52 Wwi. A. Norton on the Variations 



year, after than before midnight. It appears also that cold winds 

 are more prevalent, and warm winds less so, in the first and last 

 than in the middle quarters of the year. Both of these facts are 

 opposed to the idea that difference of direction of wind may be 

 the general cause of the diminution in the hourly decrement of 

 temperature as the night advances, and in the nocturnal decrease 

 of temperature from the warm to the cold months. 



If we reject the n.e. and w. winds from the list of cold winds, 

 that is, regard them as not affecting the mean temperature, in the 

 colder months, cold winds will still be as prevalent after midnight 

 as before : and if we reject the w. wind, in the warm months, 

 from the list of warm winds, it will still be true that warm winds 

 will prevail no more in these months after midnight than before. 

 But upon these suppositions the cold winds will become more 

 frequent in summer than in winter in the proportion of 31 to 22, 

 and the warm winds less frequent in the proportion of 27 to 33 : 

 and both cold and warm will occur with about the same degree 

 of frequency in summer as in autumn. It is to be observed, 

 however, that the relative cooling effect of northerly winds in 

 different seasons is not in exact proportion to their relative fre- 

 quency, for the relative force of the wind is to be taken into ac- 

 count. Now it appears on examining the curves showing the 

 force of the wind in the different quarters of the years 1843-4-5 

 at Philadelphia, that the force of the wind at night is from two 

 to three times greater from September to April than from March 

 to October. If we connect with this the facts, that the strong 

 winds in winter are most frequently from the n.w., and that, as 

 shown by the first of the above tables, n.w. winds are much 

 more frequent during the former than the latter of the above 

 mentioned period, it will be seen that there is little room to doubt 

 that the greater nocturnal decrease of temperature in the summer 

 than in the winter, must be due to some other cause than the dif- 

 ferences, generally subsisting, between the directions of the wind 

 in these seasons. If there be any lingering doubt upon this point, 

 it will be removed, if we reflect that the law of variation of the 

 nocturnal decrease of temperature from one season to another, 

 which we have been considering, is as true for one place as for 

 another. It is found to hold good, with occasional partial excep- 

 tions, at all places, both in this country and Europe, where mete- 

 orological observations have been made. In fact, this law is 

 essentially connected with the general fact that the loss of tem- 

 perature, by nocturnal radiation, is equal to the daily rise of tem- 

 perature, both in January and July. So universal a law cannot, 

 it is believed, be dependent upon the direction of the wind, since 

 the prevalent winds are often very different in different localities. 



The question of the influence of the force of the wind follows 

 next in order. Fig. 12 exhibits the mean variations of the force 

 of the wind during the year 1844. 



