It is commonly assumed that reactions in the silicification of land plants take place at low to moderate diagenetic temperatures when the solvent for the silica (H₂O) is in the liquid stability field. The early Permian forest of Chemnitz, buried by rhyolitic pyroclastic deposits ca. 290 Ma, may be an example of silicification at elevated temperatures above 100 °C by siliceous H₂O vapor. Many independent observations support this theory: the presence of low-density (gaseous) inclusions in primary α-quartz, the impregnation and partial replacement of silica phases in the wood by fluorspar, the preservation of relict organic material in the form of the high-temperature mineral anthracite, and the close proximity of the fossil forest to an eruptive center, the Zeisigwald Caldera. We have designed an experimental apparatus that allows silicification to be simulated by silica-bearing H₂O vapor. Water was reacted with rhyolitic obsidian at 150 °C for several days to take up silica, then passed through the parenchymatous stem tissue of Dicksonia antarctica in the form of a hot, silica-bearing steam. The reactions taking place in the organic tissue are documented. Amorphous silica gel was found deposited in vapor-treated cells, suggesting that steam can be efficient in transporting aqueous silica species and depositing them into stem tissue. These experiments cannot duplicate every detail found in the natural examples in Chemnitz, but they do underline how important it is to derive the temperature conditions at which the natural silicification reactions took place.