1820.] a Barometer for measuring Heights. 173 



temperature was proper. This place is easily found on the scale 

 of inches ; so that by screwing the slide till the one end come to 

 the proper place on this scale, the other end will of course take 

 the right position on the scale for determinate distances. By 

 adjusting the slide at the foot and top of the mountain, and 

 counting the divisions between its positions, the approximate 

 altitude of the mountain is obtained. As the finding of the true 

 altitude does not depend on the barometer but the approximate 

 only, it is not necessary for me to say any thing about that, as 

 the method is given in every work in which mensuration by the 

 barometer is treated of. 



I shall conclude this letter by making a few remarks concern- 

 ing the most favourable times for measuring with the barometer. 

 Calm weather is indispensably necessary for this operation, 

 because air in motion has not the same perpendicular pressure as 

 when at rest, but its gravity diminishes as its velocity increases. 

 Hence, if a stratum of air of the altitude of any mountain moves 

 with such a velocity as to cause its perpendicular pressure 

 almost to vanish, then the column of mercury, being supported in 

 no sensible degree by that stratum, but by the atmosphere above 

 it, will suffer no apparent change in bemg carried from the foot 

 to the top of the mountain. Clear unclouded weather is also 

 necessary, for clouds by their shadows and different temperature 

 disturb the regular progression of the atmosphere. Mr. Greato- 

 rex, in an account which he gives in the London Philosophical 

 Transactions for the year 1818, of a measurement he made of 

 the height of Skiddaw, ascribes the erroneous results he 

 obtained by the barometer wholly to the effect of clouds. 



Beside the above, there is still another circumstance worth 

 attending to, which is the height at which the barometer stands ; 

 when it stands high and continues steady, then is the most 

 favourable time for measuring with it. When it is low, it is 

 unfavourable ; for the fall may be produced by some approaching 

 wind, giving a buoyancy to the atmosphere before it, or by 

 winds in the higher regions, which circumstances would disturb 

 the geometrical progression of the atmosphere, and hence render 

 a correct measurement impossible. 



If all these favourable circumstances which I have mentioned 

 are possessed, and a good barometer used, the results, at the 

 very least, are equally to be depended on as those obtained by 

 trigonometry. Yours, &c. 



Jambs Allan. 



