28 Dr Brewster on the Register of the Thermometer kept 



III. On the relation between the Mean Temperature of the 

 24 hours, and that of any single Iwur, or any similar pair 

 of hours, SfC. 



It was long the practice of meteorologists to observe the 

 thermometer three times a-day, on the supposition that the 

 mean of these three observations gave the mean temperature 

 of the 24 hours. Observations of this kind are still continued 

 in many parts of Europe. To the following short table of 

 some of these, I have added the deviations from the mean tem- 

 perature, as computed from the results of the preceding tables : 



Deviation from 

 Morning. Afternoon. Night. Mean Temp, of day. 



Edinburgh, 8 h Maximum. 10 h + 0°.3f6 Professor Playfair. 



Williamstown, 7 2 9+0 .510 Professor Dewey. 



8 1 6+1 .225 Proposed by the Phil. 



Soc. of New York. 



As three observations made every day, are not convenient 

 for many meteorologists, who are engaged in professional pur- 

 suits during the day, it became desirable to select those two 

 hours, the mean of whose temperatures approached nearest to 

 that of the whole day. The following times have been used 

 in this country, and many of them give results that differ very 

 considerably from the mean temperature of the 24 hours : 



Deviation from Mean 

 Morning. Afternoon. Temp, of Day. 



Hawkhill, - 8 2 + 0°.982 



Gordon Castle, -8 2 +0 .982 



Kinfauns, - 8 10 —1 .114 



Ditto, - 10 10 —0.122 



Leadhills, - 6 1 —0.134 



Isle of Man, - 9 11 —0.838 



Royal Society, London, 9 3^ +1 .453 



9 3 +1 .526 



9 2i +1 .511 



— 84 3 +1 .273 



8| 24 +1 .258 



— 8 3 +1.013 



. 8 2+0 .982 



7 3 +0 .641 



7 2+0 .610 



Royal Society, Edinburgh, 10 10 —0.120 



74 84 —0 .805 



9\ 84 .000 



