on the RaUs of Chronometers, 



41 



meter was in the magnetic meridian, the rate having changed from 

 — 3".5, the daily aberration in its detached state, to — 3".0. On 

 turning the time-keeper, however, a quadrant, that it might occupy 

 the situation of E, Fig. 2, the rate became — 3".3 ; and on giving 

 to it the position of E, Fig. 3, a decrement of — 1".9 resulted, the 

 rate amounting in this situation to — 5".2 ; but on moving the 

 chronometer through another quadrant, that the balance might 

 occupy the position of E, Fig. 4, an increment of + 2".3 was pro- 

 duced, the rate being — 2".9. When detached, the rate of the 

 time-keeper was found to be — 2".2. Hence it appears that the 

 application of this time-keeper to the iron mass, occasioned both an 

 acceleration and a retardation of rate ; the greatest increment cor- 

 responding to the position of E, Fig. 4, and the greatest decrement, 

 to that of E, Fig. 3. This completed the first course of expe- 

 riments. 



Chronometer E in a Vertical Plane passing through the Centre of the Iron 

 Shell, at right Angles to the Magnetic Meridian, and to the West 

 of it. 



Distance of the 

 Vertical Axis of 

 the Chronometerj 

 from Centre of 

 Shell. 



Elevation of the 

 Centre of the 



Face, jibove the 



Centre of the 



Shell. 



Detached 

 Rate. 



Rate in 

 Fig. 1. 



Rate in 

 Fig. 2. 



Rale in 

 Fig. 3. 



Rate in 

 Fig. 4. 



Detached 

 Rale. 



11.5 Inches. 



Inches. 



-3".5 



-3''.0 



-3".3 



-5".2 



-2" .9 



-2".2 



A second course of experiments was now undertaken with the 

 same chronometers placed in other situations round the Iron Shell, 

 and also with the addition of the time-keeper F. 



For this purpose, the chronometer A, instead of being placed to 

 the north of the ball, as in the former course of experiments, was 

 now placed to the South, as in A, Fig. 7, with the centre of its 

 balance in the magnetic meridian of the attracting mass. The 

 detached rate of the time-keeper was -t-12".8, it having retained 

 the impulse communicated to it in the former set of experiments; 

 and, the result first presented by it in the new position was-1- 12".4, 

 the rate having undergone a trifling declension ; but, on moving 

 the chronometer into the situation denoted by A, Fig. 8, the line 

 joining the centres of the balance and chronometer being in this 



