Many of the bedrock walls exposed in Wind Cave exhibit prominent boxwork, which consists of paper-thin intersecting veins of calcite that project into the cave as resistant fins, and from which the intervening bedrock has been removed by weathering. Most of the calcite veins are roughly perpendicular to the bedding, but some lie along bedding planes and others show splayed patterns. Vein spacing ranges from a few centimeters to several decimeters. The veins are similar in composition and origin to the matrix of the angular breccias, but they formed along narrow fractures resulting from stresses produced when anhydrite was hydrated to gypsum. The intervening pockets, from which the bedrock has fallen out, average about 10 cm deep, although in places their depths exceed 1/2 meter.
The bedrock is less resistant than the boxwork veins not simply because it is less crystalline, but because it has been considerably altered to a friable sand consisting of calcite crystals held together at their contacts by a sparse cement of secondary quartz. The calcite crystals in the bedrock are mostly pseudomorphs after gypsum, and the quartz is the remnant of an early matrix that formed around former dolomite crystals. Much of the original bedrock was apparently removed by H2S-H2SO4 dissolution processes, which left many millimeter-sized pores. These porous zones readily weathered away during cave development, as well as later when they were in contact with the moist cave atmosphere. Oxidation of metallic sulfides in the cave atmosphere has colored the weathering zones in places with a variety of bright reds and yellows.
Boxwork veins are limited almost entirely to dolomite beds within the Madison. Petrographic analysis shows that the calcite veins have replaced former gypsum, and that they predate the late Carboniferous paleokarst (Palmer & Palmer 1989). This process has been observed in various stages of completion at the contact between gypsum and dolomite beds in the Permian San Andres Formation of New Mexico, USA.
Boxwork has also been erroneously attributed to other processes. The traditional explanation is that it consists simply of more resistant calcite veins projecting from the bedrock. Palmer (1981), noted that the boxwork was concentrated only around the caves. So he originally assumed that the veins were precipitated after the cave formed, where supersaturated infiltrating water was able to lose CO2 through the porous dolomite and precipitate calcite in the fractures. This mechanism is valid for certain calcite veins in young carbonate sediment, but it is not appropriate in the Black Hills caves.
The origin of the veins has also been attributed to calcite precipitation in fractures as the result of dedolomitization during enlargement of the caves during the Tertiary Period (Ford & Bakalowicz 1983; Millen & Dickey 1987). However, this is not a feasible mechanism either.
The boxwork veins are truncated by the late Carboniferous paleokarst features, and therefore greatly predate the Tertiary enlargement of the caves. In addition, it is not clear how calcite can precipitate within the fractures of a dissolving dolomite surface.
From A.N. Palmer & M.V. Palmer. (2000). Speleogenesis of the Black Hills Maze Caves, South Dakota, USA. In Klimchouck, A.B., Ford, D.C., Palmer, A.N., and Dreybrodt, W. (eds). Speleogenesis: Evollution of Karst Aquifers. National Speleological Socity. p. 274-286.