as determined hy Chronometers at Sea. 253 



gen, the longitude of which was determined by a great many ob- 

 servations of the distances of the sun and moon for several days 

 with Troughton's eight-inch sextants and reflecting circles. The 

 rates in the column entitled " Mean Rates at Sea," are deduced 

 by dividing the difference of the errors by the interval. 



The rates in the column entitled " Mean Rates on Shore," or 

 more properly what they would have had, are means between the 

 rates of chronometers on shore before leaving England, and those 

 obtained at Spitzbergen; and although a mean between the 

 rates of chronometers obtained at different times, may not accu- 

 rately be the mean rate they would have had during the interval 

 of those times, from the continued variation to which they are 

 subject; yet, upon comparing the two last columns together, of 

 the rates thus deduced, it will be perceived, that in all the chro- 

 nometers their gaining rates had either been increased, or their 

 losing ones diminished on ship-board, or, in other words, they 

 had all been accelerated. 



Nor is this acceleration peculiar to high latitudes ; it was ob- 

 served very soon after the chronometers were put on board in 

 the River, particularly in Nos. 3 and 8, which, upon arriving 

 at Shetland, were found to have gained instead of losing rates, 

 which they had in London. 



This acceleration was very soon perceptible in the chronome- 

 ters taken out by the Hon. Captain Phipps, made by Kendal and 

 Arnold. Mr. Lyons, who accompanied him, landed at Sheer- 

 ness Fort, and found the longitude by them to be SO' 0" E. 

 which is about 13' W. of the true longitude, as determined iu 

 the Trigonometrical Survey. 



The same occurrence took place last summer (1819). The 

 longitude of a place in one of the Orkney Islands, as determined 

 by three chronometers made by Arnold, two of them belonging to 

 myself, the other to Lieut. E. Home, R, N., who accompanied 

 me, was 6' 40" W. of the longitude determined by the difference 

 of Ai of stars E. and W. of }) . 



Again, in the trial of Mr. Harrison's time-keeper, in 1764, the 

 longitude of Barbadocs by the watch was 10' 45" more to the 

 westward than that determined by astronomical observations 

 made by the persons sent out for that purpose. 



Soon after this trial, the commissioners of longitude agreed 

 with Mr. Kendal, one of the \vatchinakers appointed by them to 

 receive Mr. Harrison's discoveries, to make another watch on 

 tlie same construction, which went considerably better than Mr. 

 Harrison's. Mr. Kendal's watch was sent out with Captain 

 Cook in his second voyage towards the south pole and round the 

 world, in the years i 772-^3-2 and 5, *< when the only fault 



found 



