28 



RECORD AND DISCUSSION OF TEMPERATURES. 



Accordingly the mean is reached at 7 ]l -.7 A. M. and 7 h -.6 P. M; the maximum 

 at 2 P. M., and the minimum at 1 A. M. The mean range equals 3. 69, a quan 

 tity necessarily smaller than the mean amplitude for the whole year as given above. 

 Maximum mean hourly difference or change 0.62 between 7 h - and 8 h - A. M., and 

 0.45 between the same hours P. M. 



We now return to the last vertical columns of the general monthly abstracts, 

 and examine the 



Observed Hours of Mean Daily Temperature. The following table contains the 

 hours of the day when the temperature equals its mean daily value, made out for 

 each month of the year. Also shows their difference, or the &quot;critical interval.&quot; 

 These values are derived directly from the general abstracts of observed tempera 

 tures, and, for the first five months, are mean values derived from two sets of 

 observations in the first and second years. 



In the following table, I have exhibited the greatest absolute changes of tempera 

 ture observed between any two successive hours, between the highest and lowest 

 of any day of 24 consecutive hours, any month, and for the whole year. 



The following values are the absolute maximum and minima observed : 



Minimum in winter, 1853-54 ... . 66. 4, observed on Feb. 5th 



Maximum in summer, 1854 ... . +51.0, July 23d 



Minimum in winter, 1854-55 . . . . 65.5, Jan. 8th 



Absolute maximum difference, 11 7. 4 Fahr. 



Mean Monthly Temperatures and Annual Variation. In the following recapitula 

 tion of the mean monthly temperatures, the values have been taken directly from 

 the general abstracts. The means for the months of February, March, and April, 

 of 1855, have been added by means of the table given in Vol. II. of the Narrative, 

 p. 425, corrected from ten comparisons of monthly means between 14 and 34, 

 so as to refer the quantities to the same system of corrections as used in the present 



