228 ilir R. Threlfall, On an automatic mechanical [Mar. 2, 



of this is that the mercury will not cling to the points while they 

 are rising from it; and that therefore the beat will be much 

 sharper than in the case of platinum where an increase of resistance 

 must precede the break of contact. For many purposes this is an 

 advantage as it diminishes the intensity of the self induction 

 spark. Since however in this apparatus the efficiency of the 

 regulator depends on the defmiteness of the point of contact, any- 

 thing tending to diminish this must be more or less injurious. 

 Since this was written I have discovered that the aluminium burns 

 away so rapidly that, balancing disadvantages, platinum is on the 

 whole to be preferred. 



The action of the regulator will now be easily understood. A 

 battery continually maintains a current round the electro-magnet : 

 and through a high resistance inserted in circuit between the 

 battery and the magnet. 



When the potential rises above the desired point, the attracted 

 disc by means of its contact makes short-circuits the high resist- 

 ance and allows an increased current to flow round the electro- 

 magnet. This will tend to put an increased brake on the copper 

 disc and the rotating parts of the apparatus will instantly slow 

 down. I say instantly, because if india-rubber bands be used to 

 transmit the motion between the various pullies we shall only 

 have to take into account the momentum of the copper disc and 

 replenisher and not that of the motor : at least during short inter- 

 vals of time. Now the copper disc maybe made pretty thin and the 

 other parts are not very large, so that a practically instantaneous 

 change of velocity may be obtained: it is only a question of having 

 a large enough battery—say two Groves', a fairly high resistance 

 to be short-circuited, and long india-rubber bands. 



For a given change of potential the attracted disc will be more 

 sensitive the nearer it is to the fixed disc. The method of using 

 the regulating arrangement is therefore as follows. The apparatus 

 to be maintained at the constant high potential is charged up to 

 this potential as indicated say by a portable electrometer ; and by 

 means of the Weber suspension and the adjustable support of the 

 mercury cups the attracted plate is lowered so as to make contact 

 when the potential is slightly increased. The limit of the sensi- 

 tiveness of the apparatus depends on the critical distance of the 

 plates, and this depends on the law of stretching of the spring. As 

 a general rule in a long spring the extension will not be far from 

 proportional to the stretching force. Sir William Thomson is in 

 the habit of using the torsion of a wire in some of his instruments 

 instead of a spring. In this case however we wish to make the in- 

 strument as sensitive as possible always, and not as in electrometers 

 so as to have a constant relative sensitiveness : we must therefore 

 have an adjustment for working at different distances between the 



