Supplement to ''Nature,''' August 25, 1923 



I I 



the gas molecules stream by it with a velocity equal and 

 opposite to that of the a-particle. Now the maximum 

 velocity of an -a-particle from radium C is equivalent to 

 that gained by an electron in falling freely between a 

 difference of potential of about 1000 volts ; so that 

 the electrons comprising the molecules of air or other 

 gas have a velocity of translation numerically equal to 

 this. For brevity, it is very convenient to speak of 

 this velocity or energy as that due to a " looo-volt " 

 electron. 



When the electrons in an atom pass close to the a- 

 particle^ one of them may be removed from the parent 

 atom by the coUision, energy being required for this 

 process. The ionisation potential for oxygen or nitro- 

 gen is about 17 volts, which is a very small quantity 

 compared with the energy of translation of a 1000- volt 

 electron. 



If we consider the forces involved between an a- 

 particle and moving electron as of the ordinary electro- 

 static type, the electron will describe a hyperbolic orbit 

 round the nucleus, the angle of deflexion of the path 

 of the electron resulting from the collision depending 

 on the nearness of approach of the electron to the 

 nucleus. On ordinary dynamics, the electron will 

 never be captured in such a coUision if there is no loss 

 of energy by radiation. If capture for some reason 

 results from the collision, it means that an amount of 

 energy corresponding to at least a looo-volt electron 

 has in some way been got rid of. This loss of energy 

 may be supposed to be due to some interaction between 

 the a-particle and colliding nucleus with its attendant 

 electrons, or to the loss of energy by radiation during 

 the collision. The first supposition seems at first sight 

 plausible, for we know that the innermost electrons of 

 oxygeji or nitrogen are strongly bound and require 

 energy of the order of 500 volts to remove them from 

 the atom. But there is one very strong and, it seems 

 to me, insuperable objection to this view. 



I have found that the deflexion in a magnetic field 

 of a pencil of a-particles passing through a suitable 

 pressure of hydrogen is similar to that shown in curve 

 Fig. 4 for air. This shows that the a-particle passing 

 through hydrogen captures electrons of energy about 

 120 volts to about the same degree as in air. Now we 

 know that the electrons in the hydrogen atom or mole- 

 cule are lightly bound, and an energy of not more than 

 a 30-volt electron, suitably applied, would entirely 

 separate the component nuclei and electrons in the 

 hydrogen molecule. In the case of hydrogen, therefore, 

 we cannot hope to account for the requisite loss of 

 energy, which for the experiment considered is about 

 100 volts. If these experiments with hydrogen are 

 correct, and are valid for all velocities of the a-particle, 

 we are driven to conclude either, that some unknown 

 factors are involved in the capture, or that the loss of 

 energy of the electron must be ascribed to radiation. 

 In such a case, capture of an electron may be regarded 

 as the converse of the photo-electric effect, where radia- 

 tion falls on matter and swift electrons are ejected 

 from the matter. In the case under consideration, 

 swift electrons are shot towards a charged nucleus and 

 an occasional electron is captured with the emission of 

 energy in the form of radiation. On such an hypo- 

 thesis the radiation of energy from an a-particle passing 



through a gas due to the frequency of capture is very 

 great, amounting to about 3 per cent, of the total 

 energy of the a-particle. This seems to be an unex- 

 pectedly large amount, but cannot be ruled out as im- 

 possible in the present state of our knowledge. 



In the discussion of this very thorny question, I 

 have confined myself mainly to the case of capture by 

 the swift a-particle, where the difficulties of explanation 

 are much greater than for capture at slower velocities. 

 Our information is at present too incomplete to give a 

 decisive answer, but there seems to be no doubt that 

 the unexpected frequency of capture of electrons by 

 swift a-particles raises many new and interesting 

 questions of the nature of the processes that can occur 

 in collisions between electrons and matter. 



I need scarcely say that the phenomena of capture 

 and loss are not confined to the a-particle, but are 

 shown by all charged atoms in swift motion through a 

 gas, and were long ago observed in the case of positive 

 rays. On account, however, of the high velocity of the 

 a-particles and the ease of their individual detection, the 

 process of capture and loss can be studied quantita- 

 tively under simpler and more definite conditions than 

 in the case of the electric discharge through a gas at 

 low pressure. 



On this occasion I have devoted my attention to 

 ■ the most recent additions to our knowledge of the life 

 history of the a-particle. This knowledge has been 

 obtained from the study of the rapid interchange of 

 charges when an a-particle passes through matter. I 

 have only incidentally referred to the numerous colli- 

 sions with electrons along the track of the a-particle 

 which result in dense ionisation. I have omitted any 

 consideration of those rare but interesting encounters 

 in which an a-particle is deflected through a large angle 

 by a close collision with a nucleus. I have omitted, 

 too, the still rarer encounters which may result in a 

 disintegration of an atomic nucleus hke that of nitrogen 

 or of aluminium. We have seen that an a-particle has 

 an interesting history. Usually it is retained as an 

 integral and orderly part of a radioactive nucleus for 

 an interval of more than a thousand million years. 

 Then follows a cataclysm in the radioactive nucleus ; 

 the a-particle gains its freedom and lives an independent 

 life of about one hundred milHonth of a second, during 

 which all the incidents referred to in this lecture 

 occur. 



If we are dealing with a dense and compact uranium 

 or thorium mineral, the a-particle after acquiring two 

 electrons and becoming a neutral helium atom may be 

 imprisoned in the mineral as long as the mineral exists. 

 The occluded helium can be released from the mineral 

 by the action of high temperature, and after removal of 

 all other gases can be made to show its presence by the 

 characteristic brilliant luminosity under the stimulus 

 of the electric discharge. In the circumstances of such 

 an experiment, only small quantities of helium are 

 liberated. Large quantities of helium, sufficient to 

 fill a large airship, have, however, been isolated from 

 the natural gases which escape so freely from the 

 earth in various parts of ('anada and the United 

 States. It is a striking fact that every single atom 

 of this material has in all probability had the life 

 history here described. 



