20 MEMOIR OF DR. DALTON, AND 



are certainly needful as accessory or confirmatory, leaving" 

 what to the eyes of many is a want of finish, so that others 

 have been needed to complete what was in reality suffi- 

 ciently complete had it been laid out entire as it existed in 

 his own mind. 



As an example of his style, at p. 97, we find 

 "It appears from the observations (see table p. 15) that 

 the mean state of the barometer is rather lower than higher 

 in winter than in summer, though a stratum of air on the 

 earth's surface always weighs more in the former season than 

 in the latter; from which facts we must unavoidably infer 

 that the height of the atmosphere, or at least of the gross 

 parts of it, is less in winter than in summer, conformable to 

 the table p. 80. There are more reasons than one to con- 

 clude that the annual variation in the height of the atmos- 

 phere, over the temperate and frigid zones, is gradual, and 

 depends in a great measure on the mean temperature at the 

 earth's surface below, for clouds are never observed to be 

 above four or five miles high, on which account the clear air 

 above can receive little or no heat, but from the subjacent 

 regions of the atmosphere, which we know are influenced by 

 the mean temperature at the earth's surface; also, in this 

 respect, the change of temperature in the upper parts of the 

 atmosphere must in some degree be conformable to that of 

 the earth below, which we find by experience increases and 

 decreases gradually each year, at any moderate depth, ac- 

 cording to the temperature of the seasons. 



" Now with respect to the fluctuations of the barometer, 

 which are sometimes very great in twenty-four hours, and 

 often from one extreme to the other in a week or ten days, it 

 must be concluded, either that the height of the atmosphere 

 over any country varies according to the barometer, or other- 

 wise that the height is little affected therewith, and that 

 the whole or greatest part of the variation is occasioned by a 

 change in the density of the lower regions of the air. It is 



