chronometers by the proximity of masses of iron. 383 
twelve inches from the vertical through the centre of the dial, 
and its centre about the same depth below the plane of the 
balance, and the rate thus obtained will be a very close ap- 
proximation to the ship rate of the instrument, provided care 
be taken, when it is removed on board, to keep it out of 
the immediate action of any partial mass of iron. The plate 
for this purpose should be a double one, such as I have de- 
scribed in my “ Essay on Magnetic Attraction/’ and if it 
weigh about ^lbs. it will be sufficient to prevent any partial 
action. 
It should be observed, that the plate is meant as a substi- 
tute for the iron forward ; and therefore the chronometer, 
when on board, should be placed in the same direction in 
reference to the ship’s head, as it had with respect to the iron 
plate when its rate was determined ; that is, if the 1 1 o’clock 
mark of the dial be turned towards the iron plate on shore, 
then must the same be turned towards the ship’s head when 
taken on board. 
Experiments on the detached parts of a chronometer. 
As some of the results of the preceding experiments were 
not precisely what I had anticipated, nor quite consistent with 
the ideas I had formed of the nature of the action between 
the iron and the balance, I was desirous of making some 
experiments on the detached chronometrical parts, in order, 
if possible, to trace the irregularity to its source. 
Having mentioned my wish on this subject to Mr. Frod- 
sham, he very cordially and earnestly entered into my views, 
desirous, not of avoiding, but of meeting openly every diffi- 
culty which presented itself in the construction of such a 
