INSTABILITY IN THE ATMOSPHERE 319 



However the electrification of the precipitated vapour may ultimately be accounted 

 for, there can be no doubt of the fact that at least as soon as cloud is formed the 

 particles are electrified ; and what I have said as to the immense rise of potential 

 as the drops gradually increase in size remains unaffected. I have tried various 

 forms of experiment, with the view of discovering the electric state of vapour mixed 

 with air. For instance, I have tested the vapour which is suddenly condensed when 

 a receiver is partially exhausted ; the electrification of cooled bodies exposed to moist 

 air from a gas-holder ; and the deposition of hoar-frost from a current of moist air 

 upon two polished metal plates placed parallel to one another, artificially cooled, and 

 connected with the outer and inner coatings of a charged jar. All have given results, 

 but as yet too minute and uncertain to settle such a question. These experiments are 

 still in progress. It appears probable, so far, that the problem will not be finally solved 

 until experiments are made on a scale much larger than is usual in laboratories. 



A great thunderstorm in summer is in the majority of cases preceded by very 

 calm sultry weather. The atmosphere is in a state of unstable equilibrium, the lower 

 strata are at an abnormally high temperature, and highly charged with aqueous 

 vapour. It is not easy, in a popular lecture like this, to give a full account of what 

 constitutes a state of stable equilibrium, or of unstable, especially when the effects 

 of precipitation of vapour are to be largely taken into account. It is sufficient for 

 my present purpose to say that in all cases of thoroughly stable equilibrium, a slight 

 displacement tends to right itself; while, in general, in unstable equilibrium, a slight 

 displacement tends to increase. Now, if two cubic feet of air at different levels 

 could be suddenly made to change places, without at first any other alteration, and 

 if, on being left to themselves, each would, under the change of pressure which it 

 would suddenly experience, and the consequent heating or cooling, with its associated 

 evaporation or precipitation of moisture, tend to regain its former level, the equilibrium 

 would be stable. This is not the case when the lower strata are very hot, and fully 

 charged with vapour. Any portion accidentally raised to a higher level tends to 

 rise higher, thus allowing others to descend. These, in consequence of their descent, 

 tend still farther to descend, and thus to force new portions up. Thus, when the 

 trigger is once pulled, as it were, we soon have powerful ascending currents of hot 

 moist air, precipitating their moisture as cloud as they ascend, cooling by expansion, 

 but warmed by the latent heat of the vapour condensed. This phenomenon of 

 ascending currents is strongly marked in almost every great thunderstorm, and is 

 precisely analogous to that observed in the centre of a West Indian tornado and of 

 a Chinese typhoon. 



When any portion of the atmosphere is ascending it must be because a denser 

 portion is descending, and whenever such motions occur with acceleration the pressure 

 must necessarily be diminished, since the lower strata are not then supporting the 

 whole weight of the superincumbent strata. If their whole weight were supported 

 they would not descend. Thus even a smart shower of rain must directly tend to 

 lower the barometer. [A long glass tube, filled with water, was suspended in a 

 vertical position by a light spiral spring, reaching to the roof of the hall. A number 

 of bullets hung at the top of the water column, attached to the tube by a thread. 

 When the thread was burned, by applying a lamp, the bullets descended in the 



