416 HORARY VARIATIONS OF THE BAROMETER. 



hours after the pun'.'^ passage of the meridian. These vapors, which change their 

 name and become dew in the morning, are so abundant under thf tropics as to 

 indicate a large conversion of watery vapor into liquid water. Now this water, 

 when it was in a state of vapor, exerted on the barometrical cistern a pressure 

 additional to the proper pressiire of the air, and contributed, in a certain measure, 

 to maintain the mercury at the height which it had attained ; but if, afterwards, 

 and when the air is about to descend because of its refrigeration by contact with 

 the earth, the watery vapor is converted into cloud, [serein,) and no longer 

 oceiipies more than a very small fraction of its primitive volume, this progressive 

 diminution of the quantity of water which was in a state of vapor at the moment 

 when th^ air began to assume a descending movement, cannot but operate as a 

 lessening of the pressure which the descent of the column of air exerts on the 

 barometer. 



It would be difficult to estimate with any precision what part, as regards the 

 ascension of the mercury in the morning, should be assigned to the evaporation 

 of the d^vf, and equally so to estimate how much the maximum of the evening 

 should fall below that of the morning from the effect of the production of the 

 evening vapors. The difficulty is the greater hecause here we arc to consider 

 not only a weight to be added to the weght of the atmospheric column or to be 

 deducted from it, but chiefly because when dew is in question, account is to be 

 taken of the elasticity which the vapor of water acqviires, as soon as it is formed, 

 by its mixture' with the air already warmed by contact with the surface of the 

 earth. Nothing, how^tver, forbids our making a supposition, and this I have 

 done by inquiring how much a stratum of dew, represented by a height of two 

 miUimHtres of water, .a supposition, I think, involving nothing excessive,) would 

 cause the barometer to ascend, by the fact alone of its vaporization ? The re- 

 suk has been found to be a height comprised between one and two tenths of a 

 millimetre. If we allow an effect of equal extent, but in an opposite direction, for 

 the diminution of the mercurial column at the evening maximum of nine o'clock, 

 there will result three-tenths of a millimetre, or very nearly so, as the quantity 

 by which the maximum of the morning would surpass the maximum of nine 

 o'clock in the evening; admitting, as has been just supposed, that the dew of the 

 mornmg, like the damps of the evening, represents a stratum of liquid water 

 two millimetres in height. 



It was t'tated at the commencement of this article that, in the course of 1826, 1 

 took occasion to d'scuss the meteorological tables of the whole of the year 1825 

 and of a part of 1824, as they had been given in the Annals of Physics and 

 Chemistry. I distributed the days first into categories indicated by the state of 

 the 'sky, and next into categories determined by the direction of the whid which 

 prevailed during these different days. Let us first observe to what result the 

 former operation led : 



Eighty days of a sky entirely overcast. The mean height of the barometer, at =""• 



tht moaient of the maximum of 9 o'clock in the morning, was 756. 82 



The mean height, at the moment of the maximum of the evening, was 756. 20 



Difference 0.62 



One hundred -and sixteen days of a cloudy sky. Maximum of 9 o'clock in the 



mornmg 758. 28 



Maximum of 9 o'clock in the evening 757. 72 



Difference 0.56 



Forty-eight days of a sky perfectly clear, or presenting only small clouds, white 



and thin Maximum of the morning 759. 87 



Maximum of the eveaing 758.53 



Difference 1.34 



