(322 Sir Napier Shaw [March 10, 



logical Office for July 27 and 29 which are represented in Figs. 7 

 and 8 respectively. The order of magnitude which is indicated is 

 quite reasonable, and for the one occasion on which the two can be 

 compared the agreement turns out to be exact. That may be fortuit- 

 ous, but we may take advantage of the circumstance to use the 

 combination of the figures for the wind in the stratosphere, and 

 the horizontal temperature gradient at 13 kilometres to compute 

 the latitude of the place of observation Avith an accuracy that may 

 lead us to reconsider the common remark that Meteorology is not an 

 exact science. 



The same equation applied to the troposphere, assuming normal 

 values for temperature, gives correctly the rate of change of velocity 

 luith height as shown in the corresponding diagram. 



5. The j^ermanence of vertical motion about a vertical axis in the 

 atmosphere, which is inclicated hy the long travel of cyclonic depressions. 



From Equation C applied to the stratosphere, it follows that a 

 circulation in the base of the stratosphere with a given horizontal 

 temperature gradient, such as is found there, will have only a limited 

 extension upwards. With a wind velocity of 20 metres per second, and a 

 horizontal temperature gradient of 5' per 100 kilometres, the extension 

 will be 1 • 4 kilometres upwards, so that the vortex will be covered by 

 a cap in which the velocity gradually falls off to zero within a very 

 limited height. 



For the extension downward the calculation is more complicated, 

 but the computed change of velocity is very small, so that the vortex 

 must be regarded as reaching the ground, and it would appear that 

 a vortex extending throughout the troposphere terminating with a 

 cap in the stratosphere is a possible reality. 



Thus the hypothesis of an atmosphere in which the wind velocity 

 is everywhere adjusted to balance the pressure distribution, enables 

 us to explain many of the ascertained facts that have been disclosed 

 by the investigation of the upper air, and strongly supports the 

 hypothesis that the pressure distribution at the surface is controlled 

 by the stratosphere, and only modified locally by the local convection. 

 Against the control of the distribution of pressure by the upper 

 atmosphere may be urged the formation of anticyclones over the 

 relatively cold areas of sea and land, especially the winter anticyclones 

 of the great Continents of the Northern hemisphere. For the local 

 result of surface-cold we have to bring into account the eddy-motion 

 of the surface layer. Some examples are given in the Meteorological 

 Report of the voyage of the Scotia in 1913 by CI. I. Taylor, pubUshed 

 by the Board of Trade in 1914. The formation of horizontal lines 

 or rolls of low cloud in a steady air-current is probably another 

 example of eddy-motion due to the friction of the surface frequently 

 in evidence. 



