590 



SCIENCE. 



[Vol. II., No. 39. 



condition of the atmosphere when it lies at rest, 

 examining it especially with regard to the kind 

 of equilibrium that then exists, and the changes 

 necessary to produce a tendencj- to motion. 



AVheu the air is at rest, it is normality densest 

 and warmest next to the earth's surface, and 

 becomes thinner and cooler at successive alti- 

 tudes above it. It is denser below because 

 the earth's attraction pulls it down, and com- 

 presses the lower layers b}' the weight of the 

 upper ones. It is warmer below, mainlj" be- 

 cause the air gets nearly- all of its heat by 

 contact with, or radiation from, the warm earth, 

 and not directly from the sun's raj's, which pass 

 through it with but little obstruction. The 

 average rate of upward cooling, determined by 

 manj' observations on mountains and in bal- 

 loons, is about one degree F. for every three 

 hundred feet of ascent. In this restful con- 

 dition let us take a block of the drj' air (the 

 effect of the presence of water-vapor will be 

 considered with the storms at sea) from the 

 earth's surface, where the temperature is, sa}', 

 60° (fig. 1 ) , and lift it up three hundred feet, 

 to where the temperature is 

 one degree less, or 59°. The 

 block of lower air expands as 

 it rises, because it is pressed 

 on by less atmospheric 

 weight, — less, at least, by 

 the weight of three hundred 

 feet of air ; and, in thus ex- 

 panding, it is cooled mechan- 

 icall3'. It has been shown 

 that this mechanical cooling 

 of an ascending mass of dry 

 air amounts to one degree F. in a hundred 

 and eighty-three feet of ascent, whatever its 

 initial temperature ; so that in this special 

 case the block is cooled bj" 1.6°, and its tem- 

 perature is reduced to 58.4°. Now, let us 

 compare it, when thus expanded and cooled, 

 with an equal-sized block of air beside it, 

 whose temperature is 59°. Evidentlj-, of these 

 two blocks of the same volume, and at the same 

 pressure, the cooler will be the heavier. The 

 block brought up from the surface, and now at 

 a temperature of 68.4°, will weigh more than 

 the air at 59° beside it, and hence it will tend 

 to sink ; and it must sink all the way down 

 to its original level before it finds any air as 

 heavy as itself. In this imaginary experiment 

 we have disturbed the arrangement of the nor- 

 mal, quiet atmosphere ; and the disturbed mass 

 returns to its original position as soon as freed 

 from the constraining force. Such an atmos- 

 phere is therefore in a condition of stable equi- 

 librium, like a rod hung bj^ its upper end, which 



300 F€€T 



60' 



is opposed to an}' change in its position, and, 

 when displaced, tends to return to its original 

 attitude. 



Evidently, when a whirhviud springs up in 

 the calm air of a desert, as is so often the case, 

 the atmosphere cannot possess this normal sta- 

 bility : for then there would be no temptation to 

 an}' such disturbance ; the air would prefer to 

 stand as it is. Before the wliirhvind can arise, 

 there must have been a change to a condition 

 of unstable equilibrium, in which the air, like 

 a rod balanced on its lower end, is ready to 

 move on small provocation ; and we have now 

 to look for the cause of this change. To be 

 guided properlj' in the search, the conditions 

 necessary and antecedent to the formation of 

 the whirls must be examined. They are, that 

 the whirls occur generally in level, barren, 

 warm regions, in quiet air, and onlj' in the day- 

 time after the sun has risen high enough to 

 warm the sand}' ground, and the air next to it, 

 to a rather high temperature. As the first and 

 second of these conditions may be present at 

 night as well as by day, it must, without doubt, 

 be the heat from the sun that disturbs the 

 quiet equilibrium into which the air tends to 

 settle, and, b}' warming the lower layers, 

 causes a departure from the ordinary stable 

 condition of rest. 



Let a case be supposed : the sun has warmed 

 the lower air of the first example to a tempera- 

 ture of 90° (fig. 2) , while the air three hundred 

 ieet above the desert sands 

 * has, in virtue of its diather- 



mance, risen onl}' to 70° ; so 

 that there is now a difference 

 of twenty degrees between 

 these two lajers. If we here 

 repeat the experiment of car- 

 rying a block of surface-air 

 to a height of three hundred 

 feet, it is again mechanically 

 cooled 1.6°, so that its tem- 

 perature is reduced to 88.4° ; 

 and now, comparing it with 

 an equal volume of adjoining air at 70°, the 

 latter is evidently the heavier, and therefore 

 the block of air brought up from the surface, 

 instead of tending to sink, as in the first case, 

 tends strongly to rise farther, and continue 

 the motion given to it. In other words, the 

 air is now in a condition of unstable equi- 

 librium : it is ready to upset and re-arrange 

 itself. The lower layer maj' be compared to a 

 film of oil balanced beneath a quiet sheet of 

 water : a little disturbance would cause the two 

 liquids to change places, and the oil would rise 

 through the water, draining itself upwards. 



300 freer 



90" 



