Nlpher — Local Magnetic Storms. 63 



violent. In Plate XIX the general direction of the wind 

 was at right angles to the needle. On the next day the 

 wind had shifted about 90°, and its direction was parallel 

 to that of the needle. The compensating adjustments of 

 the needle were the same in both cases, and the wind was 

 no more violent on the second day than on the first. 

 Plate XX. 



The vertical lines represent the amplitude of oscilla- 

 tion in scale divisons. The lower end represents the 

 scale reading to the left of the zero, and the upper end 

 of each line represents the scale reading to the right. 

 The observations of Plate XIX covered an interval 

 of 1 h. 37 m. 38 s., during which time 290 to and fro 

 oscillations occurred. The scale divisions were in milli- 

 meters, each having an angular value of 3.4 minutes of 

 arc. The middle line represents the mean of the extreme 

 readings of the scale. In Plate XX one hundred con- 

 secutive oscillations are represented. 



The needle was made of tool steel having a diameter 

 of l^ inch and a length of 3.5 inches. It was mounted 

 in a wire clutch, and with it was placed a brass rod of 

 about the same dimensions. This arrangement was 

 adopted in order to tone down some of the irregularities 

 shown in the vibrations of a highly magnetized needle 

 made of a single small wire of steel, such as knitting 

 wire. 



The needle with small moment of inertia follows the 

 variations in the earth's field more closely, but summa- 

 tion effects are occasionally obtained with the needle 

 of larger moment, and the amplitude of vibration is 

 thereby greatly increased. 



These results are in all cases obtained when the earth's 

 field is partly compensated, so that the changes are a 

 large fraction of the resultant directive force. 



It is of course understood that these observations do 

 not indicate any change in the intensity of the earth's 

 magnetic field, due to linear transfer of ionized air as 



