37° SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



Hence follow the following important consequences, which 

 are not always sufficiently kept in mind : 



"For equal values of the gradient G the gradient-acceleration 7- 

 is inversely proportional to the density of the air, hence it increases 

 with increasing temperature and with diminishing pressure." 



Of course at the surface of the earth this difference in the density 

 of the air is not generally of much importance, especially so long as 

 we consider only a small part of the surface. But if we do not thus 

 limit ourselves then its influence can in extreme cases certainly 

 amount to more than 30 per cent. 



Assume, for instance, that at some place on the earth's surface 

 the temperature is 37 C. and the barometric pressure 7io mm , such 

 as can happen in whirlwind storms, but that at another place we 



T 



have — 33 C. and 78o mm , then the values of the quotient 7. at the 



P 

 two localities will have the ratio 44/31, so that for equal gradients 



the gradient-acceleration in the neighborhood of the highest pressure 

 will amount to only 70 per cent of that within the low pressure. 



Even for smaller regions, such as those covered by our ordinary 

 weather charts, this influence can be considerable. 



For instance, assume that in the center of a depression of 715mm 

 the temperature is i2°C, but in that of a maximum of 77 5 mm on the 

 same weather map the temperature is — 33 C, then the ratio of the 



T. 



two values of r is as 100/77 an d for equal gradients the accelerations 



at the two localities would have the same ratio. 



Since now the examples here chosen, although slightly exagger- 

 ated, reproduce the conditions that are ordinarily observed in low 

 areas with warm centers and high areas with cold centers, therefore 

 we perceive that in general "the gradient-accelerations" near baro- 

 metric maxima would be greater than would be expected from 

 the gradient itself. It is only in the case of depressions with cold 

 centers that a partial compensation of the rarefaction of the air 

 due to the low pressure is brought about by the low temperature. 

 Therefore the lesser density of the air, such as generally occurs in 

 the lows, contributes still further to increase the velocity of the wind 

 in this region independent of the closeness of the isobars, i. e., inde- 

 pendent of the strong gradients. 



"Since now in front of the cyclones, where the air flows in from 

 the equatorial side, the temperature averages higher than in the 

 rear, so also in general for equal atmospheric pressure, i. e., along 

 any one isobar, the density of the air in front of the depression is 



