176 Measurement of Heights by the Barometer, [Sept. 



of secondary importance. It is always well placed wherever the 

 barometer is placed. Its variations are of no consequence, pro- 

 vided the temperature of the barometer undergoes the same 

 changes ; but this accordance is the thing of which we must be 

 careful to assure ourselves, and this is the dilRcult point. It is 

 very seldom that the two instruments, joined together as they 

 usually are, will preserve this accordance when the temperature 

 undergoes great and frequent variations. We may diminish the 

 sensibility of the thermometer ; but in thus preventing its indi- 

 cations from outstripping the rapidity of the changes which the 

 temperature of the barometer undergoes, we must take care not 

 to approach the point where, on the other hand, it may expe- 

 rience a retardation ; and in all cases there is a wide field open 

 to doubt and conjecture. 



We should commence our operations by disposing the instru- 

 ments in a convenient manner, and should then allow them time 

 to lose the heat which they have contracted in carrying, and to 

 acquire, each in its proper manner, the local temperature. This 

 time is considerable for barometers cased in wood, which become 

 unequally heated in the hands, or on the backs, of those who 

 carry them. The heat thus acquired never distributes itself uni- 

 formly, and is dissipated with extreme slowness. Often an hour 

 does not suffice to bring the barometer to an agreement with the 

 thennometer, and with itself. More than a quarter of an hour 

 is not necessary for barometers mounted in copper, and this is 

 one reason for giving them the preference. 



The time thus occupied, however, is not lost : we can make 

 trial of the position: it may not always be suitable for the baro- 

 meter : if it be too much exposed to wind, we can seek for a 

 shelter. At the summit of a mountain, a strong wind has other 

 inconveniences besides that of agitating the instrument : it rises 

 on the acclivity against which it directs itself, and forms an 

 .ascending current, which bears up the column of air, and 

 depresses that of the mercury. In such a situation the thermo- 

 meter should be left; but the barometer should be removed 

 from this part of the acchvity, and if the summit has not suffi- 

 cient extent, we should leave it, and seek for a calm under the 

 shelter of the opposite side, only being careful to allow for the 

 quantity of our descent : then we shall await the moment of 

 observation in considering the changes of the thermometers. 

 We shall remark attentively how the one is afiected in the free 

 air, and the other in the mounting in which it is inclosed ; whe- 

 ther they undergo considerable and frequent variations, or 

 whether they reduce themselves by slow degrees to a stationary 

 point. In this last case, we shall have no occasion to doubt the 

 success of the operation : in the former, we shall examine the 

 causes of the variations : we shall take an exact account of 

 the nature and influence of every accidental circumstance. 

 We shall repeat this examination after the observation in order to 



