22 PRESIDENTIAL ADDRESS. — SECTION A. 



elements did not vary from place to place in any regular pre- 

 dictable manner. The hope that irregularities would disappear 

 as the observations became more exact has had to be given up, 

 and at the present day we may be quite certain — to quote Bauer : 



— "When you see perfectly regular or smoothly tlowing Imes you may rest 

 assured that they have been either smoothed out or that they depend upon 

 but very few data. Instead of the irregularities being the abnormal features, 

 they are the normal ones, and regularities are, in fact, the abnormal 

 features." 



The results of Thorpe and Riicker's* work in the British 

 Isles, and of Carlheim Gyllenskold's t in Sweden, first 

 brought before magneticians the extremely irregular character 

 of isomagnetic lines reproducing the values of magnetic ele- 

 ments at a given epoch. One of the first effects of the clear 

 perception of this was the introduction of suitable names for 

 the different kinds of lines. At the present day it is customary 

 to speak of true isomagnetics, in which all the observations — 

 after suitable reduction to the epoch — are taken into considera- 

 tion, and terrestrial isomagnetics derived from the former, 

 either by calculation or by a graphical method ; these are to a 

 great extent ideal, and depend on the method of production. 

 The maps which are now thrown on the screen show the in- 

 crease in complexity due to a more intensive survey; the first 

 gives the isogonics for France (epoch 1885), from observations 

 at 80 stations ; the second gives the isogonics for the same 

 country (1896) from observations at 617 stations; the third map 

 shows the true and the terrestrial isogonics for Great Britain 

 (epoch 1886). 



After a survey, with the stations as numerous as in 

 Moureaux's second survey of France, or Riicker and Thorpe's 

 surveys of the British Isles, the magnetic state of the region 

 under consideration can be further studied. One of the best 

 known methods of doing this is by the method of districts, 

 first used by Dr. van Rijckevorsel, J and fully developed by 

 Riicker and Thorpe. v^ The two latter observers divided the 

 British Isles into nine districts, and obtained for each of these 

 a mean station by taking the means of the latitudes and the 

 longitudes of all the stations in a given district. The value of 

 any magnetic element at such a station was obtained in a 

 similar manner. In this way nine mean stations were deter- 

 mined, each with its own mean values for the different magnetic 

 elements. General formulae, expressing the value of an 

 element as a function of the latitude and the longitude, were 

 then constructed, reproducing the values of that element at the 

 mean stations; from such formulae the terrestrial isomagnetics 

 were derived. Once a suitable formula of the above nature has 



♦Philosophical Transactions, Series A, Vol. 181, p. 53, Survey of the 

 British Isles for the epoch Jan. i, 1886. Riicker and Thorpe. London, 189:. 



t Memoire sur le magnetisme terrestre dans la Suede meridionale par C. 

 Gyllenskold Kongl. Svenska Vetenskaps-Akademiens Handlingar. Bandet 

 27, N : 07. Stockholm, 1895. 



I Report on a Magnetic Survey of the Indian Archipelago, by Dr. van 

 Riickevorsel. Part I., p. 17. J. Muller, Amsterdam, 1879. 



§ Phil. Trans. Loc. cit. 



