374 DR BREWSTER on the Register of the Thermometer kept 



The summer curve descends regularly from midnight till 4 

 o'clock in the morning, when the coldest time of the day occurs, 

 and it ascends with great regularity till 3 o'clock, when it 

 commences a very rapid descent to its minimum ; the total 

 mean range being about 8. 61. 



The winter curve, on the contrary, has a gentle rise from 

 1 A. M. till 2 A. M. It then descends till 6, when it commences 

 its ascent, reaches its maximum at 2, and again descends, but 

 more slowly than it rose, the greatest difference of temperature 

 being about 3.86. By examining the individual curves which 

 compose the winter group, especially in the means of 1824 and 

 1825, the rise of temperature after midnight, and its subsequent 

 fall, will be very apparent, an effect which never takes place in 

 the curves of summer. 



The difference of character in the curves of April and Octo- 

 ber deserves to be noticed. Although these months are consi- 

 dered as giving nearly the mean of the year, and therefore as re- 

 sembling each other in temperature, yet there is a singular dif- 

 ference in the mode of its distribution. In October the mornings 

 and evenings are comparatively warm, while in April these 

 times of the day are remarkably cold. April, in short, unites 

 the low temperature of a winter month, with the great range of 

 a summer month ; while October unites the temperature of a 

 summer month, with the low range of a winter one. 



II. On the Determination of the two times of the Day, when the 

 Mean Temperature occurs. 



I am not aware of any observations made in our climate, by 

 which the hours, when the mean temperature of the day occurs, 

 could be determined. It has generally been believed that it oc- 

 curs at 8 o'clock in the morning ; and Professor PLAYFAIR not 



