600 ON THE RESULTS OF THE MAGNETIC DECLINATION. 



alterations required in the apparatus or in the building, and 

 other accidental hindrances. The remaining twenty-nine ob- 

 servations were made, but were subsequently rejected as un- 

 certain, on discovering that the free motion of the bar had been 

 impeded by spider threads. Accidents of this nature were 

 noticed in the Resultate for 1836. When a pecuharity in the 

 reading leads to a suspicion of such an occurrence, it is easily 

 tested by putting the bar in vibration, when the rapid decrease of 

 the arc, and the unusually short time during which the bar con- 

 tinues to oscillate, afford incUcations of the existence of a dis- 

 turbing cause. Should the thread be broken, on the bar being 

 set in motion, its previous influence may still be inferred, should 

 the direction of the bar, as concluded from the elongations, 

 differ much from its position before it was put in motion. It is 

 desirable that the freedom of the bar should be frequently 

 proved in this manner. 



