Molecular diffusion, which acts to smooth out gradients in the
concentration of the diffusing quantity, is usually thought of as having
a stabilizing effect. However, the fact that heat diffuses about one
hundred times faster than salt in water allows diffusion to be the
cause of a curious type of instability called salt fingers.
Consider the scenario where warm and salty water overlies cool and fresh
water. Now suppose a small parcel of fresh water is displaced upward into
the salty water. Since the heat diffuses so much faster than the salt,
this parcel soon becomes warm while remaining fresh. Thus, it is buoyant
relative to the water around it, and it rises even further. That is, the
system is unstable to vertical motions. (Note that a similar argument
holds for a salty parcel being displaced downwards.) In practice, many
of these parcels tend to go unstable at the same time, producing
long, finger-like structures. Hence, the name "salt fingers".
Experimental Set-up:
A large beaker (1 litre capacity) was filled to approximately 3/4 capacity
with cool (10 C) tap water. Half a gram of salt was dissolved in 200 ml of
hot (60 - 75 C) water, and a small amount of dye was added. The hot
solution was then carefully poured onto the surface of the cool water in
the beaker. Mixing as a result of the pouring action was minimized by
pouring into a small floating container with a sponge bottom which allowed
the hot solution to seep out onto the surface of the cool water. After a
short period of time, salt fingers appeared spontaneously.
Results:
See below for images and/or movies of salt fingers captured during the
experiment. It was observed that the instability is very sensitive to the
amount of salt in the upper layer (i.e. not very much salt is required
to make the instability appear, and too much will cause the instability to
occur immediately). Also, it was noted that the onset of instability is
most easily seen when the water is motionless, and for this the use of the
floating sponge-bottomed container was crucial.
Experiment performed and written up by:
Paul F. Choboter, Sept. 98,
Thanks go out to Adarsh Mehta for her help in conducting the experiment.
Page constructed by:
Bruce R. Sutherland, Apr. 98,
