October 28, 1915] 



NATURE 



241 



obtained from laboratory researches ; from types 

 to M, the stars show spectral lines characteristic 

 )f the "arc"; from B2 to F the "spark" lines are 

 shown; and in the earliest divisions O to Bi the 

 lines are for the most part unattainable in the labora- 

 ftory, except by the most powerful electric discharge. 

 [These last may be described as "super-spark" stars. 

 'The evidence fully establishes a physical continuity 

 corresponding to the Draper sequence. Prof, Fowler 

 concluded by referring to the theory of Sir Norman 

 Lockyer, and to Prof. Russell's hypothesis, which is 

 closely akin to it. Accordinf^ to these, the stars fall 

 into parallel series of ascending and descending tem- 

 perature respectively, so that the hottest stars (types 

 B and O) come midway in the order of evolution. 



Sir Frank Dyson set forth the evidence, based on 

 the observed luminosities and densities of the stars, 

 which led Russell to the view that the red stars 

 represent both the earliest and latest stages of evolu- 

 tion. It seems necessary to believe that some of the 

 M stars are extremely diffused bodies, and others 

 very dense. The Draper sequence gives the order 

 of temperature; but the order of evolution is that 

 of Increasing density, not necessarily that of decreas- 

 ing temperature. Prof. Eddington pointed out that 

 _the actual calculations of stellar density in particular 

 ises compel us to admit that two stars having the 

 snsities respectively of water and of air can yet 

 show the same type of spectrum ; this seems to 

 remove one of the chief objections brought against 

 Russell's theory. The adoption of this theory would, 

 however, play havoc with the regularity of many 

 statistical results which have hitherto seemed orderly 

 and Intelligible. Father Cortie referred to the 

 changes of a spectrum in a Nova, which seem to 

 show that in these cases the nebular and type O 

 spectra come, not before, but after the B and A 

 stages. Prof. Nicholson and Mr. Merton discussed 

 the nature of the Wolf-Rayet spectrum (type O). In 

 his reply. Prof. Fowler pointed out a serious difficulty 

 of Russell's hypothesis, that we have no celestial 

 spectra which can be regarded as bridging the gap 

 between primordial nebulous matter and the intensely 

 bright giant stars of type M. 



Thermionic Emission. 

 On Friday Prof. O. W. Richardson opened a dis- 

 cussion on thermionic emission. He said that he 

 would confine his remarks to the emission of nega- 

 tive ions from hot bodies, which seemed to be an 

 intrinsic property, whereas the emission of positive 

 ions was not permanent and could be traced to impuri- 

 ties. The emission of negative Ions increases rapidly with 

 temperature according to the formula C = ATii?-'''/T, 

 where T is the absolute temperature. It was found, 

 however, by H. A. Wilson and by Langmuir that 

 the constants depend greatly on the experimental 

 conditions. The temperature-law would follow 

 theoretically from purely physical considerations. On 

 the other hand, a similar law may be deduced if 

 the effect is due to chemical action. In the experi- 

 ments of Haber and Just the emission of electrons 

 from alkali metals was observed under the action of 

 water vapour and other agents. This is the only 

 known case of the kind, and the distribution of energy 

 among the emitted electrons suggests that it Is some- 

 thing different from thermionic emission. On 

 the question of the distribution of energy among 

 the electrons during chemical action, Prof. Richard- 

 son had gained the impression from his experiments 

 that it did not conform to Maxwell's law, but was 

 more analogous to the laws governing the photo- 

 electric effect ; the facts of thermionic emission re- 

 quire Maxwell's law. As the best possible test 



NO. 2400, VOL. 96] 



between the physical and chemical theories, he had 

 recently conducted experiments on a tungsten wire 

 with special precautions against impurities ; the 

 emission was found to be much too great and per- 

 sistent to be accounted for by residual gaseous im- 

 purities. He concluded that the action was not 

 chemical, nor could it be a photoelectric action of the 

 temperature radiation ; and the physical theory seemed 

 the most satisfactory. 



Dr. F. A. Lindemann discussed the bearing of the 

 experiments on the chain theory of electric con- 

 duction. Mr. E. Newbery criticised the use of tung- 

 sten in the crucial exf)eriment on account of its 

 great chemical activity at high temperature. Dr. 

 J. A. Harker discussed the bearing of experiments 

 on the electric arc under high pressure. 



Miscellaneous Papers. 



There was no subdivision of the Section this year, 

 and the papers selected for reading were taken in 

 full Section. Technical subjects were therefore 

 avoided, and even the mathematical papers were such 

 as would appeal to the members generally. Mr. 

 G. H. Hardy greatly interested his audience with a 

 paper on prime numbers. He gave an historical 

 survey of the investigation of the distribution of 

 primes, with particular reference to the theorem that 

 the number of primes less than x approaches asymp- 

 totically to x/logx for large numbers. This theorem 

 was first conjectured by Legendre, but was not proved 

 until 1896. An explanation of the importance of the 

 Rlemann-^ function in this connection was given. 

 The paper is to be printed in extenso In the report. 

 Prof. A. N. Whitehead, In a paper on space, time, 

 and relativity, gave an account of the philosophical 

 difficulties connected with space and time so far as 

 they concern mathematicians. His remarkably 

 simple method of arriving at the fundamental equa- 

 tions of the principle of relativity was especially valu- 

 able. Mr. A. A. Robb and Mr. H. R. Hass'd took 

 part in the discussion, the former explaining his 

 method of logical development of the subject based 

 on the Idea of points arranged in "conical order." 



Prof. W. H, Bragg's account of X-rays and crystal 

 structure showed the power of the new methods of 

 determining the arrangement of the atoms in a 

 crystal. Instead of attempting to summarise this 

 remarkable paper, we may refer the reader to Prof. 

 Bragg's Bakerian lecture in the Phil. Trans, (vol. 

 ccxv., p. 253), which covers similar ground. Prof. 

 J. C. McClennan gave an account of his production 

 of single-line spectra of cadmium and zinc, shovv-Ing 

 that the wave-lengths of the lines are connected 

 with the ionisation potential by a relation depending 

 on the quantum theory. Sir J. Larmor discussed 

 the decomposition of the irregular vibrations con- 

 stituting white light into regular trains of waves, 

 when a grating or prism is used; for the prism, he 

 gave an explanation based on the difference between 

 wave-velocity and group-velocity in a dispersive 

 medium. 



Meteorology was represented by a paper by Mr. 

 F. J. W. Whipple on the mechanism of cyclones, in 

 which an account was given of the observed distribu- 

 tion and pressure at different heights, and the 

 dynamical connection between the pressure-gradient 

 and the inflow and outflow of air was discussed. In 

 presenting the report of the committee on seismology. 

 Prof. H. H. Turner stated that the work of plotting 

 on a map the earthquake epicentres observed by 

 Milne had now been completed. The epicentres were 

 found especially on two great circles cutting at right 

 angles. He referred also to Improvements made in 



