456 eepoet — 1878. 



to the vibrator diverts from a regulating cell of the battery a minute stream of 

 exciting fluid ; but if the vibration ceases this small stream passes on to the battery, 

 and so increases its power until the vibrator is again set in motion and diverts the 

 exciting fluid. Thus the current is continually restored to the right point, and the 

 impulse continues to be the same. 



(b) If the electro-motive force varies, the current can be maintained constant 

 by causing the resistance to vary in the same ratio. This can be effected by a very 

 light pendulum, having the same period as the main pendulum. Any variation in 

 the current wili change the arc of the light pendulum, before the heavy pendulum 

 is sensibly affected, and this variation of arc can be made to vary the resistance in 

 a solid or fluid portion of the current. 



Two vibrators may be employed, with counterpoises slightly different. If the 

 current is in the normal state, the light vibrator acts, but not the heavy one. If 

 the current be too strong, both vibrators act ; if too weak, neither. • The action of 

 both may be made to introduce resistance or diminish electro-motive force. The 

 inaction of both may diminish resistance or increase electro-motive force, so as to 

 restore the current to its normal condition. 



(c) The variations of the arc can be confined within very narrow limits by the 

 following contrivance. A permanent magnet attached to a pendulum, by its 

 attraction for two small magnets attached to a balance or unstable equilibrium, 

 causes the latter to oscillate so as to make contact when the pendulum is at the 

 extremity, and break contact when at the middle of the arc ; the current so opened 

 passes through an electro-magnet placed below the pendulum, which terminates in 

 an armature. The current being made 1 slightly stronger than is necessary for a 

 given arc, its oscillation increases until the permanent magnet attached to it brings 

 into action a balance similar to the former, but placed somewhat further from the 

 centre of oscillation. The action of the balance opens a second path ' for the 

 current, which is thus shunted from the electro-magnet, diminishes the impulse 

 given by the latter, and so reduces the arc again. 



It is, therefore, concluded that if by any of the modes above mentioned, or any 

 other system of rheostal, the current be maintained constant, the impulses will be 

 equal, and the arc must be constant ; or, again, if the variation of the arc be kept 

 within very narrow limits, the time of oscillation will remain constant, and the 

 pendulum will remain isochronic. 



13. The Temperature of the Earth Within. By William Morris. 



The state of this question amounts to this : that we know, nor, by the methods 

 hitherto pursued, can ascertain absolutely nothing of the heat of the earth within.* 



The following is a method to ascertain the temperature of the earth within, 

 and, furthermore, to establish a permanent system of observation on the tem- 

 perature and ejectric state : — 



Place one of two suitable uncompensated chronometers (provisionally so-called) 

 at the bottom of a (special) bore. This chronometer is to be actuated from surface 

 apparatus electrically, so as in its inaccessible, buried position, to be maintained 

 permanently in motion, and in turn to control electrically a clock at the surface 

 which will indicate to an observer the time (that is, revolutions in given time), 

 kept below ; being susceptible to temperature, the movements of this chronometer 

 will be inter-dependent with the temperature below. The second of the pair, 

 placed on the observatory wall, with a standard thermometer hanging by it, will 

 indicate time (revolutions in given time) inter-dependent with the known tem- 

 perature of the observatory. Knowing the temperature of the observatory, and 

 comparing the inter-dependent time registered by the observatory chronometer 

 with that of the chronometer at the bore-bottom as represented by the clock 

 electrically controlled therefrom, we can tell readily whether the temperature of 

 the observatory is higher or lower than that affecting the buried chronometer. 



The electric cables can be used to lower the chronometer into the bore. 



* See 'Nature,' No. 447, vol.xviii. May 23, 1878. 



