APRIL 4, 1921 proceedings: philosophical society 163 



stituted 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 electromagnetic induction. This probably accounts for at least a part 

 of the discrepancy between the results obtained by the two methods. 



In new experiments performed under superior conditions in the non-mag- 

 netic experiment building of the Department of Terrestrial Magnetism, by 

 the magnetometer method, considerable improvements have been made in 

 the completeness with which the earth's field is compensated, in the elimina- 

 tion 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 elimina- 

 tion of eddy current effects and has other advantages. Eddy current effects 

 have been more thoroughly studied by rotating copper and otherwise. The 

 work is still in progress and other tests remain to be applied. 



The most extensive observations have been made on a rod of Norway iron 

 (for which observational curves were shown) . Many observations have been 

 made on a rod of cobalt, and some on rods of cold-rolled steel. All the rods 

 gave values about one-half of 7.1 instead of 7.1, or even less, as in the ex- 

 periments on iron by the method of electromagnetic induction, thus again 

 indicating an effect of positive electricity or else indicating that negative 

 electricity alone is involved, but has, for the motions responsible for mag- 

 netism, a smaller value of m/e than that determined in known experiments. 

 Satisfactory experiments on nickel have not yet been made. 



It is interesting to observe that, while all the sources of trouble have not 

 yet been removed, the method is so sensitive that, in the later part of the 

 night, when extraneous magnetic disturbances are least, the effect can be 

 measured at even very small speeds. Curves between the scale readings for 

 right and left handed rotations and the time were exhibited showing clearly 

 the effect for cobalt at the speed 3/4 revolution per second. 



The paper closed with a reference to recent experiments on the converse 

 effect (rotation by magnetization). These experiments, on the theory which 

 has been adopted by the investigators, but which involves an uncertain as- 

 sumption with reference to the seat of the reaction to the resultant electron 

 momentum produced on magnetization, also appear to indicate an effect of 

 positive electricity or else the participation of negative electricity with a 

 value of m/e different from that hitherto known. 



The paper was illustrated by lantern slides, and was discussed by Mr. 

 Humphreys. 



S. J. Mauchly: Results of atmospheric electric observations made during 

 solar eclipse of May zg, igig, and summary of similar observations. 



The only atmospheric-electric observations made within the belt of totality 

 during the total solar eclipse of May 29, 1919, appear to have been those by 

 observers of the Department of Terrestrial Magnetism at Sobral, Brazil. 

 These observations were in charge of Mr. Andrew Thomson, of the Depart- 

 ment of Terrestrial Magnetism, who made measurements of the electrical 

 conductivity of the air due to positive and negative ions, respectively. Mr. 

 Thomson was assisted by Mr. Antonio Lima, a native of Brazil, who had been 

 educated in the United States, and who made potential-gradient observations 

 under Mr. Thomson's direction. 



