METEOROLOGY. 495 



Ocean it has in fact been observed that above the trade winds more air 

 is flowing at high altitudes (toward the pole) in that hemisphere in 

 which winter is prevailing. Toynbee says that during the northern 

 winter and spring the upper clouds above the trade winds move north- 

 ward more frequently than in summer and autumn. The reverse is 

 true during the winter of the southern hemisphere. 



" If the temperature of the tropic zone is subject to periodic changes, 

 as Koppen has shown is probable, since periods of greater intensity of 

 solar radiation follow parallel with the periods of greater or less fre- 

 quency of solar spots, then must similar changes also exist in the gen- 

 eral atmospheric circulation. At the time when the temperature of 

 the tropics attains a maximum the descent of air in the upper strata 

 from the tropics towards the pole is greatest and the circulation most 

 intense. The low pressure in the equatorial belt will experience a fur- 

 ther diminution, while the high pressure in the middle latitudes will 

 increase, and in the circum polar regions the atmospheric pressure, in 

 consequence of the increase in the rotating ring (vortex ring), will fur- 

 ther diminish. The whole energy of the atmospheric movements will 

 increase, but what influence this may have upon the temperature 

 and precipitation in the middle and higher latitudes cannot easily be 

 deductively inferred. On this point only investigations based upon 

 sufficient observations can give satisfactory conclusions. 



" Recently, Blanford has called attention to a relation of this char- 

 acter between the distribution of pressure in the tropics and in higher 

 latitudes, in that he shows that the periods of low pressure in Indo- 

 Australian tropical regions coincide with periods of high pressure on 

 the northern Asiatic continent. Especially is this the case in winter. 

 The air that flows at high altitudes out of the tropics streams towards 

 and accumulates where the descent in the higher strata (the thermic 

 gradient) is greatest. Since high atmospheric pressure (a maximum 

 barometer) is associated with greater and permanent clearness of the 

 sky, therefore also with greater radiation of heat outwards and cool- 

 ing of the lowest strata of air in winter (origin of a local pole of cold), 

 whereby conversely the thermic gradient and the inflow of air from 

 above is again increased — it is thus quite possible that periods of great 

 warmth and low pressure in tropical regions should correspond with 

 periods of great winter cold in high latitudes. 



"The discovery of such relations between the temperatures of the 

 tropical zone and the general weather of the middle and higher latitudes 

 is one of the most important present problems of meteorology and com- 

 parative climatology. We have already seen that the influence of the 

 tropics is actively opposed to that of the temperate zone of either hem- 

 isphere in its winter season in the ratio of 10 to G.5. The mean condi- 

 tion as to temperature in the tropical zone on which depends the energy 

 of the upper currents of air which become the prevailing ones in the 

 temperate zone must, therefore, have the greatest influence upon the 



