1823.] the Barometer to the Measurement of Heights* 263 



ponding observations should be made at the same times, and in 

 the same number. The observers will thus see whether their 

 instruments have proceeded in concert ; whether their changes 

 have been correspondent ; whether their variations have taken 

 place in the same directions. If they should be of opposite 

 kinds, we shall suspect that the local influences have taken the 

 place of the variations of the atmosphere, and we shall suppress 

 the observations which reciprocally condemn each other. 



Such are my methods of proceeding. They have often 

 brought my measurements to a degree of precision which leaves 

 nothing further to wish. I recommend the same care, the same 

 precautions, to those who wish to try the merits of the formula, 

 and especially to those who may wish to correct it. 



All this is, I allow, minute and difficult, and this is not, per- 

 haps, the idea we usually form of the nature of barometric 

 measurements. We probably wish that there should be nothing 

 but what is easy in the use of instruments which we employ so 

 commonly ; yet what method of measurement is there which 

 has not its uncertainties, its unfavourable times, and even 

 greater difficulties? On the side of the barometer there is 

 always the advantage of simplicity of apparatus, quickness of 

 operation, facility of calculation, the most varied and extensive 

 applications, and a much less dependance on circumstances 

 which put obstacles in the way of using other instruments. 



I will now reduce into a brief summary the requisite conditions 

 for the measurement of heights. 



1. To employ instruments which correspond; are well con- 

 structed ; verified with care ; and rigorously compared. 



2. To choose stations as good as the nature of the places will 

 admit. 



3. To allow as little horizontal distance between the two 

 observers as possible ; but subordinately to the suitableness of 

 the stations. It may be several leagues without being too 

 great, if the difference of level be considerable, and if there be 

 not between the two stations any eminence which rises above 

 both. The proximity of the stations, on the contrary, will cause 

 more inconvenience than advantage, if the lower barometer is 

 badly situated. 



4. To make observations always simultaneous, and exclu- 

 sively at noon, or between the hours of eleven and one. 



5. To choose in general a time when the air is calm rather 

 than when agitated ; but not to fear wind when it is gentle and 

 regular. It then renovates the local mass of air, and reduces the 

 thermometers to the temperature of the atmosphere. 



6. Not to fear a cloudy sky, except when it threatens stormy 

 weather. The suppression of the solar radiation is favourable to 

 the observations, especially if they are made in places freely 

 exposed to the air, and if the instruments have no shelter. 



7. To avoid rain, storms, and violent winds ; and to be dis- 

 trustful in uncertain weather, when approaching changes are 



