A great deal of physical deterioration of rock art is attributable to the effects of moisture as discussed above: the deposition of various mineral skins (silica skins, rock varnish and similar ferromanganous laminae, oxalates, carbonates and others); the deposition of water-soluble salts (as surface efflorescence, interstitially or as subflorescence which effects surface spalling); the removal of soluble salts and subsequent granular or massexfoliation. Here, however, we are primarily concerned with temperature or insolation-induced stresses; hydration (for instance of a clay matrix) and subsequent massexfoliation; brush or forest fire spalling; the release of tectonic stresses; kinetic damage by fluvial transport; and damage by lightning strike. Several of the processes involved in these preservation hazards cannot realistically be arrested, but some palliative action may be possible.

Where physical processes are attributable to moisture, a site’s hydrology can often be modified or managed indirectly as detailed above. Fracture by insolation or Kernsprung can only be averted by planting shade trees, but in the arid regions where this is most relevant such measures might not be feasible. Tectonic stresses in rock structures are unavoidable, unless geophysical damage is humanly caused (through mining or water extraction). Similarly it would be unrealistic in most cases to seek to arrest lightning damage at topographically prominent sites, although it is certainly possible to do so. Petroglyphs occurring in the seasonal flood zone of rivers are exposed to the abrasive action of sand and coarser material, which can be as effective in obliterating rock art as sand blasting. Again, there are no pros-pects of saving these petroglyphs in the long term, short of diverting the river. Where the petroglyphs occur on soft rock in high-velocity fluvial conditions, as at Siega Verde (Spain), it needs to be accepted that they will not survive the next few millennia.

However, damage from brush and grass fires is avoidable by keeping fuel away from the art panels. Back-burning is practised at some Australian sites, for instance in the Grampians. Fire is a particularly effective form of physical deterioration, because it focuses on the surface zone of rock, and depending on the availability of fuel, whole rock art panels can be stripped off in minutes. Therefore fuel control is one of the most important measures in the management of rock art sites.

The salt deposits caused mostly by capillary or gravitational water within the rock can occur either within the rock’s fabric, or on the surface. In the latter case they are considered unsightly by site managers and visitors. Since these precipitates are readily water soluble, their removal is possible by compresses or poultices (Schwartzbaum 1985). Sheets of long-fibre, wet strength tissue paper are applied with water and a flat bristle brush, from the bottom of the area to be treated to its top. The poultice is intended to soak up the dissolved salts. However, salts may retreat into the porous rock as well upon mobilisation, so the technique must be applied by experienced conservators. The same applies to the poultice treatment of pictogram panels in which specific solvents are introduced to target such unwanted substances as dust, soot or body grease.

At a number of sites, massexfoliation has led to programs of consolidating disintegrating panels by grouting and other physical means. The perhaps largest project of this type was at Mootwingee, New South Wales (Probert and Wallace 1970; Chestnut 1972), where in 1973 it was attempted to arrest the exfoliation of a massive layer of sandstone on a steep slope. Destabilised by the disintegration of the supporting substrate, large blocks of capstone were gradually sliding due to their weight, a condition aggravated by heavy visitor traffic. Large pins of stainless steel were installed in the gaps between blocks, with wooden and later plastic space blocks. This did not arrest deterioration, but introduced new factors leading to further remedial action in the form of grouting experiments. This interfered with the site’s hydrology, and after catwalks were installed the site was eventually closed to tourism.

Small exfoliating rock fragments of the engraved pavement were occasionally fixed into position with epoxy resin at Mootwingee. The same method has been employed at another site in Australia, Trotman’s Cave in the Great Sandy Desert. Here, J. Clarke reattached exfoliated fragments of a pictogram panel with epoxy resin, using only spots of this material to ensure that moisture could not be trapped in an impervious layer of resin. The same approach has been recommended to affix a thin exfoliating skin on sandstone at Shishkino, central Siberia (Bednarik 1992b). In Canada, polyester resin was used to arrest congelifraction of a metamorphosed limestone pavement near Peterborough. A large separated rock slab at Deer Corral, British Columbia, was reattached by grouting with a mixture of sand and acrylic resin (Kennedy 1979).

Exfoliating rock surfaces at rock art sites have been impregnated with various synthetic sealants, for instance in India and the United States, generally with negative results. Rather than alleviating the deterioration of porous rock, it is accelerated by the application of substances that produce an impervious skin, such as water repellents and resins. Interstitial moisture in rock must be free to move and evaporate. Therefore consolidants to be used must be pervious to moisture. Spry (1981) recommended the use of silicon esters (alkoxysilanes), silicones (polymerised alkoxysilanes) and acrylics or related polymers for building masonry. However, the first do not bond well with rock and penetrate pores poorly, especially in hot surface conditions, while the silicones are adversely affected by ultraviolet radiation (Rosenfeld 1985: 61-3). Experiments conducted by Clarke (1978) suggest that the key to successful treatment is to get enough resin into the pigment layers to make them water repellent without rendering them impermeable. Such methods have been used only experimentally in Australia, and no long-term results are available from them. It seems reasonable to conclude that there is no known ‘safe’ method of stabilising rock surfaces by concealing or impregnating them, be it with silicon esters, silicones or silanes.

The failure of such treatment has been reported from other regions, such as Siberia (Bednarik 1992b, 1995a; Bednarik and Devlet 1993). At the major sandstone site of Shishkino, exfoliation had been arrested with a preparation of eroded sand and tetraethoxysilane (C8H20O4Si) which failed structurally within a year of application. Cracks developed between the repair grout and the rock — or to be more precise, within the weathering zone of the rock, 1 or 2 mm beneath the interface. order nolvadexSo the net result was an acceleration of weathering. Similarly, the repair of rock varnish (Elvidge and Moore 1980) has been described as yielding unsatisfactory results (Bock and Bock 1990) and should thus be avoided.

After observing how rock paintings can survive very well under silica skins, Watchman has investigated the possibility of synthetically creating thin transparent epilithic silica deposits to protect paint residues. He attempted to use an artificial silica gel consolidant in the early 1990s (Watchman 1995).

A viable form of structural repair of rock art sites concerns the rapid erosion and retreat of specific rock strata in vertical exposures bearing rock art. This can lead to the collapse of better-cemented rock above once it loses its support. The method of arresting such impending damage is to install a parcel of grout, together with an elastomeric membrane compensating for stresses and with provision of draining gravitational water from the rock’s interior.

Most other tectonic adjustments in a rock mass, be they caused by seismic or other factors (e.g. oscillating aquifer level), are more difficult to deal with, and any action depends very much on the specific circumstances at a site. Nevertheless, remedial action of providing support structures for rock components of a site is certainly possible in many cases. It can take a number of different forms as required by site conditions and resources. Similarly, aeolian erosion in the form of sand blasting, which affects numerous sites, can be reduced considerably. Artificial barriers, including vegetation, reduce such effects considerably. Some sites have been buried under sediment or other material, for this and other reasons (e.g. congelifraction). Examples are the fragile sandstone petroglyphs at Mt Cameron West, Tasmania (Blanks and Brown 1991) and the granite pavements at Tanum, Sweden (Bertilsson and Magnusson 2000: 100).

REFERENCESBibliography of Rock Art Conservation