718 KANSAS CITY REVIEW OF SCIENCE. 
cylinder of steel, bored through from end to end. One end of this cylinder is 
stopped by a glass plate on which a fine scale is etched. In the other end is put 
a lens, the principal focus of which is at the scale. This magnet is hung in a stir- 
rup attached to the fiber, in such a way that the scale is viewed through the lens, 
by means of the telescope. Hence, when the telescope is in focus for a star it 
will also be in focus for the scale of the magnet. 
The observations are made as follows: The verniers of the horizontal scale 
of the instrument are set on 0°, and the telescope is turned on some distant ob- 
ject, as a church spire. This object is taken as the starting point for angles. 
The lower plate of the transit is then clamped, and the upper one loosened, so 
that when the telescope is turned, the scale reading of the vernier changes, show- 
ing the number of degrees swept over. The telescope is set on the pole-star 
when it is furthest east or west in its path around the pole. It is then only neces- 
sary to turn west or east a small angle, (which is given in the astronomical tables) 
and the instrument will be pointed due north. The vernier reading of due north 
being thus known, we know the number of degrees between the church spire and 
north. . Hence, on any successive day, we could set the telescope on the spire, 
and turning this known number of degrees, the instrument would be pointed 
north. The telescope is then directed upon the magnet scale, and its reading 
taken. (Call this ‘‘A.”) After this the magnet is turned upside down, so 
that the scale is seen inverted. (Call this reading ‘‘B.””) The scale reading is 
generally different from the former. If the telescope is turned to the point on 
the magnet scale, midway between these scale readings ‘‘ A” and ‘‘B,” it is 
found that a reversal of the magnet does not change the reading. ‘The telescope 
is then directed upon the magnetic axis, and lies in the magnetic meridian, being 
pointed to magnetic north. The angle between this direction and true north, 
read on the verniers of the horizontal circle, gives the desired variation. It is 
however well known that the position of the magnetic needle varies during the 
day, moving over an angle which is seldom less than four or five minutes, and 
which often amounts to a quarter of a degree or more. Hence, the mean posi- 
tion of the magnetic axis is determined for each day. 
In order to see the results of such determinations it is customary to record 
the results reached at each station, at the proper places on a map of the region. 
Having done this, itis easy to drawn full degree lines through points having the 
same variation, just as we might draw an zsothermal line through points having the 
same temperature. ‘These lines in Missouri were found to be much more irregu- 
lar than had been supposed, and it was easily seen that there was a marked rela- 
tion between the contour of the surface and the position of the needle. It 
appears that the needle tends to set at right angles to river valleys, and that this 
tendency is inappreciable when the valleys are in general east and west or north 
and south, and that the tendency is greatest when the valley makes an angle of 
45° with these directions. This points to a disturbance in the earth-current sheet, 
due to unequal conducting power in the moist valleys and on the dry hills. The 
e 
