July 5, 191 7] 



NATURE 



365 



logical problems of the first importance is pointed 

 out by Sir Napier Shaw in the preface, and we 

 trust that the wished-for extension of stations into 

 wide areas not yet represented in the Reseau 

 Mondial will materialise before long-. 



The work concludes with an excellent biblio- 

 graphy of the 160 memoirs consulted by the 

 author. There are numerous illustrations, includ- 

 ing plates showing the seasonal isotherms, and 

 isopleths for four typical stations. This book will 

 remain a standard work of reference on Brazilian 

 meteorology for many years to come. R. C. M. 



LETTERS TO THE EDITOR. 

 [The Editor does not hold himself responsible for 

 opinions expressed by his correspondents. Neither 

 can he undertake to return, or to correspond with 

 the writers of, rejected manuscripts intended for 

 this or any other part of Nature. No notice is 

 taken of anonymous communications.] 



The Radiation of the Stars. 



The law TocM'-p» given by Prof. Eddington in 

 Nature of June 14 is so neat that one feels reluctance 

 to cast any doubt on it, but since it is notoriously 

 difficult to destroy a scientific error which has once 

 obtained a fair start, I feel that whatever is to be 

 said against the law ought to be said now. 



The law is derived from Prof. Eddington 's supposed 

 equation, radiation-pressure := gravity. It must first 

 be noticed that the two sides of this equation are of 

 different physical dimensions (ML-*T-^ and LT-" 

 respectively), so that neither the arguments on which 

 the equation is based nor the laws derived from it can 

 carry much conviction. If p,i and pn denote gas- 

 pressure and radiation-pressure, the true equation of 

 equilibrium is 



d 



dr 



(Po+pK)=-pg, 



1^ being gravity, and when po is neglected in com- 

 parison with pR, g does not become equal to /«, but to 



p dr' 



It is true that in his original paper on this subject 

 {M.N., November, 19 16) Prof. Eddington obtains a 

 '.relation of the form pK = Rg, where R is a constant, 

 !but this equation is not based on the physical concep- 

 tions put forward in the article in Nature. The out- 

 iward flow of radiation from a star of radius r and 

 sfFective temperature T is, of course, 4ffr^Xo-T^. Where 

 does the energ\' of this radiation come from? Prof. 

 iEddington assumes that it is a transformation of 

 i-adio-active energy, initially emitted at a uniform 

 I -ate 4iTe per unit mass throughout the whole star. The 

 I'tar is assumed to be in a steady state, so that the rate 

 pf generation of energy, 4-Me, is equal to the rate of 

 amission 4n-rVT^, whence o-T'^Me/r'. 



This equation is of the form p^—'Rg, and on 

 eliminating r by means of the relation M=A-/3r^, we 

 ind that TcoMt^*. (Prof. Eddington's complicated 

 inalysis, boiled down to its essentials, comes merely 

 I o this.) But it now appears that the supposed law 

 s not, as might be thought from a casual reading of 

 |he article in Nature, a deduction from already 

 mown laws of Nature; it is a transformation of the 

 ery special assumptions that all stellar matter is 

 qually and uniformly radio-active, and that this 



NO. 2488, VOL. 99] 



radio-activity is the origin of all the energy' radiated 

 from a star. 



The second supposed law, that the total emission 



of radiation from a star depends only on its mass, is 



a still more thinly veiled repetition of the same 



I assumptions. The radiation is assumed tobe4n-M«, and 



; e is assumed to be a universal constant. When these 



' assumptions have been made, it is scarcely worth 



announcing as a new law that the radiation depends 



only on M. 



I venture to think that few physicists will be ready 

 to accept these assumptions; in any case it is impor- 

 , tant that they should be clearly stated. It may be 

 I remarked that if the assumptions are true for giant 

 j stars they might be expected . to be equally true for 

 I dwarf; they lead, however, to the laws given by Prof. 

 I Eddington, such as that of absolute brightness de- 

 pending only on mass, which are palpably untrue for 

 dwarf stars. J. H. Jeans. 



I London, June 16. 



Mr. Jeans 's criticism of the dimensions of my 

 equation, "radiation-pressure=gTavity," is clearly 

 only a verbal matter. It may be preferable to expand 

 the sentence so as to read, '* force on material due to 

 radiation-pressure = force on material due to gravity.** 

 But statements that radiation-pressure is x times 

 gravity have been commonly made in connection with 

 the theory of the repulsion of comets' tails, and I 

 thought the reader would have no difficulty in inter- 

 preting my statement in the same sense. 



The rest of Mr. Jeans 's letter is based on a mis- 

 conception. He states that I assume e (the rate of 

 liberation of energy per unit mass) to be a universal 

 constant; that is not the case, and consequently his 

 account of the mode in which my principal results 

 arise is erroneous. My results are not " a trans- 

 formation of the very special assumptions that all 

 stellar matter is equally and uniformly radio- 

 active. . . ." 



I should be glad to take the opportunity of correct- 

 ing a misprint in the article : p. 309, col. 2, 1. 18 from 

 bottom, /or M pi reaif M V- A. S. Eddington. 



Cambridge, June 19. 



Protection from Glare. 



Efficient protection of the eyes from glare is a 

 subject of considerable imf>ortance at the present time, 

 but unfortunately a great deal of misconception has 

 arisen in regard to it. Most glare protectors are 

 designed for conditions of unusually strong illumina- 

 tion ; generally speaking, for daylight. Many indus- 

 trial operations also demand the use of light filters. 

 We will deal first with the problem of protection in 

 sunlight. 



A great deal of experimental work has been done 

 recently on the physiological effects of ultra-violet 

 radiation in the eye, a quartz-mercury lamp being 

 used as a source, which is especially rich in its emis- 

 sion of the shorter wave-lengths. The results, un- 

 questionable and undoubted, are that with long-con- 

 tinued exposure serious harm may result from the 

 absorption of these shorter waves by the refracting 

 media of the eye, but with low intensities of radiation 

 regular exposures produce no permanent effect. Turn- 

 ing now to the effects of daylight on the eyes, we 

 recognise that few cases occur in which the symptoms 

 point directly to the influence of ultra-violet light as 

 distinct from the effects of ordinary strong illumination ; 

 but the problem of nerve strain arising from g'are 

 is ever present. 



Quite low intensities of light will produce glare 



