Oscillations of the Horizontal Needle. 9 



lar to the plane in which the vibrations are intended to take 

 place : These are to receive the fine index wires, which are easily 

 retained in the slits after a slight pressure, or tap of a hammer. 

 It should now be nicely poised on a fine suspension-silk at- 

 tached to a small loop, inserted through the vertical central 

 hole, so as to remain exactly horizontal. The horizontal position 

 may be readily ascertained at any time by a very simple method. 

 Let the bar, Fig. 5, be held near the surface of a reflecting fluid, 

 such as mercury, or, what answers equally well, near the surface 

 of water, in which a little indigo is dissolved, to give it colour ; 

 an image of the bar will then be perceived in the fluid. When 

 the fines a b, c d, formed by the bar and its image, are parallel, 

 then we may be assured that perfect horizontality is obtained. 

 The parallelism of the lines is easily detected by the eye. 



15. After the horizontal position has been thus accurately de- 

 termined, two points are to be impressed on each side of the cen- 

 tre, at an equal distance from the centre and ends. Two small 

 sliders of platinum, a b, Fig. 4, of equal weight, are then closely 

 fitted, one on each arm. These sliders have small central holes 

 drilled in them, and bear a very small proportion to the whole 

 mass. They are at first placed so as to bring the central holes*im- 

 mediately over the points impressed on the bar, and in this state 

 the horizontality of the whole is again observed as before. 



16. The bar being now properly tempered throughout, is fi- 

 nally polished by filing, and, after being again examined as to 

 position when suspended, is rendered magnetic. It is then ex- 

 posed for a short time to the temperature of boiling water, and 

 after this to a temperature as low as zero of Fahrenheit's scale. 

 We have then a magnet calculated, as far as possible, to retain 

 for any ordinary temperature an invariable magnetic state. We 

 should now place the bar thus prepared under a piece of strained 

 paper, and having sifted some iron-dust over it, by means of an 

 extremely fine sieve, examine carefully if the centre of the mag- 

 netic curves falls about the centre of suspension. Each face of 



VOL. XIII. PART I. B 



