DISTRIBUTION IN THE ATMOSPHERE OVER ENGLAND. 271 



eddies were supposed to be formed is often non-existent, for balloons after rising to a 

 givjit height frequently fall near to their starting point or to the westward. 



It must be borne in mind that the velocity of the air in a vertical direction, save 

 over small areas as in thunderstorms, is very small. It is more conveniently measured 

 in feet than in miles per hour, and the principle of the equation of continuity shows 

 that it cannot exist at all close to the earth's surface. 



The information that has been gathered from the English ascents about the 

 distribution of pressure and temperature in cyclones and anticyclones is shown in a 

 graphical form in fig. 1. 



The figure may roughly be taken as a section of the atmosphere stretching from a 

 region of low (29'00 in.) to a region of high (30'50 in.) pressure, but with the 

 following qualifications. As in all such diagrams the vertical scale is out of all 

 proportion to the horizontal, but also the horizontal scale is not the same in different 

 parts. The horizontal distances represent differences of pressure at sea-level, but 

 since the barometric gradient is never uniform, equal distances on the diagram do not 

 represent equal geographical distances. The gradient is always slight in anticyclonic 

 regions, and steepest in the intermediate regions, and hence the right-hand part 

 requires to be opened out. An attempt is made to correct this in fig. 2, where the 

 horizontal distances are meant to represent geographical distances. How to open out 

 the scale on the sides is a matter of judgment, but the amount of enlargement is 

 based on the fact that the mean temperature shown along any even kilometre height 

 should coincide with that given in Tables II. and III. The distance across may be 

 taken roughly as 1000 miles. 



Fig. 1 obviously fails in this respect ; it is only meant as a representation of the 

 facts disclosed in the tables previously given. In both figures the full lines represent 

 the section of the isobaric surfaces by a vertical plane, and the dotted lines the 

 sections of the isothermal surfaces. The broken lines show the departure of the 

 temperature + or at any point from the mean value for that height. In drawing 

 the lines irregularities that seem to 1x3 due to an insufficient number of observations 

 have been ignored, and the lines are not carried to the ground because the distribution 

 there depends upon the season. If we take a temperature below the mean to indicate 

 that the air there has recently ascended and conversely, the following points appear 

 from the diagram. There is an ascending current in the cyclone starting from close 

 to the ground and reaching up to the isothermal. As the height from the ground 

 increases this current is extended over a larger area and reaches a greater height, so 

 that roughly it forms the frustrum of a cone with its apex downwards. In its outer 

 parts it reaches to about 12 km., and seemingly the air that rises in the centre 

 spreads out still rising obliquely just under the isothermal. Of course, this refers to 

 the conventional cyclone ; the actual cyclone is unsymmetrical, but if we could get a 

 series of simultaneous observations in a straight line running from a low pressure to a 

 high-pressure area, say 1000 miles distant, although there would be many local 



