Mr. Fox on Irregularities in the Magnetic Needle. 311 

 120°, it made 80 vib. in 64-1". Arc at first 90=,&ended withl6°-5 

 104 ... do. 641-5 ... do. ... 18 



85 ... do. 64]'5 ... do. ... 18 



52 ... do. 644-25 ... do. ... 34 



A warm blanket was then thrown over the slate box in 

 order to communicate warmth to its top and sides, the sup- 

 porting granite being at 52°. 



At the temperature of 60°, it made 80 vibrations in 645"-5. 

 Arc at first 90°, and ended with 38°. 



After the needle had been held a short time in the hand, 

 in order to warm it, 



It made 80 vibrations in 642"-2. Arc at first 92°, and 

 ended with 22°. 



In these experiments it appears that the number of vibra- 

 tions was mostly rather increased, and the arcs diminished, 

 when the bottom of the box, or the needle only, had its tem- 

 perature augmented. When merely the sides of the box were 

 warmed, the result was different ; and when the heat was ap- 

 plied as uniformly as possible to all parts of the box, the irre- 

 gularities of the vibrations were less considerable than when 

 the bottom of it only was heated ; and merely touching the 

 latter with the hand tor a short time frequently produced con- 

 siderable derangement in the action of the needle. The effects, 

 however, were often so different when all circumstances ap- 

 peared to be alike, that it seemed desirable to investigate the 

 subject further. 



For this] purpose I suspended a very slender needle in a 

 copper case, and placed the latter on supports in a basin, into 

 which I poured warm water till it reached the bottom of the 

 case. An extraordinary agitation of the needle then took place, 

 its vibrations sometimes amounting to ten or fifteen degrees 

 on each side of the meridian, occasionally stopping, and then 

 starting again, and I'requently shifting its centi'e of vibration 

 backwards and forwards on either side of zero; and this mo- 

 tion continued more or less, till the water* had approximated 

 to the temperature of the room. At first I fancied the agita- 

 tion of the needle might be owing to electricity, but subse- 

 quent observations have induced me to attribute it to currents 

 of air, rapidly rising and descending in the box containing it. 

 These effects were produced by any heated substances put 

 under the needle, and at the distance of several inches, or even 

 a foot, when the heated body was large; but when it was held 

 above the box containing the needle, the influence was com- 

 paratively inconsiderable. In the course of these experiments 

 the needle was suspended in close boxes of metal, slate, and 



