Silica content of living rhizomes from the perennial salt marsh rush Juncus roemerianus had values of 0.34, 0.20, and 0.60% of dry weight in three morphologically distinct populations along the Mississippi coast and was directly related to available silica content of the soil (29.7, 17.0, 169.6 mg/100g soil, respectively). On the other hand, living leaves had about the same average silica content (0.93, 0.87, 0.9070 or dry weight). The silica content of living leaves varied from 0.142% in younger leaves to 1.520% in older ones. Similarly, rhizomes also increased in silica content with age, varying from 0.137% in younger portions to 1.030% for older ones. Mature leaves collected in October all had a higher average silica content (0.737%) than those collected in April (0.413%), indicating that silica content also increases over the growing season. Decomposed leaves (dead-standing) had a relatively high silica content o f 1.8170, obviously reflecting a loss of organic matter and soluble minerals. Roots contain considerable silica, but reliable results were not possible as the soil could not be completely removed from them. Petrographic microscope studies showed that the silica was clear, colorless and isotropic with a refractive index of 1.43, all properties typical of the mineral opal. No olquartz was present, as occurs in some species of Juncus. The silica was deposited in a sheet made up of small, irregular phytoliths arranged in rows lengthwise in the lcavcs. Ash percentages were much highcr than those for silica and no definite conclusions could be drawn from their variation. In comparison to the maximum silica content of leaves from Juncus interior (3.21%), the concentrations found in leaves of J. roemerianus were relatively low.