370 C. Abbe — Atmospheric Radiation of 



and prolong the meiobars and therefore the overlapping peri- 

 cyclones of Ferrel or vortices of Helmholtz, thus eventually 

 intensifying the pleiobars. 



{g.) According to these views the primary factors in the 

 origin and maintenance of local rains, water-spouts, thunder- 

 storms and tornadoes are local buoyancy, expansion, thermody- 

 namic cooling, condensation and latent heat. On the other 

 hand in extensive storms of the temperate and arctic zones it 

 is the slight depression, due to a whirl in the general circula- 

 tion, that determines where a storm or low area will originate. 

 Finally in tropical hurricanes the general' circulation must 

 decide whether an initial local disturbance shall die away or 

 shall be fed and developed into an extensive hurricane. 



(h ) So mobile is the air that two conflicting currents cannot 

 produce any considerable excess or deficiency of pressure and 

 such a mechanical heaping up, or compression, as would pro- 

 duce a barometric gradient of 0"l mm per degree is immediately 

 followed by a movement in the air that will relieve the excess. 

 Therefore, as Ferrel distinctly shows, we must not expect that 

 any obstruction to the moderate currents of the general circu- 

 lation can explain such great barometric elevations as occur in 

 the high areas of the United States or Europe. A slight 

 mechanical excess of pressure may possibly initiate a high just 

 as a deficiency does a low, but the further development of the 

 pressures must depend upon other considerations. Among 

 these latter Ferrel so long ago as 1856 showed that just as the low 

 pressure depends upon the centrifugal forces due to a cyclonic 

 whirl so the high pressure depends, at least in part, upon a 

 centripetal force due to an anti-cyclonic whirl. 



Again Ferrel has stated that if the cold of the lowest layer 

 in the high area extends upward then the weight of this dense 

 air would help to explain the high pressure. This suggestion 

 needs careful consideration in view of the observed presence of 

 warm air in the upper layer where cold air was supposed to 

 exist. Although Ferrel maintained that the air in the high 

 areas must on the average be colder and therefore be descend- 

 ing by virtue of its own gravity yet he has not made clear the 

 details of the process by which descending air can become 

 cooler and denser. 



(*.) The existence of warm air above the cold layer near the 

 ground has long been recognized for moderate elevations dur- 

 ing clear nights ; its influence upon the atmospheric refrac- 

 tion is apparently often very notable in astronomical observa- 

 tions. The fact of its existence has also been brought out, 

 even for the daytime, by many balloon observations notably 

 those of Glaisher in England and of Mr. Hammon in the 

 United States (see the Am. Met. Jour., Feb., 1891.) To the 



