According to Mr. Osmanagic, the famous "stone balls" (kamene kugle) can be found only in Bosnia, Mexico and Costa Rica. However, as I have already mentioned here, and here, these particular phenomena are in fact much commoner than generally realized, and can be found on every continent, including Europe. Recently, for example, I discovered [1] that they are also present in France, in Saint-André-de-Rosans in the Hautes-Alpes.

Almost the first sight to greet visitors to Saint-André de Rosans in the village square, with its Mairie, its cenotaph for the dead of World War I, and the little inn (where, if people remember to book the day before, Monique (fr) will serve them a delicious meal...), is this bizarre object:

During a stroll around the edge of the village, visitors will see others dotted here and there in people’s gardens, and dozens more in a large outcrop a few hundred meters further along the D425 leading to the Col de Palluel:

At a slightly higher point along this road, visitors will find themselves near the hill known as Serre d’Autruy:

There, a small track off the road on the right leads to a sandstone plateau, where there are what must be at least several hundred balls, on the surface of the plateau itself:

or protruding from the escarpment face:

What is the origin of these balls, which the local inhabitants sometimes wrongly refer to as "dinosaur eggs"?

Before considering the geological explanation, let’s begin with a few observations in the field. The balls’ diameter is always important: at least a few feet, the largest sometimes measuring nearly two yards. Their shape, always rounded, can vary from an ovoid to a perfect sphere. These balls are clearly fixed in a layer of sandstone, which completely rules out any human intervention; the photograph below illustrates the point of contact between a ball and the rock that encases it:

and some more examples of balls solidly encased in a layer of sandstone:

Even if they show some evidence of frost weathering, the balls are clearly more resistant than the actual layer of rock that encases them. Over time, therefore, the action of weathering gradually wears away that rock layer, leaving behind a conspicuous hollow:

Sometimes, they roll quite far down the marl gully beneath the sandstone escarpment:

As long as the balls, in their sandstone matrix, enjoy some degree of protection from weathering, their surface is quite even:

But, as soon as they are exposed to the harsh climate of the region (during winter, at 800 m in this part of the Hautes-Alpes, the temperature frequently drops below zero), and because of the almost total lack of soil and vegetation, the balls’ surface rapidly degrades, and the effects of frost means that the entire ball will gradually splinter:

Here, for instance, we see the contrast between the lower part of the ball, which, until a short time before, was completely encased, and the upper part, exposed to protracted weathering:

Finally, we should note that these balls frequently display a structure characterised by more or less concentric layers:

and that their composition seems to be very similar to that of the encasing sandstone, from which they differ only in colour - frequently darker - and in hardness.

The sandstone bank in Saint-André de Rosans in which these balls are to be found was formed over 110 million years ago during the Secondary era; more precisely, during the Lower Cretaceous; and more precisely still, during the Upper Aptian. At this time, long before the uplifting of the Alps took place, the whole region (the "Baronnies") lay on the slopes of a submarine "trough", known to geologists as a "fosse vocontienne" or "Vocontian Trough": on this subject, see the excellent work published in 2003 by Gérard Friès and Olivier Parize, "Anatomy of ancient passive margin slope systems: Aptian gravity-driven deposition on the Vocontian palaeomargin, western Alps, south-east France" [2], from which comes most of the information included in the present text concerning the Rosans sandstones. This submarine slope, extending between the shallow seas on the continental shelf of that era (the present day Vercors and Vivarais) and the deep oceanic basins of the former Tethys, were subjected to concentrated sedimentation, mostly marls (the Aptian and Albian "Blue Marls", resulting from the deposit of mostly fine muds and silts) of which one example can be seen here:

But, on this slope, various violent disturbances regularly took place, in the form of submarine landslips and avalanches, sometimes for tens of miles. Particularly in the valleys and canyons that dissected the slope, this resulted in local depositions of enormous quantities of coarser sediment. During one of these events, an avalanche of sand filled the channels incised into the underlying marl, and travelled nearly thirty miles down the slope; and it was this, according to G. Friès and O. Parize, that was responsible for the formation of the Saint-André de Rosans sandstones.

These authors also confirm that the sandstones resulting from the lithification of this sand avalanche are composed mostly of grains of quartz, with a significant percentage of glauconite (responsible for the frequent greenish colour of the sandstones), and siliceous and carbonate débris (shells). An important feature of these sandstones is their mass, and the almost complete lack of internal sedimentary structures [3].

Friès and Parize refer to the sandstone balls as "diagenetic doggers", or diagenetic concretions. This is to say that their formation took place after the deposition of the sand, but before the completion of lithification, that is, during the long period when, under the growing pressure and heat resulting from their burial, the sands underwent the processes of cementation and compaction that changed them into sandstone. During this period, there were many instances of matter redeposition and crystallization within the sediment that was undergoing a process of lithification, and the crystallization was enhanced by what must have been a pre-existing "seed", such as calcareous debris or shell, capable of "attracting" carbonate crystallization. Starting from this "seed", the ball progressively develops and crystallizes in equal measure in every direction, given that the sediment is relatively homogenous, lacking any significant rupture or heterogeneity capable of hindering this regular growth. This development process is the reason for the concentric layers of the ball’s structure.

Although largely dismantled because of erosion following the uplifting of the region, this particular massive Upper Aptian sandstone bank is to be found in various parts of the Baronnies, where sandstone balls similar to the ones at Saint-André de Rosans have been seen. The existence of the balls is attested in Bourdeaux, Arnayon, La Charce, Bevons, Châteauneuf-de-Bordette, and Condorcet... I myself have seen some in Saint-Ferréol-Trente-Pas, where the sandstone layer, forming an outcrop above a place called Le Monestier, regularly "releases" balls of various sizes that the inhabitants harness for decorative purposes:

But nowhere can you find balls as beautiful and in such numbers as the ones at Saint-André de Rosans. If you want to travel back in time 110 million years, and if you can appreciate the marvel of the fossilized sandy avalanche and its surrealistic landscape, then you should go for a stroll around the Serre d’Autruy!