cases the 

 interval is less than the average, in others inuch greater. 

 In fact, bv observing the intervals between the entrance 

 of a large number of o particles, we should be able to 

 determine accurately the " probability " curve of distribu- 



