Geological Survey Professional Paper 3 The Geology and Petrography of Crater Lake National Park

HYPERSTHENE-DACITES.

In all the descriptions of the rocks of Crater Lake heretofore published the rocks which are here designated as dacites have been called rhyolites. This is not to be wondered at, as these rocks present all the outward characteristics of rhyolites, and even a microscopic examination does not at first give strong ground for the change of name. Previous descriptions have been based largely upon field observations. But, in spite of the fact that comparison may frequently be made and is made in this paper between these dacites and rhyolites of well-known occurrence, a careful study of the mineralogical characteristics, supported by the chemical analyses, has led the writer to the conviction that the change of name is justified. As will appear later, the determination of these rocks as dacites is further strengthened by resemblance to other dacites in the adjacent regions of northern California.

As will be seen to be the case with the basalts the dacites of Crater Lake are somewhat local in their distribution. As they are chiefly confined to distinctive lava flows that, with one exception, represent the latest eruptions to be seen on the rim, and whose outlines can be definitely traced, it is possible and desirable to treat each of the dacite flows separately. Beginning with the most conspicuous flow, that of Llao Rock, and passing around the lake toward the east, the dacite flows will be described under the following heads: Llao Rock, Grouse Hill, Cleetwood Cove, Wineglass, Cloud Cap, Sun Creek. Outside of these six areas, as well as within the same, there occurs abundant dacitic material, in the form of bombs and tufaceous matter, that will also receive separate consideration.

As is usually the case with dacites, the groundmass of these rocks presents a most remarkable variety of structures, and upon the variations in the groundmass most of the distinctions are to be based. To a certain extent this is true of the phenocrysts, at least as far as their relative abundance is concerned. Upon the whole, however, the Crater Lake dacites, irrespective of the extreme variations in the groundmass, are characterized by the presence of well-defined phenocrysts. These minerals are labradorite, hypersthene, and brown hornblende, all of which are usually present, also augite and olivine, which may occasionally be seen.

MINERAL COMPONENTS.

FELDSPAR.

While phenocrysts of feldspar are fairly abundant in all these dacites, and in some specimens very numerous indeed, in no case could orthoclase be detected. Not only is it true that the feldspars turn out to be plagioclase, but plagioclase of a decided basic variety, usually labradorite. In form they do not appear to differ materially from the similar phenocrysts in the andesites. There are to be seen comparatively large, stout crystals with the basal pinacoid and brachypinacoid, the two half prisms, and an additional dome or pyramid; also crystals that are more nearly rectangular in cross section and usually smaller in size. The stout crystals with more elaborate forms are probably older than the others, at least they appear to have greater extinction angles and to belong to a more basic plagioclase. These crystals are usually not numerous enough in any one thin section to allow the accurate determination of the maximum extinction angles; the extinction angles given below, therefore, although they are the maximum observed, probably do not indicate feldspars as basic as those actually present. Sections cut perpendicular to the twinning plane, and therefore showing symmetrical extinctions, gave the following maximum extinction angles, viz: 30°. 101, 31°. 104, 30°. 197, and 33° in No. 103. All four of these specimens came from the dacite of Llao Rock. No. 101 is a vitrophyric dacite from the southern edge of the flow, and is almost identical with No. 102, of which a chemical analysis will be found on page 140. No. 104 is a spherulitic variety of a vitrophyric type, and Nos. 197 and 103 are approximately holocrystalline types. These measurements indicate a plagioclase at least as basic as labradorite, and probably more so. The above extinctions represent, however, not the whole crystal, but the inner part. The outer shell often gives much smaller angles. For instance, in No. 101 the margin gives an extinction angle of 24° is six degrees less than at the center, but even this is not too small an extinction for labradorite. The measurements given for the plagioclase phenocrysts of the Llao Rock flow do not differ materially from the observed extinction angles in the other dacite masses.

Zonal structure is very strongly developed in these larger plagioclase phenocrysts, and is particularly conspicuous in sections cut approximately parallel to the brachypinacoid. In such sections the zonal banding, as seen in polarized light, indicates that the crystals in the earlier stages of growth had simpler forms. For instance, in a section parallel to the brachypinacoid, showing externally traces of the basal pinacoid, prism, and two domes, the central core shows only the basal pinacoid and one of the domes almost at right angles to the first-named form. In such sections the zones of different extinctions usually shade gradually into each other, so that from the center outward the extinction angle becomes less and less oblique to the trace of the basal pinacoid. Almost always, however, there are to be seen one or more quite sharply defined shells with rather abrupt difference in extinction. Furthermore, this abruptness of change from one shell to the next is not infrequently accentuated by the fact that the extinctions do not change regularly from the center outward, but oscillate more or less. In other words, the plagioclase consists of concentric shells that alternate between less acid and more acid feldspars. As far as observed, the actual center is nearly always the most basic portion of the feldspar, but the shell immediately surrounding this center may be more acid than the next succeeding one. The alternation of more and less acid shells is not usually sharp enough to admit of positive measurement. In the two following cases the measurements were sharp enough to justify recording.

In No. 102 occurs a section of plagioclase that is cut approximately parallel to the brachypinacoid and that shows the alternation of zonal shells very clearly. This may be seen illustrated in fig. F of Pl. XIV (p. 76). The three most conspicuous zones are marked 1, 2, and 3, from the center outward. The crystal forms that could be identified by means of the cleavage cracks are the basal pinacoid (001) and the prism (110). Two other plagioclase crystals are grown into this one—one almost at right angles, to be seen on the left side of the figure; the other, in the upper left corner of the figure, appears almost to continue the outlines of the main crystal. These two crystals do not appear to be in twinning relationship to the other. The extinction angles, as measured to the trace of basal pinacoid, as well as the corresponding percentage of the anorthite molecule, are given below for the three zones:

1 = -12° per cent An = basic andesine.
2 = -21° per cent An = labradorite.
3 = -5° per cent An = basic oligoclase

The percentages of An are given as corresponding to extinction angles measured on a section exactly parallel to the brachypinacoid. That this is not quite true of this section is proved by the fact that the angle of equal illumination for the three zones is +29° the required angle of +34°. As these extinction angles do not indicate quite as basic a plagioclase as do the extinction angles on symmetrical sections, it is probable that a section cut exactly parallel to the brachypinacoid would give still larger extinction angles than are here indicated. But, at least, these measurements suffice to prove that in this case the inner portion of the crystal is not as basic as is the intermediate zone.

In No. 101 was seen another section of plagioclase, cut similarly to this one, and also showing similar extinctions. Like the example given above, 1 is the center and 4 the margin.

1 = -22° labradorite.
2 = -14° basic andesine.
3 = -16° basic andesine.
4 = -5° basic oligoclase.

A comparison of the refractive powers of the plagioclase phenocrysts in the rocks containing these two crystals shows that they have higher refractive power than has the adjacent Canada balsam, i. e., higher than 1.540. This would indicate that they are more basic than oligoclase, even at the edge of the crystal.

The plagioclase phenocrysts are frequently broken and the fragments scattered through the glassy groundmass. They also show frequent corrosion, but, unlike the andesitic plagioclases, they do not often contain abundant glass inclusions; at least, the crowding with glass inclusions and their distribution in an intermediate zone is not characteristic. An exception must be taken to this statement, however, in favor of the dark-colored secretions that receive special treatment further on in these pages. Inclusions of slender apatite needles are common; also zircon crystals in short colorless or slightly brownish prisms may be seen, but very sparingly.

The plagioclase that belongs more particularly to the groundmass seems to be oligoclase, on account of the very small extinction angles. A more detailed description of the groundmass feldspars will be found in connection with the description of the different dacite flows.

Orthoclase and quartz are also entirely confined to the groundmass and will be discussed later. Tridymite is not common. It occurs in the customary clusters, apparently filling small cavities (121) also to some extent in the fluidally arranged lithoidal dacites.

HORNBLENDE.

This is a very characteristic ingredient of the Crater Lake dacites. While it is almost entirely wanting in the andesites, it is conspicuously present, although in only occasional crystals, in the dacites, or, at least, in most of them. Out of twenty-eight thin sections prepared from different rock specimens, hornblende could be found in all but five. True, it is to be seen in a few of the thin sections in very small crystals, and in none of the rocks does it assume a prominent role as far as quantity is concerned, but it is always very easily recognized by its peculiar color and characteristic pleochroism. It occurs always in long slender prisms, several to many times as long as thick. The prism invariably, and sometimes also to a slight extent the clinopinacoid, is strongly and sharply developed. Terminal faces are often wanting, and when present appear to be the usual flat basal or pyramidal faces. They are frequently minute and are rarely over 1.5 millimeters in length. These hornblendes may all be classified as belonging to the brown variety, although there is really a great range in color. They may be further divided into, first, brownish-green; and second, brownish-red varieties.

The first, or brownish-green hornblende, is much more common than is the second variety. It may be studied in Nos. 102, 117, and 118. Pleochroism is very strong. c = dark olive green with usually a trace of brown, b = dark greenish brown, a = lemon yellow to greenish yellow. At times the rays vibrating parallel to c are almost a pure deep olive green. In any event brown is most conspicuous in the rays vibrating parallel to b. This hornblende has a very unusual absorption. in that the absorption parallel to b is greater than that parallel to c, thus, b>c>a. In some cases there appears to be very little difference in color and absorption between b and c, but wherever there is a marked difference the absorption is as given above. Professor Rosenbusch^a refers to an observation by A. Osann of a hornblende phenocryst in an andesite from Hoyazo, Cabo de Gata, which has not only the same absorption but also nearly the same colors as have these hornblende phenocrysts from Crater Lake. The pleochroism as given is as follows: a, light greenish yellow; b, greenish brown; c, dark greenish yellow; and b>c>a.

---

^aMikroskopische Physiographie, 3d edition, Vol. I, 1892, p. 558.

---

The extinction angles for these hornblende crystals are very small, 6° or 7° being the maximum observed. Further optical properties, as far as could be observed, appear to correspond with those of the similar hornblende that forms a large part of some of the so-called secretions to be described later in this paper. Twinning parallel to the orthopinacoid is very common.

The dark brownish-red variety of hornblende is to be seen in Nos. 104, 110, 112, and 114; it also occurs less abundantly elsewhere. In form it does not differ from the above-described variety. The pleochroism is fully as marked, but the colors are very different, c = dark brownish red, b = reddish brown, a = yellow, with c>b>a. The absorption is normal. The deep brownish-red color of the rays vibrating parallel to c are very striking and characteristic. This color is sometimes almost blood red and reminds one strongly of the color of hematite in very thin scales. The extinction angle does not differ materially from that of the first variety.

In one or two cases both of these varieties of hornblende may be seen in the same thin section (104). They are both of them usually perfectly sharply crystallized and unaltered. Only in rare cases can a partial resorption with development of black rims be noted (124), and, from the very few sections observed, it is not possible to state whether both of the varieties may have such resorption rims, but it appears as though this were the case. As far as can be seen, there appears to be no difference in the relative ages of these two varieties. They are the youngest of all the phenocrysts. The greenish-brown variety, at least, contains inclosures of plagioclase, hypersthene, and augite. The hornblende occasionally appears with other minerals in the form of nests and then is not as apt to occur in slender prisms as in more irregular grains.

HYPERSTHENE.

This mineral is not as abundant as it is in the andesites, but it is never entirely lacking. It occurs in the same forms—namely, prism, two pinacoids, and flat terminal faces, in the same habits and with the same color, pleochroism, and inclosures of glass, etc., as it does in the andesites. The reader, therefore, is referred to the description of this mineral as given under the andesites on pp. 78 to 82. It could not be noticed that the pleochroism is less pronounced or the colors less deep than they are in the andesitic hypersthenes, so that there is no reason for considering these orthorhombic pyroxenes as enstatite rather than hypersthene. The crystals are mostly very sharp, although they may at times show some rounding of the corners. In only one case was a resorption noted, accompanied by the development of a dark, blackish-red, opaque rim (112).

AUGITE.

This mineral is to be seen in about one-third of the thin sections studied. It is much less abundant than hypersthene and is inclined to occur in more or less irregular grains rather than in sharp crystals, although the latter are by no means lacking. In color it is a pale green without noticeable pleochroism. It has been noticed to inclose magnetite and apatite. It is younger than hypersthene and also, as a general thing, younger than plagioclase, but is older than hornblende.

OTHER MINERALS.

Olivine was observed only in the older secretions that accompany these dacites and will be referred to under that head. Apatite and zircon occur as occasional inclosures in the different phenocrysts, as does also magnetite. The first of these, however, is not easily found. Its color is sometimes a distinct brown, and the form roundish prismatic.

The order of crystallization for the above minerals, leaving out these last-named accessory minerals is as follows: Hypersthene, plagioclase. augite, and hornblende, with some variability as to the plagioclase, as this mineral is at times older than hypersthene.