DEPARTMENT OF TERRESTRIAL MAGNETISM. 313 



and concentricall3^ The construction necessitates that the successive layers 

 in each solenoid start from points 180 degrees apart. When each coil has two 

 or more layers it is probably best to wind all layers of one coil with the same 

 integral number of turns, and all layers of the other coil with this same number 

 of turns plus or minus one-half turn. It is then clearly possible to join suc- 

 cessive layers of the two solenoids systematically in series by very short 

 connectors at the ends, in such a way that the current goes alternately up a 

 layer of one solenoid and down a layer of the other, and that the small effects 

 of each pair of short connectors are practicallj" canceled near the center of 

 the field. Difficulties are encountered in making the windings precise close 

 to the ends, but slight irregularities there are of little consequence. 



In a large double solenoid constructed in the laboratory of the Department 

 of Terrestrial Magnetism the coils are wound on tubes of bakelite-dilecto, 

 which is known to have excellent magnetic and mechanical properties and to 

 insulate so well that any kind of current may be used. It can be worked 

 readily with a diamond tool. 



Note on electromagnetic induction and relative motion. S. J. Barnett. Phys. Rev., 

 p. 527 (Mar. 19, 1920). 



A further discussion of the results obtained in an investigation pubUshed 

 in the Physical Review, August 1918, p. 95. 



Further experiments on magnetization by rotation. By S. J. Barnett. Abstract of paper 

 read before Philosophical Society of Washington, Oct. 9, 1920. 



This paper is a report of progress in the first part of a general investigation 

 of the relations between magnetization and rotation designed to extend the 

 earlier work in this field, to obtain more precise results, and especially to find 

 out whether positive electricity is partially responsible for magnetism. 



In earlier papers it has been shown that, if the slowly moving electron 

 known from other experiments is alone involved in the Amperian vortices, 

 the rotation of a magnetic substance at an angular velocity of one revolution 

 per second is equivalent to placing it in a magnetic field of strength 7.1 X 10 "'' 

 gauss directed along the axis of rotation. Two series of experiments made in 

 1914 and 1915, by a method of electromagnetic induction, gave 3.6 and 3.1, 

 respectively, instead of 7.1, apparently indicating that negative electricity 

 was chiefly responsible, but that positive electricity also was involved. 

 Another series of experiments in 1916 and 1917 by a magnetometer method gave 

 the same sign as before, but gave numbers approximately 5 and 6 in place 

 of 7.1, thus indicating much less definitely an effect of positive electricity. 

 A few experiments made more than two years ago at the Ohio State Uni- 

 versity, where the other work was done, with copper substituted for the 

 magnetic substance, indicated that a part of the effect in the magnetometer 

 experiments was due to eddy currents, the effects of which appear to have 

 been completely eliminated in the work done by the method of electromag- 

 netic induction. This probably accounts for at least a part of the discrep- 

 ancy between the results obtained by the two methods. 



New experiments were performed under superior conditions in the non-mag- 

 netic experiment building of the Department of Terrestrial IVIagnetism by the 

 magnetometer method, with which alone the remainder of the paper deals; 

 considerable improvements have been made in the completeness with wliich 

 the earth's field is compensated, in the ehmination of mechanical and magnetic 

 disturbances, and in other ways. The polar position of Gauss, which was 

 earlier thought impracticable, has been substituted for the equatorial position, 

 as it makes less difficult the elimination of eddy-current effects and has other 



