September 8, 1923] 



NA TURE 



;6i 



experience suffered by Huxley when he lectured at 

 the Royal Institution on the cerebellum. At the end 

 of the lecture, a devout hearer approached to inform 

 him that she had understood and enjoyed the lecture 

 — with the exception of one point — was the cere- 

 bellum inside or outside the skull ? After I have 

 filled 24 columns of your valuable space to prove that 

 Huxley was altogether right when he denied that 

 use-inheritance played any part in the evolution of 

 man — or of any other animal — Prof. MacBride, after 

 reading these columns, turns round and practically 

 asks me if I have heard of Kammerer ! 



If Prof. MacBride will be so good as read my 

 Huxley lecture again, he will see that I neither 

 affirm nor deny the doctrine of use-inheritance What 

 I have denied, in as clear terms as are in my voca- 

 bulary, is that Lamarckism — whether of the original 

 1809 vintage or of the brand bottled in 1923 by Prof. 

 MacBride — has had no part in the evolution of man. 

 To give my reasons for this conclusion would compel 

 me to inflict on the readers of Nature a repetition 

 of my Huxley Lecture. Here I must content myself 

 by saying that Lamarckism gives no explanation of 

 man's developmental history, none of his anatomy ; 

 it leaves the ancestral forms of man, such as we 

 know of from the discovery of their fossil remains, un- 

 explained ; it cannot explain the characters which 

 differentiate one racial type of modern man from 

 another. In brief, the tenets which Prof. MacBride 

 clings to with such fidelity cannot serve the purposes 

 of even a working hypothesis for the modern anthro- 

 pologist. 



Prof. MacBride is good enough to suggest that I 

 should be staggered did I know of certain facts with 

 which comparative embryologists are familiar. Well, 

 I do sometimes make little excursions into the realms 

 of invertebrate embryology and frankly confess I 

 am staggered by the fact that men who are familiar 

 with the developmental histories of invertebrate 

 animals can have any belief of Lamarckism as a 

 factor in evolution. Arthur Keith. 



Solar Activity and Atmospheric Electricity. 



Dr. Bauer's courteous attempt (Nature, August 

 II, p. 203) to reconcile our views respecting the con- 

 nexion he believes in between sun-spots and atmo- 

 spheric electricity calls for a reply. I should first 

 explain that we differ as regards even the connexion 

 between sun-spots and terrestrial magnetism. Appar- 

 ently we both accept the relation 



R = o+6S . . . . (I) 



beTtween R, the range of the regular diurnal variation 

 for the year, and S, the sun-spot number. Here a 

 represents the range for no sun-spots and 1006 the 

 increase in range for a sun-spot frequency of 100. 

 The value of 1006/0 varies with the magnetic element 

 and with the station, but is usually in the neighbour- 

 hood of 0-8. The further relation mentioned by Dr. 

 Bauer, p. 204, " an increase of 100 in the sun-spot 

 number would correspond to a decrease in the in- 

 tensity of magnetisation of the earth of about o-i 

 per cent.," is not a result I consider proved. If it 

 were true, there should be a decided ii-year period 

 in the secular change. Claims to have established 

 such a period have been made, but seem to me to 

 have broken down. Quite recently failure to detect 

 the phenomenon at Paris, one of the most satisfactory 

 stations, has been announced by M. A. Angot (Ann. 

 de I'Institut de Physique du Globe, Paris, 1923, p. 288). 

 But if Dr. Bauer and I are not exactly at one on this 

 point, we are at least agreed that the influence of 



NO. 2810, VOL. I 12] 



sun-spots on the absolute values of the magnetic 

 elements is exceedingly small, if not zero. 



Coming now to the potential gradient of atmo- 

 spheric electricity. Dr. Bauer claims to have estab- 

 lished a substantial spot influence both on the ampli- 

 tude of the diurnal variation and on the mean value 

 for the year. In the Physical Society paper to which 

 he refers (Proc. Phys. Soc, London, vol. 35, p. 129), I 

 attempted to check the alleged sun-spot influence 

 both for the diurnal range and the absolute value by 

 means of formula (i). In the case of the absolute 

 value, R represented the mean value of the potential 

 gradient for the year. In addition to results from 

 the Ebro Observatory, on which Dr. Bauer had 

 mainly relied, I employed data from two periods of 

 years at Kew, determining a and b in all cases by 

 least squares. Except in one case the value found 

 for loobja was positive, but it was much below o-8, 

 and the values found for the correlation coefficients 

 were too small to warrant the conclusion that a true 

 sun-spot influence had been made out. 



In his recent letter Dr. Bauer does not impugn the 

 accuracy of my mathematical work. What he does 

 is to employ instead of (i) a formula of the type 



R=a'+6'S-f-c'T 



(2) 



where S is now the difference of the sun-spot number 

 from its mean value, and T the time in years counted 

 from the middle of the period. We may, I think, 

 treat it as a mathematical certainty that the observa- 

 tional results must be expressible exactly by a formula 

 of the type 



R-a'-6'S=/(T). 



What Dr. Bauer has found is that for one particular 

 period of years /(T) = c'T gives a good result at certain 

 stations, notably Ebro and Eskdalemuir, which he 

 considers good, and a less good result at other stations, 

 Potsdam and Kew, which he considers inferior. He 

 would no doubt get a still better result if he put 



f{T)=c'T +d'T:\ 



But is the goodness of fit in such a case any evidence 

 of the real existence of a sun-spot influence ? There 

 might, for example, be an excellent fit with b' =0. 



There may admittedly be special conditions in 

 which something is to be said for a formula of type (2). 

 As I showed some years ago, the absolute value of 

 potential gradient at Kew, and presumably else- 

 where, is affected by the visibility (purity) of the 

 atmosphere, potential falling, as the visibility rises. 

 If the purity of the atmosphere at a station improved 

 at a uniform rate, potential gradient would naturally 

 fall, and it might be a proper course to apply a cor- 

 rective term c'T, with c' negative as found by Dr. 

 Bauer at the Ebro, Eskdalemuir, and Kew. Again, if a 

 station went on applying an invariable factor for the 

 reduction to an infinite plane, while the factor was 

 really altering owing to continuous deterioration of 

 the insulation or other instrumental cause, a corrective 

 term c'T with c' negative might be justifiable if the 

 rate of deterioration was constant. 



The reasons assigned by Dr. Bauer, p. 203, for 

 considering Kew an inferior station are the large size 

 of Cg/Cj, the ratio of the amplitude of the 12-hour to 

 the 24-hour Fourier wave, and the high mean value 

 of potential. Now I can imagine another critic 

 holding — and with equal reason — that a low value of 

 c^lc^ and a low mean value of potential gradient are 

 both symptoms of inferiority either in the site or in 

 the apparatus. He might even suggest that the mean 

 values at the Ebro, 86i;/m in 192 1 and j6vlm in 1922, 

 are outstandingly low. 



If a high mean potential gradient is a sign of 



K 2 



