100 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



plained easily by recollecting that the clouds form a sort of 

 covering above the surface of the earth, preventing its cool- 

 ing in the winter and its warming in the summer. And, 

 again, the clouds come generally from the adjacent ocean, 

 and bring with them in winter a temperature warmer than 

 the interior of the continent ; therefore in winter a clear, 

 cool day is followed by a cloudy, warm day; and the obser- 

 vations show that if, during the whole year, the heavens re- 

 mained covered, the mean annual temperature of St. Peters- 

 burg would not change from year to year. In order to pre- 

 dict the minimum of temperature with some probability, it 

 is necessary to take into account the consideration of the hu- 

 midity of the air, for if it happen that, with a clear sky, the 

 atmospheric humidity is considerable, then in the evening, 

 when the temperature lowers, the vapor may condense, and 

 its latent heat become free, and delay the fall of temperature. 

 The minimum temperature during the night is generally 

 but very little different from the temperature of the dew- 

 point as observed the previous evening. 'WilcVs Reperto- 

 rium filr Jlet., vol. iii. 



THE MEAN TEMPERATURE AT HELSINGFORS. 



The great toil attending the careful reduction of a series 

 of observations, extending over many years, renders it always 

 a pleasant task to note the completion of so laborious and so 

 useful a work as that recently published by Krueger, who 

 has endeavored, for Helsingfors, to extend the work per- 

 formed a long time ago by Hallstrom. The observations 

 made at the Magnetic Observatory at Helsingfors, during the 

 years 1848 to 1856, were recorded every twenty minutes, and 

 give one a very complete insight into the daily variation of 

 temperature, and the times of maxima and minima, together 

 with their relation to the mean temperature. This mass of 

 material lias been partially reduced by Xordenskjold, and 

 Krueger has completed the work by determining the co-effi- 

 cients of the harmonic series by which it is customary to 

 represent the average temperature of any station. He finds 

 that the results of sixty-eight years of observations may be 

 represented within an error of less than one thirtieth of a de- 

 gree Centigrade, by a formula in which the sines and cosines 

 of the first, second, third, and fourth multiples of the temper- 



