THE BAROMETER, 



fore that the quantity of air composing such a column will be 

 one-third of the quantity composing a column extending from 

 the lower station to the summit of the atmosphere. 



If the atmosphere were uniformly dense, the barometer would 

 supply a most easy and simple means of determining its actual 

 height. 



In the example just given, the column of air between the two 

 stations would weigh one-third of the weight of a column extend- 

 ing from the lower station to the summit of the atmosphere ; 

 and, if the air were uniformly dense, it would follow, therefore, 

 that the entire height of the atmosphere would be just three 

 times the height of the upper above the lower station. But, 

 owing to the circumstances already explained, which produce a 

 gradual rarefaction of the air as the height increases, it follows that 

 the heights of columns of air are not proportional to their weights. 



If the only cause which produces a gradual rarefaction of the 

 air as we ascend in the atmosphere were that which has been 

 just stated, namely, the weight of the incumbent air, it would 

 not be difficult to find a rule by which a change of altitude might 

 be inferred from observing the change of pressure indicated by a 

 barometer. Such a rule has been determined, and is capable of 

 being expressed in the language of mathematics, although it be 

 not of a nature to be rendered intelligible in an elementary and 

 popular treatise. 



15. But there are other causes affecting the relation of the 

 change of pressure to the change of altitude. The density of any 

 stratum of air is not alone affected by the incumbent pressure of 

 the superior strata, but also by its own temperature. If any 

 cause increase this temperature, the stratum will become more 

 rarefied, and with a less density will support the same incumbent 

 pressure ; and if, on the contrary, any cause produce a fall of 

 temperature, it will require a greater density with the same pres- 

 sure. In the one case, therefore, a change of elevation which 

 would be necessary to produce a given change in the height of 

 the barometer would be greater than that computed on theoretical 

 principles ; and in the other case it would be less. The tempera- 

 ture therefore forms an essential condition in the calculation of 

 heights by the barometer. 



Formulae have been contrived, partly by theoretical principles, 

 and partly from observation, by which the difference of height of 

 two stations may be deduced from the observations simultaneously 

 made at them on the barometer and the thermometer. To apply 

 such a rule it is necessary to know, first, the latitude of the place 

 of observation ; secondly, the heights of the barometer and ther- 

 mometer at each of the two stations, besides some other physical 

 184 



