the Construction of Galvanometers, 6fc. 239 



upper end of the spindle, and tui'ns round on it as on an axis: it may be 

 secured in any position by means of a minute nut which screws it against a 

 shoulder on the spindle. Yet the total weight is trifling. One of my compound 

 needles, thus constructed, weighs but 2;^ grains, including the nut; its bars 

 are distant from each other 4- of an inch ; each is 2^ inches in length, and 

 .^ inch in width ; their sides are parallel. The best material for the bars is 

 the mainspring of a small Geneva watch ; it is very thin, and retains magnetism 

 a long time. When saturated with magnetism, it no doubt dissipates a portion 

 in a few days ; but it retains about one-third for any length of time, and this is 

 sufficient for ordinary purposes. It is obvious that the bars must be filed per- 

 fectly straight, and placed parallel to each other. 



The following is the process which I employ for magnetizing the compound 

 needle when great sensibility is not required. I caused a horse-shoe magnet of 

 considerable power to be made, the limbs of which, from one extremity of the two 

 polar faces to the other, are of such extent and thickness that the compound 

 needle intended to be magnetized can rest its whole length on the faces. The 

 limbs of the horse-shoe should approach each other within half an inch. In 

 this way the compound needle will rest one edge of each bar on the terminal 

 faces of the magnet, being placed exactly between the two limbs, and in the 

 same situation that its keeper generally occupies. The compound needle will 

 become saturated with magnetism in four hours. Before it is removed, the 

 keeper of the horse-shoe should be put on laterally, close to the extremities, to 

 facilitate the removal. 



The needle is now capable of two different applications. If it be intended 

 to measure powerful deflective forces, it is ready for that purpose, so far as its 

 state of magnetism is concerned. But as the two north poles are together, the 

 needle has acquired a dip, and it is necessary to balance this exactly by the 

 counterpoise sliders already mentioned. The needle, when properly placed in 

 the galvanometer, being not astatic, it will give short, quick oscillations, will 

 require no small voltaic energy to deflect it to 90°, and will return to zero witli 

 great precision and promptitude. It is now in its least sensible form. 



Should greater sensibility be required, we have only to turn the upper bar 

 round on its axis until it be precisely reversed, and to remove the counterpoise 

 towards the centre. To reverse the bars with precision is no easy matter, as 

 will be seen hereafter. 



