90 



H. W. Willson — Magnetic Field 



the positive sign indicating excess over the normal value which 

 is assumed to be that at the southwest corner of the building, 

 being nearly the mean value for rooms 13-16 and differing by 

 one fifth of one per cent from that in the tower sixty feet from 

 the ground. The season of the year made it impracticable to 

 make a satisfactory comparison with points out of doors. 



It is obvious that, as was inferred from the first observations 

 on the piers in rooms 13-16, over much more than nine tenths 

 of these four rooms the greatest variation from the value as- 

 sumed as normal did not exceed one half of one per cent. In 

 room 12, however, the variations were so great that a complete 

 survey was not attempted ; the difference between the two 

 piers for instance being four per cent. The disturbance in 

 room 14 is such as would be produced by the iron stove on the 

 second floor above the point Gr, if magnetized in the direction 

 of the dip, about 60°. The symmetry of the disturbance in 

 rooms 12 and 17 led to a closer examination of the steam pipes 

 against the east side of the eastern wall forming the boundary 

 of the non- magnetic wing of the building. Of these there are 

 three groups. At A, B and C, fig. 2 are pairs of pipes reaching 

 from the basement to the second floor, and at A is a third pipe 

 reaching from the basement to the third floor. 



Th« strength, of the free north magnetism at a point five 

 feet above the first floor was found to be such as to give at a 

 distance of 10 feet a field of .025 H for C and .04 H each for 

 A and B. To compute the effect of these pipes upon the 

 directive force over the whole ground to be covered, seemed at 

 first sight an undertaking of considerable magnitude, involving 

 the determination of the distribution of magnetism not only 

 throughout their whole extent, but also in that of the whole 

 system of iron steam pipes in the basement and of a large 

 radiator connected with A upon the third floor, and consisting 

 of five parallel horizontal pipes each fifty feet in length. As a 

 matter of curiosity, however, the following investigation was 

 made. 



Upon the platform of the elevator at E was placed a simple 

 magnetometer read with mirror and scale, the readings of 

 which as the elevator car was raised or lowered served directly 

 to determine the difference of the magnetic declination at 

 different points in the vertical line above E. These differences 

 referred to a point 10 ft. above the first floor were as follows : 



4 ft. 



above 



basement 



floor, 



3°-32 W 



9 " 



u 



a 



u 



3*00 W 



4 " 



a 



first 



a 



1 '15 W 



10 " 



a 



a 



a 



o-oo 



4 " 



a 



second 



u 



1-18 E. 



10 " 



u 



a 



a 



1-21 E. 



4 " 



u 



third 



a 



1-81 E. 



