334 CARNEGIE INSTITUTION OF WASHINGTON. 



sideratioiis, therefore, the apparent discrepancy between his conclusions and 

 those of E. H. Nichols is greatly reduced by this circumstance. 



In conclusion, reference is made to certain experiments by E. H. Nichols 

 to test the electric field in the immediate vicinity of the cap of the electrom- 

 eter. It is pointed out that the results so obtained, if driven to their logical 

 conclusion, would indicate that even in the capped form of instrument there 

 is a considerable error, a result inconsistent with the conclusions which E. H. 

 Nichols draws from his other experiments and recorded data. 



On the magnetic and electric fields which spontaneously arise in the vicinity of rotating 

 conducting spheres. W. F. G. Swann. 



There are several ways in which small magnetic and electric fields may be 

 conceived as arising in the vicinity of a rotating body. Thus, centrifugal 

 force will make the free electrons move away from the axis of rotation, until 

 the electrostatic forces brought about as a result of this phenomenon restore 

 equilibrium. The rotation of a body, a sphere for example, may accordingly 

 be expected to result in the generation of magnetic and electric fields from 

 this cause alone. Again, in so far as the electrons are pulled by gravity, they 

 will tend to move towards the center of the sphere and will do so until the 

 electrostatic forces, etc., resulting from this action restore equilibrium. This 

 effect also provides a reason for the existence of a magnetic field. Analogous 

 remarks result from a consideration of the potential-gradient which arises in 

 a sphere, from causes of the nature of the Thomson effect when the sphere is 

 hot at the center and cold at the surface. 



A very little consideration of the order of magnitude of effects such as 

 those cited above will show, without exact mathematical analysis, that they 

 can play no appreciable part in the explanation of the Earth's magnetism, 

 and in the paper the results of the above effects are calculated simply as a 

 matter of physical interest. 



The three effects, centrifugal force, gravity, and temperature gradient, in 

 the sphere (Thomson effect) are considered separately, for it is easy to see 

 that when all act together the solution is practically that obtained by adding 

 the individual solutions. 



The distribution of electricity throughout the sphere and on its surface 

 becomes determined by the following conditions: (1) the resulting electrical 

 forces within the sphere must balance the mechanical force arising from the 

 phenomena under consideration; (2) Poisson's equation must be satisfied 

 throughout the sphere; (3) the total volume charge plus the surface charge 

 must be zero. When the charge distribution has been obtained, the electric 

 and magnetic potentials may readily be calculated. 



Tables 7 and 8 summarize some of the main conclusions for a sphere of 

 unit-specific inductive capacity. Table 7 gives the values of the horizontal 

 and vertical intensities He and Zp at the equator and pole, respectively, two 

 cases being cited, that of a sphere of the density of copper and of radius 

 20 cm., rotating 100 times per second, and that of a sphere of the size and 

 mean density of the Earth, rotating with the Earth's angular velocity. H is 

 measured in the direction of increasing colatitude, and Z is measured upwards 

 from the surface of the sphere, the direction of rotation being supposed to 

 take place from west to east. 



In the calculation of the contribution arising from the Thomson effect 

 certain assumptions have to be made as to the theory of this effect. For 

 the details of these the complete paper must be consulted. The fields calcu- 

 lated are for the case where the temperature gradient is confined to a thin 

 shell near the surface of the sphere, and the total fall in temperature amounts 

 to 5000° C. 



