170 DYNAMIC METEOROLOGY AND HYDROGRAPHY. 



interpret the chart as representing vertical specific momentum in the height where 

 the pressure is one unit smaller than at the ground, or vertical velocity in unit 

 height above the ground. If we use the latter interpretation, the numbers added 

 to the curves give the vertical velocity in millimeters per second at the height of 

 100 meters above the ground, and in centimeters per second at the height of 

 1000 meters above the ground. The chart will give the correct picture of this 

 part of the vertical velocity, provided that the chart of plate XXXII represents 

 the average horizontal motion for the sheet between the ground and these heights. 

 For a wind which has the regularity of the monsoon, it is not improbable that the 

 observations at the ground give the character of the motion up to considerable 

 heights. But decided exceptions exist. Thus, if the line of convergence in the 

 Ganges valley existed unchanged to the height of 1000 meters it should give here 

 a vertical velocity of 9 cm. per second, and a corresponding greatly localized precipi- 

 tation might be expected. But as Sir John Eliot's chart of precipitation for July 

 does not show any sign of this, we have a strong reason for believing that the line of 

 convergence is a local phenomenon limited to the lower layers. (Compare section 134.) 



Comparing the plates XXXIII and XXXIV, we see that the free vertical motion 

 has a certain tendency to be of an opposite sign to the forced vertical motion existing 

 at the ground. The addition will therefore in most places give a reduced vertical 

 motion. Lower down the forced vertical velocity is the stronger of the two. But 

 at the height of 1000 meters both are of about the same order of magnitude and as 

 we proceed farther upward the influence of the ground will constantly recede to 

 the background. 



From the two charts XXXIII and XXXIV we can derive charts for the total 

 vertical velocity at any constant height above the ground by graphical addition. If 

 we wish to have the total vertical velocity at a given height above sea-level we 

 must, before the addition, perform the graphical multiplication of chart XXXIV by 

 a chart which represents the height from the ground to the given level. It is interest- 

 ing to draw such charts of total vertical motion and to compare them with charts of 

 average precipitation like those found in Eliot's Atlas. But in a case like that before 

 us no complete accordance should be expected. We have referred already to one 

 departure, the reason of which is easily understood. Another cause of departures 

 is this: In spite of its great regularity the monsoon- wind shows changes from day 

 to day, causing corresponding changes from day to day in the distribution of the 

 vertical motion. For this reason there will from time to time appear ascending 

 motion and consequently precipitation in places where the average motion is descend- 

 ing and where no precipitation would appear if there were no departures from the 

 average motion. 



201. North America, 1905, November 28, 8 a. m. Instead of average motions 

 we shall henceforth consider actual motions. 



Plate XXXV represents the field of pressure and of mass in the lowest atmos- 

 pheric sheet above North America, November 28, 1905, 8 a.m., time of 75th meridian. 

 The single lines give the absolute topography of the 1000 m-bar surface, and the 



