THEORY OF CYCLONES VON BEZOLD 351 



(B) Cyclonal rotation with gradients directed outward, i. e., 

 with anticyclonal distribution of atmospheric pressure. These con- 

 ditions are met within the upper strata of cyclones with warm cen- 

 ters. Here the gradient is directed outward but the curvature of 

 the orbits of the particles of air must be cyclonal up to very con- 

 siderable altitudes, since the movement of rotation that the mass of 

 air has under the ordinary conditions brought up from the lower 

 strata cannot immediately disappear. 8 



But under these conditions the whirl cannot be centered, since 

 the equation appropriate to this case, viz: 



p. + p< + r-o (3) 



cannot be fulfilled; hence perfect equilibrium must prevail, that is 

 to say, each of these three quantities must be equal to zero. 



(C) Anticyclonal rotation with gradients directed outward hence 

 with anticyclonal distribution of pressure. These are the conditions 

 that we usually meet with in the lower part of the anticyclone. 



In this case the equation of condition for the centered whirl is 



Pc-Pi + r = t> (4) 



Although theoretically this is not impossible yet this equation may 

 still be practically meaningless, since the relations in the lower 

 part of the anticyclone are always such that an agreement of wind 

 direction with the isobars is not imaginable. Thus there remains at 

 best only the very highest portions of the cyclones with warm cen- 

 ters, in which, indeed, anticyclonal distribution of pressure must 

 prevail and where perhaps anticyclonal movements of the atmos- 

 phere can also be present, provided that this system extends so 

 far upward that the moment of rotation in cyclonal direction as 

 brought up from below is already completely consumed in over- 

 coming the resistances. 



But since we have no basis of facts for the investigation of this 

 question it will be better to lay it entirely aside. 



8 The scheme of atmospheric motions in the upper portion of a cyclone 

 deduced by Clement Ley (Quarterly Journal Met. Soc. London, 1877, III, 

 p. 437) from observations of the cirrus clouds corresponds to this presentation 

 of our second case. We attain the same result if we think of a cyclonal move- 

 ment in which the paths of the air particles are more and more straight- 

 ened out by the forces p c , p{ and r all acting in the same direction until 

 finally the paths are curved in the opposite or anticyclonal direction while 

 the velocity of the outflow increases and at the same time the whole system 

 is carried eastward in the great whirl of the polar region. 



