Irving Langmuir's Ball Lightning Tube

*                                                                     *
* This article from Wayne Strattman at MIT.  The source is unknown,   *
* perhaps it's from Scientific American from the 1920's or 30's.  The *
* photos are unrecognizable, having gone through many generations     *
* of photocopying.  I know of no modern attempts to replicate this    *
* device.                                                             *
*                                         -Bill Beaty  10/95          *
*                                                                     *



BALL LIGHTNING MADE IN THE LABORATORY

The history of modern physics is replete with experiments on gaseous
conduction.  The field seems especially rich in possibilities, chiefly
perhaps because we are there concerned more directly with that fundamental
factor in electrical phenomena- the electron.  Gasses under atmospheric
conditions have comparatively low densities; that is; their molecules are
placed further apart than in solids or liquids, and the electrons have
larger spaces over which to roam aimlessly or to fly rapidly, impelled by
electromagnetic forces.  In rarefied gases, the motions of the electrons
are far more vigorous than in everyday conditions. 

Free for a moment from the influence of the atom, the electron will
display unusual properties and give rise to phenomena whose beauty and
novelty contribute much to the prevalent interests in electrical
conduction through low pressure gases. 

Observe, for instance, the photograph in Fig. 1.  That tenuous electrical
discharge, so beautifully simple, is coupled with myriads of agile
particles that move about, impatiently swaying now under the influence of
electric, now under magnetic forces.  This is no ordinary discharge
through a vacuum, but one actuated by some, as yet unknown, forces.  A
moment before the picture was taken, the arc was stretched motionless
through the tube, glowing with a quiet purplish red color.  Then through a
momentary interruption of current, the arc gains life and tears itself
away from the tube with the writhing motion of a snake, while from a
tungsten filament at the base of the tube melted tungsten is sputtered 
and  (Continued on page 255)

(cont)
bright blue flashes are sent up the vertical stem.

This spectacular and beautiful electrical discharge which gives promise of
leading to a grater understanding of vacuum phenomena was produced in the
research laboatories of the General Electric Company, by Dr. Irving
Langmuir, C.G. Found, and A.F. Dittner, prominent Americal physicists. 

And yet at first glance the apparatus in which these curious phenomena
take place seems to differ only in shape from the ordinary two-electrode
vacuum tube, and essentially is similar to the argon-filled tungar
rectifying bulb.  Fig. 2 shows a sectional view of it.  The large glass
cylinder is 15cm (5.9 inches) long and 10cm (3.9 inches) diameter, and
contains a single loop tungsten filament at each end.  To this cylinder is
welded a slender glass tube 50 cm. (20 inches) long and 3 cm (1.2 inches)
in diameter at the upper end of which a disc electrode is mounted. 

Now the filament is supplied with current at a low potential and is
brought to incandescence at a very high temperature, about 2500C (4532F)
Then 250 volts are applied between the plate and the filament, but as yet
no glow appears in the tube.  The ionization in the tube is insufficient,
but can be much increased by approaching a terminal of a high frequency
coil to the glass tube.  This done, an electric arc of the characteristic
argon color appears in the tube.  The arc draws one ampere and the
potential from the plate to filament drops to 25 volts. 

Nothing unusual is involved in this action, but let the filament circuit
be opened for only one-half second and a sequence of most remarkable
electrical phenomena ensues.  The momentary decrease in filament
temperature causes a rise in the voltage across the arc by reducing for an
instant the electronic emission from the filament.  The voltage rising to
about 100, causes a sputtering of the filament and small quantities of
tungsten are shot into the arc.  Though the tungsten so emitted is only
about .000001 grams, its effects are astonishing.  The arc which before
the "sputtering of tungsten" filled the tube and was insensible to the
influence of a magnetic field, now detaches itself from the walls and can
be attracted or repelled by a permanent magnet held near the tube. 

The arc now has an altered appearance.  It has a central reddish core,
which bears a positive charge and is about 1cm in diameter.  In this red
column positively charged argon atoms are moving and vibrating under the
influence of the electric field between the plate and the filament. 
Immediately surrounding this region is a thin dark space, and beyon this a
bright yellow sheath of glowing gas.  This sheath is negatively charged
and gradually disappears, the central part of the arc increasing in size
until it almost fills the tube.  In place of the yellow "skin," a thin
layer of negative charges remains.  This, together with a positive layer
around the reddish arc, forms what is called an "electric double-layer";
that is, two invisible sheaths extremely close to each other, the inner
one bearing a positive while the outer a negative space charge. 

If an ordinary horseshoe magnet is brought near the tube, the arc is
deflected, as is any conductor carrying a similar current.  At the same
time the yellow skin appears on the opposite side of the arc on the side
not in contact with the wall. 

As the magnet is brought nearer, the yellow skin becomes brighter and
thinner, and curiously begins to act like a liquid.  Slowly, tiny drops of
golden yellow liquid fire are formed.  They move along the surface, only
to break away and fall, molten spheres of bright white light into the arc. 
By regulating the intensity of the magnetic field, these droplets or
globules, ranging from a few tenths mm up to 5 or 6 mm in diameter (about
the size of a pea) can be made to form slowly and detach themselves singly
from the skin of the arc.  By proper combination and longitudinal and
transverse fields the globules may often be made to move upwards or
downwards in the arc parallel to its axis for distances ou to 5 or 10 cm,
2 to 4 inches.  Under certain conditions the globules have been observed
to move very slowly so that their motions through the arc could be easily
followed by the unaided eye.  But more often they move with the velocity
of about a foot per second, and thus appear as brilliant lines or
filamentary streamers.  See Fig. 3 where several streamers with nearly
parallel paths are shown. 

These streamers are formed by the influence of the magnetic field, and
vary in appearance with the magnitude of the current flowing through the
arc.  If this current is varied, the motion of the streamers will be
correspondingly affected.  Thus, by superimposing an alternating current
on the direct current fed to the anode, that is, the disc electrode, the
streamers of individual globules will move in a sinusoidal path.  That is,
they will appear as a sine wave which is characteristic of alternating
currents.  The streamers will reproduce accurately the wave shape of the
current even (Continued on page 284)

(cont)
at frequencies up to 1000 cycles.  A similar effect is produced by using
alternating current for heating the filament.  In either case the globule
streamer assumes a wave motion, travelling across the arc in beautifully
sharp curves, rather than in straight lines. 

If the arc current is maintained, this effect continues for hours before
the minute portion of tungsten ceases to act.  The effect can then be
brought back by again cutting out the filament current for a moment and
thus sputtering another trace of tungsten into the argon vapor.  If the
arc current is shut off for about 40 seconds, the effect disappears- the
tungsten has been deposited and the streamer discharge will not start
again until more of the metal has been sputtered from the filament. 

According to this theory, tungsten atoms, negatively charged, roam about
in the spaces outside the arc.  These atoms entering the "electric double
layer" lose their charge and collect on the positive sheath, forming small
globules.  At these points on the double layer the negative sheath is
indented, and when sufficient tungsten accumulates, this indentation
becomes sharp and will extend into the arc until the globules tear away. 
By the nature of their formation the globules will thus have a positive
sheath on the outside and a negative sheath on the inside; and arrangement
the reverse of that found in the arc proper. 

These glowing detached globules seem to have characteristics similar in
many respects to those that have been described as belonging to ball
lightning.  It is perhaps not certain that ball lightning is anything more
than a psychological phenomenon, but if it has objective reality, it may
possibly be due to causes similar to those which give rise to the globules
described above.  The ions of a highly ionized gas, such as the
electrified atmosphere, recombine on solid particles forming small spheres,
the solid particles being retained within the ball by their charges and
the electric field being retained at the surface of the ball.  This theory
is very effectively supported by the marked resemblance between ball
lightning and the globules of liquid fire produced in Dr. Langmuir's
experiments. 


Fig. 1
Mr. Found of the General Electric Experimenting with the streamer
discharge, has the miniature ball lightning formed in the tube under
perfect control.  Under the influence of the horseshoe magnet, these balls
of fire can be made to move up and down the tube. 


DRAWING OF VACUUM TUBE

Fig. 2
Above: A cross-sectional view of the discharge tube in which the
phenomenon of ball lightning was observed.  The tube is filled with argon
at very low pressure, and with the filament heated to incandescence, a
reddish arc is formed between the plate and the terminal. 


Fig. 3
Left: The agitation in the argon arc which is writhing like some strange
electric snake, is caused by a sputtering of melted tungsten from the
incandescent filament at the base of the tube. 


Fig. 4
The argon arc shown at left is deflected with a magnet.  The transverse,
filamentary streamers are in reality small globules of incandescent
tungsten moving rapidly across the arc.  The globules of liquid fire are,
in formation and behavior, very similar to the incandescent spheres of
ball lightning.  These unusual effects are produced by only .000001 gram
of tungsten emitted by the filament. 

Fig. 5-1
Fig. 5-2
Fig. 5-3
Fig. 5-4

In a tube filled with argon gas small globules of liquid fire have been
formed, which bear a strong resemblance to ball lightning.  A magnet draws
the argon arc to one side of the tube and negatively charged tungsten
atoms emitted by a filament move into the arc (Fig. 5-1) Near the edge of
the arc they cross a region called the electric double layer bearing
opposite charges on the two sides.  Here the atoms lose their charge and
collect in groups (Fig. 5-2) The neutral tungsten atoms thus form small
masses of liquid tungsten on the border of the arc, and the layer of
negative charges bends into the arc (Fig. 5-3.)  This indentation continues
until becoming very sharp the tungsten globule becomes detached and forms
a rapidly moving miniature ball lightning (Fig. 5-4.)