2 94 



NATURE 



[November 7, 1912 



wave amplitude would be reduced to about one-lhird 

 of that at the origin. It is clear, therefore, that to 

 propagate waves for any considerable distance through 

 the earth's crust the specitic resistance would have to 

 be as high as looo megohms per c.c. Also it is 

 evident that unless the resistivity of the air as pro- 

 duced by ionisation is less, say, than 20,000 megohms 

 per CO., or considerably less than 10-° electromagnetic 

 units, the ionisation cannot be the cause of the day- 

 light absorption. 



I believe no one has observed so low a resistivity 

 for air near the earth's surface, or even a thousand 

 feet up, as 10" C.G.S. electromagnetic units per c.c. 

 The point, however, requires further investigation. 



On certain assumptions this atmospheric resistivity 

 can be determined by elevating a captive balloon to 

 the required level by a wire, and measuring the poten- 

 tial of the balloon when the wire is insulated, and 

 also the current flowing through the wire when the 

 wire is earthed. The ratio of the current in amperes 

 to the potential in volts gives us the total conductivity 

 S of the air. If C is the electrical capacity of the 

 balloon, and if a is the conductivity of the air, then 

 it can easilv be shown that S = 4'rC(r (see J. A. Flem- 

 ing, " Principles of Electric Wave Telegraphy and 

 Telephony," second edition, p. 725). Accordingly 

 P = 4TC/S = 47rCI/V, where I is the total electric cur- 

 rent flowing along the wire to the earth, and V is 

 the potential of the balloon. 



Some useful measurements of this kind made by 

 Messrs. A. J. Makower, W. Makower, W. M. Gregory, 

 and H. Robinson, in 1910 ("An Investigation of the 

 Electrical State of the Upper Atmosphere made at 

 Ditcham," see Quarterly Journal of the Roval Meteoro- 

 logical Society, vol. xxxvii., October, igii) showed 

 that the total resistance from a certain kite at a height 

 of 1400 ft. on a certain day was 1000 megohms. 

 Assuming a kite capacity of 100 electrostatic units, 

 this corresponds to a resistivity of the air equal to 

 125 X 10" ohms, which seems rather small. A plane 

 electric wave would, however, have to travel nearlv 

 6o,oon km. in the air to have its amplitude reduced 

 to i/E. Hence even this conductivity is not enough 

 to account for the observed attenuation by daylight 

 of long radio-telegraphic waves. 



Unless, therefore, there is very much greater atmo- 

 spheric conductivity than one billionth of a mho 

 d bimho) per qx. at or about 1000 ft. or so above 

 the earth's surface, it does not appear as if air con- 

 ductivity caused by ionisation due to ultra-violet light 

 could account for the diminished amplitude of radio- 

 telegraphi( waves during davtime. 



The careful measurements of the air conductivitv 

 at various heights, and at various hours of the dav 

 and night over sea and land, would give us valuable 

 data for further testing this matter. 



Sommerfeld suggests that the daylight reduction 

 is due to the increase in the value of the coefiRcient h 

 for the air bv the production of conduction due to 

 ionisation. The effect of this is lo increase the value 

 of the " numerical distance," and therefore to reduce 

 the intensitv both of the surface and the space waves. 

 Ml- supnorls this view by an observation made bv 

 Ebert, who states that he measured the conductivitv 

 of the air in brightest sunshine at a height of 2500 m. 

 during a balloon ascent and found it to be twenty- 

 three times greater than at the ground level." 



It is possible that some part of the effect mav be 

 due to actions taking place quite close to the sending 

 antenna. This view mav be supnorted bv the interest- 

 ine observations made durlne the nearlv total solar 

 eclipse on April 17 last, on the effect of the tempornrv 



'■^ See Attn, dcr Pftvs., vol. v., p. 724, 190T. 



NO. 2245, VOL'. 90] 



diminution of daylight on the strength of radio-tele- 

 graphic signals. 



Whilst visiting the Eiffel Tower station at Paris, 

 Commander Ferris, who is in charge of this station, 

 informed me a slight increase in the strength of the 

 signals at distant receiving stations had been noticed 

 at the time of greatest observation of the sun at Paris. 

 Also in Denmark, Mr. H. Schledermann stated in 

 a letter to The Electrician that observations made 

 between the Royal Dockyard station in Copenhagen 

 and the Blaavands Huk Lighthouse on the North 

 Sea, at 300 km. distance, showed that durin"; totality 

 the signal strength was increased. 

 ■ Also in England, Dr. Eccles noted an increase in 

 the strength of atmospheric strays and signals from 

 I Clifden observed in London during greatest obscura- 

 tion. These observations show that even a partial 

 I diminution of the sun's light is suflicient to increase 

 1 the strength of radio-telegraphic signals, possibly by 

 I an action on the air between the stations, and especially 

 on that near the transmitting station. 



I suggest that it would be well worth while to erect 

 temporary transmitting stations on and off the line 

 of totality during the future total solar eclipses, for 

 ! the purpose of observing the effect on radio-telegraphic 

 I signals sent out from them to other receiving stations. 

 Another possible explanation of this daylight diminu- 

 tion has, however, occurred to me which I should 

 like to submit to you. It is well-known that sound 

 is better heard when the wind is blowing from the 

 source to the observer than when it blows in the 

 opposite direction. It is also known that there are 

 curious vagaries in sound transmission whereby loud 

 sounds are heard sometimes better at great than at 

 short distances. These effects were explained by Sir 

 George Stokes as due io the fact that the velocity of 

 sound is greater when moving with the wind than 

 against the wind. Now, owing to friction and other 

 causes, the velocity of the wind is generally greater at 

 a height ab.ovc the earth's surface than at the ground 

 level. Hence, if a sound wave is travelling outwards 

 from a centre against the wind, the upper parts move 

 more slowly than the lower parts of the wave front, 

 and hence the ray direction is tilted up and the sound 

 passes over the observer's head. 



The suggestion I venture to make is that when the 

 upper layers of the air are ionised, the ions act as 

 condensation nuclei for water vapour, and the presence 

 of these numerous water spherules gives the upper air 

 a larger dielectric constant.'^ Therefore an electric 

 wave moves more slowly in it than in non-ionised air. 

 Hence, if a plane wave is moving parallel to the 

 earth's surface and the upper layers of air are ionised 

 by light, the greater velocity of the wave front at the 

 lower levels causes it to slope backwards and the 

 direction of the ray is elevated, so that it may pass 

 over above the receiving antenna and not affect it. 



.Accordingly, if in the upper region of the air the 

 ionisation of the air by ultra-violet light increases the 

 dielectric constant so as to retard slightly the wave 

 velocitv, the wave front would be tilted backwards 

 and at long distances the waves might pass so far 

 above the receiving antenna as to weaken very greatly 

 the signals^ 



This tilting up of the ray will occur when the 

 ionisation of the upper air has taken place over a 

 part of the interval between the stations. It will be 

 most pronounced when the greatest difference exists 

 between the dielectric constant at the earth level and 

 that at the level a few miles up in the air. At very 

 large distances, say 2000 miles, an extremely small 



"See Sir I. J. Ttionison. " Coiidiiclin.. In O.-ise^," p. 517. who sayi tlia' 

 .lir exposed to uItra-\ iolct light may lie regarded as full of exiremely minute 

 drops of water. 



