632 



APPENDIX. 



[No, IV. 



It has already been remarked, in the narrative of occurrences at Fort Enterprise, 

 that we found the errors of the sextant materially changed, that the glasses had lost 

 their parallelism, and the graduated arches of the sextants were warped, owing to 

 the contraction of the brass during the intense frosts of the winter ; and it is to the 

 error produced by this cause alone, that I attribute the very great difference between 

 the results of our observations on each side of the moon, in the spring of 1821. This 

 difference exceeded two degrees with each of the sextants. The observations of 

 0E. I being 112° 3' 49" W., and of those of O W. 1) , 114° 12' 35" W. From this 

 circumstance, it is obvious that the observations then obtained, of O W. j> , could 

 not, with any regard to accuracy, be used with those procured on the opposite side 

 in the autumn of 1820. 



Advantage was taken of every opportunity to ascertain the rates of the chronome- 

 ters, during the descent of the Copper-Mine River, and in the subsequent journey 

 along the coast ; and whenever we were assured, by good observations, of their hav- 

 ing altered, a correction was then made. At other times, their rate of going was 

 usually seen by comparison with 1733. 



On the return of the Expedition to Wilberforce Falls, the chronometers were found 

 to have altered their rates so much, as to give the longitude of the first cascade in 

 Hood's River forty miles westward of what they had done twenty-four days before. 

 Other rates having been procured during our stay near these Falls, a proportional 

 allowance was made, to be applied to the results of all the observations by the chro- 

 nometers, between Goulburn's Islands and Point Turnagain. The book, containing 

 the longitudes thus corrected, having been lost by the accident of September 14th, 

 as detailed in the Narrative, we could not use them in constructing the map ; but as 

 the latitudes observed between these parts agreed so nearly with what the reckoning 

 gave, as deduced from the bearings of the different points and estimated distances, I 

 thought it best to lay down that part of the coast from dead reckoning corrected by 

 latitudes, rather than by the uncorrected longitudes, as we perceived the shape of the 

 land would have been much distorted by using them. Point Turnagain is, therefore, 

 placed in 109° 25' W., instead of 110° 5' 15" W., its longitude by the chronometers. 



