S26 Mr Galbraith's Barometric Measurements of Heights. 



made at the top and bottom, which was not done here, though the 

 barometer was observed to stand steady during the interval. 



From the method of graduating the sliding scale of the sym- 

 piesometer adopted by the inventor, the divisions are not quite 

 theoretically exact. In fact, there are only such a number of 

 points found by calculation and experiment in general as may 

 be thought necessary, and the intermediate spaces are then di- 

 vided into equal parts nearly, as the deviation from perfect ac- 

 curacy would, in most cases, be insensible, or at least less 

 than the errors arising from other sources, which cannot 

 easily be avoided. Indeed, the correction of part of the error 

 of graduation is in some degree obviated by a sort of ten- 

 tative process ; and, consequently, as appears both by the for- 

 mer examples and the present, the errors of the practical con- 

 clusions generally fall within the usual errors of observation ; 

 and the results may, I think, be estimated to be equal in accu- 

 racy to those by the Englefield barometer, while the instru- 

 ment itself is much hghter, and considerably more portable. 



I have been inclined to think that it requires rather more at- 

 tention to operate accurately with the sympiesometer in certain 

 cases, especially in ravines, or water-courses thickly wooded. 

 In particular, I recollect that, in attempting to measure the 

 height of the romantic banks of the Esk, at the old mansion-house 

 of Hawthornden, the first observations appeared to be tolerably 

 coiTect, when both the upper and lower observations were made 

 near the house ; but as soon as I had gone down the river side 

 to a small field surrounded with wood, by some irregular influ- 

 ence, perhaps from strata of air of different densities, or tempe- 

 ratures more loaded with aqueous vapour in that confined situ- 

 ation, I soon found that it required the instrument to be pro- 

 tected, otherwise errors to a considerable amount would be pro- 

 duced, and that such a situation was very unfavourable for 

 these operations. As my observations were made in the company 

 of a friend from Liverpool, I had not sufficient leisure to exa- 

 mine the causes of this circumstance ; but concluded, rather 

 hastily perhaps, that it agreed partly with a remark of Professor 

 Babbage of Cambridge, " That when the lower observation is 

 made in a narrow or deep valley, situated at the foot of a moun- 

 tain range, the upper observation being made on an exposed 



