O P B S E R V A T I O N S F R T B M I' E R A T UIl E . 



27 



Owing to the fact that the observations extend over one year only, the table, in 

 some instances, must necessarily contain some small irregularities. The closest 

 results are obtained from the hours G, 2, 10, which was also the case at Van Rens- 

 selaer Harbor. 



Connection of the Lunar Phases with Low Winter Temperatures. 



The apparent connection of the lunar phases with the observed temperature of 

 the air during the Arctic winter, the thermometer being below the zero of Fahren- 

 heit's scale, was long ago noticed by Arctic explorers, and was again independently 

 observed by Dr. Kane, in the discussion of whose observations I have attempted 

 an explanation of the phenomenon. In that paper, the connection of the lunar 

 phases with the serenity of the sky and the fall of snow was also discussed ; for the 

 observations now on hand, the numerical relations alone will be represented. 



Dividing the daily means of the temperature into penthemers (or periods of five 

 days), a table was formed showing the time of full and new moon and the mean 

 temperatures; and, by means of differences of the alternate means at these periods, 

 the amount by which the mean temperature is lower at full moon than at new 

 moon is exhibited in column headed A. 



