310 METEOROLOGY. 



ON THE BEST HOURS OF DAILY OBSERVATION TO FII^D 

 THE MEAN TEMPERATURE OF THE YEAR. 



BY PROF. CHESTER DEWEY. 



The mean temperature of a day is to be obtained, originally, from 

 observations of the thermometer, taken twenty-four times daily, or 

 double that number. The mean of daily and hourly observations of 

 this kind must give a close approximation to the actual mean tem- 

 perature. From a series of such hourly observations the two, or three^ 

 or four hours may be selected, which will give nearly an equivalent 

 result. 



The large Meteorological Society of Manheim, in Germany, selected 

 the hours of 7 a. m. and 2 and 9 p. m., but I could find no reason for 

 this selection in any accessible work, when I began observations in 

 meteorology in 1815, In 1816 and 1817 I made twenty-four hourly 

 observations of five days each in the different seasons ; the first of the 

 kind on record, so far as I know, being made for thirty days. The 

 mean of the 24 observations is 41°. 50 ; of 10 a. m. and 10 p. m, 41°. 45; 

 and of 7, 2 and 9 about one desfree higher. Coming so near the mean 

 I adopted those hours of the Manheim Society, for the ease of com- 

 parison with the results obtained by them. 



The mean of observations at 6 a. m., 2 and 10 p. m., gave a close 

 approximation to the mean of 24 observations ; but tlie morning hour 

 would be too early for half the year in view of many observers. 



The results of this series of observed temperature I communicated 

 to Secretary Calhoun, as he was about to organize the system of 

 meteorological observations, so successfully made by the surgeons at 

 the military posts of the United States since 1819. These hours were 

 adopted for all the posts. 



The fitness of these hours, 7 a. m., 2 p. ra., and 9 p. m, for obser- 

 vations, is sustained by the following facts : 



1. By the hourly observations for a year at Leith Fort, Scotland. 

 These give the mean of 24 daily observations, 41°. 50 ; of 10 a. m. and 

 10 p. m. very nearly the same ; and of 7, 2 and 9 about one-fourth of 

 a degree above the mean of the 24 observations. 



2. By the hourly observations at Amherst College, Massachusetts, 

 through 1839, under the direction of Prof. Snell. The mean of the 

 24 observations is 47°. 23 ; of 10 a. m. and 10 p. m., is 47°. 16; and of 

 7, 2 and 9, is 47°. 88. This last, then, is two-thirds of a degree above 

 the 24 mean. Prof. Snell shows the mean at 6 a. m.; 2 and 10 p. m. 

 is nearly the same as the mean of the 24 observations. 



3. By the " Girard Observations," under the direction of Prof. ' 

 Bache, an extensive series of several years, bi-hourly and hourly. 



The mean of 7, 2 and 9 is only three-tenths of a degree above that 

 of 24 observations, and from this last, that of 10 and 10, differs only 

 one-tenth of a degree. ; 



4. Brooklyn Heights' Observations, hourly, for 1856, by E. Meriam^ 

 esq. I have summed only the first seven days in each month. The 

 mean of 24 observations is 47°. 72, and of 10 and 10 is very near the 



