REDUCTION OF OBSERVATIONS OF HORIZONTAL INTENSITY. 41 



The observations are made by a counter, the rate of which 

 should be known, and it only then remains to make the corrections 

 for this rate of the counter. 



The two methods lead, of course, to the same results ; but the 

 table of successive reductions, spite of the differences due to errors 

 of observation, has the advantage of showing that the reduced length 

 of observations tends manifestly to increase with the time. In the 

 particular case the increase of duration was due to a rise in tempera- 

 ture of the vibrating needle. 



694. If the zero is fixed, or if it is merely displaced in conse- 

 quence of a periodic motion like that above referred to (684), we 

 may note the passages over any given division, since we always count 

 a whole number. This is not the case when the zero is displaced 

 without reference to any law, as is most frequently the case with 



Fig. 133- 



magnetised bars, owing to variations of the declination. If at the 

 moment the transit is noted, the position of equilibrium is not the 

 same as in the previous observation, an error is committed equal to 

 the time which the moving point takes to traverse the arc the zero 

 of which is displaced, and this error can only be avoided by 

 observing each time the passage on the division which corresponds 

 to the actual zero. With this object, a little before the observ?tion 

 the values of two successive elongations are called out to an 

 assistant, who, taking into account the damping, would immediately 

 deduce from them the division corresponding to zero, and the 

 reading would be made on this division. 



