1858.] Meteorological Observations on Parisnath Sill. 19 



temperature of the stations, which on the 3rd is higher than on 

 the 1st.* 



The only consequence which could possibly be inferred from this 

 would be, that on successive days the course of the daily mean tem- 

 peratures of larger masses of air, removed from the ground, does not 

 always proceed parallel to the daily means of the stations ; the 

 latter may be rising when the former are falling or remaining sta- 

 tionary, or the reverse. 



In the present case, the small number of days and the omission 

 of taking the moisture into account does not allow us to consider 

 this result as an established fact ; but the possibility of such an 



* Such a result might have been predicted from the decrease in the barometri- 

 cal pressure from the 1st to the 3rd. The height obtained by the barometric 

 formula 



H = C Log h (1 + * T) 



D 



(omitting the smaller corrections for gravity) depends mainly on the value of the 

 quantity Log _.. This value will increase, when the barometrical pressures 

 decrease, if the difference between the latter, (b — b^ remains the same. In such 

 a case, to obtain a constant value for the height, the mean temperatures should 

 decrease with the pressures. We are, on the contrary, accustomed to see the 

 temperatures at the stations rising, when the pressures are falling, and such is 

 also the case in the present instance. The barometer from the 1st to the 3rd is 

 falling and the hourly mean temperatures of the stations on the 3rd are higher 

 than on the 1st, but the barometrical differences (b — b') being, on the whole, very 

 nearly the same on both days, it follows that we must find the real mean tem- 

 peratures on the 3rd lower than on the 1st. To give an instance to which extent 

 the temperature is dependent on the value of (b — b') I will take the observation 

 at 2 p. m. on both days. The barometrical differences were on the 



1st. 3rd. 



3.833 3.837 

 and the real mean temperatures were found to be 



29°.4 28°A 

 showing a difference of one degree. To make the real mean temperature at 2 on 

 the 3rd equal to that on the 1st, or one degree higher than we found it, the baro- 

 metrical difference on the 3rd should have been 3.821, or 0.013 inches less than 

 it was. This would have required a barometrical pressure of 25.916 inches on 

 the hill instead of 25.903, a difference whicli is far too great as to have been caused 

 by an error in the reading. 



D 2 



