March 13, 1913] 



NATURE 



3i 



therefore incorrect to call two gases with different 

 densities "non-separable by any known process." 



Electrical and magnetic forces are also agents which 

 can be applied to distinguish between molecules having 

 different masses. Such forces should be ci 

 before any sweeping assertions are made. 



It is possible that Mr. Soddy wishes his statement 

 to be limited to the ordinary chemical processes, and 

 as he is trying to prove a negative, it is perhaps 

 unfair to be too critical, but one cannot help remem- 

 bering the time when neodym and praesodym were 

 "non-separable," and reflecting how many substances 

 might not be separated at the present moment if their 

 optical properties had not given us a clue. No doubt 

 radio-active tests are severe, and the chemical pro- 

 perties of the bodies in question are probably more 

 nearly equal than those of the older chemistrv, but 

 there is a vast interval between "very similar" and 

 " identical." 



Incidentally, we may reflect that these bodies which 

 are believed to be "non-separable" actually ■ 

 themselves of their own free accord in the natural 

 course of their subsequent history, but this mav only 

 prove the perversity of nature. 



According to Mr. Soddy's theory, the non-separable 

 bodies have identical spectra. This is the vital issue, 

 which, if made good by experiment, will help us to 

 overlook many weaknesses in the argument. The 

 evidence here rests entirely on one experimental fact. 

 It was shown by Russell and Rossi, and also by 

 Exner and Haschek, that a mixture of ionium and 

 thorium does not show in the electric arc lines which 

 can be assigned to ionium, the spectrum of the mix- 

 ture being identical with that of pure thorium. 

 Assuming that ionium is the only intermediate product 

 between thorium-II. and radium, the life of ionium is 

 100,000 years, and the ionium-thorium preparation of 

 Russell and Rossi must have contained about 16 per 

 cent, of ionium. But these authors also point out that 

 if the length of life is reduced to 12,000 years, the 

 preparation would only contain 2 per cent., and the 

 absence of ionium lines would be accounted for. At 

 present the radio-active evidence seems in favour of 

 the longer period, and the absence of ionium lines 

 wants explaining ; nevertheless, it seems to me to be 

 going ahead too quickly to make a sweeping assertion 

 that not only is the spectrum of ionium identical with 

 that of thorium, but that the same holds in all similar 

 cases, for the accumulated evidence of the spectra of 

 known bodies has all been in the direction of indicat- 

 ing that optical properties of absorption and radiation 

 discriminate in the most decisive manner between 

 bodies which are otherwise similar in chemical pro- 

 perties. 



Granting now for the sake of argument that the 

 bodies in question have spectra which cannot be dis- 

 tinguished from each other, it remains to examine 

 the alternative that the bodies are actually identical. 

 It is said that they have different molecular weights, 

 because one has been formed from the other by an 

 expulsion of one a and two (3 particles. This argu- 

 ment is not necessarily conclusive, as a mass equal to 

 that expelled may have been picked up again in the 

 process. It may be urged that the subsequent history 

 of these bodies shows that they are essentially different. 

 Though a strong argument, this is not quite the last 

 word, because, granting for a moment the temporary 

 identity of two systems, the particular instability 

 which determines their future may depend on their 

 past. 



Taking all arguments into consideration, we are 

 left with an interesting theory consistent with our 

 present knowledge but supported bv very little real 

 evidence. It may be presumptuous for one w-ho can 

 only claim to be an amateur in modern physics to 



no. 2263, VOL - 9 1 ] 



ni opinion, but having in a previous genera- 

 tion taken part in establishing the fact that the same 

 element can have different spectra according to its 

 molecular constitution, one cannot, without good 

 cause, accept the belief that different elements can 

 have the same spectrum. Mr. Soddy's case would be 

 much strengthened if he could adduce positive instead 

 "i merely negative evidence, and this might be sup- 

 plied if the bodies grouped together with thallium 

 lim rould be shown to give the thallium spectrum, 

 assuming thallium not to be present in the raw 

 material. Arthur Schuster. 



Manchester, March 7. 



Atmospheric Electrification during South African Dust 

 Storms. 



This short note on the variation in the atmo 

 electrical charge due to the presence of dust is not 

 intended to be exhaustive, but merely to direct atten- , 

 tion to a factor which has a very great influence in 

 modifying the positive potential gradient existing in 

 the atmosphere during fine weather. Very few ob- 

 servations as to the causes of the variations have been 

 recorded, but Prof. Michie Smith seems to have 

 observed (Phil. Mag., vol. xx., p. 456) something of 

 the same kind during dust storms in India. He notes 

 that "the negative electrification was strongest during 

 gusts of dust-laden air," and, further, "the potential 

 would often run up so rapidly that it was impossible 

 to measure it accurately, whilst during lulls it would 

 often fall almost to zero." 



I was, however, unaware of any work having been 

 done in this direction until the present year, though 

 I have been making a study of the variations in the 

 potential gradient over the high veld in South Africa, 

 and have published several short papers on the sub- 

 ject. 1 The general result has been to show that very 

 extraordinary variations are caused by the presence 

 of dust in the atmosphere, whether due to the natural 

 dust-storms or to any artificial means, such as the 

 clouds of dust raised from the mine refuse heaps 

 formed during the working of the cyanide process. 

 At all the places where observations have been taken 

 the dust is either sand or is of siliceous character, and 

 invariably has the effect of lowering the positive poten- 

 tial gradient, and if present in sufficient quantity, to 

 reverse it and give a very high negative gradient. 



During the past six months systematic observations 

 have been taken at Bloemfontein with a Bendorf 

 recording electrometer, furnished with a radium- 

 coated plate to act as collector. (The apparatus was 

 obtained by aid of a grant from the Royal Society 

 of South Africa.) The normal potential gradient in 

 South Africa is, of course, positive, but varies con- 

 siderably with the elevation. The diurnal range is 

 also considerable under fine weather conditions, and 

 during stormy weather very great deviations are 

 shown if rain is falling or dust is blowing. It may 

 be noted here that during the past eighteen months, 

 when verv little rain has fallen, the charge brought 

 down bv the rain has been invariably negative. A 

 studv of the records made by the electrometer shows 

 that three tvpes have to be considered, viz. : — (1) The 

 ordinarv fine weather record ; (2) record taken on a 

 dav when some dust is blowing ; (3) record taken on 

 a very dusty day. 



In the first case, the positive gradient rises to a 

 maximum at about 7 to 8 a.m., falls to a minimum 

 at midday, remains fairly uniform over a period of 

 several hours, and then rises to another maximum. 

 The slope of the curve is steeper for this second 

 maximum than for the first one. Fig. 1 shows such 

 a curve which was taken in July, from midnight to 

 midnight. The horizontal line shows zero potential, 

 \frican Journal of Science ; Proc. Roy. Soc. Soulh Africa. 



