METEOROLOGY. 3 1 1 



same ; while tliat of 7, 2 and 9 is 48°. 28, or greater than the 24 mean 

 by one-half a, degree nearly. 



5. At Sacramento, Cal., lat. 38 N. The 24 mean is 64°. 41, and 

 the mean of 7, 2 and 9 is 64°. 11. 



Note — By the last three it is evident that the mean of 7, 2 and 9 

 approaches nearer to that of the 24 mean, as the places have a lower 

 latitude, and an examination of the 24 hours ohservation in the Arctic 

 regions show^quite a departure of the mean of 7, 2 and 9 from the 

 mean of 24 observations. 



The 24 hourly observations give the mean of the year — 



At Halle 48°. 00 



AtGottingea 52 82 



At Padua 56 74 



On calculating the mean of 7, 2 and 9, I find that — 



At Halle '. 48 89 



AtGottingen 53 45 



At Padua 57 47 



Observations on the temperature of Salem, Massachusetts, were 

 made with much care by Dr. Holyoke for thirty-fchree years precediilg 

 1819. The hours of observation were four, viz : 8 a. m., noon, sunset, 

 and 10 p. m. By interpolating for sunset, in my series of 24 daily 

 observations, I found that the mf-an from these four hours is only a 

 little greater than that of the three hours, 7 a. m., 2 and 9 p. m. 

 Dr. Hulyoke's mean temperature of Salem is 48.68 degrees. The 

 mean heat at Leith, by the 24 daily observations, is 48.24, and by the 

 hours, 7, 2 and 9, is 48.50. 



This approximation from these four hours, and one of them variable, 

 is another unexpected result. 



Between 1842 and 1855 the observations at the military posts were 

 directed to be made at four periods of the day, viz : a little be/ore sun- 

 rise as the coldest generally ; at 3 p. m. as the hottest, and at 9 a. m. 

 and 9 p. m. as approximating the mean temperature of the day, but 

 half the sum of the observations at sunrise, S. R., and 3 p. m. was 

 to be taken as the mean heat of the day. In the preparation of the 

 "Army Meteorological Register," published in 1856, the fourth part 

 of the sura of those four observations was taken as the mean of the 

 day, because Dr. Coolidge and his associate became satisfied from 

 extensive comparison of the twenty-four daily observations, that the 

 mean of ttie four observations was nearer the twenty-four mean, than 

 those at sunrise and 3 p. m. would give. Dr. Coolidge states also, 

 that the evidence was clear from the comparison of numerous twenty- 

 four hourly observations at the posts that the mean of 7, 2, and 9, 

 was for all the posts, the nearest approximation of any hours selected 

 to the mean of the twenty-four daily observations. In 1855, there- 

 fore, the Surgeon G-eneral, Dr. Lawson, issued his circular requiring 

 a return to the original hours of observation, viz: 7 a m., 2 p. m., 

 and 9 p. m. 



In ascertaining the relative correctness of the results in the " Con- 

 solidated Tables" of the Army Meteorological Register, taken for 



