INTliODlT 



XL11I 



'I A I '.I. K X. 



Mean temperature of the air at every hour of the day, deduced from tJie term-day observation*. 



a One observation. 



ft Two observations. 



There is but one maximum and one minimum in the diurnal horary range of the tempera- 

 ture. Commencing with the spring months and terminating with those of autumn, the instant 

 of the former ranges between three hours and four hours P. M., though it more frequently occurs 

 nearer to the first than to the last-named hour. In winter, as might be inferred, the maximum 

 takes place an hour earlier, or about 2 p. M. 



The observations do not permit us to fix so definitively the hour of minimum temperature in 

 nil the months. There were no records between midnight and 6 A. M. in any of the months of 

 September, October, and November ; but during the twenty-one term-days kept, it was colder at 

 6 A. M. than at any other hour. 



In all the year the greatest heat is from 3 P. M. to 4 p. M. in February, and the greatest cold 

 from 5 A. M. to 6 A. M. of July. March is subject to the greatest diurnal range, and June to the 

 least ; though, if we take the indications of the register thermometers instead of the hourly ob- 

 servations, their mean results show greater fluctuations in November than in March, and these 

 are probably entitled to greater confidence than deductions from the two term-days of 1850 

 and 1851. 



Comparing the yearly temperatures of each hour witli the monthly temperatures at corre- 

 sponding hours, remarkable accordance is found between the April and November observations ; 

 and if the mean annual temperature were to be deduced from the observations of a single month, 

 April would give the nearest approximation ; October next. Of the hours, the mean observa- 

 tions in all the months, either at 9 A. M. or 8 P. M., differs less than 1 from the mean annual 

 temperature. 



Adopting as the mean temperatures of the several months at Santiago, those found by com- 

 bining all the tri-hourly, self- register, and hourly observations with 0411:1! weight, the excess 

 of the temperature at each hour of the day is shown in the following table : 



