Pumice

Abstract

Cold pumice floating on water slowly absorbs water into the vesicles and eventually sinks. Experiments show that some pumice can remain afloat for over 1 1/2 years. The time taken for enough water to be adsorbed to sink depends on the pumice size, initial density, the size distribution of vesicles and the connectedness of the vesicles. Hot pumice often sinks immediately on immersion in water despite having a lower density than water. Experiments demonstrate that for any pumice there is a critical temperature above which the pumice will sink. Even pumice with a density of 0.2 g/cm³ will sink if the temperature exceeds 700 °C. The critical temperature correlates well with initial pumice density with lower-density pumice requiring higher temperatures to sink. The mechanism at low temperatures (< 150 °C) involves the absorption of water by contraction of hot air within the pumice. However, at higher temperatures conversion of absorbed water to steam in the hot pumice flushes out air, and further cooling results in condensation and absorption of water into the pumice. The experiments on hot and cold pumice suggest that all the vesicles in pumice are interconnected. This was confirmed by vacuum impregnation of pumice with resins. The behaviour of hot and cold pumice indicates that the deposits of hot and cold pyroclastic flow deposits may be distinguishable. Hot deposits will contain a significant proportion of low-density pumice, whereas cold deposits will not. Pumice falling hot onto water could also sink immediately to form subaqueous pumice-fall deposits. The physical properties of pumice were further examined by a nitrogen absorption technique and by mercury porosimetry. The former method shows that pumice has a typical surface area of 0,5 m²/g, corresponding to a sheet of material of 1 m² and 0,87µm Thick. Porosimetry shows that there are often three apparent vesicle-size populations in pumice. However, the porosimetry data gives surface areas which often greatly exceed those measured by the absorption method. The calculation of surface area by porosimetry assumes that vesicles are open cylinders. The large discrepancy with nitrogen absorption data suggests that the surface areas and proportion of small vesicles are overestimated by porosimetry and that pumice vesicles have narrow entrances. The porosimetry size distributions reflect the dimensions of pore entrances rather than the vesicles themselves. A three stage degassing history was proposed by Sparks and Brazier (1982). However, the small size population of sub-micron vesicles they identified probably represent larger (≫ 1µm) vesicles with narrow entrances. The experimental data indicate that pumice can degas very quickly because of the connectedness of vesicles and high internal surface areas.