422 Mr. W. S. Jevons on the Deficiency of Rain in an 



surface. This paradox has been founded upon the fact that a 

 rain-gauge, when placed at a moderate elevation in the atmo- 

 sphere, is found to collect much less rain than if placed upon 

 the ground. As the sudden increase of rain while it falls 

 through the intervening air cannot be explained in accordance 

 with the known laws of nature, many writers have spoken doubt- 

 fully of this subject, but have brought forward only scanty or 

 palpably erroneous arguments to account for the experimental 

 results. 



2. I now hope to show that the observed differences of rain- 

 fall must be attributed to the influence of the wind upon our 

 mode of experiment. 



3. In observations with rain-gauges at different elevations, 

 the higher gauges have been placed upon the roofs of houses, 

 the summits of church-towers, or other erections which act as 

 obstacles to the wind. It is obvious, too, that the rain-gauge 

 is itself an obstacle, causing the wind to swerve aside, and to 

 change the ^direction in which, the rain-drops fall. 



4. In order to determine the curves which the wind describes 

 in meeting such obstacles, I have performed some small expe- 

 riments. A vessel is formed of two oblong plates of glass, 

 enclosing a layer of air about a quarter of an inch thick. One 

 end of the vessel communicates through a pipe with a chimney 

 or an aspirator, so that a regulated current of air may be drawn 

 through it, to represent on a small scale a section of the wind 

 moving over the surface of the earth. The curves described by 

 the currents ol air are shown very distinctly and beautifully by 

 simply holding a piece of smoking brown paper in the draught 

 of air which is about to enter the glass vessel. We may now 

 place in the lower part of the current any small obstacle to 

 represent a house or a rain-gauge placed in wind, and the curves 

 described by the air will be depicted by the streams of smoke. 



In trying such experiments, it is soon perceived that the 

 curves are similar so long as the velocity of the current changes 

 proportionally to the magnitude of the obstacle; and I am led 

 to believe that the miniature experiment will indicate the course 

 pursued by the actual wind meeting an obstacle, provided that 

 the velocity of the wind and the magnitude of the obstacle bear 

 somewhat the same proportion to each other as in the experiment. 

 From such observations I have drawn the dotted lines in figs. 2 

 and 3, PI. VI. They are intended to represent the course pursued 

 by horizontal strata of air meeting an obstacle, such as a house 

 (fig. 2), or a rain-gauge (fig. 3). Whatever may be the value of 

 the experimental method, it cannot be denied that the air must 

 move somewhat as shown in these figures. 



5. A stream of air, then, meeting an obstacle leaps over it; 



