Geological Survey Professional Paper 296 Geology of Glacier National Park And the Flathead Region, Northwestern Montana

STRATIGRAPHY (continued)

BELT SERIES (PRECAMBRIAN)

CLASSIFICATION

All of the sedimentary rocks of supposed Precambrian age in northwestern Montana belong by definition (Wilmarth, 1925, p. 108-112) to the Belt series. The series comprises a very thick sequence of beds, largely argillaceous and quartzitic but containing also much carbonate. Shallow-water features, such as ripple marks and mud cracks, are abundant. The rocks are thoroughly consolidated and partly recrystallized but retain many of their original larger features. The units within the series are commonly thousands of feet thick, with few distinctive horizon markers and many points of similarity. Much of the broad region underlain by the Belt series in the Northwestern United States has been studied only in preliminary reconnaissance as yet. In consequence, much diversity in nomenclature and correlation of units in different areas persists. The definitions of stratigraphic units adopted in the present report differ only in details from those of previous workers in the general vicinity of Glacier National Park. This fact, however, would not be clear from reading published reports. Hence, a statement, as precise as available data permit, as to the stratigraphic nomenclature employed is essential before the rocks can be described. Some units are designated by lithologic terms as assignment of formal names would be premature.

In a tentative classification of the Belt series in general (Ross, 1949), the representatives of the series in the region that includes Glacier National Park are divided into the Ravalli group at the base, the Piegan group in the middle, and the Missoula group at the top. These groups are broadly equivalent to the Ravalli, Siyeh, or Wallace, and Missoula groups of Clapp (1932, p. 22) and the Ravalli, Piegan, and Missoula groups of Fenton and Fenton (1937, p. 1875-1903). They do not differ fundamentally from the classification adopted by Campbell and his associates in their field notes or from the summary by Dyson (1949a, p. 4-10) except that these geologists do not use formal group designations. Similarly, in the paper by Clapp and Deiss (1931, p. 673-696) that deals specifically with problems of Belt stratigraphy, the Ravalli and Piegan groups are not mentioned, although no essential difference in interpretation of the stratigraphy is involved. An idea of the relations between the subdivisions of the Belt series in the vicinity of Glacier National Park and in other regions can be had from the correlation table (p. 17).

For convenience in following the discussions below, two additional tables are included. The first of these (p. 18) summarizes the ideas of Fenton and Fenton (1937) in regard to subdivisions of the Belt series in the parts of Glacier National Park that they studied. Fenton and Fenton have done the most detailed stratigraphic work so far accomplished and as a result propose to divide the formations of the Belt series into many members. At present, no basis exists for tracing most of their members on the geologic map of Glacier National Park; hence, the members are not adopted here. They are presumably significant in the particular localities where they were described, and some may prove to be widespread when further mapping is done. The second table (p. 19) gives the map units in the Belt series that are employed in the present report. In a broad way, the grouping here adopted corresponds to that of the Fentons and other recent investigators. Changes in the position of the boundary between the Piegan and Missoula groups are suggested with the aim of keeping lithologically similar beds together so far as possible. No basis other than similarity in lithologic characteristics is as yet available for the grouping of strata of the Belt series, although work done during the present investigation leads to the hope that the stromatolites and the trace element content of the different rocks will aid in this respect when further information has been accumulated.

Correlation chart for the Belt series in northwestern Montana

[Thicknesses are maxima for the different localities. Formation names whose authors are not given are in general use in the senses here employed]

Comparison of the two tables (p. 18, 19) shows that the differences are in details and that the principal ones are in the upper part of the series. In the table giving the Fentons' classification, their names for the algae are used, whereas in the second table Rezak's are used. The differences in the thicknesses given result in part from differences in the limits placed on the units and in part from actual variations in thicknesses in different localities. Differences in methods of measurements may also enter. The three formations in the Ravalli group (as that term is here used) are sufficiently distinctive, so that they can be recognized and mapped with confidence. All who have worked in the region are in essential agreement as to the limits of these three. One might query whether the Appekunny and Grinnell argillites would not better be term "formations" because both, and particularly the Appekunny, contain much quartzite.

The Siyeh limestone is a distinctive and easily recognized unit, but different geologists place somewhat different limits on the formation. The Campbell party, Dyson (1949a), and, so far as can be inferred, also Willis (1902, p. 316, 323) included beds of argillaceous and arenaceous composition at the top of the Siyeh. Fenton and Fenton, on the other hand, separate the upper clastic beds and include in the Siyeh only the rocks that are essentially carbonate-bearing—a practice which was followed in the present investigation because the clastic beds are lithologically much like the main body of the Missoula group. Recognition of the kinship of the elastic beds just above the limestone with the main body of the Missoula group has necessitated dropping the group boundary low enough, so that, in the park, the Piegan group becomes so restricted as to include only the Siyeh limestone, as that term is here used. Further discussion of these names and their significance is given in the following description of the Piegan group.

The Belt series in Glacier National Park according to Fenton and Fenton (1937)

For the purposes of this report, all of the Belt series above the Siyeh limestone as defined above belongs to the Missoula group. This procedure is the only one that permits the setting up of satisfactory map units, but the shift in the position of the boundary between the Piegan and Missoula groups proposed above departs from the ideas expressed by this writer earlier (Ross, 1949). The change is required by the fact that the clastic beds below the Shepard formation and above the characteristic limestone of the Siyeh are lithologically indistinguishable from the main body of the Missoula group.

RAVALLI GROUP

Altyn Limestone

In Glacier National Park the Ravalli group consists, in ascending order, of the Altyn limestone, Appekunny argillite, and Grinnell argillite. The Altyn is not exposed in the Flathead region, and its base has nowhere been recognized. This formation does not appear to be exposed in any other part of Montana. It is commonly regarded as part of the Ravalli group—a convenient procedure which is here adopted provisionally. However, one should bear in mind that no evidence exists as to the stratigraphic relations of the Altyn limestone of Glacier National Park to the basal components of the Ravalli group in other parts of Montana or to the underlying Prichard formation (Calkins, 1909, p. 33-42). This is one of many unsolved problems in the regional correlation of the Belt series.

The Altyn limestone forms a narrow band immediately above the Lewis overthrust along the front of the Lewis Range extending about as far south as latitude 48°20' and northward past the Canadian border. Here it immediately overlies the Lewis overthrust and much of it has been cut out by that fault. The formation is also exposed in outlying blocks such as those in Divide and Chief Mountains and on either side of Waterton Lake. The Altyn should occur near the western base of the Swan Range but has not been found there. It is either slightly below the depths so far reached by erosion or is masked by detrital material. Figure 2 shows the general character of the formation in Appekunny Mountain near the former settlement of Altyn, the type locality.

FIGURE 2.—Altyn limestone in the lower slopes of Appekunny Mountain, northeast of Many Glacier Hotel, Glacier National Park. The view is at the type locality of the formation, close to the site of the former settlement of Altyn. The Lewis overthrust is at the base of the cliffs, and the smooth slopes below are underlain by shale of Cretaceous age, which yields few outcrops. Photograph by Baily Willis (1902, fig. 2, pl. 47).

Subdivisions of the Belt series (upper Precambrian) in Glacier National Park and the Flathead region as used in the present report

In general, this formation consists of a very light-gray¹ magnesian limestone that weathers grayish orange, rendering it conspicuous on distant cliffs. The Fentons (1937, p. 1881-1885) say the thickness is 2,180-2,480 feet, and Dyson (1949, p. 7) gives the average as about 2,300 feet. Willis (1902) reported an upper member of argillaceous ferrugeneous limestone 600± feet thick, and a lower member of massive somewhat siliceous limestone with concretions, 800± feet thick. His figures appear to apply only to the middle one of three members described by the Fentons.

---

¹Rock-color chart (Goddard and others, 1948) used no a guide to color names throughout the present report, except where data on color are taken from the records of other geologists.

---

The Fentons divide their Altyn formation into three members called, in ascending order, the Waterton, Hell Roaring, and Carthew. The Waterton member of the Fentons corresponds essentially to the Waterton formation of Daly (1912, p. 50-56), which he described as underlying the Altyn formation. This unit, to which the Fentons assign an estimated thickness of 280 feet, with the base not visible, is exposed in Waterton Lakes Park in Canada but is not known to crop out in the United States. Whether the Waterton is to be regarded as a member of the Altyn or as an underlying formation, its absence in Glacier National Park reduces the exposed thickness of the Altyn there to about 2,000 feet.

The Fentons describe the Hell Roaring member as "dolomite and dolomitic limestone, variably siliceous; blue-gray to greenish-gray, weathering to gray, buff or cream; beds 2-24 inches thick. Many beds show laminae of limestone and dolomite and apparently primary dolomite nodules, associated with biostromes of Collenia albertensis Fenton and Fenton (= C. frequens of Rezak). Thickness estimated at 1,200 to 1,300 feet." They remark that, where it can be found, Collenia albertensis (= C. frequens of Rezak) is an index fossil. At Appekunny Mountain they found edgewise mud breccia and lenses of dolomitic sandstone or conglomerate in the upper part of the Hell Roaring member. They also report a zone consisting of beds crowded with Collenia columnaris Fenton and Fenton (= C. frequens of Rezak) at this locality and speak of carbonaceous films described as Beltina and Morania antiqua above their Collenia columnaris zone (= C. frequens zone of Rezak). They say that near Roes Creek, north of St. Mary Lake, the uppermost part of the Altyn that is exposed consists of buff-weathering sandy dolomite containing semirounded and angular pebbles, 2-15 millimeters in diameter. Probably the beds at this locality are in much the same stratigraphic position as those near Swiftcurrent Falls and Appekunny Mountain.

The Fentons describe their Carthew member as "magnesian limestones, dolomites, quartzites, and intermediate rocks which grade upward into the basal Appekunny. Colors range from blue gray through buff to brown and dark brownish red; bedding is thin to thick. Red beds, especially, show thin laminae. Thicknesses estimated at 700 to 900 feet." They have recognized this member only in Canada, where its type locality is, and in the northern part of Glacier National Park. They remark that near Many Glacier Hotel the basal beds of the Appekunny argillite rest on the Hell Roaring member of the Altyn. Farther south they think their Carthew member has been cut out by faulting.

During the present investigation the Altyn was studied east of Many Glacier Hotel and on the lower slopes of Appekunny Mountain nearby. The beds in these localities apparently belong to the Hell Roaring member of the Fentons, The specimen of Altyn limestone whose analysis is listed on page 55 came from near the hotel. The analysis shows the rock to be an impure dolomite. Three analyses listed by Daly (1912, p. 58-61), while differing in details, agree in showing the rock to be dolomitic, which appears to be true of most of the formation. Hence, Altyn dolomite probably would be a more precise name than Altyn limestone, but the latter is so well known that a change at this time would have little value. The analysed specimen and one from close to Swiftcurrent Falls are similar under the microscope. The main body of the rock consists of a mosaic of carbonate grains ranging in diameter from a few hundredths of a millimeter up to about 0.2 millimeter (fig. 6, A and B). A few thin layers of fine-grained carbonate parallel the bedding of the rock. So many of the grains have crystal faces that it is probable the whole aggregate has been recrystallized.

Scattered through the carbonate ground mass, and locally concentrated in layers up to about 10 millimeters thick, are larger pieces of several different kinds. Among the more abundant of these are rounded bodies that themselves consist of carbonate aggregates, mostly 0.5-2.0 millimeters in maximum dimension. Some of the more irregular of the aggregates may be pebbles, but some are undoubtedly oolites. One, illustrated in figure 6A is a cross section of an apparently cylinderical body, 0.8 millimeter in diameter, that is so clean cut as to appear very different from anything else in the section. In a personal letter dated April 2, 1952, J. Harlan Johnson states that in his opinion this body is organic although he is not sure just what it represents. He adds: "It appears to be a piece of spine and, if found in lower Paleozoic limestone, I would not hesitate to say it belonged to a trilobite or a chitonous brachiopod."

In addition to the carbonate aggregates, there are rounded to subangular clastic grains up to several millimeters in maximum dimension. These consist largely of quartz but include much feldspar, mostly alkali plagioclase with some microcline. Much of the feldspar is strikingly fresh, but some is sericitized. A few of the grains have rims of clear added material. Some grains contain both quartz and feldspar, and these seem to be bits of a rather coarse-grained granite. Other grains are very fine aggregates of quartz, some of which include rounded masses that look like silicified oolites.

The presence of witherite in small bodies formed by replacement along bedding planes in the Altyn limestone below Swiftcurrent Falls has been reported (Fuller, 1924). This must be an exceptional occurrence as the mineral has not been reported by other observers, and analyses of the rock show no especial abundance of barium.

Both Nordeng and Rezak visited the locality on the slopes of Appekunny Mountain where the Fentons found their C. columnaris zone. Numerous specimens of stromatolites are present. Nordeng in his field notes commented on the diversity of shapes present, which he interpreted as resulting from variation in conditions during growth. Rezak, with benefit of longer study of the stromatolites, including Nordeng's work, regards the apparent diversity in form as resulting mainly from differences in the angles at which joint surfaces cut the algal heads. Both Nordeng and Rezak report that the Collenia columnaris zone of the Fentons is not a persistent zone. Rezak has searched for the zone in exposures of the Altyn limestone other than those on Appekunny Mountain, and reports the zone to be well developed on Divide Mountain although at other localities it is either poorly developed or absent. In this connection it may be significant that neither of the two sections measured by Stebinger and Bennett and tabulated below contain any beds in which stromatolites are reported.

Two sections of parts of the Altyn limestone were measured by Stebinger and Bennett in 1914 and are taken from their field notes. The first of these is in the lower part of the slope at the east end of Red Eagle Mountain, south of St. Mary Lake. The other is in the mountain north of Crossley Lake near the confluence of Mokowanis and Belly Rivers. There may be some duplication in the lower part of this section because of minor overthrusts and thicknesses of units exposed in cliffs are estimated. The descriptions, including color designations, are adapted from those in Stebinger's notebook. Presumably all the rocks described as limestone in the field notes are magnesian.

Altyn limestone on Red Eagle Mountain

[After field notes of Eugene Stebinger and R. R. Bennett]

Altyn limestone near Crossley Lake

[After field notes of Eugene Stebinger and E. R. Bennett]

---

¹Note says that 150 ft of dolomite like that in the 650-ft unit above underlies the red limestone in Chief Mountain, constituting the lowest unit of the Altyn limestone seen by Stebinger up to the middle of September, 1914, and making the total exposed thickness in this part of the park 1,675 ft. an estimate 275 ft greater than that of Willis.

---

Appekunny Argillite

The Altyn limestone is overlain by the Appekunny argillite, which, like the Altyn, was originally defined by Willis (1902, p. 316-322). The Appekunny argillite is widely distributed and well exposed in Glacier National Park. It forms an irregular band in the northeastern part of the park just west of the exposures of Altyn limestone that are associated with the Lewis overthrust. Another band extends from the Canadian border southeastward in the western part of the park until it joins the band in the northeastern part of the park. There are also exposures near Waterton Lake.

Small exposures correlated with the Appekunny were noted along the road in unsurveyed T. 30 N., R. 18 W., on the west flank of the northern part of the Flathead Range, and more extensive outcrops might be found if further geologic work were done in that vicinity. Appekunny argillite is present low on the southwestern slopes of the Swan Range, mostly in unsurveyed parts of Tps. 26 and 27 N., R. 18 W., but here satisfactory outcrops are rare, and much of the formation is masked by dense vegetation, soil, and hillwash. The southern tip of the main body of Appekunny argillite shown on plate 1 (Glacier National Park) also appears on plate 2 (Flathead region).

The formation as a whole consists of dark fine grained sedimentary rock sufficiently uniform in character, so that Willis (1902) gave it the name Appekunny argillite. Sufficient lithologic variations exist, however, both in the Appekunny and in the overlying Grinnell to give support to the Fentons (1937) and Dyson (1949a) in speaking of these units as formations rather than as argillites. This reflects the fact that both in the park and in more distant parts of northwestern Montana the two units possess greater diversity in lithologic character than could have been appreciated during Willis's pioneer studies. The distinction is subtle, and final decision should be based on consideration of a larger region than that covered by the present report. The Appekunny in the region here described originally consisted of beds containing varying proportions of argillaceous and siliceous material with distinctly subordinate quantities of carbonate. No limestone was present, and pure quartz sandstone was rare. All the components are thoroughly consolidated and sufficiently recrystallized, so that names that imply some metamorphism are required for the rocks in their present condition.

The somber, commonly rather massive rocks of the Appekunny argillite can be recognized and delimited throughout northwestern Montana more readily than some of the other units of the Belt series. It is only here and there that local development of schistosity or abnormal colors in some beds introduce some uncertainty. It does not necessarily follow that the top and bottom as mapped are everywhere at the same horizon. That is, beds lithologically akin to the Appekunny may persist over a greater stratigraphic range in one locality than another. Willis (1902, p. 322) recognized this in his remark that detailed stratigraphic study may develop the fact that the Grinnell and Appekunny argillites are really phases of one great formation and that the line of distinction between them is diagonal to the stratification. To a degree, such a suggestion may be applicable to many contacts throughout the Belt series. The data summarized below are contributions toward testing such suggestions, but more information over even wider areas is needed.

The Appekunny argillite is rarely well exposed in the Flathead region. It was mapped on plate 2 on the basis that scattered outcrops and the soil between outcrops attested to the presence beneath the Grinnell argillite of argillaceous rock in which purple or red beds were essentially absent. In the few outcrops seen on the southwestern slopes of the Flathead and Swan Ranges, hard, in part somewhat calcareous, dark-greenish-gray or greenish-black argillite predominates. Some beds are bluish. No distinctly reddish beds were noted; but under the microscope, material from the Swan Range is seen to contain enough hematite to have a faintly reddish tint. Argillite from Wolf Creek on the west side of the Swan Range, although of uniform appearance in the hand specimen, has a microscopic texture suggestive of movement before the rock became firmly consolidated. Most of the original sediment consisted of a quartzose and calcareous mud in which the quartz grains probably ranged up to about 0.07 millimeters in maximum dimension and were poorly rounded. Micaceous flakes, in part muscovite, in part chlorite, are now abundant and are interpreted as recrystallized clay and other constituents of the mud. Originally laminae of finer grained material, up to about 1.5 millimeters in width, were present. These laminae are now broken into short slabs scattered irregularly through the rock. Apparently after the fine-grained laminae had acquired some cohesion but before the rock consolidated, sufficient movement took place to break up the laminae. A little pyrite is present.

In Glacier National Park the character of the formation is broadly similar. A few reddish beds were noted, but they are widely separated, and most of them are much less brightly colored than those typical of the overlying Grinnell argillite. The field notes of Campbell's men speak of red beds in a few places, but it appears that everywhere these are decidedly subordinate in amount. The restriction of the Appekunny to dark largely argillaceous rocks seemingly corresponds also to Dyson's usage (1949a, p.7, 11).

In the southern tip of Glacier National Park, most exposures of the Appekunny argillite consist of greenish-gray massive argillite, almost as green as the greenish parts of the overlying Grinnell argillite. This rock consists of a uniform quartz mosaic with a mesh of micaceous flakes, largely chlorite. The quartz grains range up to 0.02 millimeter in diameter, and all have strain shadows and fuzzy borders. A little carbonate and possibly some feldspar are present. Some dark-gray to almost black beds and a few light-colored quartzite beds are present. Throughout the formation, ripple marks are fairly common.

In the localities seen in 1950 (see index map, pl. 1), exposures are more continuous than they are in the Flathead region. Here medium-dark-gray relatively massive beds predominate in the lower part, and the upper part of the formation consists mostly of greenish-gray comparatively thin-bedded material. Both here and to the south, the amount of slaty cleavage varies in proportion to the local deformation. For example, some of the dark-gray argillite in outcrops on the slopes southwest of Mount Thompson might be described as slate. In both the upper and the lower parts of the formation in the park, some beds are so calcareous that their nature is obvious at a glance. The three analyses of samples of Appekunny argillite from Glacier National Park in the table on page 55 all record the presence of small quantities of carbonate. Samples could be selected that would show much larger amounts, but none of the beds in the formation could be termed "limestone."

The thin-bedded parts of the formation are represented by the greenish distinctly layered rock from near Many Glacier Hotel, specimen ID—13950 in the table on page 55. In this rock the laminae are 0.5-1.0 millimeter wide. Some individual grains exceed 0.5 millimeter in maximum dimension, but most are 0.015 millimeter or less in diameter. Most are subangular, but a few grains are rounded, and some have grown together, so that the contacts between them are obliterated. Quartz makes up roughly 60 percent of the rock, feldspar (mostly alkali plagioclase) constitutes about 10 percent and the rest of the rock consists mainly of micaceous minerals, mostly chlorite.

The exposures of Appekunny argillite along the Going-to-the-Sun Highway above Lake McDonald include green and dark-gray argillite. The green argillite here, represented by analysed specimen ID—21550, has graded laminae 0.1-1 millimeter wide (fig. 6c). The largest grains are about 0.1 millimeter wide and are in an intricate mat of micaceous shreds, mostly chlorite. A little plagioclase is present. The dark-gray argillite from the same locality, represented by analysed specimen ID—21650, is more irregularly and indistinctly layered than the green rock. Most of it resembles the coarser parts of specimen ID—13950. The rock contains quite a little feldspar and enough carbonate to be readily seen, and pyrite is conspicuous. The three analyses show an average content of about 67 percent of silica, rather high for an argillite. The formation includes a few beds of relatively pure quartzite.

The Fentons saw the Appekunny argillite in the area from near Two Medicine Lake northward into Waterton Lakes Park. They propose to divide it into three members. They call the lowest of these the Singleshot and describe it as containing argillite and quartzite interbedded with buff to greenish siliceous dolomite and dolomitic sandstone. The fine-grained clastic rocks are gray, gray green, reddish, and black; quartzites are greenish, pink, or white. Mud cracks, ripple marks, and mud breccias are abundant. Locally, a basal coarse-grained pinkish sandstone is reported to rest on the Altyn limestone with slight angular unconformity. The thickness of the member is said to be 300-400 feet. As the Appekunny was nowhere found resting on the Altyn in localities mapped in 1949 and 1950 (pl. 1), it may be that the Singleshot member is absent or poorly exposed in the places where relatively detailed study of the Appekunny argillite was made during the present investigation.

The second and much the thickest member of the Appekunny is termed "the Appistoki member" by the Fentons, They say it contains "gray, green, olive-brown, and rusty-gray argillite in thin minor but thick major beds, interbedded with thickly stratified, greenish, white, or pink quartzite." Flat-pebble breccias, mud cracks, and ripple marks are abundant; rain and sleet prints are present in some layers. The thickness in the Lewis Range is given as 2,000-2,200 feet, which corresponds to the total thickness of the entire formation as originally estimated by Willis.

The uppermost member is called the Scenic Point member by the Fentons and is described as containing argillite, sandstone, and "gravelly conglomerate" and as being green, purplish, buff, brown, and dull brownish red at the type locality. North and west from Scenic Point, which overlooks Two Medicine Lake, the member is reported to grade into thickly bedded coarsely mud-cracked argillite, which gives way to thick quartzite and subordinate gray and iron-stained argillite beds. Mud breccias, mud cracks, and ripple marks are abundant. The thickness of the member is given as 200-700 feet.

Consideration of the various descriptions summarized above leads to the tentative conclusion that the rocks mapped during the present investigation as belonging to the Appekunny argillite correspond essentially to the Appistoki member of the Fentons. On this basis, most of their Scenic Point member may have been mapped with the Grinnell, and their Singleshot member probably is not exposed in the localities shown on the index maps on plates 1 and 2 as having been mapped in 1949 and 1950. Further, it seems probable that this member was included in the upper part of the Altyn limestone by the geologists under Campbell. As it has a large content of carbonate rocks, such an assignment would be in harmony with the work done in the present investigation.

Willis (1902) credits the Appekunny with a thickness of about 2,000 feet. The Fentons say that in the eastern part of the mountains the thickness is 2,500-5,300 feet, and they cite Clapp (1932, p. 22) as estimating thicknesses as great as 10,000 feet farther west, presumably beyond the limits of Glacier National Park. Clapp gives a minimum thickness in the park of 3,500 feet. Dyson (1949a, p. 7) says the thickness is 3,000 feet or more. In the Flathead region south of the park, the thickness is surely as much as 2,000 feet, and on the southwestern slopes of the Swan Range, it may exceed 5,000 feet. The wide variation in the thickness of the formation suggested by these figures may result from lateral variations in the lithologic character of the rocks, a feature that, in varying degree, is found in all the units in the Belt series. Erdmann (1947, p. 129-130) says that fieldwork incident to his examination of the Bad Rock Canyon dam site on the Flathead River "has demonstrated interfingering of the gray-green Appekunny lithology with the dull purplish-red Grinnell lithology in the 2,400 feet of strata occupying the Grinnell horizon in Bad Rock Canyon," as the Grinnell is mapped by Clapp (1932, pl. 1). Erdmann's conclusion reflects the difficulty in distinguishing the Appekunny argillite from the Grinnell argillite. It is inherent in any attempt at establishing stratigraphic subdivisions in a thick sequence of broadly similar argillaceous rocks such as those that compose these two formations. A similar difficulty led Stebinger to remark in his field notes of August 13, 1914, that the Grinnell-Appekunny contact on the spur between Snyder and Sprague Creeks and Lincoln Creek is "hard to locate or cannot be so located because of the large amount of green argillite in the base of the Grinnell." On the other hand, such subdivisions are needed, if the character of the Belt series is to be understood, and can generally be made with reasonable precision when systematic areal mapping over a sufficiently extensive territory is undertaken. Bad Rock Canyon at the north end of the Swan Range (pl. 3) is outside of the area mapped during the present investigation but has been visited by the writer. The green and purplish rocks so well exposed there are essentially similar to those of the Grinnell argillite in the part of the Swan Range mapped in plate 2.

Alden in his summary of data obtained by Campbell's parties says that the Appekunny argillite near Swiftcurrent Falls is composed mostly of greenish argillite, with some interbedded reddish argillite and lighter-colored quartzite. At the base he reports a transition upwards from buff Altyn limestone to greenish-buff Appekunny argillite. At the top of the formation he notes that there is a transition zone in which beds of maroon argillite are interbedded with green argillite, increasing upwards until the maroon beds characteristic of the Grinnell argillite predominate. In the fieldwork of 1949 and 1950, this transition zone would have been included in the Grinnell.

Nordeng reports in his field notes that the Collenia columnaris zone of the Fentons (= C. frequens zone of Rezak) in the upper part of the Altyn limestone is 50-75 feet below the base of the Appistoki member of the Appekunny argillite near Appekunny Falls. No fossils have been reported from the Appekunny anywhere; so it would appear that the Fentons' Singleshot member is very thin or absent at this locality.

The following is a complete section of the Appekunny argillite measured by Stebinger and Bennett in 1914 at the eastern end of Red Eagle Mountain above the Altyn limestone. (See p. 21.)

Appekunny argillite

[From field notes of Eugene Stebinger and H. R. Bennett, 1914]

Corbett and Williams, also in 1914, measured a section of the Appekunny argillite on the mountain north of Crossley Lake which, while supposedly complete, is much thinner than the section just given. Corbett's field notes state that the measurement started at the highest limestone bed in the transition zone between the Altyn and Appekunny. This transition zone, although excluded from the section as given by Corbett, includes much quartzite and some argillite, so that it might well have been included in the Fentons' Singleshot member of the Appekunny. The section below stops at the horizon designated the base of the Grinnell argillite in Corbett's notes.

Appekunny argillite near Crossley Lake

[After field notes of C. S. Corbett and C. R. Williams, 1914]

Grinnell Argillite

The Grinnell argillite is the uppermost unit of the Ravalli group and the one that is best exposed in the Flathead region. It underlies broad areas in the southwestern parts of the Flathead and Swan Ranges. In these mountains the Grinnell, like the Appekunny, is largely obscured by the extensive cover of vegetation and soil. However, the Grinnell extends high enough to reach the parts of the mountains that have been extensively denuded by glacial action. In these places parts of the formation are well exposed. In Glacier National Park the Grinnell argillite, like the Appekunny which underlies it, is exposed along the southwestern flank of the mountain mass from the Canadian border nearly to latitude 48°30' and thence northward along the eastern mountain flank back to the Canadian border. There are also exposures on both sides of Waterton Valley.

The most distinctive feature of the Grinnell argillite is the purplish and reddish color of many of the beds. Some beds are green and some are white, but most display in varying degree the characteristic color of the formation, which is due to iron in the Grinnell. Like the Appekunny, the Grinnell argillite contains many argillaceous beds. Both formations are fine grained, but the Grinnell is somewhat the coarser. It is also more siliceous and contains less carbonate and more recognizable feldspar than the Appekunny argillite does. Two of the three samples analysed (see p. 55) contain about 70 percent of silica, and none of them contain more than a small fraction of a percent of carbon dioxide. The samples selected for analysis represent the average rock, and the differences in the composition of the two formations would be emphasized if analyses of the more strikingly siliceous and calcareous beds in each were available.

The Grinnell argillite, as can be seen from figure 3, is a fairly uniform even-bedded unit. In some exposures it has a massive appearance but, wherever it is sharply deformed, as in the outcrops pictured in figures 25 and 28, the bedding is emphasized, and individual beds are seen to be thin. The argillaceous beds that compose much of the formation are resistant to weathering, which is indicated by the presence of unweathered joint blocks far from present outcrops. On the slopes northeast of Harrison Lake, for example, such blocks are so abundantly scattered over surfaces underlain by Appekunny argillite that in reconnaissance work it would be easy to imagine that the Grinnell contacts are lower on the slopes than is actually the case. Ripple marked and mud-cracked beds and intraformational conglomerate or breccia are common throughout the Grinnell argillite (figs. 4, 5) but are particularly abundant in the middle part of the formation. Figure 4 shows broken bits of thin dark argillaceous layers enclosed in white quartzite. Figure 5, in addition to ripple marks, shows bulbous forms whose origin is not clear. They are somewhat like the forms that the Fentons (1937, p. 1912-1913) regard as channel fillings.

FIGURE 3.—Air view northeast from above Gunsight Pass toward the valley of St. Mary River, with Lake Ellen Wilson in the foreground. Shows the general character and topographic expression of the Grinnell argillite, Siyeh limestone and, in the distance, the Missoula group. Metagabbro sills and the Conophyton zone 1 are visible. One of the ridges on the Great Plains that is capped by early glacial deposits can also be seen. Photography by U. S. Army Air Corps. (click on image for a PDF version)

FIGURE 4.—Grinnell argillite on Broken Leg Mountain, Nyack quadrangle, Flathead region. Somewhat more distinctly laminated than is common because of the presence of white quartzitic layers containing included fragments of argillite.

FIGURE 5.—Grinnell argillite on Broken Leg Mountain, Nyack quadrangle, Flathead region. Shows ripple marks and bulbous masses.

Stromatolites have been recorded from the Grinnell argillite only in 2 localities: 1 along the Going to-the-Sun highway and 1 in Bad Rock Canyon. (Rezak, 1957, p. 136-137) This fact may aid in distinguishing its outcrops from otherwise similar exposures of the Missoula group in which stromatolites are far more abundant and easily recognized.

The Grinnell argillite is mostly in shades of red purple which, where of characteristic color, are distinctive enough to be recognized with confidence by anyone familiar with the formation. While the Missoula group also contains many red-purple beds, colors in that group approach reddish brown in hue. Typical beds in the Grinnell argillite are definitely more purplish than those typical of the Missoula, but the distinction is sufficiently delicate and the range in color is wide enough so that, as noted below, reliance on this feature alone is dangerous.

Even with the aid of color charts, precise designation of color in rocks presents difficulties. The opportunity for confusion is increased when descriptions by different authors are compared. Willis (1902, p. 316, 322), who defined the Grinnell argillite, spoke of it as dark-red shaley argillite, Both the Fentons (1937, p. 1887-1890) and Dyson (1949a, p. 7-8) speak of the red as a conspicuous feature. Probably all these authors used the term "red" in a general sense and ignored the purplish hues. The Fentons at the beginning of their description do say "red or purplish." Even the most distinctly red beds in both the Grinnell and the Missoula have a purplish cast.

The Grinnell argillite shows under the microscope rather less variation than might be expected from the variations in its megascopic appearance. From the appearance in thin section, it is clear that the original rock ranged from a siliceous mudstone or siltstone to a somewhat arkosic sandstone, a conclusion that is in agreement with the analyses on page 55. In the more argillaceous beds individual grains are only a few hundredths of a millimeter in diameter, but in the coarser layers the grain diameter ranges from 0.4 to more than 1 millimeter. Locally, coarse grains are irregularly scattered through a fine-grained matrix. In some argillaceous rocks the bedding is so irregular (fig. 6d) as to suggest that the original mud layers were disturbed while still unconsolidated. The argillaceous rocks now consist largely of quartz and fine flakes of mica, with some feldspar and, locally carbonate. The coarser grained rocks are similar except that micaceous minerals are less abundant and in some beds are nearly absent. Feldspar, largely alkalic plagioclase, is more conspicuous (fig. 6e) in the coarse-grained layers, but this difference may be more apparent than real. The minerals would be difficult to recognize in fine-grained rocks. Some of the grains in the coarser grained rocks are themselves fragments of fine-grained sedimentary rocks, and some of the quartz grains had been subjected to marked pressure before they were incorporated in the present rock. Many of the grains in these rocks are well rounded. Some are rimmed with material added late in the process of consolidation. This feature, shown somewhat indistinctly in figure 6E, is not as well developed as it is in many quartzites, suggesting that the process of recrystallization is far from complete. Carbonate is less widespread than it is in the Appekunny argillite. It is too scanty to be detected under the microscope in many of the rocks and is abundant in very few outside of the transition zone at the top. It is plentiful in some beds in Felix Basin, but those have been hydrothermally altered, and at least part of the carbonate may have been introduced in solution.

FIGURE 6.—Photomicrographs of rocks of the Belt series. A. (Plane light) Supposed fossil in Altyn limestone near Swiftcurrent Falls, Glacier National Park. Fine-grained groundmass is carbonate, and large clastic grains are quartz and feldspar. B. (Crossed nicols) Another view of the specimen shown in A, showing grains of quartz and one of striated plagioclase feldspar. C. (Crossed nicols) Appekunny argillite from east of Lake McDonald, Glacier National Park. shows rhythmic bedding. D. (Plane light) Edgewise conglomerate in Grinnell argillite, Tom Tom Lookout, Swan Range, Nyack quadrangle, Flathead region. Note disturbed flakes of fine grained argillite in quartzitic rock. E. (Crossed nicols) Coarse-grained Grinnell argillite containing striated plagioclase feldspar, from the upper part of the formation, Wheeler Creek, Swan Range, Nyack quadrangle, Flathead region. F. (Crossed nicols) Coarse-grained white quartzite in Grinnell argillite, Park Creek, Nyack quadrangle, Flathead region. Some of the grains show rims of added quartz. (click on image for a PDF version)

In the Flathead and Swan Ranges the Grinnell argillite crops out extensively with the Appekunny argillite below and the Siyeh limestone above, evidently in normal stratigraphic relations uncomplicated by deformation. The formation consists largely of pale- and grayish-blue-green, grayish-purple, and grayish-red-purple siliceous argillite, in part slightly calcareous, and some quartzite. In the Swan Range the lowest and thickest part of the Grinnell is dominantly pale- and grayish-red-purple argillite, nowhere well exposed. Above this is the middle member in which the proportion of red-purple beds decreases upwards, and much of the rock is quartzitic argillite and argillaceous quartzite, with thin argillite partings, generally rather dark red-purple. Some of these partings more nearly resemble parts of the argillite of the Missoula group and of the red argillite in the Grinnell formation farther north than do any of the thicker beds in the Swan Range. The outcrops pictured in figures 4 and 5 belong to this middle member. The uppermost member of the Grinnell argillite commonly consists of grayish-blue-green calcareous argillite and argillaceous limestone, constituting a transition zone below the Siyeh limestone of the Piegan group. This member contains a few red-purple beds, and the unit below it contains some green beds. Nevertheless, the distinction is sufficiently definite so that the transition zone at the top of the Grinnell has been shown on plate 2. Conceivably this transition zone, or some part of it, corresponds to the "Collenia symmetrica zone," which the Fentons (1937, p. 1894) define as the "upper phase of the transition between the argillitic and arenaceous Grinnell to the dolomitic and limy Siyeh" and place at the base of the Siyeh limestone as defined by them. On that basis the uppermost member of the Grinnell as mapped on plate 2 would become the basal unit of the Siyeh limestone. This unit is more argillaceous than any part of the Siyeh limestone of the present report. No stromatolites were found in this unit in the Flathead region. If they should be found, the probability that the transition zone belongs in the Siyeh would be greatly increased.

The Grinnell argillite in the Flathead Range has features closely akin to those in the Swan Range but is even less satisfactorily exposed. Here also, the transition zone at the top is mapped, but the two subdivisions of the formation beneath this zone were not recognized, possibly because the exposures are so incomplete.

Along the northeastern side of that part of the Middle Fork of the Flathead River below its confluence with Bear Creek, Clapp (1932, pl. 1) mapped a rather large area as being underlain by the Grinnell formation, with the Missoula group on the opposite side of the river. This interpretation, which appears to have been accepted by Erdmann (1944, p. 45), would require that a large fault lies approximately along the channel of the Middle Fork. As indicated on plates 1 and 2 of the present report, the rocks on both sides of the Middle Fork below Bear Creek belong to the Missoula group and are in the normal stratigraphic position for that unit. This eliminates the necessity for a fault in the valley of the Middle Fork. No evidence of faulting here was found during the present study, and Clapp's error presumably arose from inferences in regard to color.

In most parts of the park the Grinnell argillite consists largely of reddish-purple argillaceous rocks, which in many localities are interbedded with other wise similar green rocks. Exceptionally, especially near the top of the formation, the green beds predominate. In some places, notably low in the sequence white and pink to reddish quartzite is conspicuous. Among the localities where the color is exceptionally reddish rather than purplish may be mentioned outcrops along the Going-to-the-Sun Highway west of St. Mary Lake. Here the beds might well be mistaken for components of the Missoula group if their stratigraphic position were not known.

During the present investigation the Grinnell argillite was studied more closely in the southern part of Glacier National Park than in the area north of latitude 48°40'. Where examined, the 3 subdivisions noted in the Swan Range are probably present, but the distinctions between the lower 2 are inconspicuous. The transition zone at the top is certainly present in most localities but, where seen, is less conspicuous than in the Swan Range. This may be due to absence of prominent exposures of the unit along the lines of traverse rather than to any fundamental stratigraphic difference. Nevertheless, present data make it impracticable to map the transition zone throughout the park, and it is not shown on plate 1. Some of the men who worked under Campbell distinguished the zone in the field; others did not. Allowance has been made for this so far as possible in the compilation of plate 1, but it may have resulted in places in some inconsistency in the placing of the upper boundary of the Grinnell. Insofar as information permits, that boundary is placed at the top of the rocks that are dominantly argillaceous and below those that are dominantly carbonate rocks.

The Fentons (1937, p. 1887-1890) studied the Grinnell argillite mainly in and north of the northern part of Glacier National Park in localities not examined closely during the present investigation. The three members they propose are, as judged from their descriptions, different from those in the Swan Range. Their lowest subdivision, which they call the Rising Wolf member, contains variable white and pink quartzite beds interbedded with red argillite in layers that range from mere laminae to beds 5 feet in thickness. Ripple marks and mud cracks are common. The thickness is given as 200-700 feet. The Fentons note that this member is not everywhere clearly distinguishable. The colors of both argillaceous and quartzitic beds in it differ from one locality to another. In some places, such as the vicinity of Ptarmigan Lake, green beds are conspicuous.

The thick middle unit is called the Red Gap member by the Fentons, (1937, p. 1889) and their descriptions show it to be of varied character. It is reported to consist of argillite "in thin minor and thick major beds, dominantly red but incidentally brownish or green, interbedded with pink, white, or greenish-white quartzite, brown sandstone, and sandy argillite." A characteristic feature is the thick beds of red argillite with flat mud-crack polygons. The maximum thickness is reported as 2,800 feet, but in places it thins to as little as 650 feet.

The upper part of the formation is called the Rising Bull member by the Fentons and is described by them as containing argillite, quartzite and mud breccias, forming the initial transition between the Grinnell and the Siyeh. The mud breccias of the Fentons (1937, p. 1905-1909) correspond to intraformational conglomerate. The rocks include beds of gray, red, green, pink and white. The thickness is reported to range from 600-1,100 feet. In and west of Waterton Lakes Park, a thin flow of amygdaloidal lava is intercalated in the upper part of the member, but no lava has been found in the Grinnell argillite anywhere south of the international boundary.

The wide variations in thickness estimates and in descriptions of the character of the rocks in different localities must reflect much lateral variation in the Grinnell argillite. In the Swan Range the two lower members together, as judged from plate 2, may have a thickness of 4,000-5,000 feet, and the upper member or transition zone is about 500 feet thick in the Flathead region, probably thinning northward. The unsatisfactory exposures in that region introduce a large element of uncertainty into any estimates of thickness. Nevertheless, the formation in the Flathead region must be exceptionally thick. The thickness of the whole formation in the southern part of Glacier National Park, as judged from plate 1, is close to 2,000 feet. Willis (1902, p. 322-323) speaks of the Grinnell in and near the northeastern part of the park as 1,000-1,800 feet thick. The Fentons (1937, p. 1887) give a range in thickness of 1,500-3,500 feet. Dyson (1949a, p. 7) says the thickness varies considerably but is greater than 3,000 feet in several localities.

In 1914 the Grinnell argillite was measured and described at the same two localities as the observations on the Appekunny argillite recorded above. Most of the Grinnell on Red Eagle Mountain was studied by Eugene Stebinger, but the transition zone at the top was studied by Corbett and Williams in connection with their examination of the Siyeh limestone. A thickness of 739 feet at the base of the section measured by Corbett and Williams contains so little limestone that it is here included in the Grinnell. Presumably agreement was reached in the field between Stebinger and Corbett, so that their sections do not overlap. Corbett and Williams did all of the work near Crossley Lake, but their observations were made on 2 different days and may not represent the whole of the Grinnell argillite in the vicinity. On the second day they measured 184 feet of beds which are here placed at the top of the Grinnell. (See following table.) Possibly some beds are missing between these and the section of the Grinnell they had measured earlier.

Grinnell argillite on Red Eagle Mountain

[After Eugene Stebinger, G. S. Corbett and C. R. Williams, field notes, 1914]

Grinnell argillite north of Crossley Lake

[After field notes of C. S. Corbett and C. R. Williams, 1914]

PIEGAN GROUP

The terms "Piegan group" and "Siyeh limestone" have been used in different ways by different authors. These terms are here redefined in such a way as to make the units convenient to map. In a general way, both the Fentons (1937, p. 1890-1900) and Clapp (1932, p. 22, pl. 1) recognize 3 major subdivisions within the middle group of the 3 into which the greater part of the Belt series is divided. This group is the Piegan group as the Fentons defined it (1937, p. 1890-1892). Their name has been retained, but as already indicated its application has been restricted. The 2 upper subdiviions formerly included in the Piegan group or its equivalents are here excluded from it. The lower unit of the original Piegan group is the Siyeh limestone of the Fentons and of the present report. It is the only 1 of the 3 that constitutes a homogeneous, readily mappable unit throughout Glacier National Park and the Flathead region. Argillaceous beds that have been included by some at the base and top of the Siyeh have been separated therefrom in the present report. As pointed out in the description of the Missoula group, doubt exists as to the proper correlation of the 2 upper units of the Piegan group as originally defined by the Fentons, but these 2 units seem best regarded as parts of the Missoula group.

As far as Glacier National Park and neighboring areas are concerned, the name "Piegan group," as restricted in the preceding paragraph, might well be dropped altogether. However, in a preliminary attempt at broad correlation of the subdivisions of the Belt series (Ross, 1949), Piegan group was adopted as a convenient name for the thick sequence of beds in Montana and adjacent areas that differ in details from place to place but are characterized everywhere by a high carbonate content and the group lies between the Missoula group above and the Ravalli group below. In the present report the designation Piegan group is retained because of its usefulness in correlation throughout Montana and because the thick Siyeh limestone is expected to be divided into several units of formational rank when more detailed mapping is done. When this is accomplished, the name "Siyeh limestone" is expected to be restricted in its application or abandoned altogether.

Siyeh Limestone

The Siyeh limestone underlies broad areas in the median parts of the Flathead and Swan Ranges. These areas extend almost the entire length of the parts of both ranges that are within the Flathead region and that in the Flathead Range persists to the foot of Lake McDonald.

According to Clapp's map (1932, pl. 1) and Deiss's unpublished geologic map of the Ovando quadrangle, the Siyeh limestone is continuously exposed as far south as T. 18 N., R. 15 W., and at intervals beyond this. Study of these exposures and the relations of the limestone in them to overlying beds in the light of data in the present report should yield significant information as to the interrelations of the Piegan and Missoula groups and their components. Such a study should do much to settle the problem of the upper limit of the Piegan group in the Glacier Park region. Additional mapping will be required to determine what relation the so-called Spokane formation in Glacier National Park may bear to the Spokane argillite of the Spokane Hills, east of Helena. Part of the work needed is understood to have been already accomplished (Knopf, 1950), although the maps have not yet been published.

In Glacier National Park the Siyeh limestone forms the core of the mountain mass, although capped in places by later units. It extends from near the intersection of longitude 113°30' with latitude 48°30' northwestward past the Canadian border, widening northward. Its general character can be seen in figures 3, 6, 17, 27, and 28.

The Siyeh limestone is a crystalline carbonate rock that contains varying amount of magnesia, silica, and other impurities. The analyses listed in the table on page 55 show that it contains from 20 to nearly 50 percent of silica and significant quantities of alumina, so that it is a distinctly impure carbonate rock. Some beds are more argillaceous than those selected for analysis, but no aggregates of argillite or of distinctly argillaceous limestone of mappable dimensions are known to be present.

Nearly all of the Siyeh limestone is thick-bedded or massive as viewed from a distance, but much of it shows thin, wavey laminations on fresh fracture surfaces. Close inspection commonly reveals a fine lamination. Some of the limestone is oolitic. The color of the Siyeh limestone on fresh fracture surfaces is dusky blue or, more rarely, greenish gray, varying in value and hue (Goddard and others, 1948) apparently with variations in composition. The rock weathers in orange and brownish tones and commonly shows irregular etched markings on weathered surfaces (figs. 7, 8). These features are distinctive and easily recognizable, but they are quite as characteristic of the limestone here regarded as within the Missoula group as they are of the Siyeh limestone. They correspond to differences in the calcium carbonate content of the rock, but the origin of these small-scale differences is not understood. They include the forms termed "molar tooth" structures by Daly (1912, p. 72-76). His term is derived from resemblance to the markings on a molar tooth of an elephant and is vividly descriptive of some of the structures. However, there is infinite variety in the details of form assumed by the structures, and many have no resemblance to molar-tooth markings. The Fentons (1937, p. 1927-1929) speak of them as segregation structures and, in accord with Daly, think that segregation occurred long after induration. Such an explanation is probably correct for some of the structural features, but the characters displayed and the relation, or lack thereof, to bedding and parting planes of all sorts are so different in different exposures that it seems evident that no single explanation will fit all of these features. Perhaps some have direct or indirect relation to the life processes of primitive organisms in the original lime muds from which the present limestone is derived. Figures 7 and 8 show two examples, but the variations are too numerous to be adequately pictured or described here.

FIGURE 7.—Molar-tooth structure in Siyeh limestone in roadcut along Going to-the-Sun Highway. Glacier National Park.

FIGURE 8.—Irregular banding in Siyeh limestone in roadcut along Going-to-the-Sun Highway. Glacier National Park. These irregularities may correspond to movement in the limestone before consolidation, but they are in essentially undeformed beds.

Textural variations in the limestone on a microscopic scale are quite as diverse as would be expected from the descriptions already given. Two examples are shown in figure 9. Where special features are absent the rock is laminated, and in many laminae the grains average only a few thousandths or a few hundredths of a millimeter in diameter. Beds from one to a few millimeters thick composed largely of subrounded quartz grains are widely distributed but commonly are not abundant. A little sericite is present in these beds and in some of the limestone beds. Most thin sections exhibit irregularities in the texture of the carbonate, related to the "molar tooth" and other structures that are so conspicuous in the outcrop (fig. 9A). Most of the rock does not contain well-defined oolites, but in many there are irregularities in grain that may be derived from oolites now thoroughly recrystallized. Where oolites are preserved, some are roughly elliptical in section and 1 or 2 millimeters long (fig. 9B). These are fine-grained and preserve no internal features characteristic of oolites. The same rock may contain oolites that are circular in section and have well preserved concentric structure. Some of these are broken, and others are invaded in the outer layers by grains of clastic quartz. Most such oolites seen under the microscope are less than a millimeter in diameter and are themselves components of pebble like masses embedded in the fine-grained limestone.

FIGURE 9.—Photomicrographs of rocks of the Belt series and associated igneous rocks. A. (Plane light) Irregularly textured Siyeh limestone near head of Hidden Creek, Glacier National Park. B. (Plane light) Oolitic Siyeh limestone south of Red Eagle Pass, Glacier National Park. C. (Plane light) Red argillite of the Missoula group on Argosy Mountain, Nyack quadrangle, Flathead region. A quartzitic argillite, with fragments of fine-grained argillite in it. Grains are subangular and do not appear to have been recrystallized. D. (Crossed nicols) Purcell basalt, Flattop Mountain, Glacier National Park. E. (Crossed nicols) Limestone in the Shepard formation, Flattop Mountain, Glacier National Park. Shows oval masses of coarse-grained carbonate in a fine-grained carbonate matrix. Oval bodies may be recrystallized oolites. F. (Crossed nicols) Metagabbro from an irregular intrusive mass near the Spotted Bear airfield, Flathead region. Note interstitial micropegmatite. (click on image for a PDF version)

The stromatolite zones at several horizons in the Siyeh limestone locally provide means of subdividing that thick formation, but none of these has yet been mapped throughout plates 1 and 2. In the following descriptions the zones are designated by the names of the principal stromatolites found in each, and the nomenclature proposed by Rezak is used. This fact should be borne in mind when comparing the descriptions with those of previous writers, who used different names for the stromatolites, and consequently for the zones. For discussion of the differences in nomenclature, see Rezak's (1957) paper entitled, "Stromatolites of the Belt Series in Glacier National Park and vicinity, Montana."

No mappable subdivisions of the Siyeh limestone have been recognized in the Swan or Flathead Ranges; however, some stromatolite zones are known. These and other features are expected to permit subdivision when thoroughly detailed mapping is undertaken. By far the most conspicuous masses of stromatolites found are in the vicinity of Clayton Lake. The ridge crest surmounted by Tongue Mountain, west of the lake has almost continuous exposures of these fossils for a distance of more than a mile along the trail (fig. 10). Rezak visited this ridge in 1952 and concluded that the most abundant forms belong to Collenia symetrica Fenton and Fenton, although C. frequens Walcott is also present. On Graves Creek, southeast of Clayton Lake, outcrops of stromatolites are present near the trail but are less abundant and much more imperfectly exposed than on Tongue Mountain. These probably are C. symmetrica. An outcrop of stromatolites on the ridge above the head of Forrest Creek surely represents some form of the genus Collenia, possibly Collenia symmetrica. The Siyeh limestone in several other localities in the Swan Range has wavy laminae that may represent stromatolites, but these are much too small and indefinite to be named on the basis of present in formation. Most of the scattered outcrops of Collenia noted in the Swan Range are at or near the same stratigraphic horizon. Together they may mark a zone corresponding more or less to the Conophyton zone 1 of the park (the Collenia frequens zone of the Fentons).

Most of the mapping in the Swan and Flathead Ranges was done in 1949 before the significance of the stromatolites was appreciated and before special studies of them had been initiated. In the course of this mapping no stromatolites were recorded from the Flathead Range although some are probably present there. Nordeng in 1950 found some in highway and railroad cuts at the north end of the range. Rezak regards these as dominantly Collenia symmetrica.

In Glacier National Park the Conophyton zone 1, named by Rezak (the Collenia frequens zone of the Fentons), is so persistent and easily recognized even at a distance (fig. 3) that it has been shown on plate 1 throughout the area of the Siyeh limestone. On this basis the Siyeh limestone in Glacier National Park might be thought of as having three mapped components: this zone and the beds above and below it. However, formal division of the formation is not here attempted, partly because of the difficulty with present incomplete data in carrying it throughout the region covered by plates 1 and 2. Detailed mapping will result in discrimination of map units other than the three indicated above, at least in parts of the park. Because of lateral variations it seems probable that some of these units will not be found persistent enough to be mapped throughout the park, and it may be that few of them can be carried far beyond the limits of Glacier National Park. That is, it is anticipated that future work will result in establishing local units, some of which will be valid only over a few scores or hundreds of square miles, rather than throughout the broad region underlain by the Siyeh limestone.

The Fentons (1937, p. 1892-1897) have set up four subdivisions of the formation within the park. In ascending order these, with Rezak's revisions in nomenclature shown in parentheses, are the Collenia symmetrica zone and Goathaunt member (= C. symmetrica zone 1) Collenia frequens zone (Conophyton zone 1) and Granite Park member. They describe their Collenia symmetrica zone as the "upper phase of the transition between the argillitic and arenaceous Grinnell to the dolomitic and limy Siyeh" and add that it includes quartzite, argillite, and argillaceous dolomite that weather green, brownish, or buff, with subordinate purplish-red argillite limited to the lower 75 feet of the unit. The member is reported to have a thickness of 500-900 feet and to grade upward into the Goathaunt member. It must grade downward into the upper part of the Grinnell argillite. The part that contains the "purplish red argillite" of the Fentons and any part above this in which argillite and quartzite predominate over carbonate rocks belongs with the transition zone at the top of the Grinnell as mapped in plate 2. This matter has been commented on in the description of the Grinnell argillite.

Clearly the Collenia symmetrica zone 1 does not, as the name might imply, consist mainly of a definite biostrome in which C. symmetrica is dominant. Nordeng comments that, on the contrary, it is a zone in which C. symmetrica occurs with varying frequency, with occasional colonies of Collenia clappi Fenton and Fenton. Later studies by Rezak have led him to regard C. symmetrica and C. clappii as synonymous and to drop the latter term. Probably the zone has much lateral variation and would not be recognizable throughout Glacier National Park on the basis of its stromatolite content. Thus, if an equivalent unit is mapped in the future, it is likely to receive some name other than a paleontological one. Rezak reports that C. symmetrica may be found anywhere in the Siyeh limestone below the Conophyton zone 1, mostly in isolated bioherms in dense black and tan laminated argillaceous limestone. Isolated heads of Collenia multiflabella Rezak are present in the bioherms of C. symmetrica. Most such bioherms noted by Rezak are stratigraphically higher than the C. symmetrica zone of the Fentons; that is, these bioherms are within the Goathaunt member of the Fentons.

The Goathaunt member of the Fentons consists of the greater part of the limestone typical of the Siyeh. They say it contains limestone, dolomite, and subordinate quantities of oolite, dolomitic sandstone, and argillite and add that mud breccias, commonly containing coarse sand and pebbles, are abundant in northern exposures. They estimate the thickness as 2,000-3,200 feet, whereas they give the thickness of the entire Siyeh limestone in Glacier Park as 2,900-4,000 feet.

The Fentons speak of their Collenia frequens zone (= Conophyton zone 1 of Rezak) as composed of dark-gray crystalline to amorphous limestone in one or more massive biostromes with thinly bedded calcareous or dolomitic intercalations. They say that the biostromes consist of little except Collenia frequens ( = Conophyton inclinatum Rezak) and Collenia versiformis Fenton and Fenton (= Cryptozoon occidentale Dawson) and assign a thickness of 100-156 feet to the zone. It is clear that the zone is a mappable unit, although very thin in comparison to most of the other map units in the Belt series, and it would appear that the paleontological designation given by Rezak is appropriate. Consequently, this designation is employed in the present report. The zone within the Siyeh limestone is designated the Conophyton zone 1 to distinguish it from the similar zone recently found within the Missoula group and designated the Conophyton zone 2.

The Collenia frequens zone (= Conophyton zone 1) was studied by Nordeng in 1950 and by Rezak in 1951. Their descriptions differ in many respects from each other and from that given by the Fentons. This arises partly from differences in the character of the zone from place to place, but perhaps even more from the fact that classification of the different species of Collenia has not been standardized in the past. Nordeng reports that Collenia willisii Fenton and Fenton (= Collenia multiflabella) is the principal form in the lower part of the zone. As the Fentons note that Collenia willisii (= Collenia multiflabella) is common in their Goathaunt member, which underlies the Collenia frequens zone (= Conophyton zone 1), it may be that Nordeng extended the Conophyton zone 1 somewhat lower stratigraphically than they did. He says that the beds containing C. willisii (= C. multiflabella) are followed upward by a biostrome of Collenia versiformis (= Cryptozoon occidentale), overlain by bioherms and biostromes of Collenia frequens (= Conophyton). Above these he found in most places a biostrome 2-3 feet thick "of what is usually identified as Collenia versiformis but has a stronger resemblance to Collenia columnaris ( = Collenia frequens)." Nordeng noted much variation in the thickness and abundance of the components of the Collenia frequens zone but regarded the biostrome composed of Collenia versiformis as the most persistent and the masses of Collenia frequens as the most conspicuous.

Rezak thinks of the Conophyton zone 1 as made up essentially of three parts. At the base he reports a biostrome of Collenia frequens which is relatively thin near and east of Logan Pass but thickens westward. Directly above this biostrome he found pod-like bioherms of Conophyton, 5-22 feet wide and up to 6 feet high. The closely packed heads of Conophyton in these range from a few inches to 4 feet in diameter (fig. 10). Each of the bioherms has associated with it undistorted finely laminated black and tan limestone. The laminae are continuous from the colonies on the margin of each reef and are interpreted by Rezak as off-reef deposits. Each dips west and is overlain by other bioherms. The subzone containing Conophyton with the associated off-reef deposits is about 50 feet thick near Logan Pass but only a few feet thick at Heavens Peak Lookout. If, as both Nordeng and Rezak think, the exposure of stromatolites near the lower end of Lake McDonald is the stratigraphic equivalent of the Conophyton zone 1, this particular part of that zone in absent or obscure there. (See the stratigraphic section measured at that locality by Nordeng and given on p. 40). Rezak notes that the uppermost part of the Conophyton zone 1 consists of a layer 1-12 feet thick in which Cryptozoon occidentale predominates but which also contains Collenia multiflabella. The Fentons report that their Granite Park member, the top unit of their Siyeh formation, contains magnesian limestone, oolite, argillite and quartzite, with colors that range through gray, greenish gray and brown. The thickness is given as 280-900 feet. This is one of the transitional zones between formations of the Belt series in which consistency and uniformity of definition and mapping are difficult to attain. In at least some places, the argillite and quartzite grouped with this member by the Fentons are probably included in the calcareous argillite at the base of the Missoula group as mapped on plate 2 and the southern part of plate 1.

FIGURE 10.—Conophyton inclinatum in Siyeh limestone, roadcut above the big switchback on Going-to-the-Sun Highway. Glacier National Park.

FIGURE 11.—Stromatolite on Tongue Mountain, Nyack quadrangle, Flathead region. Presumably Collenia symmetrica or a similar species.

The Fentons say that their Granite Park member contains many algal bioherms. These include large colonies of Collenia willisii (= Collenia mulitflabella) and also masses of Collenia frequens (= Conophyton) and of Collenia versiformis (= Cryptozoon occidentale). Some of these stromatolitic masses may be low enough in stratigraphic position so that they merge with the Conophyton zone 1 as shown on plate 1. This statement applies particularly to the large biostrome described by the Fentons as 6.4 miles along the highway westward from Logan Pass and 1.4 miles east of the loop, or big switchback, which was included in the Conophyton zone 1 in the mapping done in 1950 (pl. 1). Nordeng notes that a zone at the top of the Siyeh limestone is composed entirely of Collenia willisii (= Collenia multiflabella) and is present over much the same area as the Conophyton zone 1 that has been mapped. He adds that this Collenia willisii zone (= Collenia multiflabella zone) is distinguished by the presence of alternating gray dolomite and brown argillite layers. Rezak also found biostromes up to 6 feet thick and isolated bioherms consisting dominantly of Collenia multiflabella at Logan Pass and along the Garden Wall in the uppermost part of the Siyeh limestone. In addition, Cryptozoon occidentale is present here.

In the Swan and Flathead Ranges the Siyeh limestone has not been measured but has an apparent thickness of at least 5,000 feet, as judged from plate 2. Erdman (1944, p. 48-55, 88-95) supplies details in regard to the Siyeh limestone in the northern part of the Swan Range, including partial sections and chemical analyses. He estimates that the Siyeh limestone on the southeast slope of Teakettle Mountain in T. 31 N., R. 20 W., is 4,550 feet thick (accurate within 5 percent).

The Fentons, as already noted, estimate the thickness in Glacier Park as 2,900-4,000 feet, but these estimates include clastic rocks not here included in the Siyeh limestone and would, therefore, be high by several hundred feet. Comparatively detailed stratigraphic sections of the Siyeh limestone in the park, adopted from the field notes of geologists under M. R. Campbell, followed by a partial section made by Nordeng in 1950 and one measured by Rezak in 1951 follow. The most reliable and detailed section appears to be that pieced together from measurements made by Corbett and Williams in three different localities near Red Eagle Creek. (See p. 41.) As the separate measurements are each tied to recognizable horizon markers, they have been compiled below with confidence that they represent a fairly accurate picture of the entire formation in the part of the park close to Red Eagle Creek. In accord with present concepts, the red and green argillite beds measured on the upper part of Almost-a-Dog Mountain and originally listed by Corbett and Williams as the upper part of the Siyeh are not included. It seems more appropriate to regard them as part of the Missoula group. Similarly, it may be that a small part of the lowest unit that is listed below should be regarded as belonging to the transition zone at the top of the Grinnell argillite and, consequently, eliminated here. Similar modifications have been introduced in copying the two other sections listed below from the field notes of geologists in Campbell's parties. Thus, the three sections as here given represent with a fair degree of accuracy the complete thickness of the Siyeh limestone, as the term is used in the present report, in the northeastern part of Glacier National Park. It will be noted that the average thickness on this basis is close to 2,000 feet. Equally complete measurements in the western part of the park are not at hand, but would probably show a greater average thickness. In comparing the sections listed below with generalized statements as to character of the Siyeh limestone given above and in other publications, it should be borne in mind that nomenclature of color, texture, and other features is so far from standarized that descriptions of the same rock by different geologists may differ markedly.

Siyeh limestone along Garden Wall from Collenia undosa zone in lower part of Missoula group down to Conophyton zone 1 in the Siyeh on Going-to-the-Sun Highway, about 7.7 miles west of Logan Pass

[Measured by Richard Rezak, 1951]

Siyeh limestone along Going-to-the-Sun Highway near southwestern end of Lake McDonald

[Measured by Stephan Nordeng, 1950, for the present study. Rezak's revised names of stromatolites given in parenthesis]

Siyeh limestone near Red Eagle Creek

[From field notes of C. S. Corhett and C. E. Williams, 1914]

An approximate idea of the Siyeh limestone on the west side of Mount Cleveland may be gained from the following section adapted from the field notes of E. M. Parks. Thicknesses given in this section are estimates based mainly on barometric readings.

Siyeh limestone on Mount Cleveland

[From field notes of E. M. Parks, 1914]

Another composite section of the Siyeh limestone is given below. Most of this was measured by Corbett and Williams on the ridge between the North Fork of Belly River and Mokowanis River but the upper part was measured by Stebinger and Bennett.

Siyeh limestone near Belly River

[From field notes of Eugene Stebinger and H. R. Bennett, 1914]

MISSOULA GROUP

Correlations and Divisions

All the Belt series above the components already described belongs to the Missoula group. The assemblage, where not eroded, is very thick, and its mappable divisions differ markedly within short distances. In a broad way the Missoula group consists of red and green fine-grained clastic rocks with some intercalated limestone, mostly impure. Among the clastic beds all gradations between argillite and quartzite exist. Ripple marks, mud cracks, and other evidence of shallow-water deposition are plentiful. The limestone, especially the relatively pure beds, has close lithologic similarity to the Siyeh limestone described above.

Many names and descriptive terms have been applied to components of the Missoula group in Glacier National Park and other areas in northwestern Montana. It is not the purpose of the present report to attempt stratigraphic correlations throughout the broad region in which the group crops out; but as names from distant localities have been applied in previous work in Glacier National Park, some discussion is required. As already indicated, the top of the Piegan group and the base of the Missoula group are placed in the present report at the top of the definitely calcareous beds in the Siyeh limestone and at the base of beds so predominantly argillaceous that their resemblance to the dominant part of the Missoula group is obvious. This places the base of the Missoula group in and near Glacier National Park stratigraphically lower than has been done in any previous publication. The decision to do this arose from the relationships of the Missoula group in the southern part of the Flathead region to beds farther south, and, in turn, from the relationships of those beds to the Missoula group in its type locality near Missoula. So far as mapping within the park is concerned, the procedure provides convenient map units and avoids long-range correlations. It seems the only logical procedure so long as lithologic character remains the basic criterion for stratigraphic subdivision and areal mapping in the rocks of the Belt series. Some uncertainty is introduced in the interpretation of data recorded by geologists who used other limits for the divisions of the Piegan and Missoula groups. The principal difficulty is in the interpretation of the field notes of Campbell's men for areas not checked during the fieldwork of 1949 and 1950 (shown on index map of pl. 1). Any inconsistencies that may have resulted are small in terms of a map of the scale and detail of plate 1.

The argillaceous rocks above the Siyeh limestone of the present report correspond essentially to the Spokane formation of the Fentons (1937, p. 1897-1898) and to the red and green argillite band in the Siyeh of Clapp (1932, pl. 1). Clapp and Deiss (1931, p. 691-693) agree with the Fentons in correlating the argillite "band" with the Spokane formation although they do not apply it quite so specifically to rocks in Glacier National Park. The correlation of this unit in Glacier National Park with the Spokane formation of the Spokane Hills (Walcott, 1899, p. 199-215) and other localities in the general vicinity of Helena (Pardee and Schrader, 1933, p. 11, 125-126) must remain questionable, at least, until much further work is done in the broad region between Glacier National Park and Helena. No map now available brings the Spokane formation from the Helena region into the vicinity of the Flathead region, and further, neither the Spokane, of the Fentons, in the park nor units above it can be traced satisfactorily southward even as far as the Flathead region, Consequently, the name "Spokane" is not here employed, and the beds thus designated by the Fentons are regarded as an unnamed part of the Missoula group.

The argillaceous beds immediately above the Siyeh limestone in the northern part of the park are overlain by flows of the Purcell basalt, with some intercalated sedimentary rocks. Data summarized by the Fentons (1937, p. 1903) suggest that in Canada similar lava occurs at widely separated horizons in the Belt series, so that correlations in localities distant from Glacier National Park would have to be made with caution. In the park, flows belonging to the Purcell basalt are succeeded upward by the Sheppard formation of the Fentons (1937, p. 1899-1900), now spelled Shepard because of a decision of the U. S. Board on Geographic Names (Wilmarth, 1938, p. 1980). This unit can be safely mapped only where it is underlain by Purcell basalt, but it is there an easily recognized map unit and the only assembly of sedimentary beds constituting a formation within the Missoula group to which a formal stratigraphic name is attached in the present report. The Shepard formation is equivalent to the Sheppard quartzite of Willis (1902, p. 316, 324), but on the whole, it contains much less quartzite than he supposed. The Shepard formation, which is largely dolomitic, is the equivalent of the upper part of the Siyeh of Clapp (1932) as nearly as can be judged by comparing his small-scale map with plate 1 of the present report. If so, Clapp and Deiss (1932, p. 691) suggest correlation with the Helena limestone of the Helena region. The same objections apply to using Helena in Glacier National Park as those given above in regard to the Spokane formation. Willis (1902, p. 316, 324) proposed the name "Kintla argillite" for the part of the Belt series above the Shepard formation in the northern part of Glacier National Park. However, he observed a thickness of only 800 feet and, as he saw no beds above his Kintla argillite, assigned no upper limit. Daly's (1912, p. 81-82) use of the name is essentially the same as Willis'. It is now known that the Missoula group extends for thousands rather than hundreds of feet above the Purcell basalt. As there is no way of separating the Kintla argillite of Willis from similar beds at horizons above those he saw, the name is not useful at present.

In the southern part of the park and in the Flathead region, no subdivisions of the Missoula group have received formal stratigraphic names. Some names have been used by other workers in areas south of the Flathead region, but none can be used in the Flathead region with confidence because of the marked lateral changes in lithologic character that are a feature of the group. The most complete sets of names are those that have been proposed for the vicinity of Missoula, the type locality of the group, and for areas to the northeast and north. Near Missoula, division of the group into five formations was proposed by Clapp and Deiss (1931, p. 677-683). Clearly these cannot be recognized far from that area, for when Deiss carried his studies into the Saypo, Ovando, and Silvertip quadrangles (pl. 3) short distances away (Deiss, 1943, p. 211-218), he was able to recognize only the Miller Peak argillite, the lowest of the five divisions near Missoula, and proposed three new formations above this basal unit, the Cayuse limestone, Hoadley formation, and Ahorn quartzite in ascending order. While there are significant differences in character and thickness between the divisions of the group in the Silvertip and Missoula areas, Deiss (written communications, 1950) tentatively regards the part of the Miller Peak argillite (the basal formation of the group) present in the Saypo, Ovando, and Silvertip quadrangles as equivalent to the upper 1,000 feet of the Miller Peak argillite in its type locality, which is south of Missoula. He thinks that the Cayuse limestone (1,000 ft) is probably equivalent to the lower part of the Hellgate formation and that the Hoadley formation (4,100 ft) is probably equivalent to the upper half of the Hellgate formation and the whole of the McNamara formation. The Ahorn quartzite (2,100 ft) he regards as equivalent to the lower part of the Garnet Range formation which near Missoula is 7,600 feet thick. As one example of the lithologic changes implied by such correlations, it may be pointed out that Clapp and Deiss say the lower 1,600 feet of the Garnet Range formation is made up of brown and green-gray to gray thin-bedded siliceous micaceous coarse-grained quartzite, with argillitic and coarse-grained quartzitic sandstones near the base. Above this 1,600 foot unit is 600 feet of black-gray to dark-blue-gray sandy micaceous argillite. In contrast, the supposedly equivalent Ahorn quartzite is described as consisting of 1,700 feet of pink thick-bedded quartzite with occasional beds of red sandstone overlain by 400 feet of green and red-gray thin-bedded argillite with a few intercalated thin beds of fine-grained sandstone. Unfortunately no maps showing the distribution of these various divisions of the Missoula group have been published.

Comparison of the data summarized above with descriptions of the rocks in the Flathead and Glacier National Park regions given below shows that there are significant differences between beds in the southern part of the Flathead region and those reported farther south as well as differences from place to place within the two regions covered by the present report. Some of the differences in the character of the uppermost beds of the Missoula group in different localities are explained by Deiss (1935, p. 95-124) on the assumption of marked differences in the amount of removal of the upper part of the group by erosion before deposition of Paleozoic strata began.

Even if this assumption should prove to be correct, there are differences in the character of the beds within distances of a few miles that cannot be accounted for in this manner. It must be remembered that wherever the contact between Precambrian and Cambrian strata in this part of Montana has been observed, the beds are essentially flat and parallel to each other. Deiss (1935, p. 102) calls attention to two localities where angular discordances are reported, but both are far from the areas here considered. Hence, the explanation of differences in character as a result of differences in depth of erosion would have to account for differences in the amount of erosion amounting to thousands of feet within a few miles without any appreciable slopes having been produced. Drastic erosion in post-Missoula time would surely have been recorded in observable discordances at the contact with the beds deposited on the eroded surface. In the absence of known discordances the logical assumption is that there are such marked lateral differences in the components of the Missoula group that formation names applied in areas south of the Flathead region cannot be applied safely to any of the divisions of the group within that region.

For present purposes the whole of the Missoula group except units locally discriminated because of distinctive lithologic features is referred to simply as the main body. When the group is better known, more formal names will be needed, and correlation of some of the map units of the present report with already named formations in other regions may be possible. The main body, as that term is here used, is described first because it comprises much the largest part of the Missoula group and in some places constitutes the whole of that group that is present. High in the sequence in the southern part of the Flathead region, two small but distinctive units are mapped separately but described with the main body. These are greenish-gray argillite and pale pink quartzite. From a regional standpoint they do not appear to have enough significance to warrant separate description. On plate 2 and the southern part of plate 1, a discontinuous basal unit that consists dominantly of green calcareous argillite is mapped separately. Description of this unit follows that of the main body. The unit is probably present also in the northern part of Glacier National Park. If so, it has been included in that area with the main body. The green calcareous argillite, although at the base of the Missoula group where mapped, does not correspond either in character or thickness to the Miller Peak argillite, which Clapp and Deiss (1931, p. 677-683) place at the base of the group. The description of the green calcareous argillite is followed by that of the Purcell basalt, which comprises thin lava flows intercalated in the Missoula group mostly at a single horizon; and this in turn is followed by the description of the Shepard formation. Finally, a description is given of the bodies of carbonate rich rocks, other than the Shepard formation. These form more or less lenticular bodies at several horizons in the Missoula group, and some may eventually be approximately correlatable with such named formations as the Cayuse limestone of Deiss or the Shepard formation of the present report, although not actually coextensive with either. These carbonate-rich bodies bear much resemblance to the Siyeh limestone and are here referred to as limestone in conformity with established custom regarding that formation. Some of these limestone lenses are so nearly at the same relative horizon as the Shepard formation that they must be nearly equivalent thereto. In agreement with this suggestion, a few of them were called Sheppard in the field notes of parties under Campbell. Attempts in 1950 to trace mappable links between the Shepard formation in the northern part of the park and limestone masses farther south were unsuccessful.

Main Body

The greatest part of the Missoula group consists of an assemblage of grayish-red, purplish-red, brownish-red, and grayish-green dominantly argillaceous rocks that differ in details from place to place. Some quartzite, argillaceous quartzite, and argillaceous beds with different proportions of calcium and magnesium carbonates are included. This diverse assemblage, whose division must be deferred until much more detailed studies are accomplished, is provisionally called the main body of the group. The unit is one of the most widespread in the region. It occupies the northeast slopes of the Swan Range, and of the Flathead Range, and extends eastward to the trace of the Lewis overthrust in the middle of the Marias Pass quadrangle. It is conspicuous on the slopes on the northeast side of the Middle Fork of the Flathead below Bear Creek, and probably constitutes the principal rock in the unmapped areas on Hungry Horse and Firefighter Mountains west of Wildcat Mountain. The Apgar Mountains and the area on and south of Desert Mountain are composed of it. Remnants of the same unit extend, in patches isolated by erosion, from Almost-a-Dog Mountain northwestward through the median part of the park past the Canadian border west of Waterton Lake. Some idea of the general appearance of the formation can be gained from figure 3.

The argillite characteristic of this unit is softer and less siliceous than most of the Grinnell argillite. In most places reddish beds predominate, but green ones are rarely entirely absent, and some beds are gray. One mass of green argillite is so conspicuous a feature of Chair Mountain that it was mapped separately. Lithologically identical rocks are present in other parts of the main body but have not been distinguished on the map. Many are in bodies too small to show at the scale of the map. Distinctly purplish rocks, so common in the Grinnell argillite, are almost entirely absent in the Missoula group, a feature which in most localities is sufficient as a basis for discriminating between rocks of the two formations with considerable confidence. However, as noted in the description of the Grinnell argillite, the rocks of that unit, especially where exceptionally reddish, may bear sufficient resemblance to some of those in the Missoula group that confusion is possible where stratigraphic relations are not evident. Although most of the main body is argillaceous, quartzite, of varying degrees of purity, is widely distributed. In places near the top of the group, it is the dominant rock.

One distinction between the beds of the Missoula group and those at lower horizons is that the older rocks seem slightly more metamorphosed than those of Missoula age. In the Missoula group the original shapes of most of the clastic grains are preserved to a greater extent than in the Grinnell argillite. Some are subangular and, locally, even sharply angular, but many are well rounded (fig. 9c). The components of the coarser beds are more perfectly rounded than those of most of the fine-grained layers. Some of the rounded grains are coated with thin films of sericite. Some recrystallization has taken place as some quartz grains are intimately interlocked and some plagioclase grains have clear, added rims. In some of the argillite, thin very fine-grained layers were broken and twisted around before consolidation. The composition of the clastic rocks of the Missoula group is so nearly identical with that of comparable beds in the Grinnell argillite that no distinction can be made on that basis.

Here and there within the dominantly argillaceous main body of the Missoula group, certain beds weather a rusty brown that is similar to parts of the Siyeh limestone and to limestone beds above the Siyeh. Some of these rusty beds are impure limestone in assemblages so intimately interbedded with argillite that they were not mapped separately. Most of them, however, do not contain enough calcium carbonate to be detected by field tests. In a few places argillaceous beds contain small calcareous bodies that appear to be in whole or in part made up of stromatolites (fig. 15). Some are irregular lenses several feet long; others are concretionlike bodies commonly a few inches in diameter but alined within a single bed or assemblage of narrow beds for distances of tens or scores of feet along the strike. Some of these bodies are obviously calcareous, but many appear to be partly or wholly siliceous. The first variety is well displayed on the trail leading to Spruce Lookout, while the latter are common on Patrol Ridge, where together the outcrops probably represent a stratigraphic thickness of some scores of feet at least.

Ripple marks are very common in the argillite of the Missoula group, and mud-cracked surfaces are fairly so (figs. 12-15). The ripple marks tend to be smaller, less accentuated, and somewhat more uniform than those in the Grinnell argillite. In the Grinnell argillite, cross ripples and certain bulbous forms of obscure origin are rather common (fig. 5). Intraformational conglomerate is fairly plentiful but is a less common and striking feature than in the Grinnell.

FIGURE 12.—Argillite of the Missoula group with sand-filled mud cracks. Upper Twin Creek, Marias Pass quadrangle, Flathead region.

FIGURE 13.—Cusped ripple marks in argillite of the Missoula group, head of Grouse Creek, Marias Pass quadrangle, Flathead region.

FIGURE 14.—Clay spalls in argillite of the Missoula group, below Mount Bradley, Marias Pass quadrangle, Flathead region.

FIGURE 15.—Laminated argillite of the Missoula group with a small stromatolite, in railroad cut nearly opposite the mouth of Coal Creek, Nyack quadrangle, Flathead region.

Wherever sufficiently detailed observations have been made, variations within short distances in the components of the main body of the Missoula group are manifest. The scattered data available are summarized below.

On the east side of Cruiser Mountain, Masursky in 1949 found somewhat less than 1,000 feet of ripple-marked and mud-cracked red-purple and yellow-green argillite with a few quartzite beds. Next above he noted two narrow sills, grouped as a single sill on the map because of limitations of scale. Above the sills quartzite beds become increasingly numerous and thicker until at the top the rock is almost exclusively quartzite, which is in part purplish red, in part almost white. The color differences are not diffuse, but occur in speckles, shreds, and bands. The thickness of beds of the Missoula group above the sills is roughly equal to that of the exposed beds below. Unfortunately, available data as to the lateral extent of the quartzitic rocks on Cruiser Mountain do not permit representing them on the map. Most of the Precambrian rocks above the sill that is mapped are quartzitic. Near the crest of Cruiser Mountain, the beds of the Missoula group are overlain by nearly white rather coarse-grained quartzite regarded as belonging to the Flathead quartzite of Cambrian age. The lithologic difference is sufficiently well marked so that Masursky felt that the top of the Precambrian rocks could be fixed with confidence, but he found no evidence of discordance or unconformity at the contact.

On Chair Mountain, about 5 miles northwest, the uppermost beds of the Belt series are distinctly different from those just described. At this locality, exposures on the lower slopes on the Dolly Varden Creek side are poor but consist mainly of purplish-red ripple-marked argillite. On the ridge crest, light-red to pink crossbedded quartzite is overlain by several hundred feet of green argillite, which, in turn, is overlain by Flathead quartzite. The green argillite is increasingly gritty upward. This green argillite is distinguished on plate 2 but was not found in any locality other than Chair Mountain. In most places the horizon at which it might be expected has been removed by recent erosion or has been faulted out.

On Argosy Mountain most of the beds of the Missoula group are quartzitic. Most are reddish, but some are pink to nearly white, and some purplish red and green argillite is present. The beds on this mountain are so much deformed that it would be difficult to determine the proportions of the different kinds of rocks present. The lower slopes on and near Union Peak are underlain largely by red and green, locally gray, thin-bedded argillite. Near the ridge tops some light-red and nearly colorless quartzite is exposed. On and southwest of Union Peak, pinkish quartzite is so conspicuous that it is mapped on plate 2. There is much white, reddish, and greenish quartzite on Lodgepole Ridge, but like that on Argosy Mountain, it is so intermingled with argillite that it could be distinguished only on a large scale map.

Rather pale-reddish or pinkish coarse-grained quartzite crops out along the South Fork of the Flathead River close to the southern boundary of plate 2 and attracts attention in roadcuts and cliffs along the stream for some distance south of that boundary. Presumably this quartzite is near the top of the Missoula group, but the base of the block of Cambrian rocks east of the river is not exposed; consequently, the character of the rocks at the contact is not determinable.

The localities described above are in the upper part of the Missoula group. Erdman (1944, p. 56-59) in connection with his study of the Hungry Horse dam site measured the lower part of the group near the dam site. The descriptions in the section below are somewhat abbreviated from those in his report. He also gives a very detailed section of the unit near the top of the generalized section in which carbonate rocks are abundant.

Generalized section of lower part of Missoula group in secs. 4, 5, 8, 9, T. 29 N., R. 18 W.

[Adapted from section by C. E. Erdman (1944, p. 56-57)]

In the part of Glacier National Park south of exposures of the Purcell basalt, the following section of the lower part of the Missoula group was measured in July 1914 by C. S. Corbett and E. S. Williams. The top of this section is at the base of a limestone body thought by Corbett and Williams to be long to the Shepard formation. In the absence of the lava, this cannot be regarded as a definite correlation.

Lower part of the Missoula group on the upper slopes of Almost-a-Dog Mountain

[Measured by C. S. Corbett and E. S. Williams, July 30, 1914]

The Fentons (1937, p. 1897-1898) report the following section of the lowest part of the main body of the Missoula group in the basin above Hole-in-the Wall Falls in the northern part of the park. The sedimentary beds in the section represent what they called the Spokane formation, and others have referred to as the "red band in the Siyeh" or similar expressions. If the greenish calcareous argillite distinguished at the base of the group in the Flathead region had been mapped in the northern part of the park, the lower 75 feet of the section and possibly some of the beds above that would have been included in it.

Partial section of the Missoula group in Hole-in-the-Wall Basin

[After C. L. and M. A. Fenton, 1957. p. 1897-1898]

In August 1914 E. M. Parks noted that 700 feet of alternating reddish and greenish argillite beds underlies the Purcell basalt on the west side of Mount Cleveland. A more detailed section of the beds in the same locality, obtained by Eugene Stebinger and H. R. Bennett in September 1914 is given below. Stratigraphically, this corresponds to the section in Hole-in-the-Wall Basin just given, but there are many differences in detail. Similar differences would be found wherever comparable observations were made.

Lower part of the Missoula group on the ridge between the North Fork of Belly River and Mokowanis River

[From field notes by Eugene Stebinger and H. R. Bennett, Sept. 20, 1914]

In the lower part of the Missoula group near Logan Pass, Rezak found thin and individually non-persistent biostromes characterized by Collenia undosa Walcott. They are especially well developed near the east base of Mount Oberlin where eight biostromes of Collenia undosa with occasional heads of Cryptozoon occidentale and Collenia symmetrica were noted. Each biostrome is associated with alternating layers of green argillite and pink limestone. Similar biostromes are exposed 0.2 mile east of the loop on the Garden Wall, 0.6 mile east of the lower end of Lake McDonald and along U. S. Highway No. 2, 1.2 miles south of Walton. Biostromes that are probably to be correlated with these are present along the railroad west of the mouth of Coal Creek (fig. 15), along the Garden Wall near Granite Park, and on the northwestern slope of West Flattop Mountain. Similar forms are exposed in scattered outcrops along the ridge crest south of Baldhead Mountain. All these appear to be below the horizon of the Shepard formation although that unit is not mapped in their vicinity. Presumably they correspond to the stromatolites recorded by Stebinger and Bennett in the section given above.

The part of the main body of the Missoula group above the Shepard formation that remains uneroded in the northern part of Glacier National Park appears to be less than 1,000 feet thick and has been little studied. This is the unit which Willis (1902, p. 316, 324) called the Kintla argillite. The description below is adapted from one by Daly (1912, p. 81-83) and includes data from Willis' report. The name was originally taken from exposures north of the international boundary northeast of Upper Kintla Lake, Mont. The rocks are deep-red argillaceous quartzite and siliceous fissile shale or argillite with some white quartzite and occasional calcareous beds. Casts of salt crystals, ripple marks, and mud cracks are abundant. Some beds in the lower 60 feet at the head of Kintla Creek, in Canada, are lithologically identical with those characteristic of the Shepard formation. Above these is a 40-foot lava flow, which Daly says is lithologically like the Purcell lava and the flow in the Shepard formation but which is much less persistent than the Purcell lava itself. Evidently Daly used the name Purcell only for the thicker and more persistent masses of lava, but it has come to be applied to all flows of similar composition and of approximately similar stratigraphic position. The argillite contains angular grains of clear quartz, fresh microcline and microperthite, cloudy orthoclase, and a little plagioclase in an argillaceous cement containing much hematite. The section below represents, according to Daly, the Kintla formation at the type locality.

Columnar section of Kintla formation

[Quoted from R. A. Daly, 1912, p. 82]

The Fentons (1937, p. 1901) give a somewhat similar composite section for the Kintla formation at and near the type locality, quoted as follows:

The descriptions given above taken in connection with plates 1 and 2 show that most of the main body of the Missoula group has been eroded from Glacier National Park. Even in the Flathead region there are few places where Paleozoic beds remain resting on the Missoula. None of the sections measured represent more than a fraction of the original thickness of the group. Inspection of plate 2 indicates that the total thickness of the group is surely over 10,000 feet and may well be nearly twice that figure. Thus the aggregate thickness of the main body and of units mapped separately in the Flathead region is probably fully as great as the 18,000 estimated by Clapp and Deiss (1931, p. 677) for the type locality of the group.

Greenish Calcareous Argillite

The foregoing descriptions apply to those parts of the thick Missoula group that have not been separated as yet into formations or other subdivisions. The components of the group differ from place to place. In the Flathead region a nearly continuous basal zone, transitional with the Siyeh limestone below, is termed "greenish calcareous argillite." Similar rocks are present locally in Glacier National Park but have not been mapped separately there. The lithologic designation chosen for the unit reflects the fact that much of it reacts readily to dilute acid. However, the carbonate-rich beds in it, as in other components of the Belt series, do contain significant quantities of magnesian carbonate. Those parts of the unit that are especially rich in carbonate resemble the Siyeh limestone in appearance. It is correlated with the Missoula rather than the Piegan group because it is dominantly argillaceous and many of its beds are lithologically identical with some of those high in the Missoula group. Nearly everywhere there is a sharper change in lithologic character at the base than at the top of the greenish calcareous argillite. In places, in fact, the upper boundary is necessarily somewhat arbitrary as green beds are interbedded with the reddish ones higher in the sequence.

The unit is exposed on both sides of Quintonkon and Wheeler Creeks and on and northwest of Pioneer Ridge. It is thus nearly continuous at the base of the Missoula group in the Swan Range. Even in the valley of Graves Creek, where it is not mapped, some beds may be present. In the Flathead Range it has been mapped only in the vicinity of Sheep Creek and opposite the mouth of Harrison Creek. Some green beds are present near the base of the Missoula between these localities, and more detailed study might result in mapping some of the basal unit here.

The basal green calcareous argillite is well displayed in the Belton Hills but has not been traced farther north in Glacier National Park. It was not recognized by members of Campbell's parties and was not sought in the northern part of the park during the fieldwork in 1950. Such data as are at hand indicate that in most places north of the latitude of the Belton Hills it is thin or absent. Calcareous green rocks that might be correlated with it are present near the base of the slopes at the south end of Flattop Mountain, but the three measured sections of the lower part of the Missoula group include relatively few beds that resemble it. Where well-displayed in the Flathead region, the greenish calcareous argillite is estimated to be 500-800 feet thick, but in many localities its thickness is much less. It seems probable that there are few places in the northern part of the park where its thickness exceeds 100 feet.

Purcell Basalt

The name "Purcell lava" was originated by Daly (1912, p. 161-163, 207-220) for a large body of lava which he regarded as a persistent horizon marker. He says that the lava has been found in the vicinity of the international boundary from the border of the Great Plains to the eastern summits of the Purcell Range. He places the Purcell lava immediately above his Siyeh formation and units that he equated with that formation. As the composition appears to be that of an altered basalt, the unit is best termed "Purcell basalt" (Wilmarth, 1938, p. 1746).

Daly remarks that "magnesian strata characterized by the peculiar molar tooth structure" become prominent "at a horizon about a thousand feet or more below the Purcell Lava." Because the top of the Siyeh limestone, as here restricted, is in places almost as far as that below the base of the lava, the correspondence is reasonably close. As noted above, Daly apparently intended to exclude such minor flows as those in Glacier National Park from his Purcell lava, but common usage has extended the name to include the minor flows of similar composition and approximately similar stratigraphic position. Nevertheless, it should be noted that the presence of lava that looks like the Purcell basalt is not in itself sufficient for stratigraphic correlation of associated beds. In the park, most of the lava is immediately below the Shepard formation, but in the vicinity of Boulder Peak, some is well above that formation, and the small masses west of the Flathead River may be even higher stratigraphically. On the other hand, similar lava is reported (Fenton and Fenton, 1937, p. 1887-1888) to be intercalated in the Grinnell argillite in Waterton Lakes Park, Canada. This lava is not regarded by the Fentons as belonging to the Purcell basalt, and its exact character is not recorded. There appears, nevertheless, to be sufficient resemblance so that confusion would be possible in localities where stratigraphic relations are not otherwise clear.

Lava flows correlated with the Purcell basalt are exposed at Granite Park, around the periphery of Flattop Mountain, and thence at intervals northward to the Canadian border. They are present on Cathedral Peak, Mount Cleveland, Porcupine Ridge, near Kootenai Peak, near Brown Pass, Boulder Peak, and north of Upper Kintla Lake. On Boulder Peak the flows are at two horizons, separated by argillaceous and calcareous beds. Two small exposures of similar rock have been recorded on the west side of the Flathead River west of the Apgar Mountains.

In Glacier National Park the lava is an irregularly amygdaloidal rock most of which has conspicuous pillow structure. The color on fresh fracture is greenish gray to almost black, locally somewhat purplish. Weathered surfaces are stained brownish. According to the Fentons (1937, p. 1903-1904), in the vicinity of Granite Park the formation includes two major flows—

"the first 30-42 feet in thickness, the second 18 feet thick. The basal 20-25 feet of the lower flow contains ellipsoidal pillows, 10-25 inches in diameter, separated by cherty inclusions; the lavas surround detached masses of modified argillite 2-12 feet thick, the base being irregular. The upper 10-17 feet is a massive, ropy lava. The second main flow is massive and amygdaloidal, containing mud inclusions, steam tubes, and irregular cavities—evidences of subaqueous extrusion."

Near Fifty Mountain Camp the pillows are especially conspicuous and range up to several feet in maximum diameter. The Fentons (1937, p. 1903-1904) speak of 150 feet of lava on Mount Kipp, 175 feet on Cathedral Peak, 200-220 feet in Hole-in-the-Wall Basin, and 275 feet of lava west of Boulder Pass. Above Pocket Lake on Boulder Peak, they counted 8 flows in the upper 100 feet of the lava.

Daly (1912, p. 207-220) says that the lava is much altered but evidently a basalt. In the outcrops in Canada that he saw, much of the rock is a devitrified glass with numerous feldspar phenocrysts. It contains labradorite and much chlorite, leucoxene and an undetermined micaceous mineral. The analysis below is from the freshest of the specimens collected during Daly's investigation. It comes from a thick body of lava in the McGillivray Range about a mile south of the international boundary. This range is west of the Kootenai River and therefore much west of Glacier National Park. Analyses of basalt from Glacier National Park are not at hand, but they would probably be broadly similar to this one.

Purcell basalt

[From R. A. Daly, 1912, p. 209, analyst Prof. M. Dittrich]

All the lava in the park appears to be too altered for precise determination. Finlay (1902, p. 350-351) says that the lava he saw had normal diabasic texture and was composed principally of augite with less abundant idiomorphic plagioclase with the habit of labradorite. He was unable to determine the feldspar accurately. The small amount of olivine originally present in the rock was altered to serpentine and chlorite.

At the south end of Flattop Mountain, the rock is even more thoroughly altered. Traces of diabasic texture remain, and there are remnants of phenocrysts that may have originally been augite and olivine. Some of the rock (exemplified by fig. 9D) contains feathery plagioclase laths with the habit of a calcic plagioclase but with indices of refraction close to that of Canada balsam. These are crowded with specks of a secondary mineral of high index and low birefringence, presumably a chlorite. The main part of the laths now approximate oligoclase or albite-oligoclase in composition but may well have resulted by recrystallization from an originally more calcic plagioclase. Much of the rock consists of a fine-grained nearly opaque aggregate of secondary minerals that include chlorite, sulfides, calcite, quartz, and possibly serpentine. The writer's examination of this highly altered rock was kindly supplemented by Howard Powers.

L. D. Burling (1916) says the Purcell lava near Shepard Glacier is 150 feet thick and—

"composed of 6 or more flows, each of uneven and more or less ropy surface, separated by small and more or less local accumulations of shale. The lower 25 or 30 feet of the flow is composed of a conglomeration of dense, homogenous, spheroidal masses averaging 1 to 2 feet in diameter. They preserve their shape in the lower layers, being separated from each other by chert or drusy cavities, and many individuals have displaced considerable portions of the mud upon which they were rolled or shoved, even to the extent of complete burial. The bottom of the flow is therefore exceedingly irregular. Toward the top of this bed the individual spheroids yield more or less to the pressure of their fellows, and they unite to form an upper surface of moderate unevenness."

The upper part of the formation is composed of a bed about 20 feet thick which is massive but very porous. Vesicles are plentiful in the lower parts of several of the individual flows.

Shepard Formation

The Shepard formation in Glacier National Park is essentially coextensive with the Purcell basalt just described. That is, it is exposed at intervals from Granite Park northward past the international boundary. Originally Willis (1902, p. 316, 324) applied the name "Sheppard quartzite" to what he regarded as yellow ferruginous quartzite resting on the Purcell basalt along the crest of the Lewis Range in the vicinity of Mount Cleveland and Shepard Glacier between Belly River and Flattop Mountain. Because of a decision of the U. S. Board on Geographic Names (Wilmarth, 1938, p. 1980) and because much of the formation was found not to be quartzite, the name was changed to Shepard formation. The field notes of members of M. R. Campbell's parties, descriptions published by the Fentons, and observations during the present investigation are in agreement that both in the type locality and in other places in the park quartzite is subordinate to dolomitic rocks.

The formation is characterized by the fact that most of the beds in it weather to a distinctive color that ranges from pale yellowish brown to grayish orange. A typical specimen proved on analysis to be a fairly pure dolomite, as is shown in the table of analyses of Belt rocks on page 55, but many of the beds in the formation are more argillaceous or siliceous than the sample analyzed. West of Continental Creek, the Shepard contains beds of conglomerate containing lava pebbles.

As can be seen from figure 9E, the dolomite analysed is a peculiar rock. The matrix consists of dolomite grains about 0.01 millimeter in diameter, but embedded in it are numerous oval masses up to 4.0 millimeters long that consist of carbonate grains up to 0.2 millimeter in diameter. Perhaps these ovals are thoroughly recrystallized oolites.

Although in the Shepard formation as in most subdivisions of the Belt series variations in composition are plentiful, the color of weathered surfaces is so distinctive that, where coupled with the presence of lava below, the formation could be mapped with confidence. Attempts to extend the mapping of the Shepard far beyond the limits of the lava were unsuccessful. The characteristic color either disappears or is found in discontinuous beds at several horizons separated by beds that have none of the special features of the Shepard formation. The Shepard formation is reported to contain stromatolites. When these and other features have been studied in detail, it may prove possible to correlate carbonate-bearing units farther south with those now assigned to the formation.

Willis (1902, p. 316) gives the thickness of the Shepard at the type locality as about 700 feet. Parks, in his field notes, recorded 400 feet on Mount Cleveland, and Stebinger and Bennett, in their field notes, measured the following section of the formation on the slopes south of the North Fork of Belly River, where the top has been removed by erosion. The Shepard formation here overlies 115 feet of Purcell basalt, which in turn overlies the argillaceous beds of the lower part of the Missoula group measured by Willis and by Parks and listed above.

Shepard formation near Belly River

[According to the field notes of Eugene Stebinger and H. R. Bennett]

Limestone

Unnamed more or less lenticular bodies of carbonate rocks of mappable proportions are present in the Missoula group at various horizons. The largest single mass extends along the northeast side of the Middle Fork of the Flathead River from near the head of Cy Creek to a point north of Harrison Creek where it disappears under deposits of Tertiary age. This mass may have a maximum thickness of nearly 3,000 feet, but it thins at both ends. It resembles the Siyeh limestone in general appearance to such an extent that Clapp (1932, pl. 1) mapped it as part of his Siyeh group and Clapp and Deiss (1931, p. 691, fig. 3) referred to it as the upper Siyeh limestone. It is not, as persistent a feature as might be inferred from their publication, based on reconnaissance. The thinner carbonate bodies around Mount Bradley, east of Spruce Park, near Hematite Peak, and in the vicinity of Grouse and Twin Creeks represent the feathering out of the large limestone mass into the main body of the Missoula group. Bodies at higher horizons are mapped in the vicinity of Baldhead and Square Mountains. There are, in addition, detached carbonate masses near the Continental Divide from Mount Cannon to Mount Logan. Some of these last-mentioned masses are so similar in character and stratigraphic position to the Shepard formation that they were correlated with that unit in the field notes of Campbell's men. Perhaps detailed studies may trace a connection between the Shepard as now mapped and carbonate rocks farther south, through argillaceous beds with some carbonate content in intervening areas. Clearly, however, the carbonate rock in the Missoula group extends through such a wide stratigraphic range that much of it cannot be correlated strictly with the Shepard formation.

The carbonate bodies near the Middle Fork of the Flathead are somewhat better known than the outlying masses. The parts of these that are relatively rich in carbonate are closely similar in appearance and composition to the Siyeh limestone. Like much of that rock, it is a limestone with some magnesian carbonate and some clastic impurities, whereas the characteristic rock of the Shepard formation is more nearly a dolomite. The thinner bodies mapped as limestone and parts of the larger ones, especially along their margins, are more impure than typical Siyeh limestone. These impure beds contain different proportions of argillaceous and siliceous material, and some of them react feebly or not at all to field tests with dilute acid. They resemble the green calcareous argillite that in places has been mapped at the base of the Missoula group. All of the rock mapped as limestone in the Missoula group weathers with the colors characteristic of the Siyeh limestone. This is the feature that guided the sketching of parts of contacts in the field.

So-called segregation structures of one kind or another are conspicuous on weathered surfaces. Some of these are identical in appearance with the molar tooth structure for which the Siyeh limestone is noted. Others are so irregular and angular in outline that they bear no resemblance to the markings on teeth. Far more intensive study than could be given in the various reconnaissance examinations that the region has received will be required before these structural features will be understood.

Stromatolite zones are fairly plentiful in the limestone bodies of the Missoula group but have received less study than those in the Siyeh limestone. During the summer of 1952, Rezak discovered 2 hitherto unknown stromatolite zones in a large limestone lens approximately 6,000 feet above the base of the Missoula group. The larger and stratigraphically higher of the 2 zones is about 100 feet thick. It was mapped by Rezak in and near the southern part of Glacier National Park but has not been looked for farther south. Because it closely resembles the better known Conophyton zone 1 in the Siyeh limestone, it is mapped on plates 1 and 2 as the Conophyton zone 2. Only two species of stromatolites occur in the zone: Conophyton inclinatum and Collenia frequens. The zone is divided into five parts that persist over the entire area in which it was examined. At the base is 23 feet of well-developed C. frequens. The next unit is 35 feet thick and consists of large heads of Conophyton inclinatum. Overlying this is 12 feet of barren black shaly limestone. Above this is 6 feet of C. frequens, and at the top of the zone there is 24 feet of Conophyton inclinatum. The Conophyton in this zone does not occur in the podlike bioherms that are so characteristic in the Siyeh limestone.

The second zone lies about 500 feet below the one described above. It is only about 35 feet thick and contains extremely large heads of Collenia symmetrica. The colonies measure up to 5 feet in height and 6 feet in diameter. Subordinate species that occur in this zone are Cryptozoon occidentale and Collenia frequens.

Both zones are well exposed in the southwestern part of Glacier Park and the adjoining area east of the Middle Fork. They may be seen near the top of Running Rabbit Mountain and along the Great Northern Railroad tracks a few miles to the east of the mountains. The Conophyton zone 2 has a prominent outcrop west of Ole Creek. The other zone, characterized by Collenia symmetrica, is well displayed on and near Mount Furlong.

CHEMICAL COMPOSITION OF ROCKS OF THE BELT SERIES

The table below shows the chemical composition of 14 samples of rocks of the Belt series in and south of Glacier National Park. The table is arranged with the sample at the left representing the stratigraphically lowest formation. The analyses show that the rocks include impure dolomite and limestone, argillaceous quartzite, and siliceous argillite. Most of them are moderately high in silica. The most quartzose beds were not analyzed as the intention was to include material fairly representative of the different formations, rather than that representative of special features. All of the samples contain some carbonate, and only 5 show less than 1.0 percent. All of the carbonate-rich rocks contain both calcium and magnesium carbonates, and the amounts of other carbonates present must be small. When calculations are made based on the assumption that all of the calcium and magnesium occur as carbonate, it is found that 3 of the carbonate rocks show excess of carbon dioxide but 3 show deficiencies of somewhat less than 3 percent carbon dioxide. Obviously in the latter, part of the calcium and magnesium is in noncarbonate minerals. As most of the carbonate rocks have rusty colors on weathered surfaces and some are yellowish on fresh fracture, some iron carbonate is present, but the percentage of FeO recorded suggests that the amount must be small. Probably some of the calcium and magnesium in all of the rocks is in silicate minerals, which would leave a little carbon dioxide available for iron or other carbonates, even in those that seem deficient in that component under the assumption that all of the magnesia and lime are in carbonates. It is a fair guess that the carbonate rocks contain 1-4 percent ferrous carbonate.

Judged by the analyses, the Altyn limestone should be more properly called a dolomite, and the basal calcareous argillite of the Missoula group is likewise dolomitic. As many beds of the basal argillite effervesce with cold dilute acid, it is probable that much of the unit is less magnesian than the sample recorded in the table. The sample from the Shepard formation has a little more lime than the mineral dolomite, but it and most of the argillitic rocks analysed are dolomitic. The two samples from the Siyeh limestone and the one from a large body of similar rock within the Missoula group are limestones with less than 8 percent of magnesium carbonate in them. Analyses of four specimens from the Siyeh limestone quoted by Erdmann (1944, p. 54) agree in showing that calcium carbonate is much more abundant than magnesium carbonate. Obviously these limestones contain much clastic material.

A large part of the sodium and potassium present in all of the rocks analyzed probably is in feldspar grains or in minerals derived from the decomposition of alkali feldspars. Some of the argillaceous rock may contain nearly 10 percent of alkali feldspar.

All of the green argillaceous rocks contain more ferrous oxide than ferric oxide, and the reverse is true of those of purplish and reddish colors. In these rocks the coloring matter is presumably ferrous silicate, including chlorite, in the green rocks and hematite in the others. One specimen of purplish Grinnell argillite yielded 5.08 percent of ferric oxide and 2.16 percent of ferrous oxide, but in most the sum of the iron oxides is about 4 percent. One decidedly purplish-red argillite sample from the Missoula group yielded only 0.78 percent of Fe₂O₃ and 0.76 percent of FeO. Clearly very little ferric oxide, properly distributed, suffices to impart a strong color to a rock. As the ferrous oxide enters into the composition of complex silicates, a distinct green color may require much more of the coloring constituent. The amount of water in all of the samples is low; that given off below 100°C is less than 0.2 percent and in most samples well below 0.1 percent. Three samples of argillite yielded approximately 3.1 percent of water when heated above 100°C, but most of the other yielded less than 2 percent. This paucity of water is one of the indications that the rocks have begun to be metamorphosed. In most other respects they accord more nearly to typical sedimentary than to metamorphic rocks.

Analyses of rocks of the Belt series

[All analyses made in the laboratory of the U. S. Geological Survey. Samples ID—15150, 15250, 15350, 15450 were analysed by Robert N. Eceber. Samples ID—13550, 13650, 13750, 13850, 13950, 14010, 21350, 21450, 21550, 21650 were analysed by Harry M. Hyman]

(click on table for a PDF version)

PRECAMBRIAN INTRUSIVE ROCKS

The Belt series throughout the two regions here described is intruded by igneous rocks of peculiar, but on the whole gabbroic, composition. In Glacier National Park north of latitude 48°28', they form narrow sills and dikes which are almost everywhere confined to the Siyeh limestone. The sills are a few score to over 100 feet thick, and most of the dikes are 10-200 feet wide. In most places only a single sill is exposed, and this is commonly a short distance below the Conophyton zone 1. However, the sill is not so strictly conformable with the bedding as to be a horizon marker of reliability closer than several hundred feet stratigraphically. In a few localities the sill departs sufficiently from conformity with the bedding to cut across the Conophyton zone 1 at a low angle. Here and there the igneous rock is not continuously exposed, and in other places a couple of sills close together are visible. The sills form black lines on cliff faces, rendered doubly conspicuous by white border zones of recrystallized limestone (figs. 3, 16). Long, thin, steep dikes in the Siyeh limestone are conspicuous from St. Mary Lake northwestward past Lake Sherburne. Shorter dikes are exposed in other localities, such as the west end of Porcupine Ridge.

FIGURE 16.—View north from the head of Avalanche Basin, Glacier National Park. Most of the rock shown belongs to the Siyeh limestone. The dark peak at the left consists of beds belonging to the Missoula group. The Conophyton zone 1 and a sill of metagabbro are visible in the upper part of the cliffs. Note the pronounced discordance between the nearly flat basin floor at the right of the view and the cliffed gorge to the left. Photograph by Eugene Stebinger.

In the Flathead region the intrusions are more irregularly scattered, more diverse in character and mostly in the Missoula group. The most conspicuous sills, are near the top of that group on the eastern slopes of Cruiser and Ringer Mountains, but there are small ones in several other places, such as Scalplock Mountain. In addition there are several irregularly shaped intrusions, such as those on and near Lodgepole Mountain and on the ridge between the South Fork of the Flathead River and Sullivan Creek (Section C—C', pl. 2).

The character of the rock in all of the intrusions is broadly similar although almost every outcrop has peculiarities in color and texture that reflect minor differences in composition. The petrographic notes that follow are based in part on data furnished by Ray E. Wilcox, who kindly examined three of the thin sections. The rocks are composed principally of titaniferous augite, largely altered to hornblende, and zoned plagioclase that ranges in composition from An₇₅ at the core to An₂₅ in some of the outermost zones. Some separate hornblende crystals may be of primary origin. Most of the plagioclase is decidedly calcic. In addition, some potash feldspar, micropegmatitic intergrowths of quartz and alkalic feldspar, minor amounts of quartz, apatite and opaque iron oxides, and such alteration products as sericite, chlorite and calcite are present. Exceptionally, potash feldspar is sufficiently plentiful to color the rock pink. The rock has diabasic texture, obscured, however, by alteration products and interstitial micropegmatite, which commonly constitutes 10-20 percent of the whole and which is locally more abundant. A photomicrograph of this rock is shown on figure 9F.

The best available description of the intrusive bodies in the Belt series close to Glacier National Park is that of Daly (1912, p. 212-255). Although most of his observations were made north of the international boundary, they agree in general with those made south of that line. He noted rather more hornblende as an original constituent than was found in specimens examined during the present investigation. Finlay (1902, p. 349), who examined the intrusive rocks in what is now Glacier National Park in connection with Willis' studies, called them diorite; but Daly, more accurately, spoke of them as "somewhat acidified, abnormal gabbro." Because many details are obscured by alteration products and the origin of such abnormal features as the micropegmatite is still open to debate, it will serve present purposes to speak of the rock as metagabbro. Much similarity exists between this rock and that called metadiorite by Gibson and Jenks (1938), but the abundance of calcic plagioclase and the presence of residual pyroxene in the rock of Glacier National Park indicate that the original rock had the composition of gabbro rather than diorite.

S. J. Schofield (1914, 1915) has given extensive information, including analyses, on sills in a part of Canada just north of the boundary and west of longitude 115° that are much thicker but otherwise somewhat similar to the sills in Glacier National Park. He thinks the sills in his area—

"represent intrusions from a single intercrustal reservoir of a series of magmas—acid magmas—which gave rise to composite sills where rock types vary in the gabbro * * *. The simple sills solidified in the usual manner of such intrusives while the acid material differentiated under the influence of gravity giving rise to composite sills."

Differentiation by gravity would not hold for the thin, steep dikes in Glacier National Park, and the other intrusive masses in that region bear no visible evidence of being stratiform. R. A. Daly thought of the peculiar composition of these rocks as a result of assimilation of quartzite wall rocks in one locality or "metargillite" wallrocks in another, but in Glacier National Park the wallrock for most of the intrusions is limestone, and no significant difference has been found between those and the intrusions in argillaceous rocks; so, his explanation, also, seems inapplicable. N. L. Bowen (1938, p. 71-74, 82-83) cites several examples of diabasic and gabbroic rocks that contain interstitial micropegmatite. He interprets the micropegmatite as one of the last constituents to crystallize in the course of the original consolidation of the rock. He regards olivine and quartz as complementary. It seems quite possible that his explanation may fit the rocks under discussion, although the data at hand are inadequate for a final decision.

R. A. Daly, S. J. Schofield, and the Fentons are agreed that the intrusive rocks and the Purcell basalt are genetically related and, hence, essentially of the same age. This opinion is based on the fact that the two are, in numerous localities, found at neighboring stratigraphic horizons and they have similar compositions. No place is recorded where a sill or other intrusive body of metagabbro has been traced into a lava flow. Even so, a genetic relation between the two is the most logical explanation now available.