LOCAL CLIMATOLOGY. 99 



cold wave coming on one day one year, and on another in the 

 next, and so on. Hence, when it shall have fallen for several 

 trnies each day, these waves, when taken together, will average 

 one another. 



And yet I am not quite sure that they would. The solstice and 

 equinox are fixed points. At the summer solstice, for example, 

 the sun has reached its northern limit, and from that time it 

 begins to decline southward, until, on the 21st of September, it 

 passes into the southern hemisphere. This is an important fact. 

 "The belt of calms, ^^ as it is called, swings back and forth with 

 the sun. This is the limit and separation wall between the north- 

 ern and the southern trade winds, and polar currents in general. 

 It determines the northern boundary of the trade winds, they 

 being felt further north in summer than in Avinter. It determines, 

 also, the latitude at which the return current passes through the 

 polar current so as to become the surface current ; this latitude 

 being some forty degrees further north in summer than in winter. 

 When now this "balance of the winds" swinofs over into our 

 hemisphere, it sets back the currents of that hemisphere, and as 

 it is constantly advancing up to the time of the solstice, it is like 

 a large vessel setting rapidly into a narrow channel, driving back 

 the Avaters until they accumulate against the barriers, and then 

 they return in a high wave and with unusual force, and then being 

 reflected back by the advancing ship, they return again to the bar- 

 rier, and so on, oscillating back and forth. When the sun begins 

 to recede from the solstice towards the other hemisphere, the 

 return wave moves with an unusual impulse and for a longer time 

 than usual. 



Now, there can be no doubt that something of this occurs with 

 the winds ; and to this I attribute not only the waviness of our 

 weather, but also the fact that about the time of the equinoxes we 

 have what are familiarly called the " equinoctial storms^ 



The reality of these storms is made manifest by the following 

 statistics showing the comparative average of the water-fall in the 

 several months : 



Inches. Inches, ' 



Jan. ._ 1.594 July 3.281 



Feb _.. 1.138 Aug 3.690 



Mar. 1.888 Sep 2.490 



Apl _ 3.375 Oct. 3.122 



May ._. 2.933 Nov 2.254 



June 3.121 Dec 2.027 



