COMMON THINGS THE ALMANACK. 



Thus it will be seen that, putting aside the consideration of 

 atmospheric refraction, day and night could never be precisely 

 equal, except in the rare case in which the equinox takes place 

 at the moment of noon. 



61. Let us now consider how these phenomena are modified by 

 atmospheric refraction, which, as has been shown, increases the 

 length of the day and decreases that of the night ; and it must be 

 observed that their effect is much more considerable than any 

 which can arise from the moment of the equinox occurring either 

 in the forenoon or the afternoon. 



On the day of the March equinox, whether day and night be 

 equal or unequal so far as depends on the position of the sun, the 

 effect of refraction will be to cause the length of the day to 

 be greater than that of the night, since its effect greatly pre- 

 dominates over any which the sun's change of position could 

 produce. 



On the preceding days, refraction has the same tendency, but 

 then the tendency of the sun's position to render the night longer 

 than the day is more considerable, and will be such as to balance 

 or predominate over the effect of refraction either one or two 

 days before the equinox. The consequence is, that the day upon 

 which the intervals of light and darkness are either exactly equal 

 or least unequal, will be either one or two days before the day of 

 the equinox. 



It may be shown precisely in the same manner that the day in 

 September on which the intervals of light and darkness are either 

 exactly equal or least unequal, will be one or two days after the 

 equinox. 



These observations may be easily verified by reference to the 

 columns of sunrise and sunset in any almanack. Take for 

 example that of 1854. 



The March equinox took place at 20 minutes past 10 in the 

 evening of the 20th. The day on which the intervals of light and 

 darkness were least unequal was the 19th, upon which the sun 

 rose at 8 minutes past 6, and set at 9 minutes past 6. 



The September equinox took place at 13 minutes past 9 on the 

 morning of the 23rd. The day and night were exactly equal on 

 the 25th, when the sun rose at 51 minutes past 5, and set at 

 51 minutes past 5. 



In these observations we have quoted the almanack as calculated 

 for London, but similar consequences may be deduced from those 

 computed for other places. 



82 



