THE MAGNETIC OBSERVATORY AT MADISON. 459 



knife-edges, there is an arrangement for lifting the instrument off of 

 its bearings, when not in use. 



The balance-magnetometer requires a delicate temperature adjust- 

 ment. For this purpose there is attached to the side of the magnet a 

 small tube containing mercury. Such is the position of the tube that 

 the shifting of the centre of gravity of the magnetometer, due to the 

 expansion or contraction of the mercury, shall just balance the ten- 

 dency of the north-seeking pole of the magnet to rise or fall with the 

 temperature. Adjusting the tube to its proper position occupied Mr. 

 Suess for five days. 



The variations of these several instruments are recorded by pho- 

 tography, each instrument, with its recording apparatus, constituting 

 a magnetograph. A cylinder, turned by clock-work, carries the sen- 

 sitive paper upon which the record is to be made. A single cylinder, 

 with its sensitive paper, suffices for both the declinometer and the bi- 

 filar magnetometer, the cylinder turning between the two instruments 

 and receiving the two records at its" opposite ends. A second and 

 vertical cylinder is required for the balance magnetometer. The 

 record of all the instruments is made in the same way. The light 

 from a German student-lamp, after passing through a narrow slit, is 

 received upon a concave mirror carried by the magnet. The mirror 

 throws a thread-like image of the slit upon two cylindrical lenses 

 fixed in the case of the recording instrument. By these lenses the line 

 of light is shortened to a dot, to be received by the sensitive paper. 



Were the spot of light stationary, a straight line would be traced 

 upon the sensitive paper, since, by the revolution of the cylinder, the 

 paper would be carried directly forward from in under the light. 

 But, by the movement of the magnet, the image of the slit is made to 

 travel back and forth along the lenses and a more or less eccentric 

 trace left upon the sensitive paper. 



In order that the trace may not go beyond the limits of the paper, 

 the magnet must be kept from swinging through more than a small 

 arc. This is effected, in the bi-filar magnetometer, by the pull of the 

 suspension skein acting against the magnetic force. In fact, owing 

 to the too great size of the glass pulley, the magnet does not swing 

 quite freely enough. 



In the other two instruments a special arrangement is adopted. 

 Surrounding the magnet and forming a closed circuit, is a rectangle 

 of four flat copper bars. Any movement of the magnet gives rise to 

 a current in the circuit, which tends to pull the magnet back again. 

 Thus, if the north-seeking pole of the magnet in the declinometer be 

 deflected toward the east, a current will be generated, running from 

 south to north along the upper bar of the rectangle, and back along 

 the lower bar. The current, in turn, acts upon the magnet, checking 

 it in its swing toward the east, so that the paper can receive the 

 entire trace. 



