ALLISON FLUCTUATION OF THE BAROMETER. 161 



clear steady night, at any time of year, he will see the column 

 decline till about 3£ a. m. ; about 4 a. m. a movement upwards from 

 this minimum begins and continues till full 9 a.m. when a maximum 

 height is attained. In 15 or 20 minutes again the column sinks till 

 3£ p. m., once more to ascend till 9. p. m., when the second fall 

 sets in ; which decline continues, as said above, until after 3 o'clock 

 next morning. This action is certain and well known. Briefly we 

 will discuss the reasons for it ; and first I think we will find at the 

 bottom of our investigations, the great controller of our atmosphere, 

 heat. Indeed we need go no farther than this agency for the direct 

 cause of the second minimum and maximum of the Barometrical 

 column. A mere superficial glance at once leads us to a correct 

 conclusion that as the heat increases during the forenoon, after 9 

 o'clock the air becomes lighter, and pressure is taken off the cistern, 

 and this will continue till after 3 p. m., when the greatest heat of 

 the day having been reached, the gradually cooling atmosphere 

 grows more dense and presses the mercury up the tube. 



But while the one wave is thus sufficiently accounted for, other 

 causes must interfere about 9 o'clock in the evening, or the upward 

 tendency would remain till the sun of the next day had restored the 

 heat and lightened the atmosphere. In short if we look to direct 

 heat alone, we have as an effect but one fluctuation during the whole 

 day. I have been careful to use the phrase "direct heat" as I 

 believe it may be proved that heat, though acting indirectly, is the 

 root of the second wave also. Now we imagine ourselves at 9 

 p. m., on any day, and looking to the diminution of heat alone the 

 barometer should still be rising ; as, speaking broadly, cold air is 

 more dense than hot air. 



But the column has reached its greatest height for the present, 

 and in a few minutes begins to descend. Why? Because in 

 addition to the gradual cooling of the atmosphere above mentioned 

 a force of descending vapour has been at work in the early hours of 

 the night pushing down upon the cistern of the barometer, and 

 greatly aiding in the elevation of the column thus doubly effected. 

 And soon after 9 p. m. this second force ceases ; the earth becom- 

 ing cooled to equal the temperature of the air, or nearly so. This 



