MEASUREMENTS 25 



set before twelve, it will be found that the point of the shadow 

 is approaching the circle from the inside. As the point 

 crosses the circle it should again be marked. The distance 

 between the two crossings on the circle should be bisected. 

 A line from the base of the rod through this point is pointing 

 north, and when the shadow of the point falls on this line 

 the following day it is twelve o'clock by the sun. The 

 error in getting the exact north point by either the short 

 shadow or equal arcs is due to the fact that a shadow is not 

 sharp and definite and it is impossible to determine correctly 

 the exact shadow of the point. The method of equal arcs 

 may be used approximately as a sun dial. After the shadow 

 corresponding to several of the hours has been located, these 

 shadows can be marked on the ground by means of small 

 pieces of wood having the hour marked on them. 



Compass. The universal method of determining north 

 is by the use of the magnetic compass. Numerous explana- 

 tions have been advanced for the action of the magnetic 

 needle and reasons for its pointing in a given direction, but 

 these explanations are not entirely satisfactory and will not 

 be considered here. The part of the compass which gives 

 direction is a slender bar of hard steel that has been magne- 

 tized and is free to rotate in a horizontal position. There 

 is a popular belief that the needle always points to the north 

 and is constant. This is not the actual condition. There 

 are only a few places on the earth's surface where the needle 

 points to the true north, and long observations have proved 

 that the needle changes somewhat in its direction with the 

 different hours of the day and with the different months of 

 the year, and that there is a slight continuous change from 

 year to year. Furthermore, it is very difficult to find two 

 compasses that will point exactly in the same direction at 

 the same time when set up under similar conditions. These 

 differences are all small, and, with the exception of the 

 one known as variation, they need not be taken into 

 consideration. 



