|
Geological Survey Professional Paper 3
The Geology and Petrography of Crater Lake National Park |
THE PETROGRAPHY OF CRATER LAKE NATIONAL PARK
By HORACE BUSHNELL PATTON.
HYPERSTHENE-ANDESITES.
(continued)
CLASSIFICATION OF ANDESITES.
While the phenocrysts fluctuate greatly in abundance and relative importance, they do not serve as well as does the groundmass as a ready means for classifying the widely different appearing varieties of the andesitic rocks. Certain structural types in the groundmass, on the other hand, recur with great frequency and are worthy of detailed description. The andesites of Crater Lake, therefore, may be divided into four main types, based upon the character of the groundmass, namely, hyalopilitic type; hypocrystalline type, with two subtypes (A and B); holocrystalline type; dacitic type.
Although each of these four types is quite distinctive, it is not to be supposed that all the examples cited in each case are equally so. Rather it may be said that they all grade into each other in such a way as to make the separation not always easy. Neither can the rocks of any one lava stream or of one locality be said to be limited to any one type.
HYALOPILITIC TYPE OF ANDESITES.
The following specimens are included here: Nos. 1, 2, 3, 4, 5, 6. 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, and 21.
This variety of andesite has mainly a very dark-gray to black, dense groundmass, and when it is not porous is apt to break in smooth, slightly lustrous surfaces. The specimens that contain less glass, as well as the more porous ones, have a much lighter color and rougher fracture. On account of the dark groundmass the plagioclase phenocrysts are usually prominent, and for the same reason the pyroxenes are hardly noticeable. Most of the hyalopilitic andesites are characterized in the sections by the presence of light to deep brown glass, which is usually very abundant and perfectly clear and unaltered (6, 9, 11, 13, 16, 17), or else this brown glass is more or less clouded by the presence of countless hosts of minute brown- to black-looking globulites (brown in 8, 15, 19, 1, 5; black in 14, 21). In case these globulites are thicker than usual, the glass base between the microlites becomes lighter or even almost colorless, but still the effect of the globulites in such cases is to lend a brownish color to the whole. In a few cases a decided change of color in the glass is to be noted, in that blackish spots appear locally in a glass otherwise brownish (12, 20). These black spots have the globulites very thick and somewhat larger, while the inclosing glass base is colorless. It is more than likely that magnetite dust is developed in addition to the customary globulites, which would explain the black color. These darker spots are similar to the whole groundmass in some of the hypocrystalline andesites.
This all-pervading glass base invariably incloses myriads of perfectly sharp and straight augite microlites; also plagioclase feldspars in slender strips of microlitic size or in larger laths and in rectangular individuals. The size of the augite microlites may vary considerably in the same thin section. Thus in No. 2, where they are uncommonly small, they vary from 0.001 to 0.003 millimeter in width and from 0.01 to 0.03 millimeter in length. In general, the width may be said to average about 0.005 millimeter, with extremes of 0.001 and 0.01 millimeter, and the length to average about 0.02 millimeter, with extremes of 0.01 and 0.08 millimeter. It will be seen that they are about five to ten times as long as wide. These microlites are very commonly free from inclosures, but perhaps more frequently inclose minute magnetite grains, or octahedral crystals of the same. The plagioclase microlites and laths vary extremely in size and form. In the majority of cases their form is either distinctly microlitic or, at least, slender lath shaped and minute (5, 14). When very small, the customary twinning is not easily distinguishable, but in the larger and broader individuals polysynthetic twinning is strongly in evidence. Interspersed with these slender plagioclases are others with short, rectangular forms that sometimes show twinning and sometimes do not. In a few rocks these rectangular forms are very numerous, and partially or almost completely supplant the lath form (11, 19, 21). Still larger plagioclases of rectangular form connect these rectangular plagioclases of the groundmass with the phenocrysts, so that a sharp discrimination is not always possible. Fluidal arrangement of all these groundmass ingredients is more or less conspicuous in all thin sections.
Intermediate stages between andesites of this type and the next are very common, and are mostly included in those listed under the following type. In one specimen, however, No. 18, collected about one mile southwest of the base of Red Cone, we have what is common, namely, a transition to the holocrystalline type. In the thin section from this rock there appears to be very little glass, but the appearance of dust-like globulitic material between the abundant feldspars of the groundmass indicates the presence of more glass than is at first apparent.
HYPOCRYSTALILINE TYPE OF ANDESITE.
This may fairly be said to be the normal type of Crater Lake andesites; at least, it is more abundantly represented than any other. It may be divided into two subtypes, which, though united by numerous connecting links, are upon the whole readily distinguished. In the hand specimen, however, this distinction between the two subtypes can not be made. Neither can the two be readily distinguished from the hyalopilitic type except in thin section. Taken as a whole, they are decidedly dark-colored rocks with a basaltic look, but the darker varieties distinguish themselves by the relative prominence of the plagioclase phenocrysts. Not a few are as dark as the darkest of the hyalopilitic andesites, but they lack the glossy luster common to some of that type. The average color is naturally lighter than in the hyalopilitic type, as they are more decidedly crystalline. In fact, not a few included under both of the two subtypes have a gray or even a light gray or drab color, break with rough fracture, and approach to the holocrystalline andesites. The pyroxenes, as usual, are inconspicuous and have a slightly resinous luster with green to brown colors.
HYPOCRYSTALLINE ANDESITE, SUBTYPE A.
This variety includes the following specimens: Nos. 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33.
The groundmass of these rocks, as seen under the microscope, is very dense. It is crowded with microlitic crystallizations similar to but more abundant than those in the hyalopilitic groundmass. Glass is probably always present, impregnating the whole, but, with one or two exceptions (26, 28), it is not at all conspicuous. In the two exceptions, which are rocks connecting this type with the preceding, the glass is readily recognizable and has a distinct brown color; otherwise the glass base appears to be almost colorless. As stated above, in discussing the plagioclase phenocrysts, these rocks, with a small amount of colorless glass base, are very apt to contain clear brown glass inclosures in the plagioclase. The marked feature of this subtype A, as distinguishing it from subtype B, is the presence of innumerable minute plagioclase microlites with their customary fluidal arrangement. These microlitic plagioclases are mostly very minute, being often no larger than the augite microlites (26). The larger, squarish, or short-rectangular plagioclase crystals, with and without twinning striae, that are to be seen commonly in the groundmass of the preceding type, are usually conspicuously absent. Hence the contrast between the plagioclases of the groundmass and of the phenocrysts is very pronounced, as connecting types are absent. In one or two specimens the appearance of faintly developed allotriomorphic feldspar (27, 29) connects this type with the holocrystalline andesites.
In the finest-grained varieties the plagioclase microlites are hardly discernible in white light. In such rocks the groundmass presents a dirty-brown or grayish-brown and minutely granular appearance, owing to the development of thickly crowded augite microlites, which are interspersed with minute grains and octahedral crystals of magnetite. The brown color is doubtless due in part to the glass base, and in part also to the presence of brownish globulites. Globulitic forms are, however, by no means as common or abundant as in the hyalopilitic type. On the other hand, both magnetite and augite in microlitic form are much more abundant in this type of andesite. The augite microlites are, as a general thing, not as sharply crystallized, although they do not differ greatly in size from the similar microlites in the hyalopilitic andesites. As the groundmass becomes more distinctly crystalline the size of these microlites tends to become larger and the form less sharp.
A few crystals of apatite of a brownish color measuring 0.1 millimeter in length and 0.015 millimeter in width were noticed. They are faintly pleochroic, with the customary absorption E>O.
Hypersthene in two generations—one in short, stout, and relatively large crystals, the other in slender prisms of much smaller size—are usually to be seen, but the distinction between the two is not as well defined as is the case mentioned in the hyalopilitic type.
In all but one case (33) these andesites appear to be very fresh, the alterations being hardly more than iron stains. In the one case referred to, the ground mass has undergone considerable alteration, which has attacked mainly the glass base and the augite microlites. The decomposition products are carbonates and a nearly homogeneous greenish substance which, in polarized light, appears to be formed of extremely minute scales that polarize light strongly. This green substance appears to be delessite.
The groundmass of subtype A corresponds closely to what Rosenbusch calls the pilotaxitic type of andesites, in that it may be described as a densely felted aggregate of plagioclase and augite microlites impregnated to a more or less extent with an inconspicuous glass base.
HYPOCRYSTALLINE ANDESITE, SUBTYPE B.
This variety of hypocrystalline andesites is represented by the following specimens: Nos. 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, and 58. It is a more abundant type, therefore, than is subtype A. As above stated, these andesites do not differ materially in the hand specimen from the subtype just described. Under the microscope the chief distinction lies in the comparative suppression of the distinctly microlitic plagioclase and in the greater abundance of the larger and broader feldspars that are not always to be sharply distinguished from the phenocrystic plagioclases. In this subtype the rock also is strongly inclined to develop allotriomorphic feldspar, and in this way, as well as in the development of more and more lath-shaped and broad rectangular-shaped feldspars, to pass into the holocrystalline andesites.
Those specimens that are transitional to subtype A (40, 43, 47, 49, 53, 54) have plagioclase laths very abundantly developed, but these are usually so much larger than the microlitic feldspars above described that they may hardly be designated as microlites. At the same time the augite microlites become larger and lose their sharp outline, and glass is hardly distinguishable. Other specimens, transitional to the holocrystalline andesites of the third type, show the development of a small amount of allotriomorphic plagioclase or quartz and contain a great amount of the rectangular feldspar forms (50, 52). It even happens that both these tendencies are manifest in the same rock (54), so that numerous small plagioclase laths in fluidal arrangement occur with a rather vague paste of allotriomorphic material.
The more normal members of this type, although subject to considerable variation in size and relative abundance of the component minerals, as well as in their structural arrangement, may be said to contain numerous phenocrysts of hypersthene and augite, and especially of plagioclase, embedded in a hypocrystalline, or at most very finely crystalline groundmass of plagioclase, augite, and magnetite, with probably a little hardly distinguishable colorless glass base. The glass base may be presumed to be present on account of the presence of a little brownish globulitic matter in the interstices between the feldspars. The augite of the groundmass rarely has as sharp forms, and is also rarely as small as in the two already described types. In a number of the thin sections studied it has a granular or roughly prismatic granular habit. The sharpness of outline seems to disappear almost regularly with the increasing size of the grain of the groundmass. The plagioclase of the groundmass is the most fluctuating ingredient both as to amount and as to size and habit. In some the slender lath-form, in others the short-rectangular habit predominates. The color of the groundmass is in most cases lighter than in subtype A, owing to the greater coarseness of grain and to the scarcity of glass. These andesites even more than those just described possess the strongly felted structure characteristic of the so-called pilotaxitic groundmass of Rosenbusch.
Mr. Diller has briefly described a hypersthene-andesite from the later lavas of Mount Shasta, in California,a that is almost identical with several of the Crater Lake andesites included in subtype B. The resemblance extends both to the groundmass and to the phenocrysts of plagioclase, hypersthene, and augite. It would, in fact, be impossible to tell from the thin section alone from which of the two volcanic areas this rock came. This andesite from Mount Shasta is No. 87 of the Educational Series of Rock Specimens collected and distributed by the United States Geological Survey, and was collected at Horse Camp, near the timber line upon the western slope of the mountain.
aBull. U. S. Geol. Survey No. 150, 1898, p. 227.
Another well-known hypersthene-andesite that resembles this type as far as the groundmass is concerned is the rock from Buffalo Peak, Park County, Colo., described by Whitman Cross.b This rock is No. 86 of the above-mentioned series. It differs considerably from all the Crater Lake andesites, not only of this but of other types, in the much greater abundance of the hypersthene and augite phenocrysts.
bIdem, p. 224.
HOLOCRYSTALLINE TYPE OF ANDESITE.
The following specimens are placed under this type: Nos. 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, and 74. These are mostly light-gray rocks with comparatively rough fracture and with relatively inconspicuous plagioclase phenocrysts. A few specimens of darker color are not thoroughly characteristic of the type.
Under the microscope these andesites do not differ very materially in appearance from those of the hypocrystalline type, when viewed in white light, except that they are much lighter in color and are entirely free from globulitic matter. Between crossed nicols, however, there is a marked difference between these andesites and all the preceding types. This difference is to be seen in the groundmass. The colorless constituents are of two distinct kinds, first, plagioclase laths not markedly different from the similar laths in the other andesites; and, second, a sort of residual product of crystallization occurring in irregular, allotriomorphic patches that have a poikilitic appearance, owing to crowded inclusions of the other groundmass ingredients. These patches vary considerably in size and in conspicuousness, but in general are from 0.2 to 0.5 millimeter in diameter. They do not show polysynthetic twinning, but very often possess a more or less undulous extinction. They are quite distinct from the plagioclase laths which lie embedded in them and appear to play the role of a glass base. The individual patches show usually simultaneous extinctions and are bounded only by meeting other similar patches.
This allotriomorphic material is in some cases undoubtedly quartz; in others it is probably feldspar. It was proved to be quartz in Nos. 63, 66, 71, and 73 by its slightly higher refractive power as compared with adjacent feldspar and by its giving a positive, uniaxial image in convergent polarized light. It is considered to be sometimes feldspar for the following reasons: First, failure to show uniaxial images; second, occasional undulous extinction; third, some of the plagioclase laths have allotriomorphic extensions that extinguish with the outer part of the lath and resemble these patches (63, 64); fourth, in a few of these andesites the lath-shaped plagioclase is nearly missing, so that the groundmass is almost entirely composed of these patches, except for the inclosed augite microlites and magnetite grains. If in this last case all this material were quartz, it would give a rock far too acidic in composition to be classed as an andesite, whereas the close association and resemblance between all of these rocks make it impossible to separate any one of them from this family.
The more characteristic of these holocrystalline andesites have a great abundance of well-defined plagioclase laths with fluidal arrangement, as well as innumerable augite microlites and octahedral crystals and grains of magnetite embedded in a base of these allotriomorphic patches of feldspar and of quartz (60, 61, 66, 71, 74). The augite microlites are, as a rule, somewhat larger and not quite so sharply developed as is the case in the other andesites. The phenocrysts of plagioclase, hypersthene, and augite differ in no essential respect from the descriptions given for all the Crater Lake andesites. Fig. C of Pl. XVI (p. 98) gives a fair idea of the allotriomorphic patches that characterize these holocrystalline andesites.
The presence of quartz in the groundmass of these rocks suggests a much more acid rock than the chemical analyses of Crater Lake andesites given elsewhere in these pages allow. As the only analysis made of this holocrystalline type is of No. 68, which is not thoroughly representative, inasmuch as the allotriomorphic patches are not very well developed, it will be of interest to compare these rocks with a similar rock collected by Mr. Diller and alluded to by Mr. H. W. Turner in his article on the Age and Succession of the Igneous Rocks of the Sierra Nevada.a The rock in question was taken from Crater Peak, California (Lassen Peak quadrangle), and is No. 1829 of the Cascade Range collection. Mr. Turner does not describe the rock, but he refers to it as a pyroxene-andesite, and gives the chemical analysis made by W. F. Hillebrand.b This analysis will be found, repeated on page 94 of this paper. Through the kindness of Mr. Diller the writer was enabled to study a thin section of this Crater Peak andesite and to compare it with the thin sections of the rocks here under discussion. Without any noticeable difference it appears to be identical with the Crater Lake holocrystalline type of andesite. This is true not only of the groundmass, with its allotriomorphic patches inclosing plagioclase laths and augite microlites, etc., but also of the phenocrysts. It will he seen, in comparing the analysis with the analyses of other andesites from Crater Lake, that this contains between 6 and 10 per cent more silica. It is, in fact, closely analogous to the dacites, such as, for instance, those from Lassen Peak, analyzed by T. M. Chatard and W. F. Hillebrand.c It is highly probable, therefore, that these andesites are, in part at least, dacites, although no well-defined quartz is to be seen in them.
aJour. Geol., vol. III, 1895, p. 410.
bIdem, p. 407.
cBull. U. S. Geol. Survey No. 150, 1898, p. 218.
No. 198 probably belongs with the holocrystalline andesites, but owing to extensive alteration its characteristics are not readily made out. The groundmass consists of much homogeneous-looking chloritic matter, with rather faint lath-shaped plagioclase and not a little allotriomorphic, colorless material that is probably feldspar. The magnetite has been altered to leucoxene and the pyroxenes of both the phenocrysts and of the groundmass to chlorite. The phenocrystic plagioclases are partially altered to carbonates and probably to kaolin. A good deal of quartz is to be seen in distinct grains, two or three of which are quite large, while the rest occur in bunches of allotriomorphic grains. As this is the only andesite from Crater Lake that shows well-defined quartz, this mineral is in this case probably secondary in origin.
These holocrystalline andesites, as is the case with the other types, are widely distributed over the Crater Lake area. There may be some significance in the fact that six out of a total of fifteen are situated at the water's edge; that is, a large proportion of them belong to the oldest lava flows exposed. No. 63, which comes from the basaltic cone called Crater Peak, south of the lake, is taken from an ejected fragment, and was doubtless torn loose from a deep-lying rock below.
DACITIC TYPE OF ANDESITE.
This fourth type of andesite is intermediate between the andesites and dacites, but seems to possess characteristics that ally the rocks here included more closely with the andesites than with the dacites. It is possible that a chemical analysis would place one or two of them in the list of dacites.
Of the six specimens placed in this group, three (75, 76, 77) have the dacitic features particularly prominent. The groundmass in these three consists very largely of a nearly colorless glass in which lie myriads of the extremely minute rod-like augite microlites characteristic of the glassy dacites (77). In two (75, 76) the groundmass consists of somewhat larger microlites of the same mineral inclosing minute black specks of magnetite, or possibly of globulites, suggestive rather of the andesites. In addition to these augite microlites very slender feldspar microlites, as well as a few other larger and broader plagioclase crystals, are sparingly developed. The glass base is rendered dark looking, especially in No. 76, by globulitic inclusions and by the presence of a very little magnetite. In Nos. 76 and 77 occur deep rusty-brown and somewhat granulated spots that tend to form around the pores and also to some extent around the plagioclase phenocrysts. The pores are apt to contain a little tridymite, partially or entirely filling them. These darker, iron-rusted spots bear a slight resemblance to the spherulitic growths in the dacites, but do not appear to possess radiated structure.
The three other specimens (78, 79, 80) are similar, so far as the augite microlites are concerned; but they contain little or no glass base, instead of which there is to be seen much more abundant plagioclase laths, as well as squarish or short rectangular feldspars with undulous extinction, or, rather, with zonal structure but no twinning; likewise small irregular or allotriomorphic colorless shreds of feldspar or, perhaps, of quartz.
The phenocrysts in these six rocks are not essentially different from those in other andesites (brown glass inclosures, abundant in plagioclase of 75). Hypersthene in two generations—the older in short, stout, and relatively large crystals, the younger in small, long, and slender prisms—is clearly developed in No. 78.
The last three rocks, without perceptible glass, are very light colored in the hand specimens while the glassy varieties are much darker.
UNCERTAIN TYPES.
With the exception of those of Wizard Island, the andesite lavas of Crater Lake do not appear to be distributed in accordance with the types herein described, but the various types occur over all the area, among both the older and the younger andesite flows. In the case of Wizard Island a partial exception may be noted in that the more thoroughly crystalline types are almost wanting. The andesites of Wizard Island are unusually black and basaltic-looking, and contain much brown glass. Occasionally they are stained a deep brownish red by iron oxide (87, 89). Some of the more glassy black portions seem to have cracked in the process of cooling, the cracked surfaces being left with a decided gloss. This was particularly noted on a specimen collected by the writer from a large block in the crater on the summit of the island.
Of the thirteen rocks from Wizard Island studied in thin section, five are of the hyalopilitic type; six of the hypocrystalline type, subtype A; two of the same type, subtype B; while only one, and this a somewhat doubtful one, is of the holocrystalline type.
Among the andesites collected by Mr. J. S. Diller on Wizard Island are ten Nos. 81 to 90, inclusive, the exact location of which is not given. Of these No. 83 has no corresponding thin section. The others do not vary greatly from the other specimens from this island enumerated above. Nos. 82, 86 and 88 are placed in the hypocrystalline type, subtype A; or possibly 82 might equally well be put in the hyalopilitic type. No. 84 is a hyalopilitic andesite. Nos. 88, 89, and 90 are so strongly impregnated with red hematite powder as not to be easily classified. They are probably of the hypocrystalline type, subtype A. No. 85 is a holocrystalline andesite with basaltic appearance. It contains small augite prisms, too large to be called microlites, and also considerable allotriomorphic feldspar in the groundmass but no poikilitic patches such as occur in most of the holocrystalline andesites.
At the head of Steel Bay, just to the east of Llao Rock, occurs an andesitic rock close to the dike from which No. 96 (page 93) was collected. This andesite is represented by No. 91, and is a nearly holocrystalline rock with very abundant phenocrysts and with a groundmass that resembles that of some of the basalts, notably Nos. 176 and 177 of the porphyritic interstitial basalts. The main features that distinguish this rock from the basalts are the absence of olivine and the very marked porphyritic development. This andesite does not closely resemble any other andesite from Crater Lake, at least not close enough to justify putting it in any of the above described types. It more closely resembles the basalts than any other rock collected on or within the crater rim.
| <<< Previous | <<< Contents >>> | Next >>> |
pp/3/part2-1a.htm
Last Updated: 07-Mar-2006