USGS Logo Geological Survey 7th Annual Report
Obsidian Cliff, Yellowstone National Park


The development of the various forms of structure exhibited in this obsidian flow may be summed up as follows:

The molten mass when it reached the surface at the time of eruption was a viscous glass, highly siliceous and slightly hydrated or holding certain gases absorbed in it. Those in the glass at the top of the mass immediately expanded, being relieved of the pressure to which they had been subjected, and the glass chilled quickly into pumice. A little lower in the mass the expansion of the gas was but slight, filling the glass with small bubbles. In the compacter glass; which cooled more slowly than the pumice, trichites and microlites formed. After these the microscopic groups of feldspar and quartz with pegmatoid and fibrous structure crystallized; this must have been under the influence of the absorbed vapors at a high temperature, since they have not as yet been produced artificially by dry fusion.

Undoubtedly particular conditions of temperature, pressure, and consequent rate of cooling obtained in different parts of the flow at different stages of its progress, but about the time it came to rest in the region of Obsidian Cliff small colorless spherulites and fibrous coatings around the granophyre groups were formed under conditions differing from those attending the production of the granophyre groups by the influence of a slightly lower temperature. In these the free silica would still seem to be quartz. In the next stage small colored spherulites crystallized with a fibrous structure in sectors of various orientation, which also appear to be composed of feldspar and quartz inclosing trichites and microlites.

With progressive cooling the nature of the spherulites passed to those whose centers are similar to the small ones just described, but the outer portion of which is formed of rays of short, attached feldspar crystals cemented with tridymite, the whole imprisoning multitudes of minute gas cavities. The absorbed vapors now began to assume the role of superheated water in their action on the surrounding silicates, which is more evident in the hollower forms, where the process appears to have been somewhat as follows: In the still viscous glass, from a center of crystallization the first frail beginnings of feldspar spread in innumerable rays, pre-empting, as it were, a sphere of the magma. The enlargement of these anhydrous microlites by crystal growth from their matrix of hydrated glass not only altered its chemical constitution by eliminating the alumina and the alkalis, but rendered it relatively more hydrous, so that, with decreasing temperature, it may no longer have been able to retain the vapors in combination.

The vapors separated from the glass were probably disseminated in minute particles through it, in some cases being able to coalesce into larger bubbles or to accumulate in certain parts of the spherule. A prolonged action of this sort permitted the separation of the original paste into feldspar, iron silicates and oxides, and residual silica, which at high temperatures would take the form of tridymite, but at lower temperatures would crystallize as pyramidal and prismatic quartz; a change of temperature or an intermediate one would cause both tridymite and quartz to crystallize.

The first crystallized feldspars were suspended in a plastic, hydrous silicate, which, from the observations of Mr. Daubree, would occupy considerably greater volume than the same siliceous glass if anhydrous, and therefore still more than its component elements when crystallized. Upon the change of the hydrous paste to a crystalline aggregate through the process above indicated there would be considerable shrinkage, the extent of which is indicated by the gaping cracks and parted segments characteristic of lithophysæ. This shrinkage was prior to the ultimate crystallization of the cementing paste, for the surfaces of the disrupted portions are coated with well developed crystals of fayalite, tridymite, and quartz. The tendency observed in spherulitic crystallization to produce concentric layers of different texture, which in the compact spherulites results in bands of greater or less density and of various colors, in the hollow varieties gives rise to the concentric shells of typical lithophysæ, by producing layers which act as nuclei about which the later-formed minerals, tridymite, quartz, and fayalite; crystallize, leaving cavities in place of the porous layers of the more solid spherulites.


Finally, the occurrence of lithophysæ in a rock, the nature of their contents, and the extent to which the cavities are developed (that is, their relative size) will depend on a number of conditions, first among which are unquestionably the character and chemical composition of the molten lava, the degree of hydration, and nature of the absorbed gases, and, second, the rate of cooling and the pressure under which consolidation takes place, for it is evident that under the same physical conditions glasses of different chemical composition will be affected differently by the same amount of absorbed gases. So far observations indicate that only the acid glasses produce these hollow spherulites. The amount of water vapor and the nature of the gases associated with it will alter both the extent of the action and the character of the minerals formed. With a given combination of these primary conditions, the results will vary with the rate of cooling and with the pressure, too rapid cooling not giving time enough for any crystallization to take place, too low a pressure permitting the gases to escape suddenly, and too high a pressure preventing their liberation in the proper state to produce the effects observed in lithophysæ, and bringing about with the consequent slower cooling a granular, crystalline structure, the grains of which inclose the imprisoned gases condensed to a liquid form.


In an earlier part of this paper attention was called to the parallel lamination of the rock, specially noticeable in the lithoidal portion. It was referred to slight local differences in the consistency or composition of the lava, which in flowing over the surface in a viscous condition spread these differing portions out in layers parallel to the plane of flow. A cause for such local differences is suggested by the following observations: In the lithoidite the layers vary in their degree of crystallization, some being glassy, others finely spherulitic, others more coarsely spherulitic and porous, while still others are quite granular and full of cavities; some glassy layers have the crystallization localized in spherulites and lithophysæ. These alternate and repeat themselves in endless succession and variety. In the obsidian the differences find expression in layers of spherulites, large and small, and bands of lithophysæ, and in layers varying in the abundance of granophyre feldspars, microscopic spherulites, microlites, and trichites; that is, in the different phases and amount of crystallization developed. Approaching the surface of the flow the same laminated condition of the rock appears in still more striking differences between the amount of microlites present and in layers of gas cavities which produce alternating bands of vesicular and dense glass and pumice, until at the surface the whole mass is thoroughly pumiceous.

It is evident that there is a difference in the amount of vapor absorbed in layers of the rock, which is shown by the varying extent, to which the layers are inflated when the pressure is removed and the confined gases are allowed to expand.

From the part which superheated water has undoubtedly played in the development of lithophysæ and the larger spherulites, from the aqueo-igneous conditions deemed necessary for the production of the granophyre groups of quartz and feldspar, and, finally, from the observed changes of plasticity produced in a siliceous glass by hydration, it seems highly probable that the differences in consistency and in the phases of crystallization here developed are directly due to the amount of vapors absorbed in the various layers and to their mineralizing influence.

Besides the lamination of the pumiceous portion of the rock, which proves that the absorbed vapors were distributed unequally in alternating layers, there are highly pumiceous or inflated spots scattered through the mass, which show that the vapors were more abundant also in certain spots than in others. The distribution of these inflated spots corresponds to that of the isolated lithophysæ and spherulites throughout the rock.

<<< Previous <<< Contents >>> Next >>>

Last Updated: 22-Jun-2009