Causes of Water Spouts. 53 
I had attached to the middle of the rod a parallelogram of varnish- 
ed pasteboard. (Fig. 3 
The vessel was filled with water to within two inches of the top. I 
placed at the bottom, a layer of blue glass, and on the water a layer of 
oil. ‘The pasteboard, which was to form the whirl, was thus situated 
obliquely in the middle of the water, at an equal distance from the 
powder and the oil. I then turned the rod at the rate of about twice 
in a second, and it was not long before two spouts were formed ; one . 
ascending from the blue glass, the other descending from the oil, and 
uniting at the middle of the axis of the mill. 
From the facts which have just been described, we see that the 
same mechanism, that is to say, the circular movement of a liquid, 
can alone produce ascending or descending spouts separately, or si- 
multaneously, following the position of the whirl; the spout will be 
ascending if it is excited on the surface ; it will be descending, if it is 
formed at the bottom of the vessel ; and finally, a single whirl will pro- 
duce an ascending spout, and another descending, if it is placed at the 
centre of the liquid column. 
It is impossible not to see a striking analogy between the result of 
_ the experiments which have just been described, and the grand natu- 
ral phenomena; in comparing them in the. particular circumstances 
which accompany them, this analogy becomes evident, and must 
produce conviction. It results from the observations of travellers, 
that spouts always take place in a calm, or at most when only a light 
breeze is stirring, and that they are never observed during a great 
storm, which would seem at first more likely to produce them. The 
reason is, that the shock of winds, which produces the primitive whirl 
is always very far from the point where the spout begins to appear. 
When we wish to raise a spout from the bottom of the vessel, we 
create a whirl at the surface of the liquid, and when we wish the spout 
to descend, we form a whirl at the bottom of the vessel. 
In the natural phenomenon, the whirl which produces the ascending 
spout is formed in the clouds, or near the clouds, and extends by de- 
grees towards the sea, favored by the inferior calm ; if a rapid hori- 
-zontal wind existed on the surface of the sea, the whirl in extending 
towards it, would be broken and scattered by the current. When 
ascending water spouts have a progressive movement, the elevated 
region of the atmosphere, where the whirl is commenced, must neces- 
sarily have the same velocity as the inferior part where the spout is 
produced ; otherwise the latter would become oblique and would soon 
