440 



NATURE 



[March 9, 1905 



These properties are irue for various relations between 

 I and E. The first approximation is I=c,E. The second, 

 introduced by I-orentz, is I=f|(E — iiB), that is, the puiarisa- 

 tion is proportional to the moving force on a moving ion. 

 Other forms allowing of undistorted pulse propagation may 

 be proposed. All give special relations between tc and ii. 

 In Lorentz's case, 



U„ = iME:;(i-«»2. (12) 



To pass to perfect reflection, reduce w to u, its least value. 

 U„ does not vanish, but has the value given by (lo), (ii) 

 still, with ui = ii. But the transmitted wave is reduced to a 

 surface film, moving with the glass. The moving force on 

 the glass is now 



F = 2/, (jc -«)/(;. + »), (13) 



and finally, if h = o, F = 2/>|. 



Here we come right baclc to the pressure of radiation. 

 It does measure the force on the glass when at rest, when 

 it reflects perfectly, and it looks as if (13) were merely the 

 form p^ + p: a little modified by the motion. But appear- 

 ances are very deceitful here, for (10) above is the proper 

 formula. 



.^s regards the distribution of F. With an actual trans- 

 mitted wave consisting of a pulse of uniform intensity all 

 through, F is entirely at the wave front. So, with total re- 

 flection, it is just under the surface of the glass. Again, if 

 Ej varies continuously in the transmitted wave, F is dis- 

 tributed continuously, to the amount BOI/ac) per unit volume. 

 What F means in (11) now is the total of this volume force, 

 i.e. the integral from the surface up to the wave front, ex- 

 pressed in terms of the momentary surface state. 



.After a pulse has left the surface there is an equal opposite 

 force at its back, so there is no further loss of energy or 

 moving force on the glass. The obscurities and apparent 

 contradictions arise from the assumption that the ether is 

 quite motionless. If we treat the matter more compre- 

 hensively, and seek the forces in a moving ether, with 

 moving polarisable matter in it as well, if this is a com- 

 plication one way it is a simplification in another, viz. in 

 the ideas concerned. There is harmonv produced with the 

 stress theory. To illustrate, {d/dt)VDB is the moving force 

 per unit volume when the ether and polarised matter have a 

 common motion, D and B being the complete displacement 

 and induction. (The variation of u is ignored here.) But 

 if we stop the ether, a part of this force becomes inactive. 

 If the rnatter is unmagnetisable, the only active part is that 

 containing the polarisation current, for that is carried along. 



Besides this electromagnetic force, there is also a force 

 due to a pressure of amount U„. But it does not alter 

 the reckoning of the moving force on the glass, because 

 the pressure acts equally and oppositely at the front and 

 back of a pulse. 



Some other illustrations of the curious action between 

 electromagnetic radiation and matter can be given. For 

 example, two oppositely moving plane pulses inside moving 

 glass. Say E, = /iw,H, one way with the glass, and 

 E,= -/iiti.H, against the glass. I'f H, = -H,, work is done 

 upon the glass when they cross, ceasing the" moment they 

 coincide, so that the energy of the momentary electric field 

 is less than the wave-energy. On separating, the loss is 

 restored. If, on the other hand, E, = -E,, work is done by 

 the glass on the waves when uniting, so that the momentary 

 magnetic energy, together with the polarisation energy, is 

 greater than the wave energy. In this second case, too, it 

 is noteworthy that the solitary waves are of unequal energy, 

 whereas they are equal in the first case. But details must 

 be omitted, as this communication is perhaps alreadv too 

 long- Oliver Heavisi'de. 



Februarv 21. 



Secondary Rontgen Radiation. 



In a paper read before the Kuyal Society on Februarv 16, 

 I described experiments demonstrating the partial polarisa- 

 tion of Rontgen radiation proceeding from an X-ray bulb, 

 and at the same time verifying the theory previously given 

 of the emission of secondary X-rays from gases and light 

 solids subject to Rontgen radiation. 



Later experiments have shown that beams of X-radiation 

 rnay be produced exhibiting a greater amount of polarisa- 

 tion than there was evidence of in the original experiments. 



This discovery has proved useful in the investigation of 

 secondary radiation proceeding from solids. 



It has been found that while the intensity of secondary 

 radiation from light substances varies considerably in 

 difterent directions owing to the partial polarisation of the 

 primary radiation, the amount of this variation diminishes 

 with an increase in the atomic weight of the radiator, and 

 ultimately is inappreciable. The radiations from air, carbon, 

 paper, aluminium, and sulphur vary in intensity in different 

 directions by a considerable amount. From calcium the 

 variation is much less, while from iron, copper, zinc, and 

 lead it is inappreciable. This must be connected with the 

 fact that the radiation from light substances differs in 

 character only very slightly from the primary, while the 

 heavier substances emit radiations differing more from the 

 primary producing them. The radiation from the heavier 

 metals was found not to consist of an easily absorbed radia- 

 tion superposed on a radiation such as proceeds from light 

 substances, and of intensity given by the law found for that 

 from light substances, but is as a completely transformed 

 radiation. This is strong evidence that the freedom of 

 motion of the electrons which permits what may be called 

 a simple scattering in substances of lower atomic weight is 

 interfered with in the heavier atoms, for we find from them 

 a more absorbable radiation in place of, not simply super- 

 posed on, a more purely scattered radiation. 



With this change in character, the polarisation effect dis- 

 appears. No special absorption of the radiation proceeding 

 from a substance by plates of the same substance has been 

 observed. 



.\ considerable variation in the penetrating power of the 

 primary radiation incident on heavy substances is accom- 

 panied by a smaller change in that of the secondary 

 (measured bv change of absorbability). 



Radiation from compounds appears to be merely a mixture 

 of the radiations which proceed from the separate elements 

 in the compound, both the absorbability and polarisation 

 effects being what would be given by such mixtures. 

 •Atomic weight, not molecular weight or density, thus seems 

 to govern the character of the radiation produced by a given 

 primary. 



These results may be accounted for by considering the 

 electrons constituting the atoms as the radiators. In light 

 atoms the electrons are far enough apart, and have sufficient 

 freedom to move almost entirely independently of one another, 

 under the influence of the primary pulses, consequently to 

 emit a secondary radiation similar to the primary, but the 

 intensity of which depends on the direction of propagation 

 with regard to that of electric displacement in the primary 

 beam. In heavier atoms considerable inter-electronic forceg 

 are probably brought into play by small displacements, and 

 the resultant acceleration of motion of an electron is then 

 not in the direction of electric displacement of the primary 

 beam, and evidence of polarisation of that beam vanishes. 

 Also there ceases to be a simple connection between the 

 time for which the electron is accelerated and that of passage 

 of the primary pulse. 



In atoms of greater weight we would expect appreciable 

 inter-electronic forces to be called into play sooner, and to 

 attain a much greater intensity than in lighter atoms. 



The precise connection between the atomic weight of the 

 radiator and the absorbability of the radiation is being 

 investigated. Ch.\rles G. Barkl.a. 



L'niversitv of Liverpool, March i. 



Dates of Publication of Scientific Books. 



1 riAVE just bought a copy of " A Treatise on Slate and 

 Slate Quarrying, Scientific, Practical, and Commercial," by 

 I). C. Davies, F.G.S., fourth edition, dated 1899 (Crosby 

 Lock wood and Son). 



To mv astonishment, I find no statistics of later date in 

 it than 1876, e.g. p. 33, statistics of 1872 and 1873, p. 58, 

 list of quarries in 1873, p. 59, production in 1876, p. 64, pro- 

 duction last year (1876), p. 170, prices of slates in London 

 last year (1S76). 



.\s the Home Oflice publishes annually a general reporl 

 and statistics of mines and quarries, and also a list of mines 

 and quarries, there is no excuse for the book being so out 

 of dale in its statistics. B. IIOBSON. 



The Owens College, Manchester, Februarv 21. 



NO. 1845, VOL. 71] 



