ON THE ENERGY OF STORMS MARGULES 539 



energy sufficient for the development of storms; hitherto it has 

 only been assumed that storms start from such beginnings. The 

 existence of unstable conditions before a storm has never been 

 demonstrated. Prohaska never found such cases in his numerous 

 studies of thunder-storms. Even in so-called calms there is still 

 enough motion to disturb a condition of unstable equilibrium. 

 The forces that are thus set free are greater than those correspond- 

 ing to the largest horizontal pressure gradients that have been 

 observed in the atmosphere. The accelerating force acting on a 

 foreign particle of air whose temperature is 7\ when surrounded by 

 air whose temperature is T, is 



(7\ - T) 



p 



and therefore, for T= 273° and T x — 274 this becomes , whereas 



273 

 the force represented by a barometric gradient of i mm mercury per 



p 

 degree of a great circle, at the base of the atmosphere is — — . The 



1000 



vertical distances in our atmosphere are small. Unstable conditions 

 can scarcely exist for any length of time over extensive areas ; they 

 would disappear very quickly ; their existence has not been demon- 

 strated nor are they probable. Where there are adjacent masses of 

 air with very large differences of temperature on the same level, 

 as in the boe-en gusts, then cold air may intrude upon the warm 

 region, and warm air flow over into the cold region, but in this case 

 the storm wind velocity arises not only by reason of the barometric 

 horizontal pressure but also directly by the action of gravity. 



§(6) Many seek for the source of energy of a storm in the latent 

 heat of condensation evolved by the formation of clouds. I will now 

 compute the available kinetic energy for the following initial con- 

 dition: 



In chamber 1 of fig. 1a let there be dry air in neutral equilibrium, 

 but in chamber 2 an ideal fictitious gas that has the property of 

 expanding only when heat is added [and not by any diminution 

 of pressure] but otherwise behaves like dry air. This latter gas 

 replaces the moisture-saturated air [of nature] and the heat added 

 during its expansion corresponds to the latent heat of condensation; 

 for neutral equilibrium the vertical diminution of temperature in 

 this fictitious gas is smaller than that in the dry air. The initial 

 condition is to be so chosen that after the removal of the screen the 

 mass 1 spreads out below and the mass 2 above; we may then omit 



