March 9, 1905] 



NATURE 



439 



wmild docide bt'lwcpii the two views would be far from easy, 

 but as I interpret I'rof. Kutherford's letter, the results there 

 f^iven do not definitely disprove the view that the a particle 

 is inilially uncharged. 



I rci ently directed attention ("Radio-activity," p. i8j) 

 to the importance of the fact that in certain well-established 

 cases there appeared to be a simultaneous production of two 

 positive charges in the disintegration of an electrically 

 neutral atom. Thus in the disintegration of the emanation 

 atom a positively char(.jed o particle is expelb^l, and the 

 residue of the atom — the matter causing the excited activity 

 — is also positively charged, and is concentrated on the 

 negative electrode in an electric field. In a rei-cnt paper by 

 Hragg (Phil. Mai;., December, 1904, p. 721), the following 

 sentence occurs :- " It is easy to see that even if the a particle 

 is imcharged when it leaves the parent body, it inOst imme- 

 di.ilely become positive, since in traversing an atom it is 

 just as likely to lose one of its own electrons as to lake one 

 away from the atom traversed." As I am unaware that 

 this consequence has received the attention it deserves, 

 perhaps I may be allowed to direct attention to its bearing on 

 the present question. There is a fundamental distinction 

 between the ionisation of the atom of a gas molecule by 

 radiant electrons or particles, and radiant atoms or a 

 particles. For in the latter case, if the atom struck suffers 

 iiinisation, the radiant atom is just as likely to be ionised 

 in the process also. The ionisation of a neutral atom 

 consists in the detachment from it of an electron which 

 forms the negative ion, the atom thereby becoming positively 

 charged and forming the negative ion. Hence the radiant a 

 particle, if uncharged initially, will become positively charged 

 on collision with the atoms of the gas or other obstacle 

 in its path, and at the same time will lose an electron. The 

 " slow-moving electrons present with the a particles," which 

 Rutherford describes as " emitted from the plates," may 

 therefore in reality be derived from the a particles them- 

 selves in the act of becoming positively charged. The fact 

 th.it they, unless deflected by a magnetic field, exactly 

 n'utralise the charge carried by the o particles seems to 

 point in the same direction. 



In further support of the view that the positive charges on 

 both the radiant particle and the residue of the atom after 

 disintegration are derived by collision with the gas molecules. 

 Prof. Rutherford's results on the distribution of the excited 

 activity in an electric field at low pressure may be cited 

 (Rutherford, " Radio-activity," p. 282). If the excited- 

 activity-malter particle gains its positive charge in its recoil 

 by collision with the gas molecules, it is to be expected that 

 at low pressures it will not become charged, and will not, 

 therefore, be concentrated on the negative electrode, as is, 

 in fact, the case. pREDliRiCK SODDV. 



The Pressure of Radiation. 

 TlEK success „l I.ebedeff and Nichols and Hull in recog- 

 nising and measuring the pressure of radiation has aroused 

 much interest in radiation pressure generally, real or ap- 

 parent. It has some interesting and sometimes somewhat 

 difficult theoretical aspects. In the first place, if the ether 

 is really absolutely at rest (this rigidity is a very difllcult 

 idea), the moving force on it has no activity, and its time 

 integral Vdb can only be called momentum out of compli- 

 ment. The force becomes active in a moving ether, with 

 interesting consequences not now under examination. The 

 present question is rather how to interpret the pressure of 

 radiation on the assumption of a fixed ether, in the measure 

 of its effects on matter which is either fixed or moving 

 through the ether. 



The following is striking in what it proves. Let plane 

 radiation fall flush upon a perfect reflector moving in the 

 same direction at speed u, a case considered by I-armor. Let 

 the energy density p = p, + pj, the incident being p,, the 

 reflected p,. Assume, which seems reasonable at first, (hat 

 />3, the pressure in the reflector, is zero, then the moving 

 force p,+p2-p3 reduces to p^ + p... Therefore 



A(w-«)-/8(z' + «)=(A+A)". (0 



because the left side is the rate of loss of energy from the 

 waves, and the right side the activity of the force on the 

 reflector. .So 



A _ ' - 2u/v 



pi I -^ 2ujv 

 NO. 1845, VOL. 71] 



-S-, say. 



(2) 



and s = Hj/H| is thi' ratio of magnetic forces in the electro- 

 magnetic case. Now (2) asserts that the reflected wave gets 

 smaller as the mirror goes faster, and vanishes when tf=\v. 

 Or if the mirror be pushed against thi- radiati(m, the re- 

 flected wave gets stronger, and the resisting force stronger 

 until ii=— ix), when it is infinite. The mirror could not be 

 pushed against the radiation faster than \v. 



An immediati; objection is that when 11 has risen to Jr, if 

 it be maintained at that speed it acts like a perfect absorber 

 to the incident energy. Moreover, since there is the pressure 

 />, left, why should it not accelerate the mirror? Hut, if 

 it does, />.; becomes negative, and s becomes imaginary. 

 Considered mechanically only, say by l'=mu, the motion 

 of m is quite determinate when u>iv, up to V, in fad. 

 But electromagnetically it means that the energy in the 

 reflected wave is negative. Now although there is nothing 

 to object to quantitatively in a continuous transition from a 

 Maxwellian stress consisting of a tension along an axis 

 combined with equal lateral pressure, to its negative, a 

 pressure along the axis with equal lateral tension, still the 

 negativity of the energy in the reflected wave causes difll- 

 culty. The stress for both the electric and magnetic energy 

 becoines of the gravitational type. That is, fikc imaginary 

 electrifications attract, and unlike repel, or matter is 

 imaginary electrification in this comparison. 'Ihe moving 

 forces and energies arc real. Hut let a real charge and an 

 unreal one co-exist,, the energy density becomes imaginary. 

 That is out of all reason in a real universe. 



We should, I think, regard (2) as a demonstration that 

 (i) is untrue, in that (.p, + p.j)ti is not the activity of the 

 force on the mirror, although p, + p.^ may be actually the 

 pressure of the radiation. In fact, in the electromagnetic 

 case, the variation of p constitutes a force on the ether 

 itself. We must find the force on the mirror in another way. 

 Let radiation fall flush upon the plane surface of a dielei trie, 

 which call glass, moving the same wriv at constant speed », 

 and let the circuital equations in the glass be 



- dillJx = ct + dl/d/, - </E/,/.t = ii = mH ; (3) 



that is, the same as for (he ether, with the addition of thi- 

 electric current of polarisation dl/dt. The reference space 

 is the fixed ether, and d/dt is the moving time differentiator. 

 Now if the relation between I and Ii is such as 10 pc-rniit of 

 an undistorted plane wave, we shall have 



E,=H7Aii, E,,= -ix7Al„ E, = ^7«[I:i, (4) 



(Incidenl) (rtnectcd) (ir.innmitttd) 



if -.) is the speed in Ihe ether, and w the wave speed referred 

 to the ether in the glass. This -u; is a function of i(. .Also, 

 the boundary conditions, 



E,-hEa = E,„ Hj + IIj = Hi„ (5) 



combined with (4), give 



liJH,={v-w)/i,v + w), HJHt = 2v/{v + w). (6) 



An incident pulse of unit depth is stretched to depth 

 (,-uiv)-' in the act of reflection; the reflected pulse is of 

 depth (v+u)(v-u)-', and the transmitted pulse of depth 

 {w-u)(v-u)-'. . 



The rate of loss of energy from the waves m Ihe process 

 of reflection is 



where the p's are the energy densities. But, by Ihe above, 

 p^v=p.^v-\-p.iW, (8) 



therefore the rate of loss of energy is 



and the moving force on the mirror is 



F=A-A-A- ('°) 



This is, in its expression, exactly the negative of the 

 previous pressure difference. It is in the direction of the 

 rise of energy density. Its amount is 



F = 2^H,Hj = z/, (z/ -«/)/(» + a/) = i/»H;-irE= = U„. (11) 



The first form in Krms of H,,Hj is useful. The second is 

 in terms of the wave speeds. The third is in terms of the 

 ethereal energy inside the glass. -All these come out of the 

 ratios IL II,, &c. Now the electric energy equals the 

 magnetic energy in the transmitted wave. Consequently 

 l'„ means the energy of the polarisation I. And the activity 

 is L„H, the convective flux of energy. 



