78 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



Another important question is that of the relation between the 

 vertical distribution of pressure and the motion of the air in the 

 vertical direction. If there are no motive forces present except 

 the difference of pressure, then equation (26) is to be applied in this 

 case and for v = o it becomes: 



r.r* --*(, + £) (29 , 



d 2 h 



It will be assumed that - = b is a constant quantity. Proceeding 



in a method entirely similar to the preceding we finally get 



_^ 

 Bo -f B <>\ g (30) 



B \B 



where B is the value of the barometer-reading B at the lower level 

 under the condition of uniform motion throughout the whole mass. 

 If, for example, it is asked how great b will become in the case of the 

 values used above for B and B when the relation B X 748.1 = 

 B X 748.0 exists (when therefore the difference of pressure is about 

 o.i mm greater in the condition of accelerated motion than in that 

 of rest or uniform motion), the result is, there will be an upward 

 directed acceleration of b = 0.007 meter per second. 



If the air has simultaneously an east-west component of velocity 

 to the amount of 25 to 30 m.p.s., then the diminution of the pres- 

 sure from below upwards will become about 0.10 4- 0.04 = o.i4 mm 

 greater than in the condition of rest. 



If in any manner whatever an increase of the vertical difference 



icssure of o.i mm should be brought about and maintained, 



for example, b)^ an upward directed removal of air at any altitude, 



then an ascension of the air must take place and by integration of 



d 2 h 

 the above equation — = b for uniformly accelerated motion, the 



velocity which a particle of air attains in passing over a distance 

 h — h of 1$ km., can be deduced. As b is assumed to be con- 

 stant there results: 



dh [ 



2b(h-h ) 

 at * 



For the above value of b = 0.007 an d f° r h~ K = I 5°° m we obtain 

 dh = 4.58 111 per second. In this vertical motion, as is known, 

 there appears again a horizontal component of motion in conse- 



