1867. 



Ill 



I Chase. 



49.744; May, G0.902; June, G9.780 ; July, 75.040; Au«:ust, 

 71.754; September, 65.043; October, 53.922; November, 42.350; 

 December, 32.132. January and July are the month.s of extreme 

 temperature, terrestrial absorption and radiation retarding the pe- 

 riods of maximum heat and cold so that they fall in the months 

 following the solstices. Averaging the temperatures at equal inter- 

 vals from January (taking the mean temperature of December and 

 February, of November and 3Iarch, &c.), we get the following re- 

 sults : 



The second of the above series of ratios (that of the difference in 

 the arc of the sun's zenith distance), is based upon the following 

 estimate of the average monthly increase of solar altitude, at all 

 places in the temperate zones : 



^lo., from winter solstice, . . 1 2 3 



Increaseof sol. altitude, 3°i W 2S°i 

 Ratio of increase, . . . .069 .255 .500 



If we allow about twenty-four days for the cumulative effects of 

 increasing heat and cold, these ratios become properly comparable 

 with the monthly ratios of temperature-variation. 



The data for the following table, with the exceptions specified in 

 the foot-notes, were drawn from Herr Dove's sixth paper, ''tJber 

 die nieht periodischen Ver'anderungen der Temperatur-Vertheilung 

 auf der Oberflilche der Erde" (Memoirs of the Berlin Academy of 

 Sciences, for 1858), and from his local corrections for periodic and 

 non-periodic variations, as inserted by Prof. Guyot, in the V. series of 

 " Tables, Meteorological and Physical, prepared for the Smithsonian 

 Institution," third edition. In determining the general ratios, all 

 observations for incomplete years were discarded, and in some in- 

 stances, where the stations involved were very numerous, I also 



