Silicified wood provides a test for the proposition that nannobacteria are responsible for some silicification in fossils. Petrified wood samples from various locations, wood types, ages, and degree and internal variation of silicification were studied to evaluate the process range.

Two Recent specimens show the progress of silicification through the sample thickness. A contemporary sample from Lyman, Wyoming has apparent splintering of lignin, which disintegrates due to oxidization and silicification. Within the sample intra-vascular spaces are filled with cryptocrystalline silica giving a strong EDAX signal, and are bordered by cell wall lignin. A sample from a log burned and then petrified in the Columbia River Gorge, Washington, shows growth of 1-6 µm euhedral quartz crystals (verified by EDAX spectra) within the intra-vascular walls. Where the wood is more indurated, the euhedral crystals persist within spaces, while the cell wall structure is replaced by cryptocrystalline silica which displays surface bumps 0.1 µm or less in diameter.

One sample from the Peacock Mine of Nevada is several million years old. The cell walls are silicified, but because intravascular spaces are not infilled, cell structure is still identifiable with the SEM. Two mineralizations occur: crystalline potassium zeolite which replaces the cell structures and displays distinct growth faces; and ‘globular’ cryptocrystalline silica in intra-cellular spaces. Each form bears 0.1 µm balls and clusters of structures that can be equated to the 0.03-0.1 µm ‘nannostructures’ reported by R. L. Folk from a variety of minerals. Many nannostructures occur in areas where crystal faces are incomplete.

Petrified palm wood from the Eocene Yegua Fm. of Texas contains euhedral quartz crystals within the monocot xylem which show nannobacterial-like chains across several faces. These chains wrap, diverge, intersect, and spiral in apparent biological fashion. We have imaged drooping chains of 0.1 µm balls, for which organic origins appear probable.