454 



NATURE 



[January i, 1920 



Helmand Kiver flows, and which is occupied, ac- 

 cording 10 season, by a large lalte or by a series of 

 lakes of variable area. Owing to the fact that in 

 flood-time the Hamun overflows, by the Shclagh River, 

 into the Gaud-i-Zirreh, "the Dead Sea of the Hel- 

 mand system," its waters do not reach a high degree 

 of salinity, and it sustains a fauna, impoverished 

 indeed, but rich in comparison with that of true salt 

 lakes. Dr. Annandale points out that although the 

 Hamun occupies part of an ancient lake-bed, "there 

 has been no biological continuity between the old lake 

 and the recent one." The present lake may even 

 have originated within historic times by a shifting of 

 the course of the Helmand River. Dr. .Vnnandale 

 describes the Cyprinid fishes of the genus Dis- 

 cognathus found "in the region and, in collaboration 

 with Dr. B. Prashad, the Mollusca. In the case of 

 the latter it is pointed out that the fauna shows a 

 mingling of Palaiarctic and Oriental types and a note- 

 worthy absence of Western Asiatic elements. The 

 fauna, however, "is a starved one, in which only 

 species of great adaptability can survive." 



PHYSICS AT THE BRITISH ASSOCIATION. 



ONE day of Section A was devoted almost 

 entirely to matters relating to wireless tele- 

 graphy. Prof. Eccles opened a discussion on 

 thermionic valves, giving a general description of the 

 history and development of the three-electrode valve, 

 explaining its rectifying property, the method of 

 heterodyne reception, and the arrangements necessary 

 to produce continuous waves. Experiments were 

 shown illustrating these uses of the valve, and the 

 way was thus prepared for the discussion of special 

 points by subsequent speakers. Prof. Fortescue 

 directed attention to the functions and properties of 

 the various parts of the valve in some detail. The 

 hot filament is the source of the electrons upon which 

 the action of the valve fundamentally depends; with 

 tungsten filaments as at present used only 4J per cent, 

 of the energy heating the filament is usefully em- 

 ployed as electron emission. This efficiency might be 

 improved by using oxide-coated filaments or higher 

 temperatures, but at present neither of these methods 

 has been entirely successful in practice. The construc- 

 tion of the grid and the question of freeing the anode 

 and containing vessel from occluded gas during 

 ;;umping were also discussed, and the importance of 

 investigating the methods of removing the last traces 

 of gas and examining their nature was emphasised. 

 Dr. Whiddington directed attention to the possibility 

 of using valves and oscillating circuits for making 

 many standard physical measurements. Thus, for 

 example, the coefficient of mutual induction can be 

 determined by observing the degree of coupling at 

 which oscillations are just started and maintained. 

 He also alluded to Prof. Eccles's example of the 

 extreme sensitiveness of heterodyne reception as illus- 

 trated by the effect of passing coal-gas between the 

 plates of a condenser in an oscillating circuit. The 

 temperature coefficient of resistance, the conductivity 

 of flames, the permeability of liquids, and other 

 quantities could also be measured by this delicate 

 method. 



In a paper entitled " A. Wireless Method of Measur- 

 ing e/m" Dr. Whiddington showed how oscillations 

 may be set up in valve -circuits, not including capaci- 

 ties and inductances. The oscillations are produced 

 by bursts of electrons from hot spots on the filament 

 of a soft valve and the periodic return of positive ions 

 from the space between the grid and anode. In the 

 special experiments described it was shown from the 

 NO. 2618, VOL. 104] 



value of ejm obtained that the ions consisted of 

 mercury. 



The report of the committee on wireless telegraphy 

 was of special interest on account of the observations 

 made on the strength of signals during the recent 

 solar eclipse. It was, however, too early to give any 

 very definite conclusions, ;ilthough it was stated that 

 Malta and Paris had received signals of increased 

 intensity during the eclipse. Bearing on the same 

 point. Prof. G. N. Watson gave a rdsume of his recent 

 work on the diffraction of electric waves, in which, 

 starting from the Heaviside-Eccles hypothesis of con- 

 duction in the upper regions of the atmosphere, 

 .\ustin's formula can be obtained as a result of cer- 

 tain simple assumptions. 



Papers were read by Prof. Horton and Miss .A. G. 

 Davies and by Prof. Horton and Miss D. Bailey 

 respectively on the ionisation by electron collisions in 

 argon and helium and on the luminosity produced in 

 the latter gas. It appears that there are two critical 

 velocities of the electrons at which radiation from 

 the atoms and ionisation occur respectively. In argon 

 these two phenomena occur at 11-5 and 15-1 volts, and 

 in helium at 204 and 256 volts. The results are of 

 great interest, but, as Dr. Goucher pointed out, thtir 

 interpretation seems still open to question. 



The phenomena of novae were dealt with in two 

 papers by Mr. Stratton and Father Cortie. In the 

 former paper the types of spectra occurring in tlie 

 course of the history of Nova Geminorum were 

 described. The observed displacements of the 

 spectral lines correspond with velocities reaching 

 2x10* cm. /sec, which are so large that electrical 

 causes are suggested to explain them. Similar velo- 

 cities were deduced from the observations on Nova 

 .'\quil3e, and, after sketching the sequence of progres- 

 sive changes occurring in the star, Father Cortie 

 conclude! that a solar eruption in a giant star situated 

 in a dark nebula would square with the observ(!d 

 spectral changes. 



In an interesting communication on the theory of 

 vision Sir Oliver Lodge put forward the suggestion 

 that the retina may be found to contain atoms in such 

 a condition of instability that impulses of the correct 

 luminous frequency can excite them and cause the 

 expulsion of electrons. A difficult but highly interest- 

 ing experiment was suggested of trying to find in the 

 retina chemical substjmces capable of emitting high- 

 speed electrons when subjected to light. 



Prof. Eddington gave an account of the observa- 

 tions which had been made at Principe during the 

 solar eclipse. The main object in view was to observe 

 the displacement (if any) of stars the light from which 

 passed through the gravitational field of the sun. 'l"o 

 establish the existence of such an effect and the deter- 

 mination of its magnitude gives, as is well known, 

 a crucial test of the theory of gravitation enunciated 

 by Einstein. Prof. Eddington explained that the ob- 

 servations had been partially vitiated by the presence 

 of clouds, but the plates already measured indicated 

 the existence of a deflection intermediate between the 

 two theoretically possible values 0-87' and 175'. He 

 hoped that when the measurements were completed 

 the latter figure would prove to be verified. Inci- 

 dentally, Prof. Eddington pointed out that the presence 

 of clouds had resulted in a solar prominence being 

 photographed and its history followed in some detail. 

 Some very striking photographs were shown. 



Following on this account Prof. Eddington opened 

 the discussion on relativity, and referred again to the 

 bending of the wave-front'of light to be expected from 

 Einstein's new law when the light passes near a heaw 

 body. It should be oossible to test experimentally this 



