some Points on the West Coast of Scotland. 303 



greatly facilitate the reading, especially at night, in observations 

 upon the stars. Such an instrument, when properly used, is 

 capable of giving good results either with regard to latitude or 

 time. 



I had also a small reflecting box-circle made by Messrs Adie, 

 which I had the pleasure of laying before this Society some 

 time ago, and which, on several occasions, I found very ser- 

 viceable, both on account of its accuracy and portability. In- 

 deed, when the arcs are read alternately on both sides of zero 

 from three verniers, I think it may be depended upon, both for 

 latitude and time, to a very considerable degree of accuracy, 

 especially when the repetitions are continued for a considerable 

 time. I should say that, by circum-meridian observations with 

 an artificial horizon, near noon or the meridian, for about ten 

 minutes before and ten minutes after transit, the results, when 

 properly reduced, could hardly, under favourable circumstances, 

 differ more than 20" or 30" from the truth, and frequently 

 much less, though such a degree of accuracy could not gene- 

 rally be expected. With regard to time, I generally found it 

 to give results not more than two or three seconds from the 

 truth, though it would require more observations than I have 

 yet had an opportunity of making, to test its accuracy com- 

 pletely. 



Of course, Kater's circle being pretty large and well divided, 

 its results will seldom differ more than 10" of arc from the 

 truth, when used with care in favourable circumstances and 

 good weather. 



The chronometer performed remarkably well, and what was 

 not a little singular, the rate continued sensibly the same 

 tliroughout the whole time of my absence, or at least the rate 

 was the same when I returned to Edinburgh, as when I left it ; 

 and I deduced almost the same rate from an interval of about 

 ten days at one place, where I remained stationary for that 

 time. 



At Edinburgh, on the 3d of August 1837, at noon, the chronometer was fast 



of mean time, . . . , . . oh C" 4s.5 



Daily rate hsittff, . . . . , . 1 .28 



From these data the following table was formed, for con- 

 veniently making the necessary comparisons. 



x2 



