TRANSACTIONS OF SECTION A. 367 



2. X-Rays and Crystal SinicUire. 

 By Professor W. H. Beagg, F.R.S., and Mr. W. L. Bkagg. 



3. Optical Dispersion as conditioned by Group Velocity. 

 By Professor Sir J. Larmoe, F.R.S. 



4. New Views of Magnetism. By Professor Pierre Weiss. 



5. The Mechanism of Cyclones. By F. J. W. Whipple. 



The distribution of pressure and temperature in cyclones in the Temperate 

 Zone has been learned from the analysis of the records from the meteorographs 

 carried by pilot balloons. Up to a height of eight or nine kilometres the cyclone 

 is composed of air cooler than its surroundings; at greater heights, i.e., in the 

 stratosphere, the cyclone contains comparatively warm air. The lower limit 

 of the stratosphere is depressed in the cyclone. This temperature distribution 

 indicates that the air constituting the lower part of the cyclone has recently 

 ascended, whereas the upper air has recently fallen, and accordingly the arrival 

 of a cyclone is marked by an outflow of air at the bottom of the stratosphere 

 and an inflow below. At the beginning of the present paper the amount of 

 this displacement of air is estimated on the assumption that there is no direct 

 exchange of heat, and it is shown that the outflow is concentrated between the 

 seventh and tenth kilometres and is about 6.5 times the net loss of air as 

 estimated by the fall of pressure at the earth's surface. 



It is pointed' out that a cyclone may be regarded as a disturbance in the 

 stream of air which flows from West to East in the Temperate Zone, and the 

 form of the isobars obtained by superimposing the permanent pressure distribu- 

 tion and the temporary cyclonic distribution is discussed. It is shown that 

 when due allowance is made for the curvature and the progressive motion of 

 these isobars, the gradient wind at certain heights is much less than it would 

 have been if the curvature were inappreciable, so that at these heights the air 

 supply from the rear to the front of the cyclone fails and the cyclone appears 

 to move under the influence of suction applied at the base of the stratosphere. 

 The explanation may be summarised as follows : — 



If the flow of air at any level were entirely horizontal and along the isobars, 

 and if changes of density were negligible, then the condition for continuity 

 would require the velocity to be inversely proportional to the distance between 

 the isobars, i.e., the velocity would be directly proportional to the pressure 

 gradient. This condition is not satisfied, however, in regions where the air- 

 trajectories are curved. The pressure has to produce the centripetal accelera- 

 tion in the curved path in addition to overcoming the tendency to turn to the 

 right, which is the feature of all horizontal motion in our hemisphere. Accord- 

 ingly the actual velocity where the isobars are curved is less than it should be 

 to secure continuity and maintain a stationary distribution of pressure. The 

 effect of curvature in reducing the velocity is greatest at the heights where the 

 winds are strongest, and therefore the suction effect is concentrated near the 

 base of the stratosphere. 



The general argument is supported by the analysis of two special cases. 



