48 Causes of Water Spouts. 
Instead of attributing the formation of water spouts to this cause, 
it would be more natural to suppose that electricity, when manifest- 
ed in their appearance, is developed by the motion of the air which 
occasions them. pis. 
But how shall we form an iden of this movement? How, from a 
cause of that nature, give a reason for two effects which appear oppo- 
sed to each other, that of descending and ascending spouts. May 
we not, without confining ourselves to conjectures without proof, sub- 
ject the phenomena to a certain extent, to experiment, and judge 
from analogy, of the nature of the movement which takes place in 
great whirlwinds, by that which we may Serve: in circular move- 
ments in other fluids, of small dimensions. 
pposition that the motion of air alone in hisbarads is 
suflicier © produce water spouts, the same motion in liquids ought 
to produce analogous effects, and if a whirling motion be excited in 
a light liquid placed on a heavier one, the latter ought to ascend in 
a spout into the former, as is the case in the natural phenomenon : 
this reasoning led me to the following experiments. 
Fig. 1. Fig. 2. 
First Experiment.—Into a cylindrical vessel ten inches high and 
four in diameter I poured water to the depth of two inches and then 
filled up the vessel with very transparent poppy oil. At the surface 
of the oil I adjusted a little mill two inches high and an inch and a 
half wide, the axis being vertical and the branches, four in number, 
