26 SECTIONAL ADDRESSES. 



intensity of the field observed was proportional to the angular velocity. 

 The direction and general shape of the magnetic field of the earth could 

 be accounted for by this gyromagnetic theory, but the intensity of 

 magnetisation produced is far too small. The estimated value is about 

 10~'' times that of the earth. However, the laboratory conditions are so 

 far removed from those of the interior of the earth that the restoring forces 

 must be very dissimilar in magnitude, and that these forces are a direct 

 factor was shown by Barnett's experiments. At present, however, the 

 effect appears to be so very small that other causes must be sought. 



Possible Mobifioation of Laws of Electrodynamics. 



The difficulties confronting such theories as an electrically charged 

 earth and the smallness of the gyromagnetic effect, have led to suggestions 

 that the field may be due to some departure from the commonly accepted 

 laws of electrodynamics. 



In 1894 J. J. Thomson pointed out that if atoms exerted slightly 

 different attractions on positive and negative electricity, then a large 

 rotating body could produce a magnetic field, and in such case the 

 intensity would be proportional to cor^ so that no laboratory experiments 

 could confirm or refute the theory. 



H. A. Wilson has considered a case assuming that the electric and 

 magnetic effects do not balance in electrically neutral matter. On such 

 an assumption the gravitational unit of matter (about 4,000 grams) which 

 attracts an equal mass 1 centimetre away with the force of 1 dyne might 

 be expected to produce a magnetic field of the same. order of magnitude 

 as the electrostatic unity of electricity. Such an assumption leads to a 

 correct ratio for the magnetic fields of the earth and sun, but an experiment 

 made by Wilson showed that if moving matter produces a magnetic field 

 like that due to a moving charge, then the mass equivalent of one electro- 

 static unit of charge is not less than 20,000 kilogrammes, and that matter 

 having a velocity of pure translation has no appreciable field. It follows 

 that if the earth's magnetic field is to be explained by some modification 

 of the laws of electrodynamics, the modification must be such as to make 

 rotation and not translation the effective motion. 



Swann, who has put forward a theory based on a slight modification 

 of the laws of electrodynamics, points out that the ratio of the magnetic 

 fields for the earth and sun would be obtained also for an expression of 

 the form DcoV^ since the ratio of the values of oi^r* differs inappreciably 

 from that of cor. Swann so modifies the equations as to provide for the 

 correct value of that part of the earth's field symmetrical about the axis 

 of rotation, and the right value for the ratio of the magnetic field of the 

 earth to that of the sun is also obtained ; in addition, there is a slow death 

 of positive electricity amounting to a disappearance of | per cent, of the 

 earth's charge in 10^° years, the corresponding surplus of negative 

 electricity thus freed, after building up the necessary electric field, passing 

 off continually as the atmospheric electric current by conduction through 

 the atmosphere. According to this theory, spheres of such size that they 

 may be used in laboratory experiments should give effects which are just 

 measurable, and Swann and Longacre have made experiments with a 

 copper sphere 10 centimetres in radius rotating at 200 revolutions per 



