DESCRIPTION OF THE TORSION GALVANOMETER. 75 



248. Let us take an example : a single pair of plates developed a current of electri- 

 city, which, when measured at the torsion balance, was found equal to 20 degrees ; on 

 subjecting this current to a secondary wire, 7 degrees passed it. Its tension might, 

 therefore, be represented by -3500. A second pair was now added in conformity to 

 the first, 31 parts passing ; but when subjected to the secondary wire, 18 were indicated. 

 The teusion had now become -5806 : in the same way, by adding three more pairs, the 

 tension rose to -6346. 



249. It must now be borne in mind, that the numerical determinations thus procured 

 are entirely conventional ; their absolute value depends upon the resistance of the 

 secondary wire, and they therefore only express the relative condition of different 

 currents. 



250. As a considerable advantage will be gained, and much repetition avoided, by 

 here indicating the mode adopted for procuring the following measures, I shall describe 

 at once some modifications and additions which are necessary in the torsion balance, 

 the instrument generally employed. 



251. The voltameter has of late come much into use in investigations of this sort, 

 but when compared with the torsion balance, the latter is much more speedy and cer- 

 tain in its indications, and should generally be preferred. In point of fact, the indica- 

 tions of the two instruments are entirely of a different character ; the magnetic needle 

 shows the quantity of electricity that is passing in each indivisible portion of time, the 

 voltameter the quantity that has passed at the end of a finite time. In the conditions 

 of the action of the one, time enters as an element, in the other it does not 



252. By applying a glass thread to the needle, the late Dr. Ritchie greatly improved 

 the accuracy and general utility of the galvanometer ; but even with that addition, 

 unless certain precautions are taken, the instrument will not work satisfactorily ; the 

 motions of the needle are too versatile, and the tremulous state of vibration into which 

 it may be thrown are insuperable barriers to accuracy of measurement. A cylindrical 

 trough filled with water is a perfect and admirable remedy for these difficulties. 



253. Another difficulty, which is very generally overlooked, is the excentric position 

 into which the thread is liable to be cast, when the upper micrometer has moved. The 

 construction of the instrument requires that the axis of motion of the upper micrometer, 

 the axis of the glass thread, the axis of the spindle carrying the needles, and the vane, 

 should be in the same vertical straight line, through whatever arc the micrometer may 

 have moved. Now it would be very difficult to accomplish this by any system of 

 adjustments. 



254. Whether the instrument is arranged with one or several needles, or whether it 

 has a coil or merely a single strap, the vertical distance from the coil or strap, when 

 the index is brought to zero, ought under no circumstances to vary. 



255. In a climate as hot as that in which the following experiments were made, one 

 of the most unpleasant deviations depends on the thread wrenching in the wax, which 

 is used to fasten it to the needles at one end, and to the micrometer at the other;- when 

 the wax softens, and the thread is moved through several degrees, it is not the free part 

 alone that undergoes torsion, but also that which is in the wax; hence arises an error as 



