HORARY VARIATIONS OF THE BAROMETER. 423 



APPENDIX TO THE FOREGOING. 



BY PROFESSOR HEXRY. 



Without wishing in the least degree to detract from the merits of the foregoing 

 interesting exposition of a phenomenon which has long occupied the attention of 

 meteorologists, I think it due to the memory of the late Professor Espy to state 

 that tlie essential principle of the explanation was given by him in the first vol- 

 ume of the Journal of the Franklin Institute, of Philadelphia, in 1828. M. 

 Yaillant has, however, without knowing of the previous explanation of Professor 

 Espy, given a more full exposition of the phenomenon by the introduction of the 

 effect of the variation in the quantity of vapor in the atmosphere.^ The following 

 is an extract from the article published by Mr. Espy : 



" The true cause of this phenomenon is the expansion, and, of course, rise of 

 the atmosphere by heat, and the contraction and consequent fall of the atmos- 

 phere by cold. Suppose just before sunrise in the torrid zone, the temperature 

 of the air neither increasing nor diminishing, and of course the air neither ex- 

 panding nor contracting, the barometer stands at thirty inches; now when the 

 eun rises at six o'clock the air will begin to be heated near the surfoce of the 

 earth, and of course, by its expanding, will elevate the superincumbent atmos- 

 phere ; and this by its inertia will react on the air below, and thus press harder 

 on the mercury of the barometer than if it were at rest, and the more rapidly it 

 is forced upwards the greater will be its reaction downwards, and of course the 

 more will the barometer be affected. It is manifest that the most rapid increase 

 of heat p.nd rarifaction of the air will take place somewhere betwen sunrise and 

 three o'clock, when the heat is the greatest, and this will evidently be near tea 

 o'clock, at which time the barometer will stand highest. Though the heat is 

 still increasing, and of course the air expanding, yet the rapidity of increase after 

 this hour is not so great, and therefore the barometer will begin to fall ; and at 

 the moment of greatest heat, when the air is neither expanding nor contracting, 

 the mercury will again stand at thirty inches high. But now the air begins to 

 contract from cold, and the mercury will therefore continue to descend, and the 

 rapidity of the descent will be in proportion to the rapidity of the contraction 

 from cold. Perhaps this effect may be more clearly understood by imagining 

 an extreme case. Suppose the lower strata of atmosphere suddenly annihilated, 

 the mercury of the barometer would be relieved from all pressure for a moment, 

 and fall down into the basin ; and if annihilation removes all pressure from the 

 mercury, a contraction of the lower strata by cold will remove some pressure; 

 therefore the mercury will fall ; at the moment, therefore, of a most rapid decrease 

 of heat, which is probably near sunset, the mercury will stand lowest, and will 

 be below the height of thirty inches, at which I have supposed it to stand when 

 the air is neither expanding nor contracting. The mercury will novr begin to 

 rise, for the rapidity of contraction diminishes from this moment, and the upper 

 parts of the air are permitted, more and more, to press upon the lower with their 

 whole weight; and even when the contraction ceases below, the upper parts, 

 having acquired a velocity downwards, are inclined to continue that motion, and 

 thus by their momentum will press upon the lower parts with a force greater 

 than their natural gravity, and thus the barometer will rise above thirty inches, 

 at which height it was supposed to stand when the air was neither contracting 

 nor expaud.ng. This effect must take place some time in the night, and it seems 

 probable, a priori, that it would be about ten or twelve o'clock. Now, as the 

 meicury at this hour stands higher than it does by the natural weight of the air 



