2 Q 8 



SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. 51 



change of condition is shown by the broken line a b c; if now it sinks 

 from some altitude that cannot be shown in the diagram and after all 

 the water has fallen from it, then the increase of temperature with 

 descent follows c d, or the adiabat of the dry stage. The nick at b 

 lies lower in proportion as the air is moister when the ascent begins 

 and it is sharper, or the first part of the curve b c rises more steeply 

 in proportion as the initial temperature is higher. 



If air alternately and for equal intervals of time rises according 

 to the la\v a b c and descends according to the adiabat c d of the 

 dry stage, then we obtain the average curve of condition by halving 

 the horizontal lines between the two curves and joining all the half- 

 way points, If t t and t 2 are the temperatures corresponding to the 

 points 7\ and T 2 , then the mean temperature t m for the altitude 

 h is given by 



* m - i & + '2) 



and the curve of average condition is represented by the median line 

 c m. 



Since now in general the currents ascending above any place will 

 have very various initial temperatures and humidities, therefore 

 the average of all must give curves whose a c branches correspond 



fig. 40 



only in general to the form of the line a c but individually show 

 great diversity. On the other hand the average curves correspond- 

 ing to the descending branch will, under the adopted assumption of 

 adiabatic descent, run parallel to the line d c but cut the axis of 

 abscissae at very different places. 



Hence as an average curve of condition of the vertical column 

 there results a curve that must have approximately the course 

 shown by m b m c in fig. 40. The lower part of this curve is dotted 

 for a reason that will be explained at once. 



In the views that have just been elucidated we had to consider 

 that at very great altitudes all the adiabats of saturated air become 



