456 THE POPULAR SCIENCE MONTHLY. 



In order to really profit by his visit, he should have gained, he- 

 forehand, some idea of what is to be observed and the manner of ob- 

 serving. This information I shall try to give the readers of this 

 article; and then proceed with a description of the observatory and 

 its equipments. 



The force called terrestrial magnetism is subject to variations 

 both in intensity and in direction. There are three ways in which a 

 varying force of this kind may be measured : 



1. It may be resolved into components acting along three axes, 

 and the intensity of these components measured. I am not aware 

 that this method has ever been applied to the measurement of terres- 

 trial magnetism ; probably, because one of the components would be 

 so small as not without great difficulty to be directly measured. 



2. We may measure its intensity along some fixed axis, and its 

 angular variation of direction from that axis, in each of two planes 

 intersecting the same. This method is frequently employed. The 

 fixed axis taken is the intersection of the plane of the magnetic merid- 

 ian and the plane of the horizon, and the angular variations from the 

 axis are measured in these planes, the variation in the horizontal plane 

 being called the " declination," and that in the plane of the magnetic 

 meridian the " dip." 



3. We may measure the intensity of its components along two 

 axes and its angular variation in direction from the plane of those two 

 axes. This last is the method in use at the observatory. The axes 

 assumed are horizontal and vertical, and their plane is that of the 

 magnetic meridian. Angular variations from this plane may be meas- 

 ured in any plane at right angles to it, as the plane of the horizon, 

 and are, therefore, changes of declination. 



The instruments used for making the measurements are the decli- 

 nometer, the bi-filar magnetometer, and the balance magnetometer. 



The declinometer consists, essentially, of a bar magnet so sus- 

 pended as to turn freely in the horizontal plane. Changes in the 

 position assumed by the bar show changes in declination. 



The magnet of the bi-filar magnetometer likewise turns in the ver- 

 tical plane ; but, while the magnet of the declinometer is free to assume 

 any position in that plane, the magnet of this instrument is pulled by a 

 constant force into a position at right angles to the magnetic meridian. 



The magnet of the balance-magnetometer, like that of the last two 

 instruments, is in a position at right angles to the magnetic meridian ; 

 but, unlike either of the other two, it turns in the vertical plane. 



The only effect of the horizontal force is to press the magnet 

 against its bearings, and were the magnet suspended at its centre of 

 gravity, the north-seeking pole would point directly downward in 

 obedience to the vertical force. In reality, the magnet is so sus- 

 pended as to assume a position approximately horizontal. The force 

 of gravity remaining constant, the magnet w T ill not change its posi- 



