About three percent of the general population have reported seeing ball lightning at one time or another.  My father described seeing a glowing red ball, about a foot in diameter, on a Halloween night, and I believe I've seen a smaller version of it myself.   There have been a number of good photographs taken of ball lightning taken in various places, including Nebraska (1933), New Mexico (1961), and Japan (1987).

 They have most often been reported as occurring right after a normal lightning strike. With some balls, no thunderstorms were present at the time.  Pressure on certain rock crystals can generate electricity.  Some scientists feel that geological forces within certain rock formations may also generate ball lightning.  Such balls of energy have often been reported in close association with earthquakes.

 There seems to be indications the balls are basically electrical in nature, somewhat similar to normal lightning.  In a Paris sighting, it was reported that a glowing red ball hovered above a tree, which suddenly caught fire.  As the ball violently exploded, it shot out several simultaneous lightning bolts.  The explosion was said to knock down several witnesses in the vicinity.

 One theory is that they are comprised of plasma.  Plasma is comprised of charged atoms, which can be contained inside a magnetic field.  Some experimental fusion reactors contain hydrogen plasma at over a million degrees Fahrenheit.  Some feel that ball lightning may be plasma contained inside a hole within the surrounding energy field, perhaps something like the electrical equivalent of a tornado.

 The balls are reported as containing very high energy levels.  When a ball the size of an orange fell into a rain barrel, a large part of the water boiled away.  Calculations showed energy required would be equal to about three pounds of high explosives.  Most balls do not appear to dim with time.  Some feel that they must somehow draw energy from the environment to replace the continual losses somehow.

      Many scientists accept that ball lightning can occur in sizes up to three feet in diameter, and a few accept sizes up to six feet.   It is not clear why even larger sizes are not generally accepted.  Glowing spheroids are most commonly reported in the range of six to fifteen feet in diameter.  Doctor Condon, chairman of the University of Colorado study, apparently felt that even larger sizes of ball lighting might sometimes exist.