48 Mr W. Galbrailh's Barometric Observations. 



As the weather was unfavourable, and the barometer change, 

 able, with only one observer, who took the observations, first, 

 at the bottom in the morning, then at the top about mid-day, 

 and again at the bottom in the evening, — of which the mean of 

 the first and last was reckoned the true height of the mercurial 

 column at the bottom, there might have been some error arising 

 from this cause. As the barometer continued to rise somewhat 

 gradually, the error from this source must likely be small. At 

 all events, whatever error attends the use of the one instrument 

 likewise affects the other, since they accord so well when used 

 together with equal care. 



I have likewise calculated the height of Ben Nevis more ri- 

 gorously by employing logarithms, and using the dew-points, 

 besides other refinements, which would increase the height to 

 about 4430 feet. As this exceeds all the heights I have ever 

 met with attributed to it by at least 50 feet, I cannot say what 

 confidence is to be placed in it, more especially, as I have been 

 informed the Ordnance Surveyors make it only about 4360 

 feet, or 70 feet less. From the great distances of their stations, 

 a very slight error in the angle of elevation, arising from the 

 variable nature of terrestrial refraction, will produce a consider- 

 able error in their results in feet. Indeed, according to Mr 13. 

 Bevan's paper in the Philosophical Transactions for 1823, 

 Part I., there are errors in the altitudes of some of the stations 

 in England, of from 50 to 100 feet, in heights of between 700 

 and 900 feet ! In this case I cannot say what confidence may 

 be placed in that of Ben Nevis, though, in our measurement of 

 Benlomond, the correspondence was as close as could be desired. 



From the present measurements, too, it appears that the 

 sympiesometer is an instrument which, when in good order, 

 may be confidently trusted as giving results, when carefully 

 used, very near the truth. The formula now investigated, ap- 

 pears to give results rather too small by about one-hundredth 

 of the whole, and this is the reason why those by the sympieso- 

 meter seem to be, on comparison, too great. If one^hundredth 

 of the height by the formula be added to itself, the final result 

 would agree very closely with the logarithmic process. 

 54. South Bridge, Edinburgh. 



