in th> ton Physical laboratory, 89 



9tigate the variations of the magnetic field in different parts 

 of the room. For this purpose I made use of the very con- 

 venient local variometer devised by Prof. F. Kohlrausch.* 



A very few observations showed that the Horizontal In- 

 tensity II, acting upon a free magnet was quite variable as 

 will be seen by reference to fig. 1 which shows a system of lino 



upon each of which the directive force is the same. The gen* 

 eral form of the disturbance of the earth's uniform field pointed 



to a source of free north magnetism in the neighborhood of the 

 \. E. corner of the room, and the stack of iron soil pipe shown 

 at S, fig. 2 was examined and found to be inadequate to pro- 

 duce so great an effect as was observed. This is not surprising 

 >ince, though of large size, it is made up of short lengths with 

 leaded joints and acts as a series of separate magnets end to end. 

 The neighboring iron steam pipes at C showed a far greater 

 amount of free magnetism, enough indeed to account for the 

 greater part of the disturbance noted. 



Feeling desirous to know more of the extent of this effect I 

 made a series of observations on each of the piers p, p y on the 

 first floor establishing the fact that in the rooms 13, 14, 15 and 

 16 the greatest deviation of the Horizontal Intensity from its 

 mean value was one half of one per cent. The differences were 

 so large, however, that I determined to make a more extended 

 set of observations, including about twenty-five points on the 

 third floor and several in the basement. The result showed that 

 besides the effect of the stove temporarily used in the room on 

 the second floor over room Xo. 14, a considerable disturbance was 

 perceptible even in those rooms farthest removed from all iron. 



Further observations on the third floor were discontinued on 

 account of the establishment there before the series was com- 

 plete of the long iron beam which was at this time put in place 

 for the work of Prof. Trowbridge on the spectra of metals.f 

 The frequent changes of position of this beam gave rise to a 

 variable disturbance which was of some magnitude on the 

 third floor though fortunately very small below. 



I then determined to make a careful survey of the first floor, 

 the only part of the building in fact where measures are likely 

 to be made which require great uniformity of field over large 

 spaces. The results of this campaign are given in fig. 1 ; ob- 

 servations were made at the points indicated by dots, the plane 

 of the instrument being five feet above the floor and two feet 

 and three inches above the tops of the piers, this being assnmed 

 a sufficient distance to prevent any disturbance from the brick- 

 work of the piers. The lines connect points at which the 

 directive force is the same ; at points on adjacent lines the 

 force differs by one quarter of one percent, of its whole amount, 



* Wied. Ann., xix, p. 130, 1883. f Proc. Am. Acad., 1888, p. 1. 



