The witness to this unusual experience is J.L., a farmer who was in his late fourties when the incident took place. The location: a small French village called La Feuillade in the department of Dordogne.

We translate from an interview conducted in 1972 by the now defunct French UFO group G.A.B.R.I.E.L.:

The investigators asked several more questions about the aspect of the spheres and came up with the following description [the specifications between straight brackets are ours]:

It formed a whole of about fifty dull spheres, grey like the ashes of wood, each with a diameter of 30 to 40 cm [1 to 1.3 feet]. They took the shape of a T-formation, with the horizontal crossbar first. In front there were at least three rows of spheres, each row measuring 2 to 3 m in length [6.6 to 9.8 feet]. Behind there were several bands with more spheres over a length of 3 to 4 m [9.8 to 13.1 feet]. These spheres may have touched one another, or they were separated by a very small space (5 to 10 cm) [2 to 4 inches]. They all moved in the same direction, like marbles rolling down a slope. All where of the same size. The apparition was silent and lasted 2 to 3 seconds.

With regard to the black, diamond-shaped cloud, the witness specified that it was "close to the ground" and "dissipated at the spot".

- G.A.B.R.I.E.L., "Les Soucoupes Volantes: LE GRAND REFUS ?", Michel Moutet  Editeur, Régusse, 1978, pp. 226-231.

- D'AIGURE, Jan, "Pleins feux sur la Creuse" in Lumières Dans La Nuit, No. 131, January 1974.

Dull grey spheres, tight together, rolling through the sky in formation? This strongly reminds us of one of nature's most spectacular cloud types: mammatus.

Mammatus are pouch-like cloud structures that usually form underneath the edge of a spreading anvil of a cumulonimbus cloud. The rounded protuberances are sometimes compared to "big puff balls". They are caused by the downward movement of the cloud that is compensated by the upward moving of dry, cloud-free air underneath.

The grouping of mammatus cells in parallel strokes or bands is a common characteristic of mammatus clouds. A fine example of such a banded structure can be found in our picture galleries.

That a cloud formation takes the shape of a perfect T, as can be seen in the drawings that accompany the report, seems odd, but perhaps we should not rely too heavily on these drawings (in fact, the investigator who drew them makes no secret of his personal belief that "UFOs" are "nuts-and-bolts craft" piloted by extraterrestrials). In any case, the description given by the witness is less categorical: he is not sure how many rows constituted the crossbar ("at least three"), nor does he recall precisely how many bands were trailing behind ("several").

One should also take into account that the interview was conducted nearly eight years after the sighting. It cannot be excluded that the witness subconsciously attributed a more structured and solid appearance to the phenomenon in the course of those years. Equally important to our interpretation of the event is that the confrontation lasted only "2 or 3 seconds", after which the witness was "blinded" and lost sight of the formation.

We can only guess as to what may have caused this temporary blindness. At 8 a.m. in October, the Sun must have been low on the horizon in the east-southeast. The map of the sighting location reveals that this is the direction the witness was looking at when he saw the formation appear over the thicket. Perhaps the sudden loss of eyesight and the effect of "stars dancing before his eyes" was caused by the rising sun piercing through gaps in the thicket. Another possible explanation may be that the witness was blinded by a lightning flash. Lightning is often observed when mammatus is present, however the witness insists that he didn't see any light, nor hear any sound. A third possibility is that the temporary blindness was related to a specific physiological condition of the witness. It is conceivable that the emotional shock, caused by the sudden confrontation with an unknown phenomenon, resulted in an elevation in blood pressure ("blurred vision" and "spots before the eyes" are common side effects of abrupt fluctuations in blood pressure). Unfortunately, we know nothing about the physical condition of the witness.

The "vertical, diamond-shaped cloud", spotted by the witness immediately after he regained his sight, strongly reminds of a whirlwind or tornado. The expanding motion, and the sudden dissipation close to the ground, point in this direction. Moreover, tornados require similar weather conditions as mammatus, namely hot air close to the ground and significantly cooler low pressure air above. Mammatiform clouds have often been observed immediately before and after the passage of a tornado.

The witness claims that he saw the black cloud in the northwest and that the spheres were travelling in a west-southwestern trajectory, but the map shows a different situation: here, the indicated direction for the spheres is west-northwest, while the black cloud is situated in the north-westnorth. This would imply that the spheres had moved more or less in the direction where the black cloud was spotted.

The biggest problem with the cloud explanation is the estimated distance ("no more than 10 m from my head", according to the witness). This points to a relatively small "object". The report does not specify if the investigators verified the witness' ability to accurately gauge distances (they could have checked this by asking him to estimate the distance to a known object and then measuring the actual distance to that object). On the other hand, the small trees, which were only about 15 m (50 feet) away from the eyewitness, should have made good reference points. Under such circumstances it seems unlikely that an observer would describe a large cloud as a formation of spheres no bigger in size than a small van.

The witness further mentioned that the spheres "moved forward, one turning inside the other". Even though such a "boiling motion" is typical for mammatiform clouds, these movements usually only become apparent in accelerated images. They are rarely perceptible in a time frame of two or three seconds. The reason for this is the altitude at which mammatus usually forms, and which is rarely less than 1,000 m (3,048 feet). At those altitudes, even objects moving at 100 km/hour (62 miles/hour) will appear as "slow" for a ground-based observer.

The conclusion we reach is that all the data, with the exception of the estimated size and distance, point to a mammatiform cloud formation. We suspect that the witness saw a rare, miniature version of this cloud type. Interestingly, there is another report in our selection of case examples in which a miniature mammatus accompanied by a mini-tornado was seen to hover less than 60 m (200 feet) from the ground. Here too pouch-like structures, animated with rapid uncoordinated movements, were observed. For details about this case, see Limelette - July 2, 1994.

Although the absence of a precise date makes it impossible to verify the weather conditions for this sighting, the description given by the witness points to an unusual weather phenomenon, in all probability a mammatiform cloud formation followed by a tornado.

In his catalogue Tornados, Dark Days, Anomalous Precipitation, and Related Weather Phenomena (*), anomaly researcher William R. CORLISS gives two references to "miniature thunderclouds".

The first one was published in Weather in 1949 (Vol. 4 : 267). It concerns an observation that occurred one evening in July, around 1919. The witness was standing about 400 yards [365 m] from the Manchester Ship Canal at Stockton Heath, England, when an highly unusual cloud formed. We quote his own words: "Gazing down the road, I saw a small black thundercloud gathering along the length of the canal, and about 30 or 40 feet [9 or 12 m] above it. It was approximately 400 yards [365 m] long and perhaps 6 feet [1.82 m] thick. As I gazed at this strange formation, a dazzling lightning flash raced through the entire cloud, i.e. parallel to the water, and a bang like the discharge of field artillery followed immediately. About 40 seconds later, another flash and report occurred; then the cloud thinned and dispersed in about four minutes".

The observer added that prior to the appearance of the cloud "The evening was somewhat oppressive and the air had become strangely still".

The second story is too vague to be of much use. It concerns a simple reference to a statement by "Academician MARCOLLE" that "lightning from a cloudlet 1 ½ feet in diameter killed a woman". No further details are given, other than that the reference was part of an article entitled "Curiosities of Thunderstorms" published in Eclectic Magazine in 1856 (Vol. 38 : 458).

(*) CORLISS, William R., Tornados, Dark Days, Anomalous Precipitation, and Related Weather Phenomena, The Sourcebook Project, Glen Arm, USA, 1983, p. 13.