84 



NATURE 



[September 25, igig 



may be within the limits of the visible spectrum, an 

 M series being, perhaps, in the infra-red. 



But how comes it that hydrogen, with onlv one 

 electron, can have a K series and an L series and 

 an M series at all? Bohr's theory suggests that even 

 a single electron may have alternative orbits— not 

 necessarily occupied; and the spectroscope strongly 

 suggests that the radii of these alternative orbits run 

 as the squares of the natural numbers 



I 4 9 16 25, etc. 

 The frequencies, or reciprocals of periodic times, 

 would then be as the inverse cube of the natural 

 numbers 



I \ h (54, etc. 

 and this is, approximately and roughly, what the 

 K, L, M series of spectrum lines correspond with — 

 with some exceptions. 



When a cataclysm occurs and an electron is 

 expelled, it is expelled, as I think, with the velocity 

 which it possessed in the atom just before it burst 

 its bonds and flew off. For the energy required to 

 fling a planet to infinity, under an inverse-square law, 

 is just double the energy with which it was already 

 moving in its circular orbit. Its own orbital energy 

 is, therefore, the quantum of energy that has to be 

 supplied in order to get a satisfactory ejection. Some 

 of it might be supplied by the falling in of other 

 particles from their original orbits; for their kinetic 

 energy therein would be inversely as the distance from 

 the nucleus. Hence if K, L, M orbits have the radii 

 1, 4, 9, three units of L energy would represent the 

 fan from L to K, and this added to the original 

 L energy would give the quadruple L energy which is 

 equal to the K energy, and able to eject a k particle. 

 Similarly, a ninefold multiple of the M energy, eight 

 units of which would be acquired by falling to K, 

 would supply that particle with the "ejection energy 

 equally well. 



Would an M particle falling to L be able to eject 

 an L particle? l-iT = s% of a K unit of energy 

 would be acquired in the fall from M to L— that is, 

 \ M units, — so altogether I of M energy would be 

 transmitted, and that, being equal to a unit of 

 L energy, ought to be sufficient. 



Hence, in general, particles may be ejected from 

 any ring, either by direct impact from outside, or 

 by accumulated disturbance of X-rays, or by a col- 

 lapse of particles from one orbit to the next ; and 

 from an immense group of atoms, as in a visible 

 speck of substance, all kinds of radiation can be 

 emitted simultaneously. 



Are we to suppose that there is only one electron 

 in each orbit, or may several of them distribute them- 

 selves over a ring in accordance with some law of 

 stability? Both alternatives are possible, and both 

 are likely to be found in Nature It seems scarcely 

 likely that a uranium atom should possess ninetv-two 

 different orbits, although it does contain ninety-two 

 electrons. Yet even this number of orbits is possible 

 within the dimensions of an atom. We need not 

 exclude the possibility as taking up too much room. 

 For, given the size of the ultra-innermost or J orbit 

 as I, the outer orbit would, on Bohr's law pressed 

 to extremes, be (92)" times that radius — say, 8464 

 times the size of the innermost orbit ; but if this 

 innermost orbit is near the uranium nucleus, which 

 mav be -^92, or, say, 5 times the radius of the 

 hydrogen nucleus, the boundary or confine of the atom 

 is some 10,000 times as far away ; leaving, therefore, 

 just room enough for the ninety-two Bohr orbits, 

 though not much more than is required. 



Hence, if there were any reason to desire them 

 separate, they could be mad e room for, without 

 endowing the atom with outlying or ever-ready elec- 



NO. 2604, VOL. 104] 



trons likely to confer upon it verv active chemical 

 properties. But, so far as I see, so manv separate 

 orbits are not likely; for there is every probability 

 that periodically, as you ascend the series, the outer 

 ring is not occupied by a single electron, but by a 

 closed, compact sort of structure of many electrons, 

 with very little outside affinity ; so that we reach 

 periodically an atom which is' chemically inactive — 

 helium, neon, argon, krypton, etc.— up to emanation, 

 or what Ramsay called niton. So little cohesion 

 holds between such atoms that they are able to exist 

 as permanent gases, in spite of the high density of 

 some of them. This, at least, appears to be the view 

 of Rutherford and Soddy. Helium only condenses to 

 a liquid when cooled down to near the absolute zero 

 of temperature. Its cohesion or intermolecular attrac- 

 tion is nearly nil. 



A Fanciful Analogy. 



If I_ attempt to compare the supposed alternative 

 orbits in an atom with the known orbits of the solar 

 system, it is mainly to emphasise, provisionally and 

 tentatively, and perhaps semi-humorouslv, the astro- 

 nomical view of the atom, and to bring out still more 

 strongly the resemblances whatever thorough differ- 

 ences there may be as well. 



I write down the squares of the natural numbers, 

 therefore, and underneath put the initial letters of the 

 names of planets, with its real distance written under 

 each in the .same units. 



Radii of 1 

 Bohr's atomic Vi 4 

 orbits J 



m 



9 16 25 36 49 64 81 100 121 144 l6g 196 

 E M Asteroids J S V 



Planetary 1 



distances/ 39 7 2 « ,5 2 20-35 



52 



95-4 



192 



The obvious suggestion is that asteroids should be 

 looked for between Jupiter and Saturn, and between 

 Saturn and Uranus ; but I would not venture to 

 predict the existence of any such bodies on the 

 strength of this analogy, because you will doubtless 

 have noticed that no analogue of the planet Venus 

 appears in the list of atomic orbits; the scheme pro- 

 vides no place for her — a lamentable omission which 

 must discredit and, I expect, condemn even the 

 analogy. Nevertheless, I make no apology for intro- 

 ducing it in order to emphasise astronomical similari- 

 ties in the possible structure of an atom. 



Quantitative Interpolation. 



On Atomic Radiation. 



Permitting, ourselves this view of the atom as a 

 working hypothesis, we have to picture each atom 

 as an attracting centre or nucleus, with a number of 

 alternative orbits in regular succession round it, but 

 not all necessarily occupied by revolving electrons. 

 The atoms of different elements differ in the number 

 of positive units in the nucleus, and in the corre- 

 sponding number of revolving negative units ; in fact, 

 the diverse chemical elements in their atomic con- 

 stitution form a definite arithmetical series with 

 common difference i. There is a discontinuity or 

 finite step in passing from one element to the next 

 in the series ; there is no continuous passage from 

 one to another; hence if the physical transition or 

 mutation ever occurs, it must be by some sort of 

 sudden convulsion. 



To extract laws for this hypothetical structure, sug- 

 gested by the labours of many workers, we may attend 

 to the different rings of one kind of atom, or we may 

 attend to the corresponding rings in different kinds 

 of atom. Each, for instance, has an innermost ring, 

 which it is convenient at present to call the K ring 



