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

HYPERSTHENE-DACITES.
(continued)

DACITIC EJECTAMENTA.

The large amount of volcanic ejectamenta that covers the surface of the crater rim in various places, especially on the northern and western sides, and that forms extensive deposits over the so-called Pumice Desert, appears to be almost entirely dacitic in character.

Some dark-colored sand, No. 134, collected on the Pumice Desert is composed of the light- and dark-colored minerals occurring as phenocrysts in the dacites, as well as of brown glass. These are plagioclase, hypersthene, hornblende, augite, and magnetite. The size of these crystals varies from about 2 millimeters downward. Plagioclase forms the most abundant of the sand particles, with forms that are usually roundish but that also may be well crystallized. Hypersthene, on the other hand, is very common in the characteristic prismatic forms with the unit prism and two pinacoids. When finely pulverized and examined under the microscope it is seen that both hypersthene and dark greenish-brown hornblende are very abundant and present in about equal amounts. Augite and magnetite are rather scarce. The latter can readily be separated from the sand by a magnet. This sand exactly resembles the coarser parts of the tuft found in the bottom of Sand Creek Canyon several miles below the crater rim (128) and in Anna Creek Canyon (129). The accumulation of the pyroxenes and of hornblende to a much greater extent than in the dacites is very noticeable in both the tuff and sand deposits.

A sediment brought up from the bottom of the lake on the west side, No. 1357 appears to be composed of somewhat similar material as the above, but as it is very fine—the largest grains being not much over 0.1 millimeter—and entirely composed of angular fragments, the percentage of the dark-colored phenocrysts, or rather of fragments of phenocrysts, is comparatively small. This is particularly true of hornblende, which is very scarce. On the other hand, glass fragments are very abundant. The nature of this sediment is not so well characterized as to make certain whether or not andesitic material is largely mixed with the dacitic ash.

The coarser ejectamenta which have been collected and submitted to the writer for investigation may be divided into three classes, viz, pumice, dark-colored secretions, light-colored granophyric secretions.

PUMICE.

No. 102, from the summit of Llao Rock, is an almost pure pumice of a pinkish white color and extremely light and porous. With a magnifying glass one can detect a very few glassy plagioclase crystals, and also still fewer dark-colored pyroxene or hornblende phenocrysts, similar to the phenocrysts in the Llao Rock dacite. Under the microscope this rock presents the customary porous, pumiceous structure, with long drawn out gas pores; but the glass differs from the vitrophyric portions of the Llao Rock flow, in that the very characteristic rod-like augite microlites are almost entirely missing. A few phenocrysts of plagioclase in crystals and crystal fragments are to be seen, also an occasional hypersthene and two or three minute brown hornblende crystals. One of the plagioclase crystals was seen to contain numerous inclosures of apatite and also several dark-brown glass inclusions, in one of which was an air bubble. No. 137 is a fragment of pumice collected from the same place, and differs only in being more discolored. It contains similar phenocrysts, and also brownish glass inclosures with gas bubbles in the plagioclase. This pumice evidently is of the same character as the dacite of Llao Rock.

No. 138 is a specimen collected from a large fragment in the Pumice Desert, the north of Crater Lake. It has a brownish-yellow color and looks as though it contained more or less clayey matter. It also has a fragmental appearance as though it were a brecciated rock composed mainly of pumice. It also contains a few small fragments that appear to be andesite. Plagioclase grains are very abundant in the hand specimen, but the thin section contains hardly any, as they have probably disappeared in the grinding. A very few minute hypersthene and augite crystals and one small fragment of reddish-brown hornblende were noted.

In addition to this ordinary pumice are also fragments of a distinctly pumiceous rock of light color in which are readily seen numerous black crystals, mostly hornblende. These hornblende crystals vary in size from one-fourth of an inch downward. Probably one-eighth of an inch in length is a fair average size. One of these fragments was collected on the northeast rim of Crater Lake, east of Round Top, and another (146) from the summit of Red Cone, while two others were collected by the writer on the crater rim just south of Llao Rock (2011, a and 2013, 2 of the private collection of H. B. Patton). The ferromagnesian minerals which are very conspicuous in the hand specimen are still more apparent in the thin section, as, indeed, are also the colorless plagioclase phenocrysts. The groundmass of these specimens does not appear to differ materially from that of the more typical pumice. It is a nearly pure glass, full of elongated air cavities and more or less stained with yellowish to brownish ferritic matter. Only a few augite microlites and but little magnetite dust is to be seen. On the other hand all the minerals that have been mentioned as occurring among the phenocrysts in the Crater Lake dacites and that are there sparingly developed are unusually numerous in these specimens. The minerals here included are plagioclase, hornblende, hypersthene, and augite. All but the hornblende appear to occur in about the same sizes as in the dacites. In the case of plagioclase there is developed a very strong tendency to inclose comparatively large and irregularly lobed and more or less connected inclusions of very deep brown glass. These inclusions are not arranged around the margin in a well-defined zone, as is so frequently the case with the plagioclase phenocrysts of the andesites, but they are more evenly distributed throughout the crystal and produce a structure that may fairly be characterized as spongiform. These spongiform plagioclases are by no means as common or characteristic as they are in the dark-colored secretions whose descriptions follow; neither do all or even most of the plagioclase phenocrysts have these inclusions, but their appearance here is important as forming a link between the dacites and the inclusions in the same. Fracturing of the feldspars is very noticeable.

Hypersthene and augite occur in the same forms and with the same general properties as in the dacites. Perhaps a stronger tendency is to be noted toward rounding of the edges, particularly at the ends. Likewise inclusions of glass and especially of brown glass are more abundant. The most interesting mineral is hornblende, whose abundance is increased relatively much more than is the case with the other phenocrysts. It also occurs in very much larger crystals than are to be seen in the dacites. Except for abundance and size, however, these hornblende crystals do not present any marked peculiarities. They show a strong tendency to develop the unit prism and the clinopinacoid, but they are usually broken into fragments like the plagioclase. In color they are similar to the hornblendes of the dacites. Greenish brown predominates, but the more reddish colors are also to be seen, notably in No. 139. Magnetite occurs rather sparingly in these rocks in the form of distinct grains inclosed in the phenocrysts as well as scattered through the glass groundmass.

DARK-COLORED SECRETIONS.

These are found as ejected fragments together with pumice. They occur as extremely dark, almost black, very porous and rough feeling bombs, and contain a great abundance of black pyroxenes and black hornblende as well as of the glassy plagioclase. They have a wide distribution, as may be seen from the following list: No. 141 comes from the Pumice Desert to the north of Crater Lake; No. 142 from near the summit of Dutton Cliff to the south of the lake; No. 143 from the summit of Llao Rock; No. 145 is a fragment from a conglomerate that overlies a small sheet of dacite just above Grotto Cove; and 2011, b (private collection of the writer) forms part of a large bomb attached to the pumice specimen, 2011, a, above described. This last was collected just south of Llao Rock. In spite of the dark color of the rock the ferromagnesian minerals are quite conspicuous. This is particularly true of No. 2011, b. where hornblende is more abundant than in the other specimens and is recognized by means of the brilliantly flashing cleavages. This mineral occurs here in crystals that measure up to 10 millimeters in length and 2 to 3 millimeters in width.

Under the microscope in thin section these rocks are seen to contain a very dark brown, glassy groundmass that incloses but few augite microlites but is crowded with air cavities. In color and general appearance the groundmass closely resembles the glass inclusions to be seen on the plagioclase crystals of the hornblendic pumice above mentioned. It forms from one-third to one-half of the rock mass, the rest being composed of phenocrysts and fragments of phenocrysts.

The plagioclase crystals belong to the type described under the dacites as occurring in rather large, plump crystals, with numerous crystal faces. Rectangular sections are rare, so rare, in fact, that they are probably only chance sections of crystals of the type just mentioned. The plagioclase crystals show a strong tendency to assume idiomorphic forms except in so far as they occur in shattered fragments. but they are generally crowded with irregular brown glass inclusions so as to assume a thoroughly sponge-like appearance, similar to but much more pronounced than the spongiform crystals in some of the pumice fragments. These ramifying glass inclusions are not confined to the interior of the crystal, but come to the surface and often appear as embayments filled with brown glass exactly like and continuing the glass of the groundmass. Figs. A, B, and C of Pl. XVII are reproductions of some of these spongiform plagioclase crystals and give a fair idea of their appearance. The plagioclase is probably all basic, and, in some cases, demonstrably so. For instance in No. 143 a honeycombed plagioclase crystal, with zonal structure, and cut symmetrically to the plane of twinning gave an extinction angle of 47° center and 37° margin. This indicates an extremely basic anorthite, more basic, in fact, than the most basic of the plagioclases whose extinction angles were measured in either the andesites or basalts of Crater Lake. Not all of the plagioclase crystals show polysynthetic twinning, but most of them show either this twinning or a zonal structure, and no reason was apparent for considering any of them to be orthoclase. They contain inclusions of ores apatite, and hypersthene

The hornblende in these dark-colored secretions is partially idiomorphic and partially allotriomorphic, with the latter tendency decidedly the stronger; that is, crystal forms may be developed on part of a crystal while the rest has a very irregular outline. The observed forms are the customary unit prism and clinopinacoid and flat terminal faces. The color is mostly like that of the greenish-brown hornblende of the dacites, and the absorption is b > c > a. The customary colors in thin section for rays vibrating parallel to the vertical axis is a deep brownish green to greenish brown, in accordance with the thickness of the section. The absorption and pleochroism parallel to b are almost exactly like that parallel to c, with the absorption usually a little stronger parallel to b than to c. In No. 145 this mineral appears to be unusually scarce and the color in rays vibrating parallel to c a deep brownish red. Those parallel to b can not be determined in this section. In No. 141 there is a peculiar mingling of the reddish with the greenish-brown colors. Most of the crystals belong to the greenish-brown variety, but others have both colors shading into each other on the same crystal. This shading of colors occurs on perfectly fresh crystals. In fact, all the hornblende crystals in these specimens are absolutely fresh, and do not show any tendency to resorption or to the development of black borders. The extinction angle of the hornblende, as measured on cleavage fragments from No. 143, is 8° to the vertical axis. The optical axial plane lies in the plane of greatest and least absorption; that is, in the plane of the clinopinacoid. A section of this mineral in No. 143 that gave a prism angle of 104°, as measured by the trace of the prism faces with the plane of the section, gave a bisectrix in the center of the field.

Plate XVII.—THIN SECTIONS OF MINERALS IN SECRETIONS IN DACITE.

FIGS. A, B, and C.—Magnified respectively 40, 25, and 15 diameters. All from Specimen No. 145. From a dark-colored secretion among the dacitic ejectamenta. Show spongiform plagioclase crystals inclosing deep-brown glass similar to that composing the groundmass from a dark-colored secretion among the dacitic ejectamenta. See page 126.

FIG. D.—Magnified 18 diameters. Specimen No. 144. From a dark-colored secretion among the dacitic ejectamenta. Shows a grain of hypersthene inclosing numerous plagioclase grains, and two irregular grains of olivine marked 0. See page 129.

FIG. E.—Magnified 50 diameters. Specimen No. 148. A light-colored, granophyric secretion among the dacitic ejectamenta. Shows crystals of plagioclase with fringes of spongiform sanidine containing inclusions of brown glass and resembling granophyric intergrowths. Illustrates a section as seen in polarized light with crossed nicols, the glass appearing black. See page 136.

FIG. F.—Magnified 10 diameters. Specimen No. 142. From a dark-colored secretion among the dacitic ejectamenta. Shows a large crystal of brownish-green hornblende inclosing crystals of hypersthene = H, plagioclase = P, and magnetite. See page 129.

The hornblende of these secretions does not contain inclusions of glass, but it is very apt to contain inclusions of the other phenocrysts so as sometimes to be filled with them, somewhat after the manner of inclusions in the hornblende of the crystalline schists. Fig. F of Pl. XVII gives one of these hornblende crystals containing inclusions of hypersthene, plagioclase, and magnetite. In other cases augite may be seen inclosed in hornblende. These minerals when thus inclosed may have idiomorphic forms, but they are more apt to be hypidiomorphic. At least their forms are not always so sharp as when they occur embedded in the glass base.

Hornblende occurs in much larger crystals than do the other minerals. It is very fluctuating in quantity, being very sparingly developed in No. 145, or at least in the thin section prepared from this specimen, and extremely abundant in No. 2011b. Fig. B of Pl. XV1II reproduces a portion of thin section of No. 2011b, showing the relative abundance of the hornblende. This is not a fair average of the whole section, but shows the hornblende somewhat more abundantly than in the rest of the section. The crystal form is unusually sharp in this rock.

Hyperstheine is by no means so abundant as hornblende, nor does it occur in so large crystals. It is more inclined to idiomorphic forms, but it also shows a tendency to contain inclosures of other minerals. Fig. D of Pl. XVII gives a section of hypersthene from No. 144 containing inclosures of plagioclase and two irregular inclusions of olivine. This is exceptional. As a rule, hypersthene is older than plagioclase and even in this case the irregular form of the plagioclase crystals indicates that their period of formation did not greatly precede that of the hypersthene. The optical properties of the hypersthene are in no respect different from what may be seen in the hypersthenes of the dacites.

Olivine is to be found very sparingly developed and always inclosed in either hypersthene or hornblende.

Augite is much less common than hypersthene. It has a pale—greenish color and but little pleochroism. It is younger than plagioclase and hypersthene and older than hornblende. It occurs both in roundish grains and in fairly developed crystals similar to those in the dacites.

The order of crystallization of the phenocrysts of these secretions is in general as follows: 1, ores and apatite; 2, olivine; 3, hypersthene; 4, plagioclase; 5, augite; 6, hornblende. At times the plagioclase appears to change places with hypersthene.

That these dark-colored bombs are in reality secretions from the dacitic magmama fairly well be established by the similarity of the mineral contents with the phenocrysts of the dacites. This is more especially true of hornblende, which, with possibly one minor exception, is absolutely foreign to the andesites as well as to the basalts of Crater Lake, but which occurs with the same colors and other properties, although very sparingly, in nearly all the thin sections of dacites studied. That these fragments were erupted at the same time as the dacites is indicated by their occurring with the pumice of the dacites and more immediately by the finding of a bomb in the midst of the pumiceous deposit south of Llao Rock consisting in part of pumice and in part of this black, glassy, hornblendic rock.

The chemical analysis of No. 143 will be found on page 140. In spite of the evident relationship of these dark-colored secretions to the dacites the analysis shows a marked difference. This is particularly noticeable in the large amount of lime and in the corresponding decrease in the alkalies. The analysis, in fact, bears a close resemblance to the analyses of the more acid basalts Nos. 158 and 189.

LIGHT-COLORED GRANOPHYRIC SECRETIONS.

These are not to be considered secretions in the sense that this term is frequently used in describing the accumulations of older minerals that form dark-colored inclusions in igneous rocks, which is the sense in which the word is used in describing above the dark-colored secretions. But, rather, these light-colored secretions appear to represent local crystallizations or differentiations of the same minerals that are to be seen in the dacites in general, but so aggregated together as to appear in totally different structural relationships. They have not been found in place or inclosed in dacite, but only as loose fragments or bombs on the surface.

Plate XVIII.—THIN SECTIONS OF SECRETIONS IN DACITE.

FIG. A.—Vitrophyric dacite and secretion in the same, from Llao Rock. Magnified 48 diameters. Specimen No. 102. A photomicrograph in while light. On the left is the dacite, consisting of a glass with a multitude of augite microlites. On the right is the secretion, composed of a felt of plagioclase laths and slender prisms of reddish-brown hornblende, which appears nearly black in the figure. The interstices of this felt are filled with light-brown glass. The lack of a sharply defined line of junction between the rock and the secretion is characteristic. See pages 105 and 110.

FIG. B.—A photomicrograph in white light. Magnified 20 diameters. Specimen No. 2011 b. Illustrates the abundance of reddish-brown hornblende in one of the dark-colored secretions. See page 129.

FIG. C.—A light-colored granophyric secretion from the dacitic ejectamenta. Magnified 20 diameters. Specimen No. 151. Shows a dark-colored brown glass and spongiform mantles of sanidine around plagioclase. See page 136.

FIG. D.—A photomicrograph in polarized light with crossed nicols. Magnified 86 diameters. Shows an enlarged portion of fig. C, giving an individual plagioclase grain with mantle of spongiform sanidine. The black portions are glass.

No. 146 is a small fragment about an inch square that formed part of a conglomerate overlying the small dacite flow immediately above Grotto Cove. Two other fragments from this same conglomerate have already been described. These are No. 139, a pumice fragment containing much hornblende, and No. 145, one of the dark-colored secretions. This association would suggest the same source as that of the dark-colored secretions. In the hand specimen this appears to be a fairly coarse grained, holocrystalline rock of very light color and composed of a whitish, glassy appearing, granular aggregate dotted with small, dark-colored grains. It is quite brittle and crumbles in the fingers. In thin section the rock is seen to consist mainly of plagioclase, which constitutes perhaps one-half of the whole or a little less, in addition to which there are hypersthene, augite, hornblende, and a very little biotite. All these ferromagnesian minerals together form but a small part of the whole. They are perhaps a little more abundant than is usual in the dacites of Crater Lake, but not markedly so. The rest of the rock forms a sort of groundmass either surrounding these older minerals or filling the spaces between them wherever the phenocrysts are close enough to touch each other. This groundmass consists in part of a deep brown and perfectly clear glass, but mainly of a beautifully sharp and characteristic granophyric or micropegmatitic intergrowth of quartz and sanidine. The presence of quartz could be demonstrated by observations in convergent polarized light, in which it gives the positive uniaxial cross, as well as by the more readily observed properties characteristic of the mineral. In the case of sanidine the demonstration is not complete, but is based upon the following observations: It is perfectly clear and colorless and contains no inclusions except minute air cavities; both the refractive power and the double refraction are less than in the adjacent quartz; its index of refraction is less than that of the surrounding balsam; in convergent polarized light it gives a biaxial image; it is entirely free from twinning striation or from zonal structure; this sanidine usually forms a mantle around the plagioclase crystals, so that either the outer part of the sanidine mantle is intergrown with quartz in granophyric fashion or, more usually, this granophyric intergrowth begins close to the plagioclase edge, with little or no free mantle growth visible; the contrast between the plagioclase and the sanidine mantle is usually rather sharp; the plagioclase very frequently shows no twinning striae, but the concentric zonal structure is always visible, whereas the sanidine mantle is entirely free from all twinning and extinguishes simultaneously. Becke's method for determining the index of refraction, as compared with that of Canada balsam, by focusing sharply on the edge of a crystal in contact with balsam and then observing the movement of the white band upon raising or lowering the focus, was found very applicable in studying this sanidine mantle. Where a crystal of plagioclase was broken and in contact with balsam it was invariably found that the portion of the crystal that showed twinning or zonal structure had a higher index of refraction than balsam, while the outer, untwinned, and unbanded portion was lower in refractive power than balsam. The same method is equally applicable in this case in distinguishing between sanidine and quartz.

Wherever the quartz or the sanidine comes in contact with the brown glass base it breaks up at the edge into an irregularly lobed fringe, with the roundish lobes interwoven or intergrown with the glass in a way suggestive of the granophyric intergrowths themselves. This interweaving of glass with the fringed edges of the crystals is so much more beautifully developed in other specimens whose description follows that a detailed description of this phenomenon will be postponed for the present. The quartz is not present except in these granophyric aggregates. It incloses air cavities similar to the sanidine.

The plagioclase in this rock does not have as sharply developed forms as in the dacites. This is true even when the surrounding sanidine mantle is disregarded. Sections suitable for determining the extinction angles were not found, but such extinction angles as were noted indicate a basic plagioclase. This is further corroborated by the fact that the plagioclase even at the edge has an index of refraction greater than Canada balsam, i.e., greater than 1.540. A few inclosures of brown glass with gas bubbles and also small apatite prisms were noted.

The ferromagnesian minerals all have poorly developed forms. Hypersthene and augite occur mostly in small prismatic grains, some singly, but more commonly in more or less parallel arranged groups, together with magnetite and hornblende and with dark greenish-brown biotite. The hornblende has a greenish-brown color and occurs quite abundantly in longish granular or prismatic individuals. Biotite is scarce, but its presence is the more noteworthy as it appears to be entirely wanting in the dacites proper.

The writer has not seen a full description of rocks similar to this, but Professor Rosenbusch^a makes brief reference to several quite analogous occurrences in Iceland, the original descriptions of which are not at the time of preparing this paper accessible to the writer. In these Icelandic rocks, which occur as inclosures in basalt and in loose fragments, are to be seen a similar mineral aggregation with plagioclase crystals surrounded with orthoclase mantles and granophyric growths.

---

^aMikroskopische Physiographie, 3d edition, vol. II, 1896, p. 590.

---

Among the Crater Lake rocks collected in 1883 there are five specimens, Nos. 147, 148, 149, 150, and 151, which are labeled as being ejected volcanic fragments from the top of the divide between Sand Creek and Anna Creek, near the south rim of Crater Lake. These five specimens are closely analogous to No. 146, just described, and in one or two cases practically identical in all essential points. No. 149, for instance, is very light colored, distinctly granular, and very friable. It differs slightly in color, in that the feldspars have a distinctly pinkish color. On the other hand No. 148 is very much darker and looks as though it were composed about equally of a white granular feldspar and of black glass.

Under the microscope the practical identity of all of these specimens with No. 146 is very evident. There are the same marginal growths of sanidine around the plagioclase crystals, and the same granophyric intergrowths of sanidine with quartz. The deep-brown glass interweaving with the sanidine is also present, but in much greater abundance in two or three of the specimens. The most striking difference between the two occurrences is in the fact that both hornblende and biotite are totally lacking here. Owing to better prepared thin sections, as well as to the opportunity to study sections from a number of similar although not quite identical specimens, some of the features not very clearly observed in No. 146 can be presented here with greater fullness and certainty.

The plagioclase does not, upon the whole, appear to be quite so basic as is the same mineral in the dacites. Twinning striations are abundant and disclose the presence not only of the albite law but also of the pericline and Carlsbad laws. But sections suitable for determining the maximum symmetrical extinction angles do not seem to be common, which is doubtless due to the fact that the rather large average size of the crystals does not allow many of them in any one thin section to be cut in the right direction. The largest extinction angle measured on a section cut at right angles to the albitic twinning plane was 27° central part of the crystal. As all of the plagioclase has a well-defined zonal structure the margin of the crystal is necessarily much more acid. Usually the sections that give symmetrical extinctions give very small extinction angles, even on the interior of the crystal. This is so often the case as to indicate that at least some and probably most of the plagioclase is oligoclase. This supposition is further borne out by the fact that such crystals, when broken so as to have Canada balsam in contact with the plagioclase, are seen to have an index of refraction almost identical with that of the balsam, which would indicate a feldspar about on the border of oligoclase and andesine. While the index of refraction thus indicates that some of the plagioclase, even in the interior of the crystal, is hardly more basic than oligoclase, in all cases where the observation was made the outer portion of the plagioclase has this degree of acidity. It seems more than likely that there are two kinds of plagioclase crystals present, a fairly basic and not very abundant variety and a more commonly developed and presumably younger variety. But in any case all the plagioclase crystals have the mantle of sanidine which, where a glass base is very abundant, or wherever this glass base comes into contact with this mantle, is very remarkably fringed or lobed, so that the roundish feldspar lobes appear to be interwoven with the glass in a most intricate way but which, in the absence of a surrounding glass, forms with quartz a beautifully distinct and rather coarsely developed granophyric aggregate. While these marginal growths around the plagioclase crystals may be seen around every individual plagioclase they do not necessarily entirely surround the crystal, nor does the fringing mantle have constant width. In this respect the mantle is very irregular and even erratic. As above stated, the interweaving of the brown glass with the fringed sanidine gives almost exactly the impression of the granophyric growth when seen in polarized light, except that a portion of the intergrowing substance remains extinguished when the specimen is rotated. Even in white light this resemblance to granophyric growths is often pronounced whenever the glass has an unusually light color, and this is the case in those specimens in which the glass base is not abundant (149), or in which the section is unusually thin. Further, this resemblance is made still more marked by the fact that at times the glass that appears interwoven with the feldspar breaks up into rectangular or otherwise polygonal forms with more or less parallel arrangement. Or, to express it in another way, the sanidine margin develops into a skeleton growth with the brown glass base filling the interstices. The variations of structure that are produced in these marginal growths, owing to constantly varying amounts of glass and the consequent appearance and disappearance of real granophyric growths, fairly baffle description. It would require a large number of drawings or photographs to give a fair conception of these almost fantastic structures. An effort has been made to reproduce some of the most striking effects in fig. E of Pl. XVII (p. 128), figs. C and D of Pl. XVIII (p. 132), and figs. A and B of Pl. XIX (p. 138).

The amount of quartz present appears to be inversely proportional to the amount of glass base. Where much glass is present it is not inclined to form granophyric growths but to occur in rather isolated, roundish, or irregular grains. Although it may have very irregular outlines, it does not break up into lobes or fringes at contact with the glass, as does the sanidine, at least not to any marked extent. In the more thoroughly crystalline specimens and therefore more thoroughly granophyric varieties, the quartz seldom occurs in separate grains. The sanidine is not absolutely confined to the granophyric margins but also occurs very sparingly, as does the quartz, in separate but irregularly formed grains.

Inclosures in the central portions of the plagioclase are not common. Glass inclusions appear to be entirely wanting, but not infrequently a crystal may contain small augite grains, or individual grains of magnetite or prisms of apatite. Minute air cavities, however, are very abundant, both in the plagioclase and in the surrounding fringes of quartz and sanidine.

It should, perhaps, be added that no additional reason can be given for the presence of sanidine in these specimens than those already given in the description of No. 146.

Plate XIX.—THIN SECTIONS OF BASALTS AND OF SECRETIONS IN DACITE.

FIG. A.—Light-colored granophyric secretion from the dacitic ejectamenta. Magnified 48 diameters. Specimen No. 148. A photomicrograph in white light. Shows the occurrence of dark-brown glass, plagioclase, spongiform sanidine, parallel growing prisms, and grains of hypersthene and augite, also a little magnetite. See page 136.

FIG. B.—Light-colored granophyric secretion from the dacitic ejectamenta. Magnified 86 diameters. Specimen No. 149. A photomicrograph in white light. Shows a bundle of parallel arranged prisms of hypersthene and augite, which can not be distinguished from each other in the photograph. See page 139.

FIG. C.—Magnified 48 diameters. Specimen No. 2029 (private collection of H. B. Patton). A photomicrograph in white light. Shows the interstitial structure of some of the basalts. See page 142.

FIG. D.—Basalt, Timber Crater. Magnified 48 diameters. Specimen No. 165. A photomicrograph in white light. Illustrates the fluidal-interstitial structure of some of the basalts. See page 149.

Hypersthene and augite are both very abundant. Neither of them occur in the customary idiomorphic crystals but either in perfectly irregular grains or—which is generally the case—in various shaped aggregates, or more especially in clusters or bunches of loosely aggregated and parallel-growing roundish and rather slender prismatic grains. Both of these minerals assume this same form and usually both of them occur together in parallel growths. All conceivable variations may be observed between isolated individuals of hypersthene or augite of the shape just mentioned and parallel aggregates of the same that contain hundreds of the small prismatic grains. Some idea of this occurrence may be derived from fig. B of Pl. XIX. Hypersthene seems to be the more abundant of the two. In color the hypersthene is not quite so dark as it is in the ordinary dacites of Crater Lake. Pleochroism, therefore, can not readily be seen in such sections as are necessary for proper study of the aggregates. The augite appears almost colorless in thin section. With these two pyroxenes there is constantly associated considerable magnetite, which occurs mostly in very minute grains inclosed in the individual pyroxene grains or clustered with them in these bunches.

The chemical analysis of No. 151 will be found on page 140, together with the other chemical analyses. While somewhat more basic than are the dacites proper, the analysis of this specimen will be seen to compare fairly closely with that of No. 114. The excess of lime may perhaps be attributable to the abundance of the pyroxene. It can hardly be due to the relative abundance of the feldspar, as the plagioclase present does not appear to be particularly basic, and the sum of the alkalies present is rather less than is the case with the dacite whose analyses are given.

In conclusion, we may say that the absolute freshness of these specimens would seem to preclude the possibility of a secondary origin for the granophyric growths.

SECRETIONS WITH BASALT-LIKE STRUCTURE.

In addition to the above-described secretions there remains a specimen collected in the bottom of the Sun Creek Canyon and already mentioned on page 123, that can not be classified with any of these varieties. The specimen in question (123) has a grayish color and is rather fine grained, but is distinctly granular and porous. In thin section the structure appears at first glance to be distinctly basaltic, with dark-brown glass in the angular interstices formed by the intersections of rather short, thick-set plagioclase laths. Both hypersthene and augite abound and occur in irregular grains not dissimilar to the basaltic pyroxenes, also in long, slender, prismatic forms. In addition to these minerals, hornblende in dark reddish-brown grains and prisms is quite abundant, and magnetite in occasional black grains. Occasionally a plagioclase or hypersthene crystal assumes somewhat larger form and appears in the role of a phenocryst. What appears to be dark-brown glass inclosing slender augite microlites and filling the interstices between the plagioclase laths is seen in polarized light to be small, spherulitic crystallizations. In spite of the general structure this rock is quite different from any of the Crater Lake basalts. Its association with dacites and the presence of abundant hornblende, as well as the spherulitic base, lead to the conclusion that this is in reality a secretion in a dacitic magma.

An inclusion in a dacite dike below Llao Rock, described on page 122, is considered to be a similar occurrence.