NA TURE 



)37 



THURSDAY, AUGUST 9, 1894. 



LETTERS TO THE EDITOR. 



[TAe Editor Joes not hold kimselj responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake 

 to return^ or to correspond ivith the writers of^ rejected 

 manuscripts intended for this or any other part of Naturr. 

 No notice is taken of anonymous communications. ] 



Wilde's Theory of the Secular- Variation of Terrestrial 

 Magnetism. 



In a recent communication to ihe Royal Society, contained 



in the Proceedings (or March, pp. 210-217, Mr. Henry 



Wilde, F. R.S., believes that he has cited additional facts in 



support of his theory of the secular variation of terrestrial 



magnetism. Wilde first proposed his theory ' to the Society, 



June 19, 1890. He had it separately printed, in addition with 



other papers, in three languages, and the pamphlet was 



scattered far and wide. It was issued in the form of a reprint 



from the Transactions (though it never appeared in the Trans- 



.'ions), and contained such surprisingly good agreements 



;ween theory and observations that several eminent men 



science, without examining the pamphlet very care- 



ily, and believing that it bore the stamp of the Royal 



- ciety, were misled into thinking that AVilde had really 



hieved some remarkable results. 



Having been engaged in the United States Coast and Geodetic 



irvey as magnetic computer for five years, and being aware 



the difficulty in reconciling the secular variation of the 



magnetic declination for various places on the earth's surface 



with each other, under the assumption of one period for the 



vvhole earth, and that at the same time the orbit described 



iring that period by the north end of a free magnetic needle 



uld be a single closed curve, I undertook to show the fallacy 



h of Wilde's theory and of the supposed agreement with 



servalions. The result of mv investigation was read before 



! Philosophical .Society of Washington, February 27, 1892, 



1 was published in (\\t American Journal of Science, vol. 



.iii., June 1892. The conclusions of that paper were : — 



(1) Wilde's theory is physically impossible. 



(2) Wilde's mechanism (magnelarium) does not agree in 

 principle with his theory as outlined. 



(3) The results from the magnelarium do not represent the 

 facts of terrestrial magnetism. 



(4) Wilde's theory is but a modification of Halley's. 



In proving (3), 1 showed that Wilde, in the case of the London 

 results, only got out of his mechanism what he had actually put 

 in. In the case of Cape of Good Hope and St. Helen.!, such ap- 

 parently good results were obtained because the law of the secular 

 variation for these two stations, within the period of observation 

 and within the large probable error of the magnelarium results, 

 is not very dissimilar from that for London. So to get a good 

 correspondence for Cape of Good Hope and St. Helena, it was 

 only necessary to turn the crank a few times more. I, moreover, 

 showed from facts slated in the very Bulletin of the Coast 

 Survey, which Wilde has cited in his late communication, as giving 

 additional proof of his theory, that in order to satisfy peifectly 

 the three stations, London, Cape of Good Hope, and St. 

 Helena, it would seem that his inner electrodynamic sphere 

 would have to make one complete differential revolution in 960 

 years (as he has deduced) for the first two stations, but that 

 for St. Helena it would have to move more rapidly and perform 

 a revolution in 7S0 years. I therefore proposed to Wilde to 

 try to turn out with his magnelarium the secular variation of 

 the magnetic declination (or stations which exhibit a vastly 

 different law from that of London, viz. stations in the United 

 States. It is remarkable that although he evidently possesses 

 records of secular-variation stations in the United States, he has 

 thus far made no published attempt to reproduce them with his 



* " That the principal phenomena of terrcslrial magnetism, and of the 

 secular changes, coulc'. be explained on the as!>umption of an electrodynamic 

 stibbt.lncc (presumably liquid or gaseous) rotating within the crust of the 

 earth in the plane of the ecliptic, and a little slower than the diurnal rota- 

 jion." A complete differential revolution of this electrodynamic subsLince 

 is performed in <,6o years, which period, according to Wilde, constitutes the 

 secular-variation period. 



magnetarium. The point insisted upon was that he shoald 

 turn out the secular variation at the various stations conjointly, 

 not separately. Is it not strange that he has made no 

 attempt in the four years since he first constructed his magne- 

 tarium, to give us the secular variation for other portions of the 

 earth than simply for the three stations given in the first com- 

 munication, and now in the last one for a station which lies not 

 far from one of these, and so has a similar law of variation? 

 Why does not Wilde read off from his magnetarium the declina- 

 tions at other stations, s..^. Paris, New York City, St. John's, 

 Newfoundland, Peking, &c. , when he turns out the secular 

 variation at London ? 



But I shall now ignore my own paper, with its conclusions, 

 and exhibit the fallacy of VVilde's theory on the basis of the 

 results given in his last communication. 



(1) The annual change of the dip at London for Ihe last fifty 

 years has been steadily decreasing, indicating the approach of 

 a minimum. A formula established by myself on the basis of 

 the London dip observations, 1576-1890, which represents the 

 observations many times better than the magnetarium results, 

 predicts this epoch of minimum dip about the middle of the 

 next century.^ This phase of minimum dip is advancing 

 steadily from the East, and has already set in in Western 

 Russia. The decrease of annual change of dip is borne out by 

 numerous stations in Europe, all indicating the near approach 

 of a minimum dip phase. Now Wilde s magnetarium results 

 give a steady increase of annual change of dip, in consequence 

 of which his minimum dip for London does not occur until 

 after the year 2200. 



(2) On p. 216 of the last communication, Wilde says : " On 

 working the inclination (at St. Helena) backwards on the 

 magnetarium chronologically, it will be seen from the Table VI. 

 that about the year 1747 the dip changed the sign from south 

 to north. As no observations were made on the dip at St. 

 Helena previous to the year 1S25, there is no record of this 

 interesting fact, nor has it hitherto been deduced from theory." 

 Unfortunately (or Wilde, we have record of observations for this 

 period, which, instead of giving a dip of o" in 1747, give a dip 

 of 9° to lo° (south end dipping) I To show how poorly his 

 results agree with observation for periods where he does not 

 possess the records himself, I give the following table : — 



NO. 1293, VOL. 50] 



Wilde only had the observations 1825-1880, aad it will be 

 seen that for this short interval the correspondence between 

 theory and observation is very good. 



But before this interval the divergence is so great as to com- 

 pletely vitiate the Wilde magnetarium results. Likewise with 

 1890, the divergence becomes marked again. The first observa- 

 tion for 1890 was made at Jamestown, the second at Longwood, 

 St. Helena. The first station agrees in locality more closely 

 wilh preceding observations than the second. - 



Wilde concludes with a reproduction of the observations of 

 declination and inclination at Ascension Island for the epoch 

 1834-1890. During this interval the correspondence is very 

 good, but he does not tell us whether he obtained the 

 results separately or conjointly with London, Cape of Good 



1 Science, vol. xx. No. 506. 



= From a paper, soon to be published, on the scciilar.vanatioii curves as 

 described, in the course of time, by a ficcly suspended magnetic needle at 

 various portions of the earth's surface, it would appear as though the 

 inclination at St. Hclcaa never becomes zero as Wilde's theory would 

 demand. 



