1835.] Meteorological Observations, 43 



One of the most interesting questions in Meteorology is 

 to ascertain the mean temperature of a place. The most 

 accurate method would be to take a great many observa- 

 tions — one every hour or half hour, and to find the average 

 of these ; but it would be quite impossible to carry this to 

 any extent ; and various plans have been proposed to obtain 

 the average temperature of the 24 hours, by means of one 

 or two observations. — Humboldt says that between the 

 latitudes of 46° and 48° the temperature at sunset corres- 

 ponds very nearly to the mean of the 24 hours. Leslie 

 supposes that the average may be reckoned from 8 o' Clock 

 in the morning ; but almost all Meteorologists agree that 

 the nearest approximation to the true daily average heat, is 

 the mean of the maximum and minimum. 



The author of the article Meteorology in the Edinburgh 

 Encyclopasdia, found from a great many observations, that 

 the mean of the temperatures at 10 a. m. and 10 p. m. 

 corresponded very nearly to that of the extremes, and con- 

 sequently to that of the whole 24 hours. To ascertain 

 whether this held good in other climates, as well as in that 

 of Britain, I have made numerous observations at Malta, at 

 Alexandria, at Cairo, at Thebes, at Darwar in the Southern 

 Mahratta Country, and at Madras, and have invariably 

 found that the mean of the two observations, at 10 in the 

 morning and evening corresponded within a few tenths, and 

 sometimes within a few hundredths of a degree, to the 

 exact daily average, and was generally, much more accurate 

 even than the mean of the extremes. In these instances 

 I ascertained the daily average from hourly observations, 

 and sometimes by taking the mean of tlie maximum and 

 minimum of every hour. I find that the true daily average 

 temperature determined at the Madras Observatory by 

 Mr. Goldingham, by means of hourly observations, on 36 

 different days, in the year 1823, differs only from the mean 

 temperature of 10 a. m. and 10 p. m. of the same days, 

 by little more than 5 tenths of a degree, of Fahrenheit, or 

 3 tenths of a degree of the centigrade. Such being the 

 case, I think we may reasonably expect that this law will 

 be found to be universally applicable, and I have accord- 

 ingly adopted these two hours for the observations of the 

 thermometer, in the accompanying register. 



