810 
ON THE DECREASE OF RAIN 
quantity of rain wTiicli enters into the rain-gauge is proportional 
to the sine of the angle of inclination of the rain.” 
Mr. Stow, in order to test his theory, instituted an elaborate 
series of experiments with a view to determine the angle at 
which rain falls. He fixed at various elevations pairs of rain- 
gauges, one of each pair being an ordinary gauge with a horizontal 
receiving surface, and the other being a gauge with a vertical 
receiving surface kept always towards the wind by the action of 
a vane. From the proportional quantities collected in each pair 
of gauges he computed the angle at which the rain fell at each 
elevation. As might have been expected, he found that the 
higher (and therefore more exposed) gauges indicated a greater 
inclination of the rain than the lower, and he clearly established 
the law that the quantity of rain collected at the different heights 
varied directly with the angle which the paths of the rain-drops 
formed with the horizon. Further, he found, in comparing the 
ratios of decrease at different periods, that the angle of the falling 
rain, depending, as he rightly said, on two factors, the force of 
the wind and the w^eight of the drops, exhibited a closer cor- 
respondence with the ratio of decrease than did the force of the 
wind taken alone. I hope to show immediately how this may 
be explained without having recourse to a theory which, however 
nicely it fits in with observed facts, is mathematically untenable. 
That the theory is untenable will, I think, appear pretty 
clearly on inspection of the diagram (fig. 1). Let the lines 
in the upper part of this diagram be taken to represent the 
paths of the drops of rain as they fall vertically through 
a calm stratum of air, and let the inclined lines represent 
the paths of the same drops as they fall through a stratum 
of moving air. Let A and B be rain-gauges of identical 
size, placed, one to catch the rain as it falls vertically, the 
other to catch the same rain as it falls obliquely. It will be 
very evident that as the diameter of the receiving surface of the 
