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

STRATIGRAPHY (continued)

CENOZOIC STRATA

The deposits of Cenozoic age in the part of Montana here described need more detailed study than they have yet received. As available data do not warrant use of formal formation names and require that unlike materials be mapped together in places, the map units chosen are based solely on characteristics readily recognizable in the field and will require modification when further work is done. The six units employed include old alluvium and associated deposits mapped in and west of the mountains described, remnants of deposits on the benches east of the mountains, two kinds of glacial deposits, modern alluvium, and landslide detritus. All are poorly consolidated or unconsolidated, and good exposures are rare except in stream bluffs and roadcuts. The probable range in age is from late Eocene to the present.

OLD ALLUVIUM AND ASSOCIATED DEPOSITS

As mapped on plates 1 and 2 the old alluvium and associated deposits consist of old valley fill of various kinds, mantled by unconsolidated deposits, mostly of glacial origin (fig. 19), all of which are grouped together on the maps. In most places the mantle is thin, and the old fill is widespread and fairly thick, but the fill is actually exposed in such limited outcrops that it would be difficult to depict the exposures accurately on maps of the scale of those in the present report. T. A. Link (1932, fig. 1) maps a small amount of Mesozoic rocks, the southern end of a large wedge, which crosses the international boundary and extends as far south as Kintla Lake. If Mesozoic rocks crop out here, they are included with the old alluvium on plate 1; a search for them made in the summer of 1951 failed to reveal any exposures. Much of the old fill is equivalent to the Kishenehn formation of Daly (Wilmarth, 1938, p. 1105-1106), but present data do not warrant use of the formal name. Eventually, the presence of beds equivalent to the Kishenehn formation and to other formations may be established. For instance, the old fill may include equivalents of the Willow Creek formation, which has not been recognized within the regions here mapped but is rather widespread just east of the area of plate 1 (Stebinger, 1916, pl. 15; Andrews and others 1944; Ross and others, 1955).

FIGURE 19.—The north side of the Middle Fork of the Flathead River below Double Mountain, Marias Pass quadrangle, Flathead region. The cut close to the river expose nearly white, calcareous beds of early Tertiary age, overlain by undeformed, younger gravel and sand. The hill slopes above expose part of the Missoula group.

The old alluvium and the mantling deposits mapped with it are found principally as valley fill along the Flathead River and its principal tributaries. A few patches of apparently similar material are present along minor streams in the Flathead River drainage basin; in the hills south of Big Meadows, for example. The deposits are extensive along the part of the Flathead River from the international boundary downstream to the Apgar Mountains. Near the northern end of these mountains, the river enters a narrow valley in which the old alluvium has not been recognized. However, the old alluvium continues southeast past the Apgar Mountains and into the valley of the Middle Fork of the Flathead. From near the mouth of Lincoln Creek to near Spruce Park, the valley of the Middle Fork of Flathead River is so nearly in line with that of the main stream north of the Apgar Mountains that from vantage points in the mountains nearby they might be mistaken for a single valley, notwithstanding the fact that the water flows in opposite directions in the two valleys. This part of the valley of the South Fork contains extensive old-alluvium deposits, which narrow upstream. Patches of the same material are exposed at intervals along the Middle Fork from Spruce Park at least as far upstream as Schafer Meadows, as well as in some localities in the valleys of other streams in the drainage basin of the Middle Fork. The old alluvium and associated deposits are abundant throughout the segment of the valley of the South Fork of the Flathead River that has been mapped. They extend distances up the principal tributaries of this part of the South Fork and are believed to be present on both sides of Firefighter Mountain and in the depressions in the vicinity of Abbott Ridge. It may be emphasized that all the deposits are confined to topographic depressions, mainly the valleys of the major streams. They extend up the valley sides to altitudes close to 6,000 feet above sea level near the international boundary and lower farther south. In a few places in the Flathead region, they reach altitudes of about 5,000 feet, but more commonly their upper surfaces in that region are nearer to 4,500 or 4,000 feet above sea level.

The old alluvium or valley fill in the drainage basin of the Flathead River consists of fairly well-bedded and, on the whole, rather fine-grained deposits of sand, silt, clay, and calcareous material, all thoroughly compacted and, particularly in the lower part, somewhat cemented. Lignite is present in many localities, and much of the clay and silt is carbonaceous. Conglomerate beds are reported in several places along the main Flathead River. Gravel and boulders are very abundant on the terraced and thickly forested flanks of the valley of the South Fork of the Flathead. Some of this coarse material is loose and belongs to the mantle over the old alluvium, but much is embedded in fairly well-indurated clay and constitutes an integral part of the old alluvium. Most of the exposures are in small partly slumped bluffs, in gullies and along trails. In the course of studies made in connection with the Hungry Horse dam site C. E. Erdmann (1947, p. 64-72) measured sections near the confluence of the Flathead River with its South Fork. At the time the work was done, the material measured was supposed to be of Pleistocene age, but later studies have shown that all except the upper part is Tertiary (Erdmann, C. E., letter of July 16, 1951). Hence Erdmann's sections, especially those on pages 66-67 of his published report, constitute the most detailed record at hand of the character of the old alluvium. They show that coarse gravel and boulders are fairly abundant in the vicinity of the mouth of the South Fork. The poor sorting and other features that originally led Erdmann to suppose that they are glacial deposits of Pleistocene age raise a question as to whether they record glacial activity in early Tertiary time. The Atwoods (1945) report that farther south in Montana glaciation took place during the Eocene.

Available data on the lignite are summarized briefly in the description of resources given early in the present report. Where lignitic coal is plentiful, most of the enclosing sedimentary rocks are fine grained. The large bluffs cut in old alluvium along the Middle Fork of the Flathead and some of its tributaries show little gravel, and most exposures are fairly fine grained and evenly bedded, although some contain gravel lenses and some large cobbles. In most exposures the old fill is tilted, in places as much as 60°. In the upper parts of some of the bluffs and in less perfect exposures in other places, the beds are coarser, less indurated, and more irregularly stratified and lie nearly or quite flat. Even in these beds, bouldery accumulations like those noted near and along the South Fork are scarce.

Beneath the mantle of younger, unconsolidated detritus, the old alluvium itself consists of accumulations of at least two different ages, separated locally by angular unconformities. This is in agreement with Erdmann's statement (1947, p. 134), that 2 or 3 distinct facies of the Tertiary rocks have been recognized. It is supported by the fact that, as noted below, fossils found in its upper part appear to be younger than those associated with the coal beds. It may be emphasized that in some exposures, beds like those containing these young fossils are deformed almost as much as any of the beds of early Tertiary age.

Available data on the fossils in the old alluvium are outlined below. Roland W. Brown, of the Geological Survey, has collected plant remains from outcrops of the alluvial material in the regions here described. The details of his own collections have not yet been published, but in a report dated October 17, 1950, on fragmentary plant remains collected by C. P. Ross and E. C. Stoever, Jr., from the south bank of the Middle Fork of the Flathead above the mouth of Paola Creek, Brown says: "The age is presumably that of better material, which I collected at a nearby locality, namely, late Eocene or Oligocene." This opinion is one of the most reliable so far available as to the age of the old fill in the drainage basin of the Flathead River, as it is based on much study by Brown. The collection was made near the exposures shown in figure 19 from rocks similar to the tilted beds shown in that view.

Two collections obtained in 1950 by C. P. Ross and E. C. Stoever, Jr., on the bank of the Middle Fork of the Flathead were examined by Teng-Chien Yen. In one he found Amnicola cf. A. truckeensis Yen, Fluminicola cf. F. yatesiana inflata Yen, Goniobasis? sp. (young, Vorticifex cf. V. tryoni (Meek), Pisidium cf. P. woodringi Yen, Sphaerium sp., and undetermined forms of gastropods, ostracodes and fish teeth. In the other he found Valvata sp., Vorticifex? sp., Menetus sp., Lymnaea sp., Sphaerium cf. S. andersonianum Hannibal, and undetermined forms of gastropods, pelecypods, ostracodes, and oogonia of Chara-like algae. Yen regards these fossils as "possibly of late Tertiary age," but they came from the same outcrop, near Paola Creek, as the plant remains that Brown regards as early Tertiary. These collections are from the locality numbered 2206 by Dwight Taylor, who assigns an Eocene age.

Tilted and compacted, but poorly cemented sandy silt in the west bank of Flathead River along the road near the mouth of Kintla Creek yielded non-marine mollusks which Yen regards as of middle to early Pleistocene age. This collection is numbered 20205 by Taylor, who regards the fossils as of Oligocene age. Yen says that the larger forms are represented only by fragments of shells and dentations on the inner wall of the aperture but that gastropods of minute size are better preserved. The forms he recognized are listed below. He comments that, with the exception of the Lymnaea sp., all are of terrestrial habitat and comparable forms range in age from Pleistocene to Recent.

Forms identified

Hendersonia cf. H. occulta (Sag)
Lymnaea sp. undet.
Succinea sp. undet.
Vertigo cf. V. tridentata Wolf.
Vallonia cf. V. gracilicosta Reinhardt
Heliodiscus sp. undet.

Invertebrate fossils collected from the Kishenehn formation in the valley of the Flathead River in British Columbia by L. S. Russell (1952, p. 125-126) yielded 6 species but they are all new; they are not of value in dating, and 3 forms were identified generically only. The assemblage is reported to be characteristic of small lakes or ponds, although two forms may have lived in streams. The age was thought by Russell to be Eocene—probably middle Eocene. Later Russell (1954, 1955) made additional collections, some of which are within the region of plate 1 of the present report (Russell, L. S., 1955). He lists molluscan fossils from seven localities and concludes that the age is latest Eocene. He also obtained mammalian fossils from 3 localities north of the international boundary (Russell, L. S., 1954). On the basis of these he decided to retain the Kishenehn formation in the uppermost part of the Eocene, but noted that it could with equal merit be regarded as basal Oligocene.

Dwight W. Taylor has recently made a study of fossils along the Flathead River and his written report dated June 6, 1957 is quoted here. He notes that the fossils can be divided into 4 general categories of the following probable ages; (1) Late Eocene (localities 6834, 12844, 14776, 20196, 20197, 20198, 20199, 20200, 20201, 20202, 20203, 20204, 20206), 2) Oligocene (locality 20205), (3) Miocene (locality 14778), and (4) Pliocene (locality 20207). The collection from Taylor's localities 20205 and 20206 are the same as those reported on earlier by T. C. Yen and cited above but, as noted above, the age determinations differ.

Taylor's report is quoted, except his introductory statement, which is paraphrased above.

The first category can be dated as middle Eocene to very early Oligocene on molluscan evidence alone, that is, the known range of Australorbis pseudoammonius. It is restricted to late Eocene only because of the mammals found by L. S. Russell in the Kishenehn formation, in southeastern British Columbia. Some of the localities may be as old as middle Eocene, but there is no evidence for this as yet.

At locality 20206 occurs a very limited assemblage (four species), with only one well preserved species, probably representing a new genus. This form is not very helpful stratigraphically, but is related to later Tertiary genera. The locality is given an Eocene age primarily because R. W. Brown (personal communication, 1957) believes it to lie lower than localities 20200-20204 in a conformable sequence. If there are faults in between these localities, then 20206 may possibly be younger, of Oligocene age. The age implications of this new genus are hardly compelling, however.

An Oligocene age seems more likely than a late Eocene for locality 20205. In addition to the genera listed in the table, there are two or three indeterminate land snails. The assemblage is thus almost exclusively terrestrial, in marked contrast to those from the Eocene, although it is geographically close to them. Furthermore, the assemblage is generally similar to that from the early Oligocene Dunbar Creek formation (in process of being proposed by G. D. Robinson) of the Three Forks quadrangle of Southern Montana in the large land snails and generally terrestrial composition. The helicinid is like an Oligocene form, although related to one in the Kishenehn formation. All these reasons are far from giving a definite age assignment, but they do suggest an Oligocene age.

The Miocene age assignment of locality 14778 is the weakest of the four dates, despite the excellent preservation of the fossils. About all one can say is that it seems to represent an assemblage different from the others. The Promenetus is distinct from known Pliocene or Recent species, and seems to combine characters of now-distinct subgenera. A Miocene or perhaps late Oligocene age is most likely, and a younger date less probable than an older.

Locality 20207 is very likely of Pliocene age. Only three species are known, but two are helpful. Promenetus umbilicatellus is a living species known as early as the mid Pliocene, although it may range further back in time. Lymnaea albiconica is surely known only from the middle Pliocene of Idaho, Wyoming, and Arizona, with a possible Pliocene occurrence in Oklahoma. Not only the species, but the assemblage and even lithologic character of the matrix are like those of the Teewinot formation of Jackson Hole (see Wyoming Geol. Assoc. Guidebook 11, 1956).

The fairly strong suggestion of late Tertiary as far north as Glacier Park is most interesting. This occurrence is well north of any others in the later Tertiary, and might give ecological information unobtainable elsewhere. Furthermore, there is now the possibility of dating events much later than one could have hoped. Further collecting, especially in these later Tertiary localities, would probably add significantly to the history of the area.

(click on table for a PDF version)

Systematic paleontology

Family Unionidae

cf. Elliptio salissiensis Russell. Localities 14776, 20199. Only one specimen was found at each locality. The preservation is poor, and hinge characters are not observable. The size and shape, however, are like those of Elliptio salissiensis as described by Russell.

Family Sphaeriidae

Sphaerium indet. Localities 12844, 20206. The poorly preserved material may be referable to S. progrediens Russell, but may also represent other species.

Pisidium indet. Localities 14776, 20206.

Family Valvatidae

Valvata indet. Localities 12844, 14778. Gyraulus procerus Russell is apparently a Valvata, judged by the original description and illustrations. The badly crushed material from locality 12844 can be identified to genus only, but may well be Russell's species.

Family Hydrobiidae

Indeterminate. Localities 14776, 20199, 20206. The material represents a small, high-spired, many-whorled species which cannot be identified generically.

Family Pleuroceridae

Oxytrema? indet. Localities 6834, 20199. So far as the poor material indicates, this might be the species reported by Russell as Goniobasis sp.

Family Lymnaeidae

Lymnaea (Stagnicola) newmarchi (Russell). Locality 20197.

L. (Stagnicola) aff. L. palustris (Muller). Locality 14776.

Species similar to the living L. palustris have been known from the Eocene of Europe for some time. L. newmarci and this species are the only American early Tertiary representatives of the palustris group, however. Discovery of these two has been made only recently, however; MacNeil was acting on the only available information when he thought this modern-looking Lymnaea indicated a mid-Tertiary age.

L. (Stagnicola) cf. L. albiconica Taylor. Locality 20207. This species, characterized by its peculiar sculpture, has heretofore been known only from Arizona, Oklahoma, Wyoming, and Idaho (see Wyo. Geol. Assoc. Guidebook 11, p. 123-125). The diagnostic sculpture of L. albiconica is not present on the Glacier Park material, but the coarse striae of these specimens suggest a relationship. Better material might show their specific identity.

L. (Stagnicola?) n. sp. Locality 14776.

L. (Fossaria?) n. sp. Locality 14776.

L. indet. Localities 6834, 14778, 20204, 20205.

Family Planorbidae

Australorbis pseudoammonius (Schlotheim). Localities 12844, 14776, 20196, 20200, 20201, 20202, 20203, 20204.

Several names have been applied previously to American specimens that represent this species, known up to now only in the Eocene of Europe and Asia. These names have been proposed without adequate consideration of variability in freshwater snails, and without comparison with the European species.

The only adequate study of variation in fossil Australorbis is that by Gutzwiller (1906, Schweizerische palaeontologische Gesellschaft, Abhandlungen, v. 32). He demonstrated convincingly the previously unappreciated amount of variation within single populations, and showed that other named species were synonyms of A. pseudoammonius. By this same token, the following also seem to be synonyms:

Planorbis convolutus Meek and Hayden, 1856
Planorbis utahensis Meek, 1860
Planorbis spectabilis Meek, 1860
Planorbis kishenehnensis Russell, 1952.

The precise range of A. pseudoammonius in America is still uncertain. There are no records in the Paleocene or early Eocene; many in the middle Eocene, a few in the late Eocene, and one in the very early Oligocene. The relatively few late Eocene records with independent dates (from fossil mammals) are as follows:

(1) Kishenehn fm., Flathead County, Mont., and southeastern British Columbia.
(2) Climbing Arrow fm., Jefferson County, Mont.
(3) Tepee Trail fm., northeastern Wind River Basin, Wyo.
(4) Unnamed unit, Beaver Divide, Fremont, County, Wyo.

Considering the small number of areas where late Eocene rocks are known, these four independently dated occurrences are a reasonably good representation. They certainly establish the late Eocene occurrence of Australorbis.

There is only one Oligocene record of Australorbis, in the Beaver Divide conglomerate member, Fremont County, Wyoming (dated by mammals). In evaluating this, however, one must recall that Oligocene mollusks are poorly known, and that the early Oligocene is even more poorly known. Outside the Three Forks quadrangle, Montana, there are only six mollusk localities assignable to the early Oligocene; these are scattered in Wyoming and Nebraska, and none is the result of thorough search. Australorbis thus occurs in one-sixth of all known early Oligocene localities. Furthermore, the Beaver Divide occurrence may be earlier than some of the other sites.

Planorbidae, new genus. Locality 20206; also 14776? This new form is related to Helisoma and Carinifex, and hence does give a later Tertiary aspect to the locality where it occurs. The collection at locality 20206 is adequate to show its distinctness from these other genera, however. A few poorly preserved specimens at locality 14776 may possibly be the same species.

Gyraulus indet. Localities 20197, 20198. This species is related to the living G. parvus (Say), and quite distinct from what Russell called G. procerus. As stated above, Russell's species is probably a Valvata.

Promenetus umbilicatellus (Cockerell). Locality 20207. This living species is known from deposits only as far back as the middle Pliocene. It may extend into the Miocene, and probably into the early Pliocene. On present evidence it definitely suggests a late Tertiary, probably Pliocene, age for locality 20207.

Promenetus n. sp. Locality 14778. The genus Promenetus can be divided into two distinct subgenera, one with a carinate periphery and the other with rounded whorls. This new species appears to be intermediate between the two groups in this feature.

Family Physidae

Physa indet. Locality 20199. This may be the same species reported by Russell as Physa sp.

Aplexa indet. Locality 14778.

Family Helicinidae

Helicinidae cf. "Triodopsis" buttsi Russell. Locality 20205. The species described from British Columbia as the polygyrid genus Triodopsis is much more probably helicinid. The heavy callus over the umbilicus, the pattern of apertural projections, size, and sculpture all agree more with the Helicinidae than Polygyridae. Peripherally carinate shells are known in the Polygyridae but are more common in the Helicinidae.

The single shell from locality 20205 is imperfect; although smaller than "Triodopsis" buttsi as described by Russell it is similar in shape and sculpture. Shells similar to that from locality 20205 are known also from the Oligocene in Sioux County, Nebraska (USGS localities 19093, 19094).

Family Pupillidae

Indeterminate. Localities 20205, 20207. The very poor material cannot be assigned even to genus.

Family Valloniidae

Vallonia indet. Locality 20205.

Family Succineidae

cf. Succinea. Locality 14778.

Family Urocoptidae

Holospira indet. Locality 20199. This species is distinct from H. adventicia Russell, described from British Columbia.

Family Ammonitellidae

Polygyrella indet. Locality 6834. Russell has not reported this genus from the Kishenehn formation, but it is known from other parts of western Montana. In the Three Forks quadrangle G. D. Robinson found it in the latest Eocene (and earliest Oligocene?) Climbing Arrow formation (USGS localities 20008, 20009).

Family Oreohelicidae

Oreohelix? indet. Locality 6834.

Localities

The first number listed is in the U. S. Geological Survey Cenozoic series, followed in parentheses by field numbers and Mesozoic locality numbers.

6834. Glacier Natl. Park, Mont. Bowman Creek about 1 mile from mouth. Sec. 14, T. 35 N., R. 21 W. M. R. Campbell, 1911.

12844. Flathead County, Mont. Sec. 33, T. 34 N., R. 20 W. West bank of North Fork of Flathead River. J. D. Northrop coll. This collection is cited by Alden (U.S.G.S. Prof. Paper 231, p. 29) as Eocene rather than Cretaceous according to W. C. Mansfield.

14776. Glacier Natl. Park, Mont. North side NE1/4 sec. 8, T. 31 N., R. 17 W. Scarp of sandstone on north side Halfmoon lake 2 miles north of Nyack Ranger Station. W. C. Alden, 1936, Alden (U.S.G.S. Prof. Paper 231, p. 27) listed identifications by MacNeil, who believed the collection to be of mid-Tertiary age.

14778. Glacier Natl. Park, Mont. Indian Ridge, north of Bowman Creek and lower part of Bowman Lake. C. L. Groghan collection (transmitted by W. C. Alden). Report by F. S. MacNeil, Jan. 10, 1938.

20196. (C 3). British Columbia, valley of North Fork Flathead River. South side of Couldrey Creek a short distance upstream from Burnham Creek. L. S. Russell, 1952.

20197. Glacier Natl. Park, Mont. Sec. 30, T. 36 N., R. 21 W. East side of North Fork of Flathead River just below mouth of Mud Creek.

L. S. Russell, 1952.

20198. Flathead County, Mont. Sec. 30, T. 36 N., R. 21 W. West side of North Fork of Flathead River just above mouth of Whale Creek.

L. S. Russell, 1952.

20199. (FM 304). Glacier Natl. Park, Mont. Near Halfmoon Lake, sec. 8, T. 31 N., R. 17 W. C. P. Ross, 1950. Shipment GC-50-26.

20200. (K. USGS Mesozoic loc. 24769). Glacier Natl. Park, Mont. SW1/4SE1/4SE1/4 sec. 6, T. 30 N., R. 16 W. Right bank of Middle Fork of Flathead River down stream from Tunnel Creek. Stratigraphic section C. C. E. Erdmann and V. K. Koskinen, 1953. Report by J. B. Reeside, Jr., Jan. 18, 1954.

20201. (L. Mesozoic loc. 24770). Same loc.

20202. (M. Mesozoic loc. 24771). Same loc.

20203. (N. Mesozoic loc. 24772). Same loc.

20204. (0. Mesozoic loc. 24773). Same loc.

20205. Flathead County, Mont. NE1/4 sec. 1, T. 36 N., R. 22 W. West bank of Flathead River along road near mouth of Kintla Creek. Tilted and compacted but poorly cemented sandy silt. C. P. Ross and Richard Rezak, 1951. Shipment GG-51-11. Report by T. C. Yen, Dec. 20, 1951.

20206. (FM 138). Flathead County, Mont. Southwest bank of Middle Fork of Flathead River above mouth of Paola Creek. C. P. Ross and E. C. Stoever, Jr., 1950. Shipment GG-50-19. Report by T. C. Yen, Sept. 25, 1950.

20207. (Mesozoic loc. 24774). Glacier Natl. Park, Mont. NE1/4NW1/4 sec. 13, T. 35 N., R. 21 W. Bowman Lake road about 1-3/4 miles east of junction with Kintla Lake road. C. E. Erdmann and V. K. Koskinen, 1953. Report by J. B. Reeside, Jr., Jan. 18, 1954."

Shells and fish remains have been collected at times from the beds under discussion. In the valley of the Flathead River, presumably near the international boundary, Daly (1912, p. 87) collected fossil shells among which T. W. Stanton identified the following: Sphaerium sp., related to Sphaerium subellipticum, Meek and Hayden, Valvata? sp., resembling Valvata subumbilicata M. and H., Physa sp., Planorbis sp., related to Planorbis convolutus M. and H., Lymnaea sp. The locality is probably a short distance up the Flathead River from that near the mouth of Kintla Creek that yielded the younger fossils commented on by Teng-Chien Yen. The beds from which the fossils identified by Stanton came were named by Daly the Kishenehn formation. On his geologic map this formation stops at the international boundary, but in his report he notes that an oil well penetrated 700 feet of soft shale and sandstone containing thin seams of coal that probably form the southern extension of the beds at the boundary. Willis (1902) described the same beds along the Flathead River but found no fossils. Campbell's parties do not record the discovery of diagnostic fossils in the Tertiary beds, but Alden in later visits did find some. In 1930 near the mouth of Tunnel Creek about 7 miles northwest of Walton, W. C. Alden and C. L. Gazin found the remains of a well-developed teleost fish which Gidley said was doubtless Tertiary and might be either Eocene or Miocene in age. In 1936 near Halfmoon Lake above Nyack Creek, Alden (written communication) collected gastropod and pelecypod shells which F. S. MacNeil identified as Helisoma sp., Amnicola sp., Lymnaea sp. (both high and low spired types), Pisidium sp., and Elliptio sp. MacNeil thought that the high spired Lymnaea was not older than middle Tertiary. He reported that the earliest known species of this group is from the White River beds and the species from near Halfmoon Lake is of even more recent aspect. It might well be that the shells from near Halfmoon Lake are similar in age to those near the north of Kintla Creek. Note that there is another and lower Halfmoon Lake southwest of West Glacier.

The fossils listed above range in age from Eocene to the lower part of the Pleistocene. The loose mantling deposits are largely related to the Pleistocene glaciers. Most of the old alluvium or valley fill has associated with it the plant remains which Brown regards as of late Eocene or Oligocene age, and this assignment is accepted, for the present, for the greater parts of the beds here mapped as "Old alluvium and associated deposits." However, deposition has recurred at intervals from the early Tertiary to the present, so that part of the map unit must be young.

TERRACE ALLUVIUM

East of the mountains certain of the flat-topped hills are remnants of formerly more extensive benches and terraces. For present purposes deposits that occur on distinct benches but are not known to be of glacial origin are mapped together. On the plains in the general vicinity of the mountain front, W. C. Alden (1932, p. 12-18, 44-45, 70-71) has discussed and mapped bench deposits that he regards as of several different ages. M. Y. Williams and W. S. Dyer (1930, p. 91-113) also give a convenient summary of data on bench deposits in Alberta and northern Montana. The field notes and maps of geologists whose work has been utilized in the present report do not permit mapping distinctions between the different bench deposits satisfactorily, and the mapping done during the present investigation was not extended into the plains area. The bench deposits are poorly consolidated or uncemented gravel and sand that were laid down by streams in erosion cycles earlier than the present one. Fossils have not been found in the bench deposits within the two regions here reported on, but the relations are such as to lead Alden to regard most of them as of Pliocene (?) and early Pleistocene age—an inference which seems in accord with all data now available although some of the bench deposits may be of pre-Pliocene age.

GLACIAL DEPOSITS

In recognition of the incompleteness of available knowledge, different kinds of glacial deposits are not distinguished on the maps. At least four kinds are present and have been shown in parts of Alden's maps, published and unpublished, but further field work is needed to map satisfactorily the different deposits. The four kinds include remnants of the deposits of pre-Wisconsin glaciers, deposits left in the mountains by glaciers of Wisconsin age, ground moraines spread by mountain glaciers that reached the Great Plains, and deposits left by continental ice sheets.

On the plains east of Glacier National Park, a diverse assemblage of highly deformed rocks is overlain by a mantle, which is thin in many places, of unconsolidated material that is largely ground moraine from mountain glaciers of Wisconsin age. This mantle is not shown on plate 1. Too little is known about it to permit the mapping of its components or accurate delineation of its limits. It is widespread between the mountains and St. Mary Ridge, but numerous outcrops of the underlying rocks are present in this area. Between St. Mary Ridge and Milk River Ridge, deposits assignable to ground moraine are scanty, and outcrops are so poor that little has yet been learned as to the pre-Pleistocene rocks. Therefore, it has seemed wise to show east of the mountains on plate 1 only those Cenozoic deposits that are indiscrete, relatively easily recognizable units. The areas shown on plate 1 as containing undifferentiated Cretaceous deposits have rocks of that character either in outcrops or under thin cover. In these areas future detailed work should yield much more detailed information in regard to the Cretaceous rocks than is now at hand.

Although little is known of the rocks in the plains east of the park, on the plains in the eastern part of the Flathead region, the limits of the morainal deposits are known with sufficient accuracy so that it has seemed desirable to show them on plate 2, even though it is appreciated that future work will result in modifications. Where Cretaceous rocks are shown in the part of plate 2 east of the mountains, Stebinger and his associates, in work already cited, have obtained so much information that the attempt to record it has been made on pl. 2.

Glacial deposits in topographically high situations on the plains east of Glacier National Park are known from the studies of Alden and his associates to have been laid down relatively early in the Pleistocene and probably in more than one stage. Alden (1912, 1932, Alden and Stebinger, 1913) has described these deposits and mapped them in a general way. From his work it is clear that Swiftcurrent Ridge, Boulder Ridge, Saint Mary Ridge, and other uplands east of the park are in part capped by early Pleistocene glacial deposits, which are now somewhat cemented and much weathered. Drift similar in age to that on these ridges must have been deposited within the mountains but has not yet been recognized there.

In unpublished manuscripts Alden speaks of the spur crests east of the upper reaches of the Flathead River (the part locally called the North Fork) as correlative with his Flaxville Plain. He regards most of the material mantling the spur crests as stream gravel but adds that in places it may be glacial drift. Any drift here would be in positions analogous to those of the early glacial drift east of the park. Surely comparably old drift must have been deposited here and in the valleys of other forks of the Flathead. Although most exposures are poor and little detailed work has been done, it is suggested that most of the drift grouped on plates 1 and 2 with the old alluvium may be of early Pleistocene rather than Wisconsin age. Most of it is in topographic locations that indicate it is older than the drift low in modern valleys. Presumably all drift mapped as such in the western parts of plates 1 and 2 is of Wisconsin age.

The drift in the northeast corner of plate 1 was deposited along the western border of one of the broad ice sheets that once covered much of North America. This drift is reported to include boulders that must have come from far to the north in Canada. In contrast, all the remaining drift in the two regions here described is composed of material like that of the bedrock in the neighboring mountains.

Most of the glacial deposits that are widespread in the mountains of both regions are of Wisconsin age. So far as has proved practicable, these have been mapped on plates 1 and 2. Many of the individual moraines, especially those in the vicinity of the cirques, are too small to be mapped or have not been recorded with sufficient accuracy to be shown. This probably applies also to all glacial deposits of post-Wisconsin age, including those of existing glaciers. In addition, in the stream valleys the products of glacial action merge with alluvial deposits that have little relation to glaciers. The deposits in the mountains mapped as of glacial origin, especially in those areas seen by the present writer, are mainly those whose surface form marks them as of that character. Necessarily, however, the distinction between glacial and alluvial deposits is somewhat arbitrary.

MODERN ALLUVIUM

In the mountains most of the streams have flood plains, the largest of which are terraced. Those flood-plain deposits that are wide enough to be shown have been mapped in all areas visited during the present investigation as modern alluvium. In most of the areas seen only by Campbell's parties, data for showing the extent of the modern alluvium are incomplete. Everywhere, the deposits of unconsolidated sand, silt, and gravel are thin, and in most valleys they are narrow. In general, the modern alluvium is later than the deposits related to the Pleistocene glaciers, but most of it is older than the present erosion cycle. This is proved by the fact that, in the mountains, the streams are in narrow gorges that have cut into and in many places through the modern alluvium and other poorly consolidated material into the hard rocks. Mostly the inner rock-walled gorges are too narrow to be shown, but some of the most conspicuous of them are represented on the maps. One example is along Nyack Creek south of Red Eagle Mountain.

LANDSIDE DETRITUS

Small landslides are fairly numerous, but only those that interfere with tracing bedrock boundaries have been mapped. These include two areas in the southeastern part of the park, one along the aptly named Debris Creek near the head of Ole Creek, and one along an unnamed tributary of Park Creek. A third area, near Dolly Varden Creek south of Gable Peak, is in the southern part of the Flathead region.