18 Meteorological Observations on JParisnatk Hill. [No. 1. 



Next the hours 1, 2, 6, 8, 9, 10 were considered correct and the 

 remaining hours were altered as the table shows, (Col. III). Fig. 4 

 in the Plate represents the corrected curves. 



Ou the 4th April no alterations were required. 



Having adopted these corrections, I now proceed to point out the 

 distinguishing features of the curves of real hourly mean tempera- 

 tures, Table V. (Fig. 4). 



In the first place they have a much smaller range than the tem- 

 peratures derived from the observations (Col. I.) The range of the 

 latter, on the 1st is 8°. 7, on the 3rd it is 9°. 5, whereas the range of 

 the real means, between 7 and 2, is on the first only 4°.l, not quite 

 half of that of the temperatures of the stations. On the 3rd, the 

 barometer had not been observed at the hour of the minimum, but 

 judging from the small hourly variation, the daily range on that 

 day must have been still smaller than on the 1st. 



It is not possible to arrive at a definite conclusion regarding the 

 absolute real daily mean temperatures on both days ; the possible 

 error in the height and the error which might arise from different 

 degrees of moisture on both days would make it uncertain. The 

 relative values of the real daily mean temperatures as compared with 

 each other, are also difficult to ascertain. The smallness of the 

 number of observations hardly allows an estimate being formed from 

 them. An estimation or interpolation for some of the deficient 

 hours could not be made with the same reliance as in the instance 

 of the mean curve of the stations, and the fitness of the combination 

 of hours which had been employed before, would have been to say 

 the least, doubtful. 



As far as a conjecture may be permitted from the average 

 hourly variation on the 3rd, the minimum temperature on that 

 day could not have been more than about half a degree higher than 

 that on the 1st. But the maximum temperature of the 3rd is a 

 whole degree lower than that of the 1st, a difference of the same 

 magnitude, but in an opposite direction as that between the maxima 

 of the curve of the stations. If therefore the real daily mean tem- 

 peratures of both days were not the same, the probability is, that 

 the real mean temperature of the 24 hours on the 3rd was lower than 

 that on the 1st. The case is the reverse as regards the daily mean 



