170 M, Ramond's Imtntctionifor the Application of [Se1>t* 



r^pect to the extent and nature of the period, we shall soon 

 attain the epoch where the correction is almost nothing, and 

 the mean does hot sensibly differ from being stationary. 



In general the result of a year may be regarded as cjuite a suf- 

 ficient approximation, and when a barometric mean js founded 

 ort two or three years* observations, we shall not risk much in 

 regarding it as decisive ; but if it be intended to serve for deter- 

 mining the relative or absolute elevation of the place, it is fur- 

 ther necessary that it should be accompanied by athermometric 

 mean dedttCed according to the same method by observations 

 in conjunction with those of the barometer. The methods most 

 commonly pursued in conducting a series of meteorological 

 observations are very far from tending to the attainment of the 

 objects just specified. The observation of the maximum and 

 minimum temperature of the day is of Httle use, and forms a part 

 of a series altogether differing from that of the pressure. 



We shall greatly simplify our labour if we can determine the 

 instant when the height of the barometer is exactly the mean of 

 its heights which correspond to the four epochs. Now this is 

 not very difficult ; it will readily be seen that this period will be 

 found at a distance from the maximum and minimum determined 

 by the ratio which exists between the oscillations of the day and 

 those of the night. I find that in our temperate climates the 

 hour of noon satisfies very well these conditions, and the coin- 

 cidence is the more fortunate as the same hour is convenient for 

 several other objects. If there be any error in the barometric 

 mean of noon, it will probably be in excess, but extremely small 

 in regard to the nature of the operation. I do not believe that 

 it exceeds '003 of an inch, a quantity which we may assuredly 

 neglect on this occasion ; with deference to the more rigorous 

 determination and introduction of it into calculation, if ever the 

 exactness of observations and observers should be carried so far 

 that a barometric mean should correspond with the atmospheric 

 pressure to the '003 of an inch. 



I make then no difficulty in regarding the mean of noon as a 

 sufficient expression for the mean atmospheric pressure corrected 

 for the diurnal variation ; and for many years my system of 

 observation has been founded on this basis. Observations made 

 at the critical periods of the diurnal variation furnish the most 

 certain presages which can be drawn from observation, of the 

 progress of the barometer. It is by the derangement of the 

 norary oscillations that the smallest changes introduced into the 

 constitution of the atmosphere announce themselves. Lastly, 

 the extent of the variation established in different places con- 

 jointly with the mean height of the mercury will establish points 

 of comparison, by the help of which we shall form a more solid 

 judgment on the ratio of the pressure of the air to its weight ; a 

 ratio which offers one of the newest questions in meteorology, 

 and one which is the most fruitful in important consequences, 



