68 MR BROUN ON THE BALANCE MAGNETOMETER, 
Observations of the time of vibration in a horizontal plane ranging through 
a period of three years, agree within 0.06, and this difference must be to a con- 
siderable extent due to alterations in the condition of the needle between the dif- 
ferent observations. 
4. The time of vibration in a vertical plane is in widely different circum- 
stances. 
These differences I shall proceed to point out. 
lst, The time of vibration in a vertical plane is found increased after the 
needle has been, by any means, vibrated through a large arc. 
The strongest evidences of this are contained in the following table; they 
were obtained either by iron having been brought accidentally near to the mag- 
net, or by the necessary removal of the box which covers it. In the latter case, 
the magnet was vibrated through large arcs by currents of air. 
One or two observations for the time of vibration are given for the periods 
immediately before and after the disturbance of the magnet. 
The last column contains the times of vibration corrected to 50° Fahr.; for 
reasons that will be shewn, it is only these that are strictly comparable. 
The observed time of vibration is generally the mean of two series, which 
rarely differ one-tenth of a second. 
Additional evidence of the above conclusionis furnished by Table II. 
