120 



SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. 51 



second for anticyclonic motion. The following table gives the 

 values of v for different latitudes in the three cases: 



By the use of tables giving values of v for different values of x, 

 T, B, and of v for different values of X, v , <p, each wind observation 

 at 1000 meters altitude was compared with the value deduced from 

 the surface isobars. The temperature correction was not applied. 



The following table gives the result of the comparisons: 



Berlin: 



January i5.7 



February 12.0 



March 



I.indenberg: 



April 



May S.3 



June 6 



July 8.4 



August S.9 



September 10.3 



October 12. 1 



November 10.6 



December 1 1 . o 



Summer 8.8 



Winter 11. 6 



Year 10. 2 



The upper wind coincides in direction very nearly with the isobars 

 at the surface, and the wind velocity observed agrees well with that 

 calculated from the pressure distribution. The differences are not 

 greater than possible errors of observation, except in spring. 



It is known that the upper wind always veers from the surface 

 wind, and the numbers in Column 7 show that in 1905 the veering 

 was considerably greater in winter than in summer. 



If the effect of the earth's surface were the same as if a frictional 

 force opposed the motion, the relation between the wind and grad- 



