ABSTRACT. The paper considers possible structure and genetic mechanisms for "gypsum nests" - very rare speleothems, growing from drying clay massifs and consisting from gypsum needles. Theoretical modelling is the only way for studying such complicated and delicate formations. The model suggested considers a "nest" as a screw-dislocated spherocrystal with its sub-individuals, separated by corrosion on some stage, and having separate growth on later stages.

The formation, called "nest" (photo 1), is found in the caves of Kugitangtou only once, as a unicum. With this, all of its main features are known in other types of formations. There are also some evidences of similar phenomenas existence in Lechuguilla cave (Guadalupe Mountains, USA) [D.Davis, pers.comm]. Both of these doesn't allow to speak about some occasional or typomorphous features, but more likely about some determined mechanism of such formations growth. Of course, these known formations are too rare, too beautiful, and too delicate for even to think about some direct studying, and nothing like was ever found outside caves. The only way left for studies is the way of theoretical modelling.

The "nest" is located on a large flat massif of alluvial clay, that's currently being dried. A part of this massif, several dozens square metres large, where the air circulation featurese are optimum, is all covered with various gypsum needles, "hedge-hogs", "cotton" masses - almost all the varieties of filamentary crystals and fibrous aggregates, known for plastic substrates. The "nest" itself appears as an empty inside funnel-shaped mass of thousands of oriented gypsum needles. The needles are approximately 0.1-0.5 mm thick, up to 40 cm long, and twist contra-clockwise around the center line off the "funnel". The diameter of the complete construction is about 30 cm, that makes it to outstand from the surrounding , specified by shorter, thicker, and chaotically oriented needles. The central hole of the nest is also funnel-shaped, and has one needle along its central line. The "nest" grows on approximately horisontal flat floor, and is inclined about 20° from the vertical direction.

The crystallisation mechanisms of filamentary crystals crusts, growing from drying clays [5,7,9], can't have other types of symmetry except spherical one. Possible disymmetry types are also restricted - only those, based on water level in the pores closeness to the surface. All the other factors, such as local air circulation conditions, humidity conditions, local clay porosity features, may cause only typomorphous features of aggregares and crusts texture. So, we must search for the source of the "nest's" symmetry not on the levels of crusts or aggregates, but on the level of individuals. Here we deal not with texture symmetry, but with structure symmetry, and it may be controlled not only by mass-transportation factors, but also by crystallization physics.

In a monocrystalline case the only reason for such regular twisting is a screw dislocation of the crystallographic network. Contra-clockwise dislocation, seen on the photo, is usual for gypsum [4]. At the same tyme, screw crystals are met not so often, on this clay massif also, and their appearence in such great quantities in one spot must have some special reason.

Probably, the only idea, having no conflicts with the statistics, is the idea, that we deal not with an aggregate, but with an individual - a lonely splitted screw crystal. If accepting this idea, the two main morphological features of the "nest" become explained - both regular needles twisting (the screw dislocation is generalised upon the whole splitted crystal), and the formation's outstanding from the surrounding (screw crystals growth speed is normally greater, than usual crystals growth speed [3,4,5]). Still several features stay unexplained. For example, filamentary crystals (the base for needles) always have active growth on their roots, and poorly connected spherolite bunches usually have it on their outer heads.

As it's known [1,5,7,9], the sulfates filamentary crystals, growing from the drying cave clays, have the following properties:

a) near the crystal "root", there is a zone of splitted growth, and after it - a zone of skeleton growth, that are controlled by the feeding physics;

b) when these zones are overlapped, dendritic needles appear in full accordance with the definition from [3], that a dendrite is a splitted skeleton crystal;

c) if the splitting grade is high, needles with properties of spherocrystals appear - with full generalization of the crystallographic network properties upon all the formation [2,3];

d) in the places with intensive seasonal humidity cycle, the needles "roots" are usually corroded;

After fixing these properties, let's consider, whether we can imagine some conditions, leading to crystallisation of something like a "nest". Knowing the cave conditions, we can soppose the fffollowing seuence of events:

1) A splitted gypsum crystal growth, having properties of a spherocrystal, and not yet extruded from the surface. Because of this, its splitted zone covers all its length. As it's shown in [7], 400%-500% relative oversaturation, needed for gypsum crystals splitting [8], is absolutely possible. The capillary pressure in pores is dozens of athmospheres, the needle physically blocks the pores around, pressing out the solution from them. When reaching the needle surface, where the pore is enlarged and the capillary pressure is lower, the solution goes across such the pressure step, capable to cause even much stronger oversaturations. There are also no theoretical restrictions on existance of such needles with screw dislocations. More, the "antholite-like packets", described in [7] and also known as "gypsum grass" [4], in reality are just splitted needles with screw dislocations, formed in very similar conditions.

2) Accidental flooding, causing 30-50% dissolution of the needle. The corrosion sculpture forms for spherolites and spherocrystals are wide spread and well known, though isn't clearly described in literature. In the case when the spheroilite center and its outside surface are equally available for corrosion, the main variant is disconnecting of the spherolite into separate radial needles - relicts of sub-individuals. Author studied lots of such corrosion effects on gypsum spherolites found in caves. With the flooding that we can suppose, this variant will be the main. The relict needles would be disconnected, but held in place by clay around.

3) Restoring of dry conditions, along with this - re-starting of the needles growth. It's evident, that when something like a described above relict exists, each relict needle of the packet would appear as an embryo for the new needle. And all of them would have the same screw dislocation, as the source needle had.

4) The solution pressing out from the nearby pores will also be restored, but with some new features. Because of large quantity of active needles on small area and impossibility of blocking their growth by other needles (the zone of concurrent growth is already passed), there will be some solution deficit. And it leads to a) limited thickness increase, and b) this increase is realised through skeleton growth, not crystals further splitting.

5) Even such restricted thickness increase still will change the packet morphology greatly. Their divirgence angle, 5-10° originally, will increase, thus converting the packet from having geometry of a cone to having geometry of a hyperboloid of one sheet. While this, the needles would be mechanically aligned to directions of generatrixes of this hyperboloid.

6) When the needles are extruded through the surface, they keep their last direction with only one additional drift - their curving under their own weight.

So, we received a genetic and structural model for some mineral bodies, similar to "nests" in morphology, and being something between 3-rd order individuals and aggregates in mineral bodies hierarchy. In spite of a lot of spatial imagination, needed for accepting a concept of "screw spherocrystal", this model is realistic enough and simple enough to have it as a working hypotheses. The last thing left to say is to note, that probably some other (not screw) bushes of oriented needles, may be also explained through this consept of splitting - dissolution - growth re-start.

1. R.Casali, P.Forti., 1969. I cristalli di gesso del Bolognese (In Italian).: Speleol. Emiliana, ser.2, v.1, no.7, p.1-24.

2. A.A.Godovikov, O.I.Rypenesh, V.I.Stepanov., 1989. Spherolites, spherocrystals, spheroidolites, corespherolites (in Russian).: Novye Dannye o Mineralakh (New Data on Minerals), vol. 36, Moscow, "Nauka", p.82-89.

3. D.P.Grigorjev, A.G.Jabin., 1975. Minerals onthogeny. Individuals (in Russian).: Moscow, 260p.

4. C.Hill, P.Forti., 1986. Cave minerals of the world. NSS, 238 p.

5. M.N.Maleev., 1971. Properties and genesis of natural filamentary crystals (in Russian).: Moscow, "Nauka", 180p.

6. V.A.Maltsev., 1989. The influence of season changes of the cave microclimate to the gypsum genesis.: Proc.10th Int.Cong.Spel.Vol. III.Budapest, p.813-814

8. G.V.Russo., 1981. Splitting of gypsum crystals (in Russian).: Zapiski Vsesouznogo Mineralogicheskogo obschestva, Leningrad, vol.110, No.2, p.167-171.

9. V.I.Stepanov., 1971. Crystallisation processes periodity in karst caves (in Russian).: Trudy mineralogicheskogo muzeja imeny Fersmana. Moscow, No.20., p.161-171