306 Captain Sabine's comparison of barometrical and 
Whence, it may be concluded that in these experiments 
the two methods of measurement have been found to corre- 
spond within the limits, which, under circumstances, may be 
attributed to accidental, and indeed to unavoidable errors. 
It may be expected that whilst on the spot, some means 
may have occurred to me of explaining the very great dif- 
ference which is recorded by Captain Phipps to have been 
found in the height of an hill in Amsterdam island, mea- 
sured geometrically by himself, and barometrically by Dr. 
Irving, in which the latter measurement exceeded the for- 
mer by 8 5 feet, in between fifteen and sixteen hundred 
feet ; the observations on that occasion were conducted with 
so much apparent care, and the difference was so great, 
as to have caused more or less doubt to have prevailed 
from that time to the present, of the equal applicability of 
the barometric formula in the higher, as in the middle and 
lower latitudes. I do not however feel better able than 
before, to conjecture in which of the operations the mistake 
originated, for such I do not doubt there must have been. 
I was desirous to have repeated the measurement of the 
hill itself, but time did not permit; judging, however, by 
the eye, in comparison with other hills on the coast, and 
especially with the one which has been the subject of this 
communication, and which was not more than a few miles 
distant, the lowest, that is the geometric result, appeared 
the most likely to have been correct; nevertheless a mistake 
of nearly a tenth of an inch, is a great amount with a baro- 
meter which was registered to thousandths ; especially as 
the tendency of probable errors is on the other side. For 
