elevated Rain-gauge, as caused by Wind, 423 



in so doing it is forced against the adjoining parallel stream of 

 air, which must also diverge from the straight direction, and 

 similarly impinge upon the next stream. But the increased 

 pressure produced by the impact causes the streams of air to 

 move more rapidly, and to diminish in thickness at the same 

 time ; and the disturbance of the streams of air will cease at the 

 point where the total decrease of size of the streams is equal to 

 the height of the obstacle. It is at least obvious that when a 

 uniform wind meets an obstacle, some parts of the air must 

 move more rapidly, just as a river moves most rapidly in the 

 narrowest parts of its channel. It is quite in accordance, too, 

 with our common experience, that an obstacle increases the velo- 

 city and force of the wind ; thus the wind is always most fierce 

 at the corner of a house, the end of a wall, or the summit of a 

 hill. 



6. We now have the whole explanation of the rain-observations 

 in question. A drop of rain in falling is influenced at once by 

 gravity and the motion of the air. It describes the diagonal of 

 a rectangle, of which the perpendicular side represents the fall- 

 ing velocity of the drop, and the horizontal side the velocity 

 communicated by the wind. In other words, we may say that 

 the tangent of the angle of inclination (from the vertical direc- 

 tion) of the path of the falling drop varies nearly as the velocity 

 of the wind. 



Now conceive two equal drops of rain falling into a current of 

 air at points where the velocity is not the same. They will not 

 pursue parallel paths, but the one drop will either approach to, 

 or recede from the other. The effect will be to increase or 

 diminish the quantity of rain falling in the intermediate space. 



To show clearly the nature of this effect, we may imagine the 

 stream of air AB in Plate VI. fig. 1, to be suddenly contracted 

 at C D to half its previous thickness, so that of course it must 

 there commence to move with double velocity. At E F the 

 stream dilates to its original size, and of course recovers its first 

 velocity. The course of equidistant rain-drops falling into wind 

 under such imaginary circumstances would be represented by 

 the oblique black lines, and it is obvious that less rain would 

 fall in the windward part of the contracted space than elsewhere. 



7. To represent a real shower of rain falling upon an obstacle, 

 we have only to conceive the drops of rain as falling through a 

 great number of strata, all varying in velocity and thickness. I 

 have thus conjecturaily drawn the full lines in fig. 2 to represent 

 the paths of the rain-drops in a shower falling through wind 

 upon an obstacle such as a house, or tower which bears upon its 

 summit an ordinary rain-gauge. In fig. 3, which is drawn upon 

 a much different scale, the rain-gauge is the only obstacle, being 



