IO DYNAMIC METEOROLOGY AND HYDROGRAPHY. 



to averages obtained by the best self-recording instruments. Greater homogeneity 

 is highly desirable. The best wind-observations which we have had at our disposal 

 are from the United States. We have obtained them partly from the published 

 weather maps, and partly from unpublished material, thanks to the kindness of the 

 United States Weather Bureau. They give the registered average wind-velocities 

 from hour to hour, and eight corresponding average wind-directions. We have con- 

 sidered these averages as defining the vector-average of the wind for the half -hour 

 epoch, though for periods of rapid changes they may differ considerably from the 

 true vector- average. Quite independently of the method of averaging, it would be 

 a great improvement to increase the number of wind-directions noted from eight 

 to at least sixteen. 



98. Preliminary Synoptic Representation of the Wind-Observations. A set 



of simultaneous observations of the wind being given, the first step in the subsequent 

 diagnostic work will be to introduce these observations in a convenient form on 

 the map. 



The most direct way will be to draw on the map a set of arrows representing 

 the observed wind-directions, and to add numbers representing the observed wind- 

 velocities in meters per second. These numbers, giving the result in quantitative 

 form, should always be introduced instead of the different qualitative signs used to 

 represent the strength of wind according to the different "wind-scales." Plates 

 XXXI, XXXVI, and LIII give examples of charts containing in this way a repre- 

 sentation of wind-observations. 



Besides representing the directions by arrows, we shall use a method of rep- 

 resenting them by numbers. This will be useful not only for purposes of regis- 

 tration, but also for quantitative work. The correspondence between directions 

 and numbers which we shall use is illustrated by fig. 32. 



The numbers defined by this figure may be used not only as names of the direc- 

 tions, but also as measure of the angles which the different directions form with the 

 initial direction, the direction toward E. We get in this manner a measure of the 

 angles by dividing the circle into 64 instead of 360 degrees. We have chosen this 

 measure of angles for our purposes by two reasons; first, 64 is the highest two- 

 figure number which is a power of 2 ; and then its tenth part, 6.4, differs only by 

 1.9 per cent from 2t or 6.28. This difference will as a rule be insignificant for us. 

 We can therefore consider the numbers 1 to 64, after division by 10, as representing 

 the angles in absolute measure. The choice of the direction E. N. W. S., i. e., the 

 direction against the motion of the hands of a watch, as the positive, and the direc- 

 tion toward E. as the initial direction, is made for reasons which will be apparent 

 later, when we shall choose our systems of coordinates and give the corresponding 

 rule of signs. 



When this correspondence between numbers and directions is used, it will be 

 found convenient to have the diagram of fig. 32 engraved on a transparent sheet of 

 glass or of celluloid. By use of this divided plate we can then easily pass from an 

 arrow to the corresponding number, or vice versa. 



