in the Magnetic Forces of the Earth. 



53 







Fig. 12. 

 Curve of the Mean Diurnal Variations of Force of Wind for 1844, in Lbs. 



Lbs. OM 2 3 4 5 6 7 8 9 10 1113 13 14 15 16 17 18 19 20 2122 23 0a 



On inspecting this curve it will be seen that the force of the 

 wind is nearly the same before and after midnight, and that its 



principal variations occur during the day. The 



law 



same law is 

 shown by the curves for other years, and for quarters of years, so 

 iar as given. The nocturnal loss of temperature cannot therefore 

 be materially modified by variations in the force of the wind, in 

 the average of months and years. The curve showing the varia- 

 tions in the average force of the wind from month to month, 

 which cannot conveniently be Riven here, indicates that the wind 

 is Highest in February or March, and is much higher during the 

 first and last quarters of the year than toward the middle of the 

 year. Since these strong winds are more apt to be from the n.w. 

 than from any other quarter, their tendency will be to cause the 

 temperature to fall more during the night, in the fall and winter, 

 than in the spring and summer, instead of less as it does in fact. 

 Let us next consider whether it may chance that the variations 

 in the nocturnal loss of temperature which I have specified, are 

 due to the cooling or heating effect of rain. The observations 

 in my possession do not furnish me with the means of making 

 a ny but a very partial examination of the relative quantities of rain 

 that fall during the first and last halves of the night. But the fol- 

 lowing facts will serve to show that the influence of rain, what- 

 ever it may be, can have no part in determining the law of the 

 decrease of temperature for a single night. I . Of twenty-one term 

 days, observed at Washington, for which the curves of the daily 

 variation of temperature are given, there are only five exceptions 

 to the general fact that the temperature falls most slowly during 

 fhe latter half of the night. 2. Of twenty-three days observed 

 in Philadelphia, in October, 1843, there are but six exceptions to 

 the same general fact. We may conclude from these facts, that 





