RECENT ADVANCES IN SCIENCE 203 



where N is the harmonic mean temperature of the outside 

 air, and C is the surface pressure. 



Exact data are not available for the height to which a 

 tornado extends, nor for the reduction of the surface pressure 

 within the core. Six possible cases are considered — namely, 

 where the reduction of pressure is 50 millibars, and the height 

 of the disturbance is 5, 10, or 15 kilometres, and where the 

 reduction is 250 millibars and the heights are the same. 



Average conditions of temperature and pressure give for 



1. 5 km. tornado C=ioo8mb. P= 534111b. N = 270 A. 



2. 10 km. tornado C= 1008 mb. P=26omb. N = 255° A. 



3. 15 km. tornado C=ioo8mb. P=i25mb. N = 230 A. 



and using equation (4) the following figures are arrived at : 



Tornado in which B = C - 50. Tornado in which B = C - 250. 

 Height. N - M. N-M. 



5 km. 2 3 A. 220 A. 



10 km. io° A. 68° A. 



IS km. 5° A. 36 A. 



These figures give the difference between the harmonic mean 

 temperature inside and outside the core, but also give with 

 sufficient accuracy the difference between the arithmetical 

 means. 



The lapse-rate for ascending saturated air is about 6° A. per 

 kilometre less near the ground than it is for dry air, and the 

 two become nearly equal at 10 kilometres height. The maximum 

 difference of temperature that could be established in this way 

 amounts to only about 30 A. Now N—M exceeds this for all 

 possible tornadoes of the class where C — B =250 millibars, 

 therefore these cannot be explained as due to saturated 

 ascending air. The class in which the reduction of pressure 

 at the surface is only 50 millibars is on a somewhat different 

 footing, and might perhaps be explained in this way; but the 

 margin is not so great as it appears to be at first sight, since the 

 atmosphere in reality seldom, if ever, has the dry adiabatic 

 lapse-rate from the ground up to a height of 5 kilometres or 

 more, and a temperature difference of 30 A. could not be 

 established in this way with a less rapid lapse-rate. 



The author suggests, as an alternative theory for explain- 

 ing the warmth within the core, that air occasionally has its 

 electrical conductivity increased over a particular region, 



