814 
ON THE DECEEASE OF EAIN 
that a local acceleration will take place at the constricted part, 
the current there having double the velocity of the current in the 
other parts of the same tube. There is thus shown, as the 
result of the obstruction, a general retardation, combined with a 
local relative acceleration. And, in theory, the same thing may 
happen with a building on which a rain-gauge is placed, or even, 
to a very small extent, with a rain-gauge itself. 
But the extent to which this cause may operate is an 
important point to consider. It must be remembered that the 
air about a building or a gauge is free, not confined within rigid 
walls, as in the constricted tube that I have supposed, or, as in 
Mr. Jevons’s own illustration of a river between its banks. 
And, bearing this in mind, I should be disposed to conclude that 
the accelerating effect of an obstacle such as a rain-gauge upon 
the velocity of the wind in its neighbourhood must be 
infinitesimally small. 
Again, it will appear from the diagram, that it is only those 
drops which are deflected along a vertical line (or a line more or 
less approaching the vertical) that will suffer the separation in 
horizontal distance which the theory supposes. And if the 
imaginary area above the obstacle be reduced in height, the effect 
will be proportionately diminished. 
Further, if the theory be true, this curious result must 
follow — that in a driving rain the fall on the leeward side of a 
house, or similar obstacle, will be greater (at least over a certain 
space) than the fall on the windward side. Having no obser- 
vations to determine this point, I can hardly use it as an 
argument, although I think it will be allowed that the result, 
if established, would be very surprising. 
But I have a more conclusive objection to Mr. Jevons’s theory 
.than any I have yet advanced. There appears to me to be an 
essential link wanting in the chain of argument by which the 
author would connect the alleged acceleration of wind with the 
