26 WILLIAM H. ALEXANDER Vol. XXII, No. 1 



This, however, is not sufficient to account for all the observed 

 facts, since it is not clear just how either the condensation or 

 the congelation could suddenly take place throughout a cloud 

 volume great enough to produce the observed effects. We 

 must, therefore, look for some other explanation, and this we 

 shall probably find, in the difference between the actual tem- 

 perature gradient of the surroimding atmosphere and the adiahatic 

 temperature gradient of the saturated air within the cloud itself; 

 or, in other words, the cause of the violent up-rush and tur- 

 bulent condition within large cumulus clouds is, presumably, 

 the difference between the temperature of the inner or warmer 

 portions of the cloud itself and that of the surrounding atmos- 

 phere at the same level. 



5. Causes of Convectional Instability. 



As we have just tried to show the sine qui non of the thun- 

 derstorm is the rapid vertical convection of moist air; the up-rush 

 must be rapid and the air must be moist; one without the 

 other is not sufficient. We may have, for example, a very rapid 

 convection over a desert region but there being no moisture 

 there will be no cloud-formation and therefore no thunderstorm. 

 On the other hand we may have air ever so humid but if the 

 movement upward is too gentle not even a cloud may result, 

 but if a cloud, certainly no thunderstorm. It is obvious, 

 therefore, that we must have both "rapid convection" and 

 "moist air." 



This leads us to a consideration of the conditions under 

 which the vertical temperature gradients necessary to this 

 convection can be established. These conditions are, according 

 to Prof. Humphreys, three in number, namely: 



(1) A strong surface heating, expecially in regions of light 



winds. 



(2) The over-running of one layer of air by another at a 



temperature sufficiently lower to induce convection. 



(3) The under-running and consequent uplift of a saturated 



layer of air by a denser layer. 



Of these three conditions, the first mentioned — "strong 

 heating surfaces" — is, for obvious reasons, of most frequent 

 occurrence over the land surfaces of the earth; number two is 

 also of frequent occurrence on land and is, perhaps, well nigh the 



