HEIGHT IN THE ATMOSPHERE, AT DIFFERENT SEASONS IN THE YEAR. 493 



cate a most rapid increase of the Difference of Temperature between February and 

 March, and a most rapid decline in November, the maximum being about May. 



That the decrement of temperature with height is most rapid in summer, 

 and least in winter, has been long known ; * but I am not aware of any attempt 

 to account for the law of its variation at different seasons. The following consi- 

 derations will probably be found satisfactory. 



If we examine the annual curves of mean temperature at Colinton and Bo- 

 nally, projected in the lower part of Plate XX., we shall find that they differ in 

 three respects, whilst there is a remarkable coincidence in their general features. 

 (1.) The entire Bonally curve stands lower on the paper than the Colinton curve, 

 because the mean temperature of any and every part of the year is lower. (2.) 

 It is a flatter curve than the Colinton curve, or the range of the thermometer is 

 less ; consequently the minima differ less than if the two curves had been similar, 

 and the maxima differ more. This is the reason why the decrement of tempera- 

 ture with height is most rapid in summer, and least so in winter. (3.) Not only 

 is the Bonally curve lower than the Colinton one, and flatter, but it is shifted to 

 the right hand, so that the maxima occur later, as weU as the minima and mean 

 temperatures. A little attention will likewise shew that a gap or vacuity must be 

 left between the curves, greatest whilst the temperature rises, and least whilst it 

 falls ; and also that the difference of the vertical ordinates of the two curves will 

 be greatest when they form the greatest ascending angle with the horizontal axis, 

 and least when the descending angle is greatest, that is, as inspection shews, in 

 May and November respectively, which agrees with the results of the uppermost 

 curve of the plate. 



The examination of these curves furnishes us with some data of the most 

 important kind for meteorology, which it is best in the first place to state, and 

 afterwards to consider how we can explain. 



The first fact is the familiar one, that mean temperature diminishes as we 

 ascend in the atmosphere. The second is, that the annual range diminishes as 

 we rise, and, at a certain height, would probably sensibly vanish. The third is, 

 that the influence of seasons begins to be felt at the plains, and is later communi- 

 cated to the mountains. The two former of these facts obtain with reference to 

 the diurnal as weU as annual variation of temperature ; the last appears to be in 

 that case reversed-! 



The shift of the annual curve, or retardation of epochs, and likewise the de- 

 creased range, is common to the strata of the air above the surface of the earth, 

 and to those of the soil beneath it. Both ultimately, no doubt, exhibit a limit, 

 first where the diurnal variations disappear, then the annual. The cause, how- 



* See my Report on Meteorology in the first volume of the British Association Reports, 

 t Saussure, Voyages dans les Alpes, tom. iv. § 2050, &c. See also Kaemtz, Lehrbuch, band ii. 

 s. 133. 



VOL. XrV'. PART II. 4 M 



