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

BASALTS.

As will be seen by consulting the map the basalts of Crater Lake comprise only the later volcanic eruptive masses that form the various cones, mounds, and flatter lava sheets to the north, west, and south of the lake. No basalts whatever are found in the walls of the crater or in the dikes cutting them. Nor are they to be seen on the immediate outer slope of the crater.

In the great majority of cases the basalts bear a very close resemblance to many of the andesites, and sometimes can not surely be distinguished as basalts. In not a few cases, however, typically developed basaltic structures are to be noted. This type will be considered first.

INTERSTITIAL BASALTS.

This type of basalt is represented by four specimens (152, 153, 154, 155) taken from the hill that rises 500 to 600 feet above the level of the plain of Diamond Lake, in the extreme northwest corner of the Crater Lake area; also by two specimens from the lava flow at the east base of Red Cone, about 1-1/2 miles northwest of Llao Rock on the crater rim (156, 157); also by four specimens taken from different places in the basaltic area west of Anna Creek in the extreme southwestern corner of the region mapped (158, 159, 160, 161).

These rocks are all rather light colored, of a grayish to brownish tint, and contain occasionally a few cavities, but are never really scoriaceous. The grains are remarkably uniform and are rather dense, but not extremely so. Under the lens they appear holocrystalline and disclose occasionally greenish to reddish olivine grains and likewise greenish grains of hypersthene. Both olivine and hypersthene have a distinctly resinous luster.

An absolutely holocrystalline structure can only occasionally be made out under the microscope (156, 152), but in all cases a glass base plays a very subordinate role. The structure is dominated by the feldspars, which occur in mostly very distinct lath-shaped forms, while the augite is largely confined to the more or less angular spaces between the feldspar laths, after the manner of a mesostasis. This structure may be termed interstitial (Rosenbusch's "intersertale struktur"). The rocks in question, however, are too decidedly feldspathic to give typically interstitial structures. In one or two cases they assume a nearly hypidiomorphic (156) and in others a porphyritic structure. The porphyritic development, however, is caused by the occurrence of phenocrysts of olivine and hypersthene, and rarely of augite or plagioclase.

MINERAL COMPONENTS.

FELDSPAR.

The feldspar is entirely plagioclase, which, on account of its great abundance, dominates the structure of the rock. It has the rather long lath form common to basalts of this class. In most cases the laths are rather irregular in outline, but where a glass base is more in evidence, as is usually the case, the edge of these lath forms may be quite sharp. In such cases the ends are cut off squarely, as though by a pinacoidal face. Wherever the absence of a glass base prevents the free development of the terminal faces the long sides of the laths are more perfectly formed than are the ends. Their length is usually several times their width. The general shape of these plagioclase crystals is tabular, as the square cross sections so characteristic of plagioclase in microlitic form are conspicuously absent. The different thin sections of these basalts present not a little variety in the size and uniformity of the feldspars in the same specimen. No. 161 will serve to illustrate one in which the feldspars are unusually uniform in size. Their average length is 0.2 millimeter. Nos. 152 and 157 show much greater variation in the size of the feldspars, those of No. 157 having an average length of 0.3 millimeter, with a maximum of 0.7 millimeter and a minimum of 0.1 millimeter: the feldspars of No. 158 have an average of 0.15 millimeter with a maximum of 0.3 millimeter and a minimum of 0.04 millimeter. Most of the plagioclase laths disclose sharply defined polysynthetic twinning, the smaller and more slender laths having two or three and the larger ones half a dozen or more stripes. In each thin section, however, there are not a few individuals that do not appear to be twinned. There is no good reason to infer that any of these represent a monoclinic feldspar. Usually such untwinned individuals have much broader forms, such as one would expect to see in a tabular plagioclase cut parallel to the largest face. The supposition that this is the case is strengthened by the frequent appearance of well-defined cleavage cracks that correspond in direction to the basal cleavage of plagioclase. These larger apparently untwinned sections not infrequently also show a well-defined zonal structure, which structure may also be seen at times on the larger crystal grains that are cut so as to show the twinning. As is generally the case in zonal plagioclases the zonal banding is best seen on the brachypinacoid. The extinction angles observed indicate a very basic plagioclase, namely, anorthite. Measurements made on sections at right angles to the brachypinacoid—that is, on sections that give equal extinction angles to the right and to the left of the trace of the twinning plane—gave maximum angles varying between 30° and 36° (the last-named angle was observed in No. 152). These measurements were made on crystals of different sizes, not only on the largest ones. In cases where the individuals were too small to allow one to determine the direction of cutting the very large extinction angles commonly obtained also point to a very basic feldspar.

With the exception of an occasional apatite needle and a little black dust, these feldspars are free from inclosures. An exception to this statement may, perhaps, be taken in that where the dust is conspicuous there is not unlikely also a small amount of glass inclosing it. Part of this black dust appears to be opaque and may be considered to be magnetite, but other particles, when highly magnified, lose their apparent opacity as well as their blackness. These last are undoubtedly globulitic matter. Such dust-laden glass is quite similar to the small amount of glass base that at times may unmistakably be detected between the plagioclase laths of these and of other basalts of this region. In a few cases (155) a small number of somewhat larger and thicker plagioclase crystals are to be seen that may possibly be considered as phenocrysts and as belonging to an older generation, but they bear little resemblance to the more characteristic phenocrysts of the andesitic basalts described further on in these pages.

AUGITE.

Augite is abundant, but is invariably much less so than plagioclase. It appears in thin section in pale-greenish or yellowish-green colors, and nearly always in small angular grains filling the spaces between the feldspar laths. Occasional exceptions to this rule may be mentioned (153) where the augite assumes in part roughly prismatic form, resembling that of the hypersthenes. It is invariably younger than the plagioclase and also younger than the hypersthene. In a few cases (156-157) contiguous or adjacent grains have simultaneous extinction, and develop into an ophitic structure through partially inclosed plagioclase laths. Augite is usually much more abundant than hypersthene, although it may become less so (152). As a general thing the augite is perfectly fresh, but brownish to deep-red iron oxide stains are common.

HYERSTHENE.

Hypersthene is a characteristic but not always abundant constituent of these basalts. It is very fluctuating in amount. Although it may be more abundant than augite, as above noted (152), it is generally not so abundantly represented, and in one case appears to be entirely absent (161). It occurs almost invariably in small, fairly well-defined prisms that are two or three times as long as wide. In the more distinctly holocrystalline varieties this mineral is almost granular, the prismatic habit being very roughly developed. On the other hand, where the structure is less crystalline and the glass base distinctly recognizable, the characteristic habit of hypersthene can readily be made out. In such cases the forms are brachypinacoid and macropinacoid with subordinate prism, terminated by a flat pyramid. In color and pleochroism they exactly resemble the hypersthenes of the andesites. In the very thin sections, however, necessary for the study of these rather fine-grained rocks, it is not easy always to distinguish between augite and hypersthene. In such cases the general form, parallel extinction, and lower double refraction are usually sufficient. In the finer grained specimens the hypersthene does not differ greatly from the plagioclase laths in size, being perhaps, on the average, somewhat smaller; in the coarser rocks they are apt to be decidedly smaller than the feldspar laths. They are nearly free from inclosures, magnetite octahedrons and an occasional apatite prism being the only observed exceptions. As stated above, they are older than the augite grains, and seem also in most cases to be older than the plagioclase.

Parallelism of growth between augite and hypersthene is a common occurrence. This is more noticeable in the basalts that are poor in augite. In such parallel growths the augite invariably appears as a slender strip on each side of a prismatic hypersthene prism, the two strips on the opposite sides of the crystal extinguishing together. Very rarely may augite be seen on the end of a hypersthene crystal. It is worthy of note that this parallelism of growth between these two pyroxenes is not universal, that is, not all of the hypersthenes in any one thin section are thus bordered by augite.

OLIVINE.

Olivine is a constant but very fluctuating ingredient. It appears to be particularly abundant in those rocks where hypersthene is either missing or at least not abundant. No. 157 well illustrates this fact. It is a nearly holocrystalline basalt, with abundant augite and olivine, but with only two or three hypersthene individuals visible in a thin section. The olivine appears either colorless or slightly yellow, but it is frequently stained blood red along the irregular cleavage cracks. In a few cases a slight serpentinization has started in. The olivine occurs usually in granular form or in clusters of grains and only exceptionally in roughly defined crystals. It is the largest constituent of these basalts and is the only one that may be classed as a phenocryst. In No. 152 the olivine appears to have suffered to some extent from magmatic resorption. This is indicated by a border of opaque ore grains. This black border does not occur on all individuals. It is best seen on the smaller ones. It may occur on one end of a crystal and be missing on the other end. A few of the smaller crystals show only a core of unaltered olivine while the greater part of the crystal has been thus altered. In this specimen the small individuals of olivine are unusually well defined and show quite distinctly the crystal shape. In such cases the granular mass of ore has retained the original crystal form of the olivine. The olivines show occasional inclosures of magnetite and also once in a while of brown glass.

MAGNETITE.

Magnetite is not present in very great abundance. It occurs in the customary grains and small octahedral crystals scattered throughout the rock and inclosed in all the other ingredients. It varies considerably both in amount and in the size of the individual crystals, but it never assumes the importance that is ordinarily expected in basalts. It also occurs, as described below, in the form of fine dust impregnating the interstitial glass.

As already mentioned, the glass base of these basalts varies somewhat in amount. It is never very conspicuous, and it vanishes almost entirely at times. In a number of cases the glass can not actually be seen, but its presence is probably indicated by the occurrence of black dust, which is probably magnetite and globulitic matter, and which occurs either in separate but very minute grains that are too small to measure or in very minute rod-like or trichitic-like growths. Apparently these opaque particles, are inclosed in the feldspars, but as they are to be seen mostly at the junction of the plagioclase laths it is more likely that they are really embedded in a very thin film of glass. None of the specimens above described is entirely free from this black powder.

This interstitial structure is brought out in fig. C of Pl. XIX (p. 138).

The chemical analysis of No. l58 will be found with the analyses of other basalts from Crater Lake on page 161. Although this is as typical an example of basalt as may be found around Crater Lake, the analysis indicates a rock closely allied to the andesites. It is distinctly more acid than is the olivine-bearing basalt (173), a description of which may be found on page 155.

A partial analysis of another interstitial basalt (162) was carried out in the chemical laboratory of the Colorado State School of Mines by Prof. R. N. Hartman. This analysis gave SiO₂=55.18 per cent. It would seem, therefore, from these two analyses that the interstitial basalts of Crater Lake have unusually high percentages of silica.

HYPERSTHENE, APATITE, AND PSEUDOBROOKITE CRYSTALS IN BASALT.

In a specimen of basalt of this type collected on the east base of Red Cone (156) occur numerous flattish cavities that measure from one-quarter to one-half inch in greatest diameter. These cavities are lined with minute crystals of four different kinds; first, hypersthene; second, pseudobrookite; third, apatite; fourth, unknown white mineral.

The crystals of hypersthene occur in very brilliant though minute tablets, transparent, and of a fine deep-brown color. When examined under the microscope these crystals give the properties of hypersthene as seen in section parallel to the brachypinacoid. Pleochroism is marked with greenish color in the direction of the vertical axis and a decided brown at right angles to this direction. The extinction is parallel and the vertical axis the direction of least elasticity. No optical image can be obtained on the large crystal face.

Owing to the extreme brilliancy of these hypersthene crystals and to the fact that they appeared to present interesting crystal forms, pieces of the rock containing these cavities were sent to Dr. C. Palache, of Harvard University for further identification and measurement. The writer is extremely indebted to Dr. Palache for working out the forms of these crystals as well as for the identification and measurement of the pseudobrookite crystals. The result of this painstaking work is given below. As will be seen by comparing his measurements the supposition that these brown, pleochroic tablets are hypersthene is abundantly proved.

The cavities also contain a very few black and not well-defined crystals which, under the microscope, appear to have a deep reddish-brown color and to show no pleochroism. Although the writer is not certain that these black crystals are the same as those identified by Dr. Palache as pseudobrookite, the probability is in favor of that supposition.

APATITE CRYSTALS.

Apatite occurs in very delicate, minute, slender needles. The identification as apatite rests upon the needle form and the fact that they possess very low double refraction and have parallel extinction with negative extension.

The unknown white crystals occur in abundant, roundish forms which, when broken, are seen to be aggregates. It is possible that they are tridymite, with which they have some resemblance, but this could not be proved.

Below is submitted the report on hypersthene and pseudobrookite crystals by Dr. C. Palache.

HYPERSTHENE CRYSTALS.^a

The hypersthene is in the form of minute tabular crystals, 0.5 to 0.75 millimeter in length, 0.3 to 0.5 millimeter in breadth, and about 0.1 millimeter in thickness. They are always attached to the walls of the cavity by an edge in such manner that double terminations never occur. The faces are bright and sparkling except a, 100, which is sometimes striated parallel to the vertical direction; but on account of their minute size the reflections are dim, and the readings, therefore, subject to considerable variations.

---

^aThe description here given is the work of Dr. C. Palache.

---

The observed forms (position of Groth, a: b: c=1.0308: 1: 0.5885) were as follows:

The most frequent combination of these forms is shown in fig. I of Pl. XIV (p. 76), which is an exact reproduction of the figure given by Schmidt^a for the hypersthene from Málnás, Hungary.^b

---

^aZeit. für Kryst., vol. X, 1885, p. 210.

^bThis figure is given by Dana, p. 349, fig. 2, but with incorrect lettering. With the axes there adopted, the letters a and b on the two pinacoids should be interchanged.

---

k, 102, also found on the Málnás crystals, was found only on a single crystal shown in fig. J of Pl. XIV, and the same was true of n, 210, a single narrow face not shown in the drawing. w, 310 was observed on several crystals and measured on two. This form was first observed by Fouqué on hypersthene from Santorin,^c but he did not indicate it as new and gave no measurements. It has therefore been rejected in most of the recent lists of hypersthene forms, but appears to have been confirmed by the present measurements.

---

^cBull. Soc. min. France, Vol. I, 1878, p. 47.

---

PSEUDOBROOKITE CRYSTALS.^d

Accompanying the hypersthene in a single cavity were several slender, needle-like crystals, having a brilliant metallic luster. Their appearance and the association in which they occurred recalled the description of pseudobrookite as found in the hypersthene-andesite of Aranyer Berg, Hungary;^e and, notwithstanding the minute size of the crystals, it was found possible to obtain approximate measurements from two of them, which seem to confirm this determination.

---

^dThe description here given is the work of Dr. C. Palache.

^eAmong others, Traube, Zeit. für Kryst., Vol. XX, 1892, p. 327.

---

The forms found are shown in fig. K of Pl. XIV (p. 76) and in the following table, together with the measurements:

The faces of the pinacoids, a and b, were bright, but the prisms were deeply striated and gave very poor reflections.

The position here adopted is that originally proposed by Koch* and retained by Schmidt, Traube, and Groth. With Dana and Goldschmidt the axes b and c are interchanged. The former seems the more natural choice, in view of the pronounced prismatic habit of the crystals and since the chemical relations of pseudobrookite to brookite do not seem to bear out the apparent form relations between the two minerals which it was the purpose of the altered position to express.

---

^*A. Koch, Min. Mitth., vol. I, 1878, pp. 77 and 344.

---

FLUIDAL-INTERSTITIAL BASALTS.

In at least two instances (152 and 155) the above-described interstitial basalts disclose a decided tendency toward a fluidal arrangement of the plagioclase laths, and thus present a transition stage between the more typically developed interstitial basalts and what may be termed basalts with a fluidal-interstitial structure. This last-named type is well illustrated by four specimens, three of which (162, 163, and 164) were collected in different parts of the basaltic area to the southwest of Crater Lake, from which most of the interstitial basalts came, and the fourth (165) from a widely separated section, namely, from the southwest slope of Timber Crater, situated about 4 miles to the northeast of the lake.

These rocks consist of plagioclase, augite, hypersthene, and magnetite with almost no olivine. Glass is also undoubtedly present, but certainly not in large amount. The fluidal structure is due mainly to the fact that the plagioclase appears in very long and slender lath form, and to the further fact that these feldspar laths have a very marked parallelism of arrangement. The structure is indeed quite suggestive of the fluidally arranged sanidine laths in trachytes and many rhyolites. These plagioclase laths are quite sharp in outline, or, at least, this may be said to be true of the sides of the crystals. They are many times as long as wide, and have an average size of about 0.1 to 0.2 millimeter. The fluidal structure is further accentuated by the fact that the hypersthene and to a considerable extent also the augite occur in rather slender prisms, which are likewise arranged parallel to the feldspars. Hypersthene is distinctly the dominating pyroxene, being usually much more abundant than augite. It occurs only in prismatic form in crystals that measure usually about 0.02 to 0.05 millimeter in length, and about one-third that amount in width. Occasionally somewhat larger individuals may be seen, but, as a general thing, their size is very uniform. They inclose small grains and octahedral crystals of magnetite. The augite is to be seen both in grains and in minute prisms, almost identical in size and appearance with the hypersthene prisms. In fact, in white light it is almost impossible to distinguish between the two, as in color, size, shape, and inclosures they resemble each other. Even the pleochroism can hardly be used as a means for distinguishing these two pyroxenes, because of the necessarily thin sections required for the proper study of such fine-grained basalts. In polarized light, however, the distinction is not difficult. Parallelism of growth between hypersthene and augite may be observed exactly as in the above-described rocks, the augite appearing in the thin section as slender parallel strips on each side of the hypersthene.

The fluidal-interstitial structure is shown in fig. D of Pl. XIX (p. 138).

The almost complete absence of olivine in the basalts in which hypersthene is unusually abundant is a further corroboration of the frequently noticed fact that the development of olivine in a basalt is not so much dependent on the chemical composition as upon the conditions of solidification. Irrespective of the fact that the amount of olivine in these and in the more typically developed interstitial basalts appears to vary inversely with the development of hypersthene, it may be remarked that the small amount of olivine present in all these basalts falls in line with the remark of Professor Rosenbusch,^a to the effect that olivine occurs most sparingly in the basalts of a hypidiomorphic or of a doleritic type. The basalts under discussion, although hardly to be called typically doleritic, are at least more allied to that type than to any other.

---

^a Mikroskopische Physiographie, 3d edition, vol. II, 1896, p. 993.

---

In No. 163 are to be seen minute, rod-like microlites of a deep reddish-brown color and almost opaque. In size they measure on the average 0.01 millimeter long and 0.001 millimeter wide. They resemble rutile, but do not give the brilliant polarization colors characteristic of such minute crystals of that mineral. In fact, it could not be shown that they affected polarized light at all. They have not been noticed in the other basalts.

These fluidal-interstitial basalts, when compared to those already described in these pages as interstitial, are closely allied to those occurring in the northwest corner of the Crater Lake area, and more especially to Nos. 152 and 155. No. 164 shows the fluidal structure much less perfectly developed than do the other three and may be considered as intermediate between the two types. No. 166 was collected at the same place as No. 164 and is probably a locally differentiated variety. It could hardly be classed here if taken by itself. The plagioclase is much less abundant and occurs in somewhat larger slender laths, that lie in all directions but are not abundant enough to interfere with each other to any great extent. The bulk of the rock consists of a feldspathic paste thickly crowded with extremely minute pyroxene prisms and with octahedral crystals of magnetite. Both augite and hypersthene are represented, but it is impossible to determine in what proportion.

These two types of basalt are, then, very closely linked together, and the transitions between the different specimens collected are much closer than are the transitions between these and the porphyritic basalts whose description follows, and still more plainly is this true as compared with the basaltic-looking andesites of this same region. The force of this statement will better be appreciated in connection with the comparison made below with the so-called hypersthene-andesite from Franklin Hill, Plumas County, Cal. We have seen that many of the basalts and andesites of Crater Lake can with difficulty be distinguished from each other, but this can not be said of the above-described interstitial and fluidal-interstitial basalts. The more marked characteristics of these rocks for purposes of comparison with the andesites is the almost complete absence of porphyritic development. This is seen both in the scanty development of phenocrysts and in the lack of a younger generation of the mineral ingredients.

These fluidal-interstitial basalts (especially No. 162) bear a close resemblance to a rock from Franklin Hill, Plumas County, Cal., which is briefly mentioned by Mr. H. W. Turner in the Bidwell Bar folio.^a In his article on The Age and Succession of the Igneous Rocks of the Sierra Nevada^b he calls it a hypersthene andesite, and describes it as follows:

"In the area of the Downieville and Bidwell Bar atlas sheets are numerous bodies of a dense fine-grained gray lava, which usually weathers with a slaty fracture, the apparent cleavage being often vertical. The rock is composed of plagioclase, augite, a slightly pleochroic rhombic pyroxene, and grains of magnetite. The feldspar, augite, and rhombic pyroxene are in the form of minute elongated prisms or laths, and this is true of the rock at widely separated localities, and the laths of all are nearly of the same size. About half a gram of No. 661, powdered and treated with HCl by George Steiger, yielded no gelatinous silica, but nevertheless there appears to be some glass present."

---

^aGeologic Atlas U. S., folio 43.
^bJour. Geol., vol. III, 1895., p. 409.

---

Through the kindness of Mr. Diller the writer was enabled to make a study of a thin section of No. 661, referred to above. The resemblance between the rock which Turner calls a hypersthene-andesite and those just described from Crater Lake is even closer than might be inferred from the above-quoted description. The differences are rather in the size of the grain than in more important matters. One sees the same delicate, uniform prisms of rhombic pyroxene and of augite, the same slender feldspar laths, and, locally developed, in the thin section the same parallelism of structure. In the California rock the pyroxenes are somewhat larger and therefore more plainly developed. Mr. Turner also gives in the same paper a partial chemical analysis, as follows:

Partial chemical analysis of hypersthene-andesite from Franklin Hill, Plumas County, Cal.

This analysis alone certainly seems to justify Mr. Turner in calling the rock he describes an andesite, but upon close inspection it will be seen that the analysis is really that of a rock intermediate between a basalt and an andesite. As far, as the silica alone is concerned the amount is more suggestive of a hypersthene-andesite than of a basalt. But the percentage is only a little over 1 higher than that given for a basalt glass (tachylite) from Säsebühl, near Dransfeld, Hanover.^a Furthermore, Mr. Diller has published a list of chemical analyses of quartz-basalts from Cinder Cone, northeast of Lassen Peak, California,^b in which silica runs as high or higher—in one case three-quarters of 1 per cent higher—than that given for the Franklin Hill andesite. Inasmuch as Mr. Iddings has fairly demonstrated^c that the presence of primary quartz in basalts is not due to the excessive amount of silica present in the magma, it may well be claimed that the rock from Franklin Hill is not necessarily too acid for a basalt.

---

^aRosenbusch, Elemente der Gesteinslehre, Stuttgart, 1898, p. 309.

^bA late volcanic eruption in northern California and its peculiar lava: Bull. U. S. Geol. Survey No. 79, 1891, p. 29.

^cOrigin of primary quartz in basalt: Am. Jour. Sci., 3d series, Vol. XXXVI, 1888, p. 220.

---

When the lime, potash, and soda in this rock are taken into account it would be impossible to distinguish as to whether the rock were a basalt or an andesite, for it would be easy to cite analyses of both rocks in which these substances are present in almost the exact relationships that are given in the case under discussion. As to whether the Franklin Hill rock should really be called a hypersthene-basalt instead of an hypersthene-andesite the writer will not assume to say, as this might well depend on surrounding associations with which he is not sufficiently familiar. In the absence of a chemical analysis of the fluidal-interstitial basalts of Crater Lake it may perhaps be assumed that the composition does not vary greatly from that of the California rock. Even so, on account of the close resemblance to well-defined basaltic types of the same region, and on account of the sharp contrast it presents to the andesites of this same region, one may, in the opinion of the writer, be justified in calling these rocks basalts. In any case, on account of the great variability in the chemical analyses of rocks we are accustomed to class under the same family name, the structural relationships would serve better as a means for determining the rock name than would the chemical analysis alone, especially in cases where that analysis is not typical.

PORPHYRITIC-INTERSTITIAL BASALTS.

This type of basalts may be said to contain well-defined and usually abundant phenocrysts, always of olivine and generally also of plagioclase, hypersthene; and augite, in a groundmass that bears a more or less close resemblance to the interstitial basalts. As a rule, however, the groundmass is inclined to contain more glass and the minerals of the second generation have more sharply developed forms than is the case with the corresponding minerals that compose the interstitial basalts.

The lava of Desert Cone, immediately north of Red Cone, may be taken as the best example of this type of structure. Red Cone also is composed of basaltic lavas that belong to this type, but not entire so, as will appear later.

The lava of Desert Cone is represented by three specimens—No. 167 from the north side, near the summit, and Nos. 168 and 169 collected close together about 1 mile farther south on the southern slope. All three are very much alike. In the specimen from the north side olivine is the only phenocryst; in the two from the southern slope, in addition to olivine, plagioclase, in a rather subordinate role, occurs as a phenocryst.

The olivine in these specimens is to be seen in very sharp idiomorphic crystals and also in crystal fragments. They are particularly well adapted to a study of the crystallographic forms and optical properties. At first glance some of the olivine crystals, especially the fragments, resemble augite somewhat, the color being a light yellowish green. Irrespective of the crystal form, however, one can usually distinguish both the crystals and the fragments by their unusually rough appearing surface and by their high interference colors, also by the absence of well-defined cleavages. They usually appear in rectangular or hexagonal shapes. The dominant forms are the brachydome (021) and a prism. In sections cut parallel to the macropinacoid this combination gives hexagonal outlines, the angle formed by the trace of the brachydome being not far from 80°. Such a section also gives a positive bisectrix, with large optical angle (168).

In No. 169 the olivine phenocrysts have very sharply developed forms and show an extensive alteration to iron oxides. This alteration product occurs both as a broad rim and as finely granulated masses scattered throughout the crystal. In the very smallest crystals the olivine has been almost completely replaced; in the larger ones the unaltered olivine rarely composes more than one-half of the entire bulk and generally much less than one-half. This iron oxide alteration product consists sometimes of perfectly black, opaque material, which is presumably magnetite; in other cases the grains are slightly transparent and are of a deep red color, suggestive of hematite. Still more commonly both magnetite and hematite abound. In spite of the sharpness of the crystal forms the olivine phenocrysts are frequently penetrated by plagioclase laths, thus indicating a very basic feldspar. (See fig. L of Pl. XIV, p. 76.) These penetrating feldspar laths do not occur in the center, but only at the margin of the crystals; occasionally, also, one may see an olivine crystal impressing its form on a plagioclase phenocryst. This, however, is by no means as common as the other case. It would seem from this that the feldspars had commenced to crystalize out before the olivines had ceased to grow. Possibly they may have begun simultaneously with the olivines.

The only other phenocryst in this basalt is plagioclase, and even this is absent from one of the three specimens (167). It is by no means conspicuous nor very markedly different from the smaller feldspathic laths that make up the greater part of the groundmass. The shape is rectangular or long rectangular. The largest extinction angles noticed on sections cut perpendicular to the albitic twinning plane was 30° in No. 168 and 29° in No. 169. This does not indicate so extremely basic a feldspar as its relationships to olivine would suggest. Probably if the plagioclase phenocrysts were more numerous sections could be found showing larger extinction angles.

The groundmass appears to be nearly holocrystalline. It is composed of well-developed plagioclase laths of hardly more than microlitic proportions, augite in minute grains, or in both prisms and grains, and abundant magnetite in mostly very minute octahedrons and grains. Hematite powder is also abundantly developed in the groundmass of No. 169 as well as in the olivine phenocrysts. In addition to the above-named minerals hypersthene has been recognized in No. 168. It is entirely confined to the groundmass and occurs in but one generation. It forms very small prisms that measure 0.06 millimeter long by 0.01 millimeter wide and that very closely resemble similar prisms of augite. They are too small to show any pleochroism and are to be recognized only by their parallel extinctions and lower interference colors. In fact, it requires some little familiarity with these rocks before the distinction between the pyroxenes can be made.

In the southwest corner of the area covered by the Crater Lake map, and about half a mile south of the road, was collected a specimen of basalt, No. 170, that is closely connected with the interstitial basalts. In a groundmass composed of plagioclase laths, augite, hypersthene, and a little dust-laden glass occur perfectly developed comparatively large plagioclase phenocrysts that exactly resemble the andesitic plagioclases; also an occasional hypersthene, augite, and blood-red olivine crystal. The plagioclase is in broad crystals that show rectangular and six-sided outlines and, as in the andesites, zonal structure. The interior is crowded with glass inclosures, while the margin is free from the same, or else both the interior and the margin are clear, while the inclosures are confined to a narrow intermediate zone. One of these phenocrysts, cut perpendicular to the brachypinacoid, gave an extinction angle of 34°.

The groundmass of this rock, were it free from plagioclase phenocrysts, would be identical with the interstitial basalts. The plagioclase laths are the dominant mineral, and appear to inclose the angular augite grains as well as the little glass present in the interstices formed by their intersection. Hypersthene, which occurs only very sparingly as a phenocryst, is more abundant in the groundmass, but as it is found here only in prismatic crystals and not in grains like the augite, it is properly not to be considered as belonging strictly to the groundmass, but rather to the older generation of crystals.

Although this rock presents an entirely different appearance from the basalt from Desert Cone, it must still be assigned to the type which has been designated as porphyritic-interstitial. It is in a way intermediate between the holocrystalline interstitial and the andesitic types.

Quite similar to this rock is No. 171, collected between Crater Peak and Sun Creek.

Still another development of the porphyritic-interstitial type of basalt is to be seen in part of the lava rocks of Red Cone. These are dark-gray, in one case red, dense, scoriaceous lavas. The phenocrysts are very inconspicuous, but may be made out under the pocket lens. They consist of yellowish to red olivine, yellow to black augite, and white plagioclase.

Under the microscope these appear to be decidedly porphyritic rocks with abundant and well-defined, though small, phenocrysts of olivine, augite, and plagioclase. The plagioclase is the most variable of the phenocrysts, as one specimen (172, collected about l mile southwest of the cone) contains almost no phenocrysts of this mineral. It occurs in rectangular, broad to narrow, lath-shaped crystals, the smaller of which graduate into the plagioclase of the groundmass. The ends of the laths are cut squarely off, as by a pinacoid, where the groundmass is distinctly glassy, but, in case the latter is more decidedly crystalline, the ends are somewhat frayed. They inclose at times a small amount of dusty-looking glass similar to that of the groundmass, but neither in shape nor in the appearance or arrangement of the inclosures do these plagioclase phenocrysts resemble those of the andesites or of the last-described basalt. Extinction angles in symmetrically cut sections, i. e., in sections perpendicular to the brachypinacoid, were measured as follows: 27°, 30°, 31°, and 34°. They are younger than both olivine and augite.

Of the pyroxenes augite is the only one of consequence. Of the four specimens studied three (174, 175, 172) contained no hypersthene and one (173) contained but one individual, which consisted of a roundish grain having a small core of hypersthene and the outer and larger part augite in parallel position. This growth of augite appears to be secondary and distinct from the more customary occurrence of augite growing around hypersthene crystals. Aside from this one case augite occurs in granular form, but even so impresses its form upon the plagioclase wherever it comes in contact. It is to be found both isolated and in nests with olivine. The color is usually greenish, but in No. 174 many of the grains have a brownish cast, or else they have greenish centers and shade into brown on the outside. It may be added that the hand specimen also discloses a few black augite phenocrysts, 2 to 3 millimeters long, that do not appear in the thin section.

Olivine is a very abundant constituent. It occurs in well-defined, sometimes in very sharply defined crystals, as well as in grains, the latter form being common where the olivine forms nests with augite and occasionally with plagioclase. In No. 174 this mineral occurs exactly as described for No. 169, one of the specimens collected on Desert Cone. It has the same clear-cut forms and has undergone the same alteration to hematite. In this case, however, the alteration has progressed still further. Not only have we the outer rim of almost opaque hematite, but the center of even the largest crystals is so thickly crowded with hematite as to leave very little clear olivine visible. (See Fig. M of Pl. XIV (p. 76.)

The groundmass of these rocks is abundant and mostly very distinct from the phenocrysts. It consists essentially of a colorless but dusty glass base that is crowded with augite microlites in the form of minute prisms and grains, also microlitic plagioclase laths and sharp magnetite octahedrons. The interstitial structure is not clearly brought out, but is produced in part by the plagioclase of the phenocrysts and in part by the plagioclase of the groundmass, there being no well-defined distinction between the two. The porphyritic structure, therefore, is much more in evidence than is the interstitial, and this rock may be considered in a sense as intermediate between the porphyritic interstitial and the more distinctly porphyritic basalts. It does not, however, bear much resemblance to that type of porphyritic basalts that are described as andesitic in these pages.

The chemical analysis of No. 173 will be found on page 161.

Two specimens, Nos. 176 and 177, collected about 3 and 4 miles, respectively, west of Red Cone, have been placed in this group of porphyritic interstitial basalts, although they are decidedly transitional between this group and the interstitial basalts proper. The greater part of the rock, as seen in thin section, constitutes a sort of holocrystalline groundmass similar to that of the more coarsely grained interstitial basalts. In this groundmass of plagioclase, augite, and magnetite occur some rather sharply defined phenocrysts of olivine, augite, and plagioclase. The two last-named phenocrysts occur also in granular form or in nests of grains, and grade off into the same minerals of the groundmass, so that no sharp line can be drawn in either case between the components of the groundmass and the phenocrysts. In the case of olivine, granular forms also occur, but these are always distinct from the groundmass in which no olivine occurs.

The olivine is slightly serpentinized, and is stained in places a deep red. It contains inclosures of magnetite. The better formed crystals of augite disclose the customary forms, namely, prism, two pinacoids, and flat terminal pyramids or dome faces. It is frequently twinned in double, triple, and multiple twins. The twinning plane is the orthopinacoid. It incloses olivine and magnetite, and, to a very slight extent, also plagioclase. The plagioclase crystals show more or less rectangular, but not very sharply cut, outlines. The period of development almost exactly coincides with that of the augite, as these two minerals often impress their form on each other. A few crystals contain minute glass inclosures, but usually they are free from inclosures, and do not bear any resemblance to the plagioclase phenocrysts of the andesites.

Hypersthene appears to be entirely wanting.

A very similar basalt, No. 178, from near the road south of Castle Creek, at the very edge of the district covered by the map, is also placed in this division.