The velocity of expulsion of the α-particles from different kinds of active matter varied over comparatively narrow limits but the value of e/m was constant and equal to 5,070. This value was not very different from the one originally found. A difficulty at once arose in interpreting this result. We have seen that the value of e/m for the hydrogen atom is 9,650. If the α-particle carried the same positive charge as the hydrogen atom, the value of e/m for the α-particle would indicate that its mass was twice that of the hydrogen atom, i.e. equal to the mass of a hydrogen molecule. It seemed very improbable that hydrogen should be ejected in a molecular and not an atomic state as a result of the atomic explosion. If, however, the α-particle carried a charge equal to twice that of the hydrogen atom, the mass of the α-particle would work out at nearly four, i.e. a mass nearly equal to that of the atom of helium.
I suggested that, in all probability, the α-particle was a helium atom which carried two unit charges. On this view, every radioactive substance which emitted α-particles must give rise to helium. This at once offered an explanation of the fact observed by Debierne that actinium as well as radium produced helium. It was pointed out that the presence of a double charge of helium-atom was not altogether improbable for reasons to be given later.
While the evidence as a whole strongly supported the view that the α-particle was a helium atom, it was found exceedingly difficult to obtain a decisive experimental proof of the relation. If it could be shown experimentally that the α-particle did in reality carry two unit charges, the proof of the relation would be greatly strengthened. For this purpose an electrical method was devised by Rutherford and Geiger for counting directly the α-particles expelled from a radioactive substance. The ionization produced in a gas by a single α-particle is exceedingly small and would be difficult to detect electrically except by a very refined method. Recourse was had to an automatic method of magnifying the ionization produced by an α-particle. For this purpose it was arranged that the α-particles should be fired through a small opening into a vessel containing air or other gas at a low pressure, exposed to an electric field near the sparking value. Under these conditions the ions produced by the passage of the α-particle through the gas generate a large number of fresh ions by collision. In this way it was found possible to magnify the electrical effect due to an α-particle several thousand times. The entrance of an α-particle into the testing vessel was then indicated by a sudden deflection of the electrometer needle. This method was developed into an accurate method of counting the number of α-particles fired in a known time through the small aperture of the testing vessel. From this was deduced the total number of α-particles expelled per second from any thin film of radioactive matter. In this way it was shown that 3.4 x 1010 α-particles are expelled per second from one gram of radium itself and from each of its α-ray products in equilibrium with it.
The correctness of this method was indicated by another, quite distinct method of counting. Sir William Crookes and Elster and Geitel had shown that the α-particles falling on a screen of phosphorescent zinc sulphide produced a number of scintillations. Using specially prepared screens, Rutherford and Geiger counted the number of these scintillations per second with the aid of a microscope. It was found that, within the limit of experimental error, the number of scintillations per second on a screen agreed with the number of α-particles impinging on it, counted by the electrical method. It was thus clear that each α-particle produced a visible scintillation on the screen, and that either the electrical or the optical method could be used for counting the α-particles. Apart from the purpose for which these experiments were made, the results are of great interest and importance, for it is the first time that it has been found possible to detect a single atom of matter by its electrical and optical effect. This is of course only possible because of the great velocity of the α-particle.
Knowing the number of α-particles expelled from radium from the counting experiment, the charge carried by each α-particle was determined by measuring the total positive charge carried by all the α-particles expelled. It was found that each α-particle carried a positive charge of 9.3 x 10-10 electrostatic units. From a consideration of the experimental evidence of the charge carried by the ions in gases, it was concluded that the α-particle did carry two unit charges, and that the unit charge carried by the hydrogen atom was equal to 4.65 x 10-10 units. From a comparison of the known value of e/m for the α-particle with that of the hydrogen atom, it follows that an α-particle is a projected atom of helium carrying two charges, or, to express it in another way, the α-particle, after its charge is neutralized, is a helium atom.
The data obtained from the counting experiments allow us to calculate simply the magnitude of a number of important radioactive quantities. It was found that the calculated values of the life of radium, of the volume of the emanation, and of the heating effect of radium were in excellent agreement with the values found experimentally. A test of the correctness of these methods of calculation was forthcoming shortly after the publication of these results. Rutherford and Geiger calculated, on the assumption that the α-particle was a helium atom, that one gram of radium in equilibrium should produce a volume of 158 cubic millimetres of helium per year. Sir James Dewar in 1908 carried out a long experimental investigation on the rate of production of helium by radium, and showed that one gram of radium in equilibrium produced about 134 cubic millimetres per year. Considering the difficulty of the investigation, the agreement between the experimental and calculated values is very good and is strong evidence in support of the identity of the α-particle with a helium atom.
While the whole train of evidence we have considered indicates with little room for doubt that the α-particle is a projected helium atom, there was still wanting a decisive and incontrovertible proof of the relationship. It might be argued, for example, that the helium atom appeared as a result of the disintegration of the radium atom in the same way as the atom of the emanation and had no direct connection with the α-particle. If one helium atom were liberated at the same time that an α-particle was expelled, experiment and calculation might still agree and yet the α-particle might be an atom of hydrogen or of some unknown substance. In order to remove this possible objection, it is necessary to show that the α-particles, collected quite independently of the active matter from which they are expelled, give rise to helium. With this purpose in view some experiments were recently (1908) made by Rutherford and Royds. A large quantity of emanation was forced into a glass tube which had walls so thin that the α-particles were fired right through them, though the walls were impervious to the emanation itself. The α-particles were projected into the glass walls of an outer sealed vessel and were gradually released into the exhausted space between the emanation tube and the outer vessel. After some days a bright spectrum of helium was observed in the outer vessel. There is, however, one objection to this experiment. It might be possible that the helium observed had diffused through the thin glass walls from the emanation. This objection was removed by showing that no trace of helium appeared, when the emanation was replaced by a larger volume of helium itself. We may thus confidently conclude that the α-particles themselves give rise to helium, and are atoms of helium. Further experiments showed that when the α-particles were fired through the glass walls into a thin sheet of lead or tin, helium could always be obtained from the metals after a few hours' bombardment.
Considering the evidence together, we conclude that the α-particle is a projected atom of helium, which has, or in some way during its flight acquires, two unit charges of positive electricity. It is somewhat unexpected that the atom of a monatomic gas like helium should carry a double charge. It must not however be forgotten that the α-particle is released at a high speed as a result of an intense atomic explosion, and plunges through the molecules of matter in its path. Such conditions are exceptionably favourable to the release of loosely attached electrons from the atomic system. If the α-particle can lose two electrons in this way, the double positive charge is explained.
We have seen that there is every reason to believe that the α-particles, so freely expelled from the great majority of radioactive substances, are identical in mass and constitution and must consist of atoms of helium. We are consequently driven to the conclusion that the atoms of the primary radioactive elements like uranium and thorium must be built up in part at least of atoms of helium. These atoms are released at definite stages of the transformations at a rate independent of control by laboratory forces. There is good reason to believe that in the majority of cases, a single helium atom is expelled during the atomic explosion. This is certainly the case for radium itself and its series of products. On the other hand, Bronson has drawn attention to certain cases, viz. the emanations of actinium and of thorium, where apparently two and three atoms of helium respectively are expelled at one time. No doubt these exceptions will receive careful investigation in the future. It is of interest to note that uranium itself appears to expel two α-particles for one from each of its products. Knowing the number of atoms of helium expelled from the atom of each product, we can at once calculate the atomic weights of the products. For example, in the uranium-ionium-radium series, uranium expels two α-particles and each of the six following α-ray products one, i.e. eight in all. Taking the atomic weight of uranium as 238.5, the atomic weight of ionium should be 230.5, of radium 226.5, of the emanation 222.5, and so on. It is of interest to note that the atomic weight of radium deduced in this way is in close agreement with the latest experimental values. The atomic weight of the end-product of radium, resulting from the transformation of radium F (polonium) should be 238.5 – 8 x 4 = 206.5, or a value close to that for lead. Long ago, Boltwood suggested from examination of analyses of old uranium minerals, that lead was in all probability a transformation product of the uranium-radium series. The coincidence of numbers is certainly striking, but a direct proof of the production of lead from radium will be required before this conclusion can be considered as definitely established.
It is very remarkable that a chemically inert element like helium should play such a prominent part in the constitution of the atomic systems of uranium and thorium and radium. It may well be that this property of helium of forming complex atoms is in some way connected with its inability to enter into ordinary chemical combinations. It must not be forgotten that uranium and thorium and each of their transformation products must be regarded as distinct chemical elements in the ordinary sense. They differ from ordinary elements in the comparative instability of their atomic systems. The atoms break up spontaneously with great violence, expelling in many cases an atom of helium at a high speed. All the evidence is against the view that uranium or thorium or radium can be regarded as an ordinary molecular compound of helium with some known or unknown element, which breaks up into helium. The character of the radioactive transformations and their independence of temperature and other agencies have no analogy in ordinary chemical changes.