1823.] the Barometer to the Measurement of Heights, 17,5 



Aug. 30, 1805. — On the summit of the Pic de Midi, between 

 10, a.m. and 1, p.m. the thermometer varied from 14° to 19°. 

 This was owing to an irregular wind. In moments of calm, it 

 stood at 16° or 17°, this being partly caused by heat of the sur- 

 face : it fell when the wind brought against it masses of colder 

 aiK; and rose to 18° or 19° when the current continued. In this 

 complication of influences, the only way was to take a meap 

 between the extremes ; for it would have been equally absurd tp 

 keep to either 14°, 16°, or 19°, for the sole reason that the ther- 

 mometer had accidentally indicated one of these degrees at the 

 precise moment when we noted the height of the barometer. 



Sept. 11. — On the Pic de Bergons, there was a brisk wind 

 subsiding at intervals. The temperature of calm was above 14°, 

 but the ground reflected much heat. It is seldom that thjs is 

 not the case with the observed temperature of a calm air. This 

 temperature, however, increased as the wind yose; it was a 

 different atmosphere which the wind drove before it, hea.ted by 

 passing over the neighbouring plains. It rose to 15v5°, but 

 when the wind continued constant, sunk to about 13°, and there 

 became stationary, which was certainly the true temperature of 

 the air under the prevailing modification, the higher tempers^- 

 tures were transient and accidental. 



; Aug. 10, 1802.— On the summit of Mont Perdu, a thermome- 

 ter, placed on the snow, sunk to — 2-5°, owing to the rapid 

 evaporation. Another suspended at the height of about 4-i. feet 

 partook of the same influence, and never rose above 4° or 5°. 

 Another at the same time suspended above a rock free from snow 

 stood at 12*5°, and one placed on the rock at 22*9°. The con- 

 tinuance of a wind (which at first brought hot air from the 

 plains, gradually reduced them all to about 7*5°, nearly the mean; 

 this was the true temperature of the air. 



It is superfluous to mention a multitude of other cases differ- 

 ing but little from these ; such as a passing shower which causes 

 a-variation in the thermometer the moment it touches it : a local 

 fog, which occasions, in the particular atmosphere of the instru- 

 ments, a cooling, in which the rest of the stratum of air does not 

 partake : the influence of the sun which raises the thermome- 

 ters ; the intervention of clouds which makes them sink ; all the 

 variations which originate in reverberations or absorptions of 

 heat, or in currents of air which are accidental and of limited 

 influence ; every thing which conspires to alter the general tem- 

 perature of unfavourable situations, such as deep valleys, and 

 even eminences above which greater heights immediately rise. 

 I have said enough to awaken and direct the atteiitio;n of those 

 who wish, and know how, to be exact. 



Such are the considerations which must guide \is in reference 

 to the thermometer for the air. We have just seen that the 

 thermometer for the barometer is the object of considerations 

 very different; for it, the temperature of the air is pnly a jn^tter 



