Prev Page--Introduction || Next Page--Characteristics
Petrography of Volcanic Ash
The type of material found in volcanic ash deposits depends on the kind of source rock emitted by the volcano or volcanoes, and the modification of this source material by selective transport and by contamination during transport or deposition. Differential alteration also is a factor that may modify the characteristics of an ash deposit. In earlier petrographic studies of volcanic ash, several characteristics have been described, among which color, particle size and shape, composition, types of inclusions or vesicles, and degree of alteration have been given most attention. These properties will be discussed individually in terms of their value for purposes of correlation, and possible effects of selective transport, contamination, and alteration will be pointed out.
Composition
Both lithology and chemical composition are generally used in the description of volcanic ash deposits. Mineral and chemical composition have been employed in determining source area (Merrill, 1885; Turner, 1892), in studying the degree of magmatic differentiation over a period of years (Falconer, 1902; Bonney, 1903), in determining the genetic relationship of two volcanic regions (Teall, 1902), and in distinguishing between beds of ash in a related series (Powers, 1941; Allison, 1945). The refractive index of glass particles has been used to estimate their chemical composition (George, 1924, pp. 364-366; Allison, 1945), to determine their source (Herold, 1937), and to determine variations within and between formational members (Bailey, 1926, pp. 118-120). The fact that composition has been considered of paramount importance in the description of volcanic ash is illustrated by the wide use in descriptive studies of Pirsson's (1915) classification of tuff, based on composition, as vitric, crystal, and lithic. More recently, Wentworth and Williams (1932, pp. 21-24) have reviewed several classifications of the pyroclastic sediments based in part on composition, but the classification proposed by them depends primarily on particle size and origin.
Minor differences in composition may prove to be of value in detailed studies. Montgomery (1895) noted that several ash deposits in Utah and northwestern Colorado gave a slight acid reaction in water, which he attributed to a sulfur compound.
Diller (1884) observed that selective transport plays an important role in determining the relative abundance of volcanic mineral grains and glass particles in an ash fall inasmuch as the thin glass particles are carried farther from the volcano by wind. This condition was shown by him to produce an increase in percentage of silica with greater distance from the vent; consequently, ash at a great distance will be more acidic than the magma from which it was derived, and ash close to the volcano less acidic if crystallization had begun at the time of eruption. In reporting an ash fall from which samples were collected at points at least 250 miles apart, Smith (1912) found no variation in the quantity of ash, but noted slight variations in the amount of magnetite present. The chemical composition cannot be expected to be diagnostic of a particular ash bed over large distances unless the vent is remote or volcanic mineral grains are absent. It is believed that selective transport had little sorting influence on the refractive index of the glass shards from the samples examined by us.
Contamination by sediments during transport or deposition or during reworking of ash may mask the volcanic mineral suite, but under most circumstances the presence of nonvolcanic minerals indicates that contamination has taken place.
Particle Size
Size of the particles is commonly noted in descriptions of volcanic ash beds, and in at least two instances (Todd, 1897; Buttram, 1914, pp. 48, 49) ash coarser than that of near-by deposits has been assigned to a different horizon. Dorell (1938) found that mechanical analyses were of value in the correlation of bentonite beds in the Pierre shale of eastern Colorado, and wrote, "The size distribution of bentonites is of value in correlation only in areas of uniform deposition."
Decrease in average size in a direction away from the source has been reported by many geologists (Barbour, 1898; Capps, 1915; Landes, 1928, p. 17; Landes, 1928a, p. 936). Buttram (1914, p. 48) observed a general decrease in size of ash particles toward the east in Oklahoma, except for samples from one locality. This led him to the conclusion that the anomalous locality represented a different ash fall--a hypothesis which was supported by additional evidence from particle shape. Particle size of unweathered ash in any one deposit is determined not only by distance from the source but also by size of particles emitted from the volcano, velocity of transporting wind, and, in the case of subaqueous deposits, water velocity.
Particle Shape
Pirsson (1915, pp. 194-198) described glass particle shape in some detail, and Ross (1928, p. 146) later wrote that glassy fragments assume three principal habits, as follows:
One type is composed of fragments of glass that once inclosed rounded bubbles and consists of curved or lune-shaped fragments of the bubble walls, or Y-shaped fragments that were formed where three bubbles were in close proximity, and of double-concave plates that formed the wall between two adjoining bubbles. A second type is made up of nearly flat glass plates that were formed by the fragmentation of the walls that inclosed large flattened lens-shaped vesicular cavities. The third type has a fibrous structure and represents pumice fragments with minute elongated vesicular cavities and the inclosing glass walls. . . .
Neither Pirsson nor Ross, however, indicated that particle shape could be used in the differentiation of ash falls. Buttram's observation of a different particle shape in one locality in Oklahoma has already been noted. He described the shape of ash particles from this locality as "angular and lenticular," and from other localities merely as "angular." Barbour (1898, p. 25) found that some ash beds in the Great Plains had flat, nonvesicular, glassy scales, and that ash from other beds seemed to be "as vesicular, as angular, and as well suited to purposes of abrasion as the best pumice of our markets." Woolsey (1906, p. 478) in his description of volcanic ash near Durango, Colorado, wrote that the particles were flat glassy scales, and were not as good as some other ash for cleaning purposes. The effect of selective transport on particle shape of volcanic glass deposits remains to be investigated. Local concentration of unbroken bubbles where ash has fallen into water is possible.
Types of Vesicles and Inclusions
The presence or absence of vesicles in glass shards is closely related to the shapes of the particles. The vesicles may be spherical, flattened with circular outline, or greatly elongated, and in some cases curved. They may be pear-shaped or round at one end and tapering at the other, and they may be arranged in groups as described by Ross (1928, p. 146). Hanna (1926, p. 94) described glass flakes from a core in Louisiana as "minutely perforated."
Mineral inclusions may be either primary or produced by alteration of the glass. Alteration products are generally recognized as nonvolcanic minerals and may not be valid means of comparison.
Degree of Alteration
The term alteration as used here refers to the effects of devitrification and weathering, which may occur singly or simultaneously. Altered glass shards are characterized by a clouded or mottled appearance of the glass and abundance of polarizing particles and inclusions, and are generally associated with cemented aggregates, the composition of which is difficult to determine by petrographic methods. The amount of alteration which has taken place is a conceivable means of correlation where deposits are of greatly different ages; if other conditions are equal, the degree of alteration should be closely related to the age of the ash. The value of the relative degree of alteration of the ash as a means of correlation may be tested by comparison of samples of known age.
Color
The colors of the deposits have commonly been described in papers concerning volcanic ash. In most instances the color description is megascopic. Color was used by Landes (1928a, pp. 932, 933) in his comparison of Tertiary and Pleistocene ash deposits of Kansas and by others in detailed descriptions of series of ash beds (Gardner, 1923; Wanless, 1923, pp. 231-236; Bailey, 1926; Allison, 1945). Colors described have included white, light to dark gray, bluish gray, brown, buff, pinkish, red, and green. Several factors that may control or modify the color of an ash deposit include original composition, particle size, introduction of nonvolcanic material in the form of either clastics or cement, and weathering or devitrification.
Original composition of the ash--Studies of recent ash falls show that there is great variation, that consecutive falls from the same source are sometimes different in color (Fry, 1912), and that there may be a progressive change in color during one fall (Flett, 1902). Most vitric ash is lighter in color than crystalline ash, although the basaltic ash known as palagonite consists of particles of brown glass (Pirsson, 1915, p. 199).
Particle size--Barbour (1898, p. 24) found that the coarser ash in Colorado, Wyoming, and Montana was dirty and ochreous, whereas the finer ash of Kansas, Nebraska, and South Dakota was white in color, thus implying that the color becomes lighter with decrease in size. Flett (1902) found no color differences between size grades in ash that fell on Barbados after the eruption at St. Vincent. Bonney (1903) on the other hand, found that the finest fraction of volcanic dust from the Soufrière showed a distinct "warm-brown" tint, whereas the sample as a whole was of a dull dark-brown color speckled with a lighter tint. Color comparisons of volcanic ash samples should be made on the same size fraction.
Contamination by nonvolcanic material--Rowe (1903, p. 7) in Montana and Buttram (1914, p. 30) in Oklahoma observed that no two volcanic ash beds had exactly the same tint, and Buttram attributed the lack of uniformity to adulteration with other substances. The presence of sand or silt may give volcanic ash a buff, yellowish, or reddish tint. Small insects trapped by an ash fall produce a brownish color. Color comparisons should be preceded by microscopic examination, and those samples or size fractions containing a large amount of foreign material should not be used.
Weathering and devitrification--Alterations of volcanic glass to montmorillonite or beidellite my give the sample a yellowish or greenish tint. Some other forms of alteration apparently make the ash whiter. Color comparisons for purposes of correlation should be restricted to fresh ash.
Prev Page--Introduction || Next Page--Characteristics
Kansas Geological Survey, Geology
Placed on web Aug. 10, 2007; originally published April 1946.
Comments to webadmin@kgs.ku.edu
The URL for this page is http://www.kgs.ku.edu/Publications/Bulletins/64_1/03_petro.html