#bear lodge

However, there are a number of issues with this model; there are shallow rocks nearby that suggest limited erosion in the 50 million years since Devil’s Tower formed, and there is little density contrast between the shallow sedimentary layers around Devil’s tower. Having one layer be less dense is a big help in getting molten rock to get stuck at a level in the crust. Finally, some details of the tower itself, such as how the top of the tower came out so flat, aren’t well explained by this model. Erosion wouldn’t necessarily carve out a flat top to the tower unless that flat top was the original top of the laccolith, but that model requires some of the tower to be eroded away to fit features of the joints.

Columnar joints form as igneous rocks cool. As they shrink, tension forces develop, eventually cracking the rocks. The cracks always form perpendicular to temperature contours, so for the cracks to be so long and vertical, the tower must have cooled top to bottom. But, as you can see in this photo, there are several complexities: the columns turn at the base, indicating the cooling at the base was not simple, and there is a boundary 2/3 of the way up the tower where the fracture pattern changes.

The boundary 2/3 of the way up the tower could be where cooling from the top and bottom met; columns would start forming at the top and at the bottom, and when they met in the middle there would be an offset boundary between them. Why then are the columns shorter at the top than at the bottom? This pattern requires either that the top was eroded away, which wouldn’t leave the flat top, or that somehow Devil’s tower cooled most rapidly from the bottom.

It is these observations, combined with other details of the local geology and comparisons around the world, that drove a team of scientists led by a geologist at the institute of Geophysics of the Czech Academy of Sciences to suggest that Devil’s Tower did not start as a body of molten rock below the surface, but instead was a thick lava flow that grew out of a maar crater.

Maar craters occur when magma rises up and meets groundwater. The heat from the molten rock triggers explosions which throw blocks of debris out of the way, leaving a crater behind. There are a number of maar craters in the area formed as part of the same volcanic complex, and there are blocks of crustal rocks near Devil’s Tower that could have been ejected as part of maar eruptions. So, starting Devil’s Tower as a maar crater is consistent with the geologic evidence in the area.

Once the explosive eruptions stopped, the magma that fed the maar could have still been there. If molten rock rose up through the edges of this crater, it could have erupted out of the crater and piled up into a thick lava flow. That lava would form either a dome if it went straight up or a coulee if it began flowing down the gentle local slope.

The top of the coulee would be flat because it was the top of a lava flow. The top of the coulee would cool straight down, creating the vertical columns at the top. The bottom of the coulee would have a complex cooling as it might cool as it flowed along the ground and circulating water could also cool that portion if it rained. This more complex cooling caused the columns near the base to turn in multiple directions, creating the unbalanced shape of the whole tower. Furthermore, because the bottom was cooled by water, it cooled more rapidly than the top, making the bottom columns longer than the top columns since the tower cooled mostly from the bottom up.

Since it cooled, most of the dome or coulee eroded away, leaving behind the piles of broken columns found in the area today.

This is an elegant explanation for the Tower’s formation and, IMO, fits the surrounding landscape better than the ideas that require hundreds of meters of sediment erosion over this time period. I’m not sure what needs to be done to further test this model, but I’m already partial to it.

-JBB

Image credit: https://commons.wikimedia.org/wiki/File:Devils_Tower_National_Monument.JPG