Hot and Thermal Springs, Sfc. 259 



crease does follow an arithmetical progression. Schmidt* has 

 arrived at the same conclusion from considerations of a differ- 

 ent nature. Observations at eighteen points in South America, 

 made exactly where Boussingault made the observations com- 

 municated below, seem also to indicate a diminishing decrease 

 of temperature.-|- But it is probable that this apparently di- 

 minishing decrease of temperature is due to another cause. 

 For, since observations by day give a more rapid, and by night 

 a slower decrease of temperature in the air, and consequently 

 the mean between the two is the expression of the real decrease, 

 we shall obtain a more accurate result the greater the differ- 

 ence of height between the points of observation is, where single 

 observations only are made, from which the decrease of tempe- 

 rature is to be calculated ; for, the daily variations of tempera- 

 ture at the lower point will have the less influence on the dif- 

 ference of height in which the temperature decreases 2°.25, the 

 greater the difference of height between the two points is. So 

 that, if observations be made at different heights at the time of 

 the daily maximum temperature, an apparently retarded decrease 

 of temperature will be found ; and, if at the time of the mini- 

 mum, it will be apparently accelerated. But by far the greater 

 number of observations at different heights have certainly been 

 made neai'er to the daily maximum than to the daily minimum 

 of temperature. We are, therefore, not justified by these ob- 

 servations, in concluding on a diminishing decrease of tempera- 

 ture from the surface of the sea to the greatest heights accessi- 

 ble to man, namely to the limits of perennial snow. Gay-Lus- 

 sac's observations during his aeronautical excursion are equally 

 insufficient to justify a contrary conclusion.§ 



" Gilbert's Annal. vol. IxiL p. 309. 



t Ibid. voL Ixii. p. 300. 



jM^moires d'Arcueil, iii. p. 580. Semanario del Nuevo regno de Granada, 

 Yol. L p. 273. ii. p. 93-341. See Ideler, p. 95. 



§ Gilbert's Annal., xxiv., p. 28. and xxxi. p. 32. Gay-Lussac himself re- 

 marks, that his thermometers did not keep pace with the temperature. And 

 how could the thermometers be expected to keep pace with the temperatore 

 of the air, considerinj^ the velocity with which an aeronaut makes his ascent ? 

 See also the remarks of Humboldt in the same, p. 34. and of Gilbert, p. 32. 



