CORRELATION

of the area

INCLUDING

KIMBERLEY, METALIHE AND COEUR D'ALENE

by

Camon Glenn Cheriton

A THESIS SUBMITTED IN PARTIAL FULFILMENT OP THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE IN THE DEPARTMENT OF. GEOLOGY AND GEOGR/PHY

THE UNIVERSITY OF BRITISH COLUMBIA APRIL 1949 SUMMARY Within the area under consideration there are two great series of strata. The lower one is known as the Purcell-Belt Series and is divided into two main groups. A widespread unconformity separates the Purcell Series from the younger and overlying Windermere Series. The Lower Purcell-Belt group consists of the Aldridge-Prichard, Creston-Ravalli, Kitchener-Wallace, and Siyeh-Striped Peak. They were deposited under marine conditions from the erosion of a western Precambrian Cascadia. The Upper Purcell-Belt group consists of the Dutch Creek, Mount Nelson and their equivalents in Canada and the Missoula Group of Montana and possibly the Priest River group of Washington. This group is separated from the Lower Purcell by a period of diastrophism marked by the intrusion of Purcell sills and the extrusion of Purcell lavas. The Upper Purcell-Belt sediments were derived from the positive areas as a result of the preceding crustad disturbance. The Purcell-Belt times were closed by large scale orogeny called the "Purcell Uplift". The north-south trending belt of Purcell mountains formed a landmass which greatly affected lower Palaeozoic stratigraphy. This positive area is commonly referrdd to as the "Montana Island". The Precambrian portion of the Windermere Series includes the Toby-Shedroof conglomerate, Irene- I3.eola volcanics and the Horsethief Creek-Monk formations. The clastic formations were derived from the Purcell Mountains and deposited on their western flank. Marine conditions arose during Horsethief Creek times. The Cambrian portion of the Windermere Series was deposited in a north-south trending geosynclinal trough which extended from the Metaline quadrangle to the Field- Golden area of the Rocky Mountains and probably beyond. It includes the lower quartzltic Hamill Group and the overlying limy and argillaceous Lardeau group. They were deposited as the shoreline transgressed south and east over the "Montana Island" and reduced It from one of high relief to one of low relief. Stages of emergence and resumed sedimentation are indicated, by upper formations of the Lardeau group. CONTENTS Page Introduction 1 Acknowledgment 3 Problems and Approach to the Correlation of the Purcell-Belt Series 4 Aldridge-Prichard 6 Oreston-Ravalli 12 Kitchener-Wallace 20 Siyeh-Striped Peak 25 General Summary and Theoretical Considerations of the Lower Purcell Belt Group 30 Upper Purcell-Missoula Group 39 General Summary and Theoretical Considerations of the Upper Purcell Group 44 Problems of Correlation of the Windermere Series 49 Toby-Shedroof Conglomerate 51 Irene-Leola Volcanics 53 Horsethief-Monk 54 Hamill Group 56 Badshot Limestone 60 Lardeau Group 61 General Summary and Theoretical Considerations of Winder^mere Series 65 Bibliography 81

Illustrations: Plate 1 37 A Plate 2 m Plate 3 74 Plate 4 80

Appendix Compiled map of the Kimberley-Metaline- / Coeur d' Alene Area Correlation chart of the Windermere and Purcell-Belt Series INTRODUCTION Cambrian and Precambrian strata, upwards of 80,000 feet in thickness, are exposed in the Kootenay district of south-eastern British Columbia. They trend south to underly large areas of Washington, Idaho, and Montana. Considerable work has been done in these areas by officers of the Canadian and the United States geological surveys, by the Idaho Bure.au of Mines, and a number of independent investigators. As a result of this work a considerable body of information has been accumulated. The following brief study is an attempt to link the Canadian and American data and interpret it without regard to geographical boundaries. In connection with the work a map on a scale of 1 inch to 4 miles has been compiled (see pocket). In a regional study of this nature many problems of sedimentation, stratigraphy,and correlation become apparent. The writer has drawn attention to several of these problems in the general summary. For example, the rusty-red weathering nature of the Aldridge- Prichard formation requires a careful analysis of the facts for a satisfactory explanation. However, time was not available for a complete analysis of most of the problems, and as a result they are presented and dismissed with but hasty generalizations and a possible answer. It is hoped that this may form a basis for future investigations. 2

The writer haa had the advantage of personal acquaintance with some of the problems involved during 2 seasons of field work in the East and West Kootenay districts of B.C. Recent palaeontological discoveries have shown that younger strata of the supposedly Precambrian rocks of the Kootenay district are of Cambrian age. Further• more, these rocks can be correlated across the great physiographic feature, "the Rocky Mountain Trench," to link with the Cambrian strata of the Rocky Mountains. This leads to the postulation of new concepts regarding the location and movement of the geosynclinal basins of deposition preceding and during early Cambrian times• On the Canadian side, the tremendous thickness of strata is divided into two series by a great angular unconformity, at the base of the Toby-Shedroof conglomerate. Formations below or older than the unconformity are included in the Purcell-Belt Series and those above, in the Windermere Series. The first part of thfia thesis deals with the Purcell-Belt Series and the second part considers the Windermere Series and its equivalent American formations. 3

ACKNOWLEDGMENTS

The writer wishes to acknowledge the constructive criticism and biostratigraphical assistance received from Dr. V.J. Okulitch of the Department of Geology and Geography of the University of British Columbia. Dr. H.C. Gunning has also made many helpful suggestions drawn from his own field experiences in this area. 4

PROBLEMS OF CORRELATION OF THE PURCELL-BELT SERIES

The greatest single handicap in attempting the correlation of the Proterozoic strata is the lack of fossils. Other than a few algal structures the Purcell-Belt Series is unfossiliferous. The formations of Lower Purcell show no sign of disconformity and tend to grade into one another so that precise boundaries do not exist. Even over short distances some formations change laterally both in thickness and lithological character; therefore, one becomes rather dubious of a regional correlation. . Some districts have been mapped accurately on a scale of one mile to the inch; other areas have been done only in a sketchy reconnaissance fashion on a scale of 8 miles to the inch. Furthermore, the lithologic units - in some areas have not been satisfactorily differentiated. To simplify the treatment of formational names, each formation has been given a composite name in which the most common Canadian and American names are linked with a hyphen.

APPROACH TO PROBLEMS OF CORRELATION OF PURCELL-BELT SERIES

Since fossils and unconformities are not present in the Lower Purcell Group it is necessary to rely upon lithology and stratigraphy for correlation. Actual continuity of a formation as traced by the geologist in the field is the ideal basis for a lithological correl• ation. This is impossible for a large area with inform• ation from isolated districts by different observers. So that it is necessary to consider the stratigraphic succession and the lithologic description of each formation of each map area, compare them, and correlate similar strata. For a graphic presentation several sections have been plotted on vertical scale of 1 inch equal to 5000 feet to form a correlation chart which is included with this discussion (see pocket). Attention is now turned to the plotted sections of the Purcell-Belt Series. A horizontal line has been drawn through the Purcell lavas in the Canadian section on top of the Siyeh-Striped Peak formation in the western Belt Series, and on top of the calcareous Wallace or equivalent formations in the eastern Belt section. It is realized that this probably does not represent strata deposited at the same time but a better datum is lacking. Those formations below this horizontal line are referred to as the Lower Purcell Group and those above as the Upper Purcell Group. The Lower Purcell Group includes, from the bottom up, the Aldridge-Prichard, Creston-Ravalli, Kitchener-Wallace and Siyeh-Striped Peak formations. The first part of the section on the Purcell- Belt Series consists of a brief description of the most 6

likely correlative formations based largely upon lithology. The description is largely verbatim from the areal geological report by the geologist to whom reference is made. The sequence of map-areas under discussion is based upon the correlation chart which consists of type sections selected from areas first in a north-south direction and then in a West-east direction, respectively along and across the trend of the Purcell-Belt Series. From north to south these sections are:.. Cranbrook Area, Nelson Area East, Boundary County, Clark Fork District, Trout Creek Quadrangle, and the Coeur D'Alene district. From west to east the sections include Kootenai County, Coeur D'Alene district, Mission Range, Missoula District, Phil)fipsburg Quadrangle, and the Belt Mountains.

Aldridge-Prichard Cranbrook Area Except in one small part of the region the Aldridge formation is the lowest stratigraphic subdivision of the Purcell Series; however, in the Cranbrook Area north of Fort Steele in the Rocky Mountains, 7000 feet of rock are exposed and have been placed by Rice (1937,page 4) below the Aldridge and named the Fort Steele Formation. The lowest member of the Fort Steele consists of hard massive quartzite and argillaceous quartzite having well developed crossbedding and ripple marks and a thickness in excess of 1000 feet. These rocks grade upward into 2000 7

to 3000 feet of banded dark grey argillite and white to grey quartzite. Above these rocks lie 2000 to 3000 feet of massive black calcareous or dolomitic argillite, in turn overlain by 300 to 500 feet of very massive, grey- green, dolomitic argillite, which in one place is topped with a bluish limestone. The Aldrj^ge formation conformably overlies the Fort Steele and appears to Rice to be at least 16000 feet thick. He describes it .y.1nri i HT] if as "grey, rusty weather• ing argillite and argillaceous quartzite, the latter occurring in beds up to 6 feet thick separated by narrow argillite partings." The Aldrfge is characteristically devoid of limestone and limy rocks. Some conglomerate beds of limited extent occur within the formation and appear to be of an intraformational type. Sedimentary structures include well preserved crossbedding, mud- cracks, and ripple marks. Usually the top of the form• ation is almost everywhere argillaceous, with narrow bands of green argillaceous quartzite alternating with grey argillite. Except toward the top of the formation, argillaceous quartzites predominate in the Purcell Range and argillite in the Rocky Mountains.

(Nelson Area East) In this area Rice (1941, page 8,) notes that the Aldri'ge is similar to that around Cranbrook. It is universally rusty weathering and consists of argillite 8 and argillaceous quartzite, the latter in beds generally about 1 foot thick, but in some places as much as 10 feet thick. Also in the lower part of the formation argillaceous quartzite and quartzite predominates and grades upward into more argillaceous sediments until the top is largely argillite with slates occurring locally.

Boundary County Kirkham and Ellis (1926, page 15), describes'the Aldri'ge-Prichard formation as being made up of pure quart- zites, argillaceous quartzites, and argillites. The quartzites are generally massive, occurring in beds as thick as eight feet. The argillites are thin-bedded, and the argillaceous quartzites have a lesser thickness than the pure quartzites. Here, again the general weathered appearance of the formation is a reddish, rusty brown. Pew evidences of shallow-water deposition were noted by Kirkham and Ellis; this fact may be due to the Boundary County section being somewhat deeper in the Aldrige as a result of the Moj^ie-Lenia fault which passes along the eastern boundary of the county.

Clark Fork District Based upon a prevailing eastward dip of 28° for 12 miles Anderson (1930, page 14) has estimated the thickness of the Prichard in this area at more than 20,000 feet. He describes the formation as composed largely of argillaceous sandstone with some shale and is more arenaceous than the 9

corresponding formation in the Coeur d'Alene district. Here again, reference is made to the upper 1500 feet which are conspicuously shaly and bear resemblance to the typical Prichard slate of the Coeur d'Alene. Minor lenses of intraformational conglomerate may be found near the middle of the section in some of the quartzite and sandstone. As with the Aldrige in the north this formation may be readily distinguished from the overlying Burke by its rusty weathering character, particularly along fractures. Where metamorphism has not been severe, sun cracks and ripple marks are preserved in many horizons throughout the formation, but these features are not as abundant as in the upper members of the Belt Series.

Trout Creek Quadrangle Gibson^Jenks^and Campbell (1941, page 367) estimate the thickness of the Prichard in the Libby quadrangle, which is immediately north of the Trout Creek Quadrangle, to be at least 9700 feet with the bottom not exposed. Dark-gray argillite, which weathers to a rusty brown, makes up most of the Prichard; interbedded light-colored sandstone, and quartzite are quantitatively less important. Towards the top, the beds become sandier and lighter in colour, but there is no well defined contact zone of the Prichard and the Ravalli. The boundary is in the middle of a zone from 200 to 700 feet thick in which gray sandstone becomes ;' increasingly abundant upward. It is also pointed out that 10 certain minor differences exist between the Prichard of this area and the type locality of the Coeur d'Alene district. First slaty cleavage is less conspicuous in the Libby and Trout Creek Quadrangles, and secondly; the Prichard in this area is somewhat calcareous, whereas no lime carbonate is reported in the Prichard of the Coeur d'Alene district.

Coeur d'Alene Ransome and Calkins (1908, page 29) estimate the thickness of the Prichard to be more than 8000 feet; how• ever, exposures are poor and the base is not exposed so that they believe the actual thickness is considerably in excess of this figure. The bulk of the formation consists of dark argillaceous material with some thin arenaceous layers, but well down in the formation there are thick beds of sandstone. Within a few hundred feet of the top of the formation indurated sandstones, sandy shales, and quartzites increase in abundance, and in the uppermost part appears a greenish-gray shale like that so abundant in the Burke formation. Accompanying this change in lithology are abundant shallow water features of deposition as ripple marks and mud cracks. The most predominant argillaceous rock in the Prichard is a variety of dark bluish argillite, which has a distinct shaly parting and its bedding is also marked by a very regular banding in darker and lighter 11

shades of blue-gray; upon weathering, this rock is stained along bedding planes and joints with reddish-brown iron oxide. The Prichard-Burke contact is drawn where the blue- gray argillite ceases to exist. Under the microscope Calkins found a majority of original angular to subangular grains of quartz with some feldspar and muscovite with interstitial fine grained sericite and quartz. Regarding the minor constituents, waterworn zircon is common, some isolated rutile and cloudy whitish particles are present. Certain other minor minerals, occurring in small clusters, are possibly of metasomatic origin. These include idiomorphic crystals', abundant "1 tourmaline, some siderite, and magnetite, biotite, chlorite and calcite.

Kootenai County Anderson (1940 page 10) considers that the thickness of the Prichard exceeds 5000 feet. He describes the formation as composed largely of bluish shale and argill• aceous sandstone, with subordinate beds of grayish sand• stone, and white quartzite. The shale, in the vicinity of large faults, generally has a prominent slaty cleavage. Near the top of the formation there appears a 500 foot white or gray quartzite member, and beds above and below consist of laminated, thin bedded shale with only minor sandstone. Toward the lower part, the Prichard becomes more quartzite and contains less of the shaly beds. 12

Anderson credits the dark, rusty appearance to the oxidation of fairly abundant pyrite in the shaly beds.

Philipsburg In the Philipsburg quadrangle Calkins (1913 page 36) found at least 1000 feet of the Neihart quartzites which conformably underlie the typical reddish-brown weathering Pritchard-like rocks. The base of the Neihart is not exposed and the lowest beds consist of. very pure white or pale drab quartzites. Farther up the section, the prevailing rock is grayish quartzite, with faint but distinct bands about half or inch thick and very thin micaceous partings from 6 inches to 3 inches apart. Near the top there are numerous interbedded layers of dark mica schist a few inches or feet in thickness. The rusty weathering Prichard has in the Philipsburg quadrangle been subjected to intense metamorphism. Prior to metamorphism the formation was largely shales with occasional beds of sandstone. These rocks are now represented by various schists and gneisses containing mica, garnet, andalusite, sillimanite, and cordierite. The quartzites are mostly near the top and the bottom of the formation which has been estimated to be about 5000 feet thick.

Creston-Ravalli Cranbrook Area Rice (1937 page 8) found the Creston in this area 13 to consist of 6300 feet of argillaceous quartzite and quartzite, both occurring in beds from 1 to 4 feet thick with narrow partings of argillite. The commonest type is massive, watery green, medium-to fine grained argillaceous quartzite. Purple and white varieties are plentiful and grade into one another along the strike. Some of the light- coloured quartzite is traversed by a network of fine purple lines. The formation is not limy but buff-weathering cal• careous beds are not uncommon, especially near the top. Ripple marks and mud cracks are abundant throughout the formation and some rain prints were observed by Rice. Green and purple are the dominant colours of the Creston.

Nelson-East Area Rice (1941 page 9) found the Creston of this area comparable to that in the Cranbrook area both in thickness and lithology. Argillaceous quartzite form^!$l the bulk and

is a fine-grained, light watery green or0purple rock in beds 2 or 3 feet thick. Rice believes the purple colour is due to finely divided hematite. The argillite present is commonly platy and like the quartzite is green-and purple. Towards the top of the formation narrow beds, pods, and"" lenses of calcareous rocks appear. Where they are" very abundant the strata is considered to belong to the overlying conformable Kitchener. Cross bedding and ripple marks are common and well preserved. Rice considers the formation somewhat coarser grained and more massive than the Under• lying Aldrige. Except where the Creston has been locally 14

mineralized it altogether lacks the rusty weathering so typical of the Aldrige.

Boundary County: Kirkham and Ellis (1926 page 16) found that the lithological changes from the Coeur d'Alene district to Boundary County in the Creston formation are not as great as in the Aldrige-Prichard. In Boundary County the lower section of the Ravalli consists of thin members of siliceous shales, gray rusty-weathering argillaceous quart• zites, and pure gray quartzites which appear to correspond to the Burke member. The middle section is chiefly pure white massive quartzite and sericitic quartzite correspond• ing to the Revett. The top section is largely sandstone with some shale having brown, purple, and green colours which seems to resemble the St. Regis of the Coeur d'A-lene section. Kirkham and Ellis found little evidence of shallow water deposition. They found two sills in the lower section and one in the upper. The section of Eastport, exclusive of diorite sills, was found to measure approximately 5000 feet thick.

Clarke Fork District Anderson (1930 page 15) was able to distinguish the Burke which he thinks might include some of the Revett and also he defined the St. Regis. The former assemblage 15 is 3500 feet thick and the latter formation measured more than 7500 f eet *m Iruckness . The Burke is composed of fine grained sandstone in massive beds or as thin 2>eds with argillites. The formation differs from the Prichard in being somewhat more siliceous and the sandstones are prevailingly gray with a greenish cast. The first appearance of the pinkish markings was taken to indicate the Burke-St. Regis Boundary. The St. Regis is readily identified by the reddish or reddish-purple colour of many of its members. The lower part of the formation is composed of quartzitic sandstones, distinguishable from the Burke due to the faint pinkish colour of some of its beds or by thin partings of red shale. The lower beds are softer and thinner, but the rocks near the centre of the formation are more massive and more resistant to erosion. Higher in the series are some beds of massive quartzites with greenish tints passing into beds with delicate purple bands. The upper St. Regis beds are alternating red and green shales and argillites with the red predominating. Both the Burke and St. Regis contain mud cracks and ripple marks throughout and are particularly abundant in top of the St. Regis.

Trout Greek Quadrangle Gibson^Jenks^and Campbell (1941 page 368) observed that the Burke formation near Coeur d'Alene is composed of gray quartzite and sandstone, and of argillite 16 similar to the Prichard. North of the Trout Creek Quadrangle the beds are less arenaceous and more argillaceous. North• ward, the Burke formation increases in thickness until in the Trout Creek area it is 2800 feet thick; however, still farther north it cannot he separated from the overlying Revett. The Revett is dominantly massive, hard, even- grained, white, locally crossbedded quartzite having individual beds commonly 1 to 3 feet thick. Wherever iron is present the beds exhibit reddish or purplish tints or weathered surfaces. In the northern part of the Quadrangle the Interbedded argillaceous sandstone increases in amount. The St. Regis formation of the Ravalli is composed mainly of shale which in places is highly coloured in shades of green and red and elsewhere is dull gray. Farther north and north east the red and green shades become less prpnounced and north of the Clark Fork the shale is gray and becomes arenaceous, calcareous, and sericitic. In the Northern part of the area the top of the Ravalli is taken at the thin beds of bluish-gray sandy argillite which overlie the shales.

Coeur d'Alene District Calkins (1908 page 82) describes the lower Ravalli or Burke formation as composed of rocks that show all gradations from nearly pure quartzite to siliceous shale. The quartzite beds are relatively more abundant in the upper part of the formation; the lower part is more 17 shaly. In the Coeur d'Alene district, the Burke grades down into Prichard and up into the quartzites of the Revett formation or middle Ravalli, thus exemplifying the characteristic nature of the Belt Series. The proportion of quartzite in the upper part of the Burke increases eastward, and to the north and east as had been pointed out, the Burke-Revett junction is indistinguishable lithologically. The typical siliceous mudstones are mostly pale greenish gray, though part of them are of a rather dark tint. Less common is a light-purplish hue, and there is in some beds anc alternation of purple and green layers. Sun cracks and ripple marks occur throughout the Burke formation. Lithologically the Burke resembles the overlying St. Regis and Striped Peak, however, both of the latter are charact• erized by brighter tints of purple and green than that found in the Burke. Calkins states that the Revett formation is lithologically the simplest and most homogeneous of the Coeur d' Alene sediments. It consists of hard, pure, rather thick-bedded quartzite, but there is also a minor proportion of rocks less purely siliceous. In general the Revett formation being about a 1000 feet thick, comprises a central portion of the Ravalli composed largely of pure quartzite, passing into less purely siliceous beds above and below, which are transitional to the overlying and underlying formations. 18

Ripple marks do not occur in the thick, massive quartzites, however, they are abundant in the more argillaceous, thin bedded rocks of the bottom and top portion of the formation. Calkins from a microscopic study reports the presence of subangular, clastic grains of quartz, a little feldspar and sericite; minor constituents include zircon, magnetite, rutile, tourmaline,and siderite. Calkins has placed the thickness of the St. Regis at almost 1000 feet and describes it as consisting mainly of thin alternating layers of argillaceous and arenaceous material. The prevailing colours are rather bright tints of green and purple, the latter colours somewhat predominating. The characteristic feature of the formation is the general presence of shallow-water structures. Some it-iTV-a-fov-ma"Tioir>al co rig\©imev-a conglomerates which are interpreted as extra formationo appear locally within the St. Regis. Calkins attributes them to local erosion of mud flats by wave action.. Microscopic study by Calkins has indicated that the characteristic colours of the St. Regis are due to hematite dust and chlorite for the red and green, respectively. The hematite seems to be in large part the speculaMte variety; the chlorite may be finely divided and evenly distributed, in nests of larger crystals, or -E intersitial in place of sericite. Kootenai County

The Burke, Revett;and St. Regis were not 19 differentiated in the map hy Anderson (1940 page 11) however, he recognized the subdivisions of the Ravalli which are so well defined in the Coeur d'Alene district. The Burke varies from 1000 to 3000 feet thick and may be classed as argillite with thin beds of fine grained massive sandstone and sericitic flaggy quartzite. Having less pyrite than the Prichard, the rock weathers a light clear gray although some of it has a pale greenish cast. Ripple marks and mud cracks are abundant. The Revett quartzite approaches 3000 feet in thickness and is difficult to separate from the Burke. The Revett quartzite is more massive and is composed largely of white and gray sericitic quartzite with only minor amounts of interbedded argillite. -tWitVie Revett The St. Regis is more distinctiveAand in the eastern part is 1000 to 2000 feet thick, but increases markedly westward to become at least twice as thick. Abundant reddish beds throughout may be confused only with the younger Striped Peak formation. The lower part is quartzitic with purple markings associated with scattered thin beds of reddish and purplish argillite, interbedded with greenish shale and argillite. The upper part is less quartzitic and consists mostly of alternating red and green shales and argillites, green argillite becoming abundant at and near the top. Ripple marks and mud cracks occur throughout. -20

Phil^ipsburg The Ravalli is about 2000 feet thick having the lower two-thirds composed of light-gray quartzite less pure than the Neihart quartzite; the upper third comprises dark- bluish and greenish shale interbedded with quartzitic sandstone. At the top there is a gradation into the Newland formation of calcareous rocks.

Kitchener-Wallace Cranbrook Area Rice (1937, page 9) describes the Kitchener as not less than 6000 feet of variously coloured calcareous , and doiomitic argillites, generally buff weathering, thin bedded, soft, and in many pieces severely sheared and contorted. Near the base of the formation the prevailing colour of the argillites is green; either a watery green similar to the Creston quartzites, or a light creamy green. Higher up in the section dark grey and light creamy grey, doiomitic argillites are more usual. In addition, Rice describes a peculiar type of rock, which is diagnostic of the Kitchener; it consists of fragments of fine-grained calcite a few inches long set In a cement of coarser calcite or dolomite with small amounts of quartz and a few flakes of colourless mica. On an exposed surface the calcite fragments weather out to produce a "Molar Tooth" structure. Mud cracks are very common and ripple marks occur in the more sandy members of the formation. 21

Nelson Area - East In mapping this area Rice (1941, page 10) did not differentiate the overlying Siyeh from the Kitchener. He.- reports that the assemblage from the top of the Creston to the base of the Upper Purcell consists almost entirely of rocks typical of the Kitchener. This unit he named the Kitchener-Siyeh and it is approximately 7500 feet thick. The formation consists predominantly of impure magnesian limestone, argillite and calcareous quartzite. The limestone and calcareous rocks, which are cream-coloured also some green, gray and purple varieties, form the bulk of the formation. All are buff weathering, magnesianjand occur in beds 1 to 5 feet thick with partings of argillite. The Kitchener-Siyeh argillite is green, dark gray, and purple, and most of it is hard and platy or slaty. These argillites are often banded and coloured in a way that suggests the Siyeh. They are however generally interbedded with the limestone. In many places mud cracks are well preserved.

Boundary County Kirkham and Ellis (1926, page 17) believe that the Kitchener-Wallace formation thickens northward from Coeur d 'Alene, otherwise there is a great similarity to of beds the type locality. They measured 4500 feet'; however, the upper member may be cut off by the Moyie Lenia Fault. The rocks are fine-grained, thin bedded, and notably calcareous. 22

They are chiefly light gray argillaceous and calcareous quartzites, and calcareous Jbanded argillites. Ripple marks and mud cracks are abundant.

Clarke Fork District Anderson (1930, page 17) states that the Wallace formation is the most heterogeneous of the Belt Series and describes them as follows: "It is composed throughout of fine grained, thin bedded rocks, which comprise green, more or less calcareous shales, blue and white banded argillites in part' calcareous, light gray and yellowish weathering calcareous quartzites, patches of pure lime carbonate, and pure quartzite." The succession from bottom to top is first; dense, flinty green argillite then repeated thin beds of greenish argillite, massive quartzite and impure limestone, near the top greenish shales and thin bedded quartzite grade into impure limestone and argillite, finally reddish sandstone, a bed of impure massive limestone and the Striped Peak formation. More than 6000 feet of Wallace are exposed.

Trout Creek Quadrangle Gibson, Jenks, and Campbell (1941, page 371) measured 10,500 feet of Heterogeneous rocks comprising the Wallace formation. Argillite and shale are dominant; ser• icitic sandstone and quartzite are much less common; dolomite and doiomitic limestone are present and are more abundant in Trout Creek Quadrangle than to the north. Calcium and Magnesium carbonates are not as widespread ; . in Trout Creek as they are in Coeur d'Alene, thus carbonates increase to the south. Attention is again drawn to the presence of the peculiar amolar-tooth" structures described by Rice and similar cellular structures described by Anderson. Theae structures, which are one and the same thing, are diagnostic of the Wallace Formation. Thin beds, which are presumably algal limestone or dolomite, are present in -the Wallace at several horizons. They are almost invariably present in the red shaly strata near the top of the Wallace.

Coeur d'Alene District Calkins (1908, page 39) believes the Wallace to be not less than 4000 feet thick. It is an assemblage of thin-bedded fine-grained rocks, including calcareous quartzites, impure limestones, and shales that are in a great part calcareous, all marked by mud cracks and ripple marks. On a lithologic basis a threefold subdivision may be made as follows: a lower division mainly of green slate, a middle division of bluish and greenish argillites, lime• stone and calcareous quartzite, and an upper one of bluish and greenish argillites, more or less calcareous without white calcareous quartzite. The top of the formation is drawn where limy beds cease to exist.

Kootenai County Anderson (1940, page 12) found not less than 5000 24 feet of Wallace which is characterized by the abundance of calcareous beds. The lower part consists of greenish arg• illites but with scattered, thin beds containing calcium carbonate. Toward the middle of the formation calcareous material increases and beds of calcareous sandstone, quartzite, and impure limestone appear. In the upper part of the formation, the calcareous material decreases somewhat and the prevailing beds are dark gray and bluish gray shale and argillite. Cellular structures are common particularly near the top of the formation.

PhilXlpsburg The bulk of the Newland formation according to Calkins (1913, page 41) consists of thin-bedded calcareous rocks which may be called siliceous impure limestones or calcareous argillites. Calkins believes that the Newland formation is more calcareous on the whole than in the Coeur d'Alene district and further, it appears that the proportion of carbonate continues to increase toward the east. Shallow-water markings found throughout the formation in the Coeur d'Alene district were only noted in the upper part of the Newland. The upper drab to greenish-gray calcareous shale may be the equivalent of the Greyson Shale. 25

Siyeh-Striped Peak Cranbrook Area Schofield (1915) limited the Siyeh formation to an argillaceous and calcareous formation above the Kitchener Rice (1937, page 10) has further restricted the Siyeh to the upper part of Schofield's Siyeh which, west of the Kootenay river, consists almost entirely of argillite and argillaceous quartzite. Doiomitic beds frequently occur in the Rocky Mountains so that the boundary location is a matter of opinion. The thickness of the formation is probably between 1000 and 2000 feet; however, a section in the Rocky Mountains appeared to be only 700 feet thick. The Siyeh consists of thin-bedded argillite and argillaceous quartzite; purple, red, mauve, olive-green, watery green, and dark grey in colour. Combinations of purple or mauve and green rocks in thin bands are common also "green striped" argillites like those in the Creston and Kitchener, are locally abundant. Mud cracks and ripple marks are very common.

Nelson Area East As has been pointed out in the discussion of the Kitchener-Wallace, in the Nelson Area East, Rice (1941, page 10) did not differentiate between the Kitchener and Siyeh. He did however, recognize the argillites and- argillaceous quartzites in the northeastern part of the Nelson Area East. Elsewhere in the area if argillite 23

suggestive of the Siyeh occurs, it is present in insignif• icant amounts only and the rocks are mainly typical of the Kitchener.

Boundary County No rocks which may be correlated with the Siyeh- Striped Peak formation are exposed in Boundary County.

Clark Fork District Anderson (1930, page 18) points out the similarity between the Striped Peak and the St. Regis of the Creston- Ravalli formation. Both formations have many reddish members however only the latter has a marked tendency toward purplish tints. Toward the west the reddish colour -the of the beds disappears and rock assumes an olive drab appearance. Anderson also mentions that generally speaking the formation is very like that in Coeur d'Alene and Montana. More than 4000 feet of Striped Peak formation is exposed in the Clark Fork district. The base of the formation is composed of several hundred feet of particularly thin bedded, nearly fissile shales, some of which are slaty and have a greenish and bluish colour which conformably overlie a thick limestone member of the Wallace formation. Above is a great thickness of reddish sandstone and shale beds, alternating with greenish shale and argillite beds. The reddish beds disappear higher in"the 27

formation and extend only 3500 feet above the base. Ripple marks and mud cracks are abundant throughout the formation.

Trout Creek Quadrangle Gibson, Jenks and Campbell (1941, page 374) divided the Striped Peak of the area to the north of Trout Creek into two members. The lower member, 2000 feet thick consists of dark red to purplish sandstone and quartzite in the northern part, but southward the beds become more shaly and greenish strata appears. The upper member of the Striped Peak has no equivalent in the Coeur d'Alene District. It consists of a light to dark gray and greenish gray argillite usually somewhat sandy and to a less extent sericitic and calcareous. The beds are mainly 1 to 3 feet thick, and some of the thinner beds weather rusty brown, and resemble the Prichard. At several localities oolitic, doiomitic limestone is interbedded with concretionary algal beds. North of Trout Creek a total thickness of 8000 feet of Striped Peak was measured however in the Coeur d'Alene District only 1000 feet of the formation which may be only pa'rt of the lower member is present.

Coeur d1 Alene District Only 1000 feet of Striped Peak of limited areal extent occurs in the Coeur d'Alene district. Calkins, (1908, page 44) describes it as mainly thin bedded rocks, including shales and quartzitic sandstones marked with 28

ripple marks and mud cracks, and mostly reddish purple and green in colour. It is similar to the St. Regis; however, the Striped Peak is less deeply coloured and the prevailing colour of the quartzitic beds inclines rather to pink than to a more bluish purple.

Kootenai County Anderson (1940, page 13) found more than 4000 feet of a very heterogeneous formation which is correlated with the Striped Peak of the Coeur d'Alene District. Along the eastern margin of the county it is very similar to the St. Regis, being largely made up of purplish and reddish coloured quartzites interspersed with reddish and greenish beds of argillite. Westward and to the north the reddish and purplish markings disappear and the formation is composed largely of sandstone or quartzite, having more or less an olive green colour.

Missoula The basal formation of the Missoula Group as mapped by Clapp and Deiss (1930) is called Miller Peak and since it overlies the Wallace would, on the basis of normal succession, be expected to correlate with the Striped Peak Formation. Near Missoula the Miller Peak appears to lie conformably on the equivalent of the Wallace formation; however, 12 miles to the west on the Lolo Fork area the Miller Peak lies directly upon Ravalli Quartzite. 29

The Miller Peak formation has a thickness approaching 2900 feet. The lower 1100 feet is composed of deep redish purple sandy argillite with intercallated thin beds of fissile, gray, sandy, mud cracked argillite. Overlying this member is 1500 feet of mixed purple and green-gray, sandy, mud cracked, and ripple marked argillite, interbedded with some massive beds of argillitic sandstone and an occasional thin bed of fine-grained purple gray argillite. The upper 3000 feet of the Miller Peak is dominantly a massive to thin-bedded argillitic sandstone, which becomes increasingly sandy toward the top.

Philipsburg Quadrangle The formation here overlying the eastern equiv• alent of the Wallace is known as the Spokane formation. The thickness probably is about 5000 feet consisting of shale and sandstone, prevailingly red where unaltered and characterized as a whole by mud cracks and ripple marks. Calkins (1913 page 45) describes the Spokane formation as having a base of red shale overlying the buff-weathering rocks of the Newland. This transitional basal phase is succeeded by 2000 or 3000 feet of shale with subordinate sandstone. Locally the lower part is chiefly arenaceous. The upper part everywhere consists chiefly of sandstone with subordinate shale. In places a thin conglomeratic band appears on a few hundred feet from the top of the formation. GENERAL SUMMARY AND THEORETICAL CONSIDERATIONS OP THE LOWER PURCELL-BELTIAN SEDIMENTS

In Proterozoic times land vegetation was absent. Under such, conditions we can suppose that the mountainous terrains of Precambrian times were subject to a somewhat greater overall rate of erosion than our present day areas of high relief. This may in part account for the stupendous thicknesses of clastic sediments of Precambrian age.

Aldridge-Prichard Red Colour of Weathered Surface The finer-grained argillaceous members of the Aldridge-Prichard weather decidedly rusty-red. Purer quartzites are^often)gray and even white weathering; however, since the upper portion is particularly dark grey, argillaceous it is the prominently rusty part. This rusty weathering nature of the Aldridge-Prichard is mentioned in all the areal geological reports and therefore may be taken as a universal characteristic. In fact it is often the main distinction from the overlying Creston-Ravalli. Regarding the Prichard formation of Kootenai County Anderson (1940, page 10) states: "Weathered exposures of the formation all have a dark, rusty appearance, largely because of the oxidation of fairly abundant pyrite in the shaly beds. This feature, together with the bluish-gray appearance of its thinly laminated, unweathered shales, helps to distinguish the Pritchard from other formations." 31

Particularly in the upper' part of the Aldridge the rusty weathered character is a striking stratigraphic feature. For example, in the Moyie district, distinctive rusty horizons toward the top can be followed completely around the long plunging anticlinal structure for some tens of miles. Since the rusty nature is restricted to certain stratigraphic sections and is uniform over such great distances then it is obviously not connected with local mineralization. It follows that the rust-producing ingredients were syngenetic. Muddy beds are very apt to have deposited with them iron compounds of one form or another. Twenhofel (1939, page 303) in speaking of black muds states: "Black muds of marine origin acquire their colour from black sulphides of iron and organic matter. Blackness due to black sulphides of iron may be expected to be lost as those sulphides change to marcasite or pyrite". Many iron compounds are thought to form in marine environments by purely inorganic processes. Sulphur of any decaying organic matter, even of very early organisms completely devoid of hard skeletons may have accumulated with the muds of the Aldridge- Prichard. If hydrogen sulphide accumulates in oxygen - deficient bottom waters then oxidation by bacterial action may lead to the release of sulphur. Trask and Wu (1930) have shown that most marine muds contain free sulphur in quantities ranging from 22 to 104 parts per 100,000. 32

We might also entertain an inorganic origin of sulphur bearing compounds in the nature of sulphates or volcanic gases. The writer has observed small blebs of pyrrhotite visible with the unaided eye in parts of the Aldridge of the Moyie district. It is suggested that the rusty weathering nature of the Aldridge in this district is due to the surficial oxidation of minute particles of pyrrhotite. In Kootenai county it may be due to the oxidation of pyrite. The writer is of the opinion that iron and sulphur bearing minerals were deposited originally with the argillaceous sediments of the Aldridge-Prichard formation. With the continual accumulation of overlying sediments, dowsinking proceeded and pressures and temperatures increased. At a certain depth, which was exceeded, the iron-bearing compounds and sulphur-bearing compounds reacted with one another and stabalized in the form of pyrrhotite. Possibly pyrite developed in the Prichard of Kootenai County. The Aldridge-Prichard formation ranges from predominantly quartzite and argillaceous quartzite to mostly shale and slate. The latter condition exists markedly in the Coeur d'Alene section and the Aldridge of the Rocky Mountains. Very ea.rly local limy conditions in Aldridge time may be indicated by the Port Steele 33

formation. Towards the west the Aldrige-Prichard grades into coarser-grained elastics with the exception of the shaly top of the formation. On the correlation chart one notices a marked thinning of the Prichard from west to east and possibly from.north to south. In the Belt Mountains apparently no Prichard was deposited. Possibly it is represented by the 700 feet of Neihart quartzite which rests unconformably on Archaen rocks. Mud cracks and ripple marks seem to be more prevalent in the argillaceous top of the formation and become less frequent in the thicker-bedded,argillaceous quartzites. In part this may be due to textural conditions of the sediment at the time of deposition and the rate of sedimentation. More likely deeper water conditions prevailed during early Aldridge-Prichard times. This is also suggested by the limy Fort Steele immediately underlying the Aldridge in the Fort Steele district.

Creston-Ravalli No important generalizations can be made concerning the Creston-Ravalli except that the predominant colours are purple and green and the rock is largely quartzitic. The colours often alternate with the beds; green may be interpreted as iron in the ferrous state and purple as finely divided hematite which is in the ferric state. Thus, changes in the degree of oxidation must be 34 postulated with the sedimentary cycles. The formation grades upward from the shaly top of the Aldridge-Prichard through argillaceous flaggy siltstones and quartzites to pure massive quartzites and thence to argillaceous rocks and the red and green shales of the St. Regis. Mud cracks and ripple marks occur throughout the Creston-Ravalli. The formation shows a tendency to thin towards the east.

Kitchener Wallace This formation consists of calcareous and dolimitic argillaceous rocks and often possesses the f "molar tooth" structure. The colour is usually some shade of green. Towards the east the limestone is more abundant and more pure. In the Belt Mountains the Newland and Helena limestones, equivalent to the Kitchener-Wallace are separated by the Greyson shale, Spokane argillite and Empire shale. These latter formations thin westward until they can no longer be recognised as mappable formations * in the area of the Coeur d'Alene mountains. Again in the Kitchener-Wallace mudcracks and ripple marks are present.

Siyeh-Striped Peak This formation consists of purple and red argillites, in some map areas rather arenaceous and everywhere containing abundant mud cracks and ripple marks. In the eastern section of the Belt Series greenish gray argillites and sandy argillites overljethe purple and red beds. 35

THE SOURCE AND ENVIRONMENT OF THE LOWER PURCELL-BELT SEDIMENTS

This problem has been one of speculation since the Beltian sediments were first studied. Walcott (1910) considers that the Purcell basin of sedimentation was enclosed and its water was fresh or brackish. His basis of argument is the apparent lack of life in the Purcell and equivalent strata and Its sudden appearance in the Cambrian; however, this fact can be readily explained by the tremendous time gap between the Purcell deposition and the Cambrian, during which the Windermere Series was deposited. Rice noted that nearly all the ripple marks are the symm• etrical or wave ripple type to which he adds that Kindle has pointed out that in tidal waters current ripple-marks predominate and that symmetrical ripple-marks are evidence of lacustrine deposition. Rice then states, "The Purcell series appears, therefore, to have been laid down in a shallow lake of great extent which was, for a part of the time at least, brackish. The basin containing this lake sank slowly and at a comparatively uniform rate during the accumulation in it of 37000 feet of sediments." However, if one considers the presence of salt crystal casts in the Upper Purcell, and the widespread calcareous Kitchener-Wallace grading down through argill• aceous and quartzitic formations to a total thickness approaching 30,000 feet over such a widespread north- 36

south trending belt; then it appears to the writer more like a normal downsinking, geosynclinal, marine environment. The exact nature may have been more that of a series of deltas extending into the sea and having expansive subaerial plains at times of maximum sedimentation. This would permit the formation of mud cracks. The subaqueous plains would certainly develop symmetrical ripple marks which would be preserved by incoming sediments. The question of the direction and position of a source for the Purcell-Belt sediments is even more obscure. In 1906 Walcott stated "The most easterly section, that of the Belt Mountains, has more limestone in proportion to arenaceous matter, and with the exception of the Neihart quartzite at the base, finer sediments; then conditions indicate that the sediments were derived from a somewhat distant source of supply." Clapp and Deiss take exception to this statement and reply, "The most cogent evidence in opposition to Walcott's concept is the actual Archaen mass upon which the Neihart quartzite lies, and the conglomerate, which Walcott himself figured at the base of the Greyson formation." Again, considering the regional characteristics of the Aldridge-Prichard namely, the tendency to thin towards the east, and its argillaceous character in the Canadian Rockies and at Coeur d'Alene; one is led to believe a source lay to the west. Also, considering the 37

Kitchener-Wallace, the limestone becomes more pure and thicker towards the east. These facts have led the writer to postulate a "Precambrian Cascadia" probably as near as the West Kootenay District. Indeed, some of the complex gneisses west of Kootenay Lake may well be remnants of this Archaen land mass. It is probable that the nature of the Lower Purcell-Belt Series reflects the physiographic cycle of Cascadia during erosion. The Aldridge-Prichard particularily toward the west and north consists of relatively thick bedded quartzitic rocks. This may be taken to indicate a terrain source of at least moderate relief consisting of granitic or older quartz rich sedimentary rocks. The top of the Aldridge-Prichard may mark a first stage of peneplanation and the Creston-Ravalli, stages of rejuvenation. In all probability the Kitchener-Wallace was derived from a low lying area and deposited in a marine environment. By this time life was present in the sea as evidenced by the algal remains of the Wallace and later formations. A palaeogeographic map (Plate li) has been prepared showing in an approximate way the possible position of "Precambrian Cascadia" and the shaded flooded areas receiving sediments. 3T A

Pa leogeog ro pin \ a Map

o-f

Lower Purcell - BeVt Times

Pldte \ 38

PURCELL FLOWS AND SILLS

The lava flows, which lie within the Siyeh formation near Cranbrook, are called the Purcell Flows. They vary in composition between andesite and basalt and are generally massive, green, fine-grained rocks, commonly amygdaloidal, and occur in flows from 1 to 30 feet thick. Booidoo tho flowo Sills occur throughout the Lower Purcell Group. They are thickest • and most extensive in the Port Steele and Aldridge formations. One of the sills has been reported to be up to 2000 feet thick. As can be seen from observing the map included in the appendix, these sills have amazing continuity; individual sills being traceable for tens of miles. Dykes are seldom over 50 feet thick and are most prevalent in the Kitchener formation. Rice (1937, page 18) gives four reasons for believing that the Purcell flows are related to the Purcell sills. He thinks that they have been intruded during, or shortly before the Siyeh epoch. The reasons are as follows: 1. The lavas and fine-grained facies of the intrusives are similar. 2. Purcell sills are the only known igneous rock in the district which could be related to flows. 3. No sills or dykes belonging to Purcell intrusives have been found above the Siyeh. 4. The sills.are very large and coarse-grained in the Port Steele and Aldridge formations and become increasingly 39 smaller and finer grained in the younger formations. This might indicate that the sills were injected while the Fort Steele and Aldridge formations were mostly deeply buried. Work in the Nelson area by Rice (1941, page 27) has caused him to question the conclusion that the Purcell intrusives and Purcell lavas are related. He states the following facts as having a bearing on this matter. 1. Sills and dykes indistinguishable in the field or laboratory occur all through the Lower and Upper Purcell. Sills and dykes closely resembling those in the Purcell also occur through the Windermere. 2. A careful search has failed to reveal any pebbles of the Purcell intrusive in either the Toby or Cranbrook conglomerates. On the other hand, no Purcell intrusives have been seen to intrude either of these formations. 3. The sills appear to have suffered all the deformation to which the sediments have been subjected.

UPPER PURCELL - MISSOULA GROUP Windermere Area Walker (1926, page 7) named the oldest formation exposed in the Winderaere Area, the Dutch Creek and he believes it to be the northern extension of the Roosville, Phillips, and the upper part of the Gateway formation. Sections in excess of 3400 feet thick of Dutch Creek Formation were measured by Walker. The Dutch Creek formation is made up of a succession of strata varying in 40 nature from slate and quartzite to magnesian limestone. The slates form the greater part of the exposed formation and are grey to almost black or green. The quartzites are thin bedded and fine grained, with a faint greenish colour. The limestones are crystalline, magnesian, thin bedded, grey in colour, and weather cream to buff. The limestones and quartzites grade from-west to east into slates and argillaceous quartzites. Resting conformably on the Dutch Creek formation, the Mount Nelson formation has an observed thickness of about 3,400 feet. The Mount Nelson comprises a succession of crystalline magnesian limestones and slates and has at its base, and also near the upper erosional surface, massive white quartzites. The lower quartzite or basal member is a massive, white granular rock having beds averaging 1 foot thick. The magnesian limestones are grey, blue, white, purple, and brick red on fresh fracture, and weather to grey, cream, buff,and purple. They are fine grained and crystalline in beds ranging from*2 inches to 2 feet. The slates are grey to black, green and purple. Salt casts and mud cracks were observed within the Mount Nelson formation. The upper quartzite appears near the top of the formation and is more massive than the lower quartzite member.

Cranbrook Area Dal^y measured 2025 feet of Gateway formation near the International boundary. This formation has been 41 correlated with the lower part of the Dutch Greek. In the Cranbrook Area the Gateway consists of red, purple, olive green, pink, greenish grey, grey, creamy green, and white argillaceous quartizte, doiomitic quartzite, and dolomite. The dolomite is universally buff weathering and is more common toward the base of the formation. Ripple marks^crossbedding and salt casts are present. Many * of the doiomitic beds show concretionary structures, pisolites, and oolites. The boundary between the Siyeh and Gateway has been drawn at the uppermost purcell lava flow. This is satisfactory in the Cranbrook Area where flows are always present; however, in other areas it is doubtful whether a satisfactory corresponding boundary can be drawn. Schofield (1915, page 36j describes the Gateway as resting conformably on the Purcell lava. The base of the formation consists of fine grained grit containing pebbles of the Purcell lava as well as a few pebbles of quartzite. This is succeeded by alternating beds of conglomerates and siliceous limestone. The conglomerates observed by Schofield never exceeded 15 feet In thickness. The limestone weathers buff and is concretionary; dolomites are also present. Interbanded with the dolomites are purple shales and grey sandstones. These are succeeded by greyish brown weathering sandy argillites in beds 1 to 2 inches thick, containing abundant casts of salt crystals. Numerous thick bedded buff weathering sandstone and quartzites are 42 interbedded with the sandy argillites.

Missoula A great thickness of sediments is exposed in the Mission Range and extend south to at least Missoula. They lie above the Wallece-Siyeh-Helena limestone and below the * Cambrian Flat Haead quartzites. These rocks have been named the Missoula Group by Clapp and Deiss (1930, page 677) who have defined five formations within the Missoula Group. From bottom to top they are as follows: Miller Peak, Hellgate, McNamara, Garnet, and Sheep Creek formations. The presence of red beds in the Miller Peak and its superposition over the Wallace formation was considered sufficient to justify discussing it in connection with the Siyeh-Striped Peak formation. The Hellgate formation is 2200 feet thick and has a basal member 100 feet thick consisting of massive red- gray coarse grained quartzite with sandy quartzite beds up to 3 feet in thickness. Overlying the basal quartzite are massive beds of fine grained siliceous gray quartzite, aggregating 300 feet with abundant ripple marks. Above the gray siliceous quartzite is 1200 feet of massive pink-gray quartzitic sandstone. The upper 600 feet of the Hellgate consists of gray-red to dull gray, massive, fine to coarse• grained, finely banded and ripple marked quartzite. For the greater part the rocks of the Hellgate formation weather a dull red buff or drab-lavender colour. 43

The McNamara formation is 3000 feet thick and consists of three members. The lower member is 400 feet thick and is composed of green-gray to purple and maroon- ; coloured micaceous and sandy argillites. The middle member consists of gray-green, coarse-grained, ripple-marked, cross-bedded, sandy quartzite. It is overlain by 800 feet of cross-bedded pinkish quartzite. The upper member is a series of bright green and red fine-grained, mud-cracked, and ripple-marked argillite. The Garnet Range formation is the thickest of the Missoula Group, and consists of 7600 feet of largely quartzites and sandy argillites. The lower part of the formation is predominantly of brown and greenish-gray micqceous sandstone and quartzite with members of sandy argillite. A 300 foot stratum of massive, coarse-grained pink, cross-bedded quartzite divides the formation into two parts. The upper 3050 feet is composed of brownish micaceous, thin-bedded and occasionally argillitic quartzite. The Sheep Mountain formation, which is the uppermost subdivision of the Missoula group, is 2300 feet thick. It consists of massive beds of red to pink-white, coarse-grained cross-bedded, purple banded, pure quartzite. The upper and lower portions of the formation are characterized by the presence of clay galls. 44

GENERAL SUMMARY AND THEORETICAL CONSIDERATIONS OF THE UPPER PURCELL-MISSOULA GROUP

The Purcell flows do not occur in the Western Purcell Series nor in the Belt Series. Thus, the Upper Purcell Group is roughly correlated with those formations overlying the Siyeh-Striped-Peak or its equivalent in other map-areas.

The Dutch Creek and Mount Nelson formations of Canada consist of about 6000 feet of slates, quartzites, and magnesiari:,limestones. Salt casts and mud cracks are present. To the south in the Missoula district of Montana, 18000 feet of quartzites and sandy argillites were deposited. Clapp and Deiss state: "The fact that the limestone forming the base of the Miller Peak is traceable into the Lolo Fork area 12 miles to the west, where it lies directly on Ravalli quartzite, makes its stratigraphic position certain, and at the same time places the Missoula group in the uppermost part of the Belt terrain." This statement indicates a possible unconformity between the Missoula Group and the Lower Belt Series. The appearance of clastic.sediments after the great thickness of Kitchener-Wallace limy rocks, and the presence of conglomerate in the Gateway formation above the Purcell flows seem indicative of the nearby emergence of a land mass. Further, the apparent unconformity on the west side of the Missoula Group is in keeping with orogeny. 45

Also, the intrusion of sills and the extrusion of lava flows suggests a period of diastrophism. Rice (1941,page 27) states: "The sills appear to have suffered all the deformation to which the sediments have been subjected." This fact would seem to support the early age of the sills and their intrusion prior to folding. No lavas occur * within the Upper Purcell Group so that it appears that this stage of crustal disturbance was completed before the deposition of the Upper Purcell Group. It now remains to attempt, from the meagre evidence available, to ascertain where the land mass was. It will be noted that the Upper Purcell strata are absent beneath the Cambrian rocks immediately west of Cranbrook. It is possible that the Upper Purcell was deposited in this area and later eroded; however one would expect some remnant from infolding or faulting. In general there is an absence of Upper Belt Strata from Kootenai County east to Missoula. These facts lead the writer to postulate a land mass between Coeur d'Alene and Missoula, extending north in a triangle as far as Cranbrook. Considering the thickness of sediments and their areal extent some doubt may arise whether the size of the land mass is sufficient to provide them. As an alternative land mass or as a supplementary contributor, we may also consider an eastern source now evidenced by the exposures of Archaen rocks in the Belt Mountains. 46

Limestone does not occur in the Missoula group whereas it is abundant in the Upper Purcell group. Argillaceous sediments are prevalent in the Upper Purcell, whereas the Missoula group is mainly quartzite. These facts imply that the Missoula group was deposited closer to the source of sediments then the Upper Purcell group. Thus, there is some justification for postulating a source of sediments in the present position of the Belt Mountains. At the close of Lower Purcell times portions of the Purcell-Belt strata were uplifted while other parts were still receiving sediments. By the end of Upper Purcell times complete emergence took place accompanied by the folding and the building of mountains. These mountains probably extended hundreds of miles north and south of the area under consideration. This period of orogeny may be referred to as the "Purcell Uplift". Pal

Upper Puree 1 I — BeH~ Series

Plate 2 49

PROBLEMS OF CORRELATION OF THE WINDERMERE SERIES

The lower formations of the Windermere Series are unfossiliferous and'so present the same problem as the Purcell Series. Fossils are present in the upper formations. In some districts metamorphism and crustal disturbance has destroyed the fossils or rendered them difficult to find. Thus, until recently, the upper part of the Windermere Series was considered to be of Pre• cambrian age because it was thought to be completely unfossiliferous. Differences in degree of metamorphism not only affect the apparent fossil content but also the lithology of the rock. Formations which are of the same age and deposited under the same conditions may appear unlike due to a different intensity of superimposed metamorphism. The tremendous thickness of coarse clastic sediments at the base of the Windermere Series varies markedly in thickness over short distances. This fact leads to difficulty when attempting a regional correlation. There is considerable evidence that part of the coarse clastic portion of the Windermere Series was deposited by a transgressing sea. If such be the case.then a formation is not everywhere of the. same age. This serves to complicate the problem of correlation and interpretation. The outcrop area of the Windermere Series has suffered greater crustal disturbance than the outcrop area of the Purcell Series. This makes it very difficult to 50 trace formational boundaries and obtain accurate data on the thickness of each stratigraphic unit. Consequently some of the information used for the basis of correlation is inaccurate. Finally, since the appearance of fossils marks the beginning of the Palaeozoic era; the problem of the Palaeozoic-Precambrian boundary arises. Transgressive relations complicate the situation. It appears that a formation may be of Precambrian age in one district and of Cambrian age in another. 51

WINDERMERE SERIES

TOBY-SHEDROOF CONGLOMERATE

Windermere Area The Toby conglomerate is the basal member of the Windermere Series and was named by Walker (1926 page 13) for the excellent exposures along Toby Creek in the Windermere Area. Since then the Toby formation has been traced south to link with the Shedroof Conglomerate of the Metaline Quadrangle. Within the Windermere area the Toby conglomerate varies in thickness from 50 to 2000 feet. The percentage of boulders to matrix is very variable, ranging from scattered fragments forming only 5 to 10 percent of the rock to a compact boulder mass. In places the matrix is largely slate through which are scattered fragments of slate and shale and occasional boulders of limestone and quartzite. In other places quartzite and limestone boulders are equally abundant and lie in a slaty matrix. Within the formation are a few lenticular beds of slate and quartzite. Walker states that the boulders can be identified with the underlying rocks of the Purcell series which it overlies with an angular discordance. He observed that many of the boulders are rounded, but many also are sub- angular and angular, indicating rapid erosion and limited transportation. 52

Nelson Area East Walker observed, a definite angular discordance in the Windermere Area however such an unconformity is not as obvious in the Nelson Area. Nevertheless, Rice (1941, page 14) observed that the Toby lies on the Mount Nelson in certain places and on the Dutch Creek in others and therefore a considerable unconformity must be present. As in the Windermere area the Toby formation varies markedly from place to place. Near Rose Pass the base of the Toby consists of greenish grey conglomerate with quartz pebbles in a siliecous cement. This conglomerate is interbedded with greenish, foliated argillite, and scattered through it are large, angular blocks of black argillite. Above this member is 30 feet of greenish schist followed by a thick bed of quartz pebbles and cobbles of magnesian limestone set in a cement of sandy mangnesian limestone. Further south the Toby Conglomerate is a fine grained arkosic conglomerate and is not unlike many members of the overlying Horsethief Creek Series. At Columbia point it has been metamorphosed and consists of 3 inch cobbles of quartz and quartzite in a matrix of green hornblende gneiss. Metaline Quadrangle In the Metaline area Park and Cannon (1943,page 7) found over 5000 feet of a coarse, poorly sorted, dingy gray brown conglomerate resting on the Priest River group. Fragments of white to reddish-brown quartzite are almost 53

equally abundant with doldmite fragments of similar colour, many of which weather buff. A few pieces of black slate or phyllite and a single pebble of granite rock were seen in the conglomerate. The fragments, which generally constitute 50 to 85 percent of the rock are embedded in a matrix of gray sandy phyllite, which weathers to a brown pitted surface.

IRENE-LBOLA VOLCANICS Nelson Area East The Irene volcanic formation occurs in the southwest corner of the map-area where it conformably overlies the Toby formation. The volcanics are now fine• grained, sheared greenstone or hornblende schist. Near the base of the formation the greenstone is interbedded with conglomerate. Rice (1941, page 16) agrees with Daly who originally interpreted the greenstones and volcanics. Salmo Area Walker (1934) failed to observe evidence as to the extrusive origin of the Irene Volcanics. A thin section showed a groundmass of pale greenish amphibole, andesine-labradorite, chlorite, and secondary quartz. Metaline Park and Cannon (1943, page 10) describe the Leold Volcanics as typical homogeneous greenstone, such as occurs commonly in series of altered basalt and andesite. Original porphyritic, diabasic, and amygdaloidal textures 54

can be recognized and even sheared pillow structures appear to be present.

HORSBTHIEF CREEK MONK The Horsethief formation was named by Walker (1926, page 14) after Horsethief creek in the vicinity of which the strata and their relations to the overlying and underlying beds are well displayed. However, Walker (1934, page 6) when mapping the Salmo area called similar rock the Horsethief Creek Series. Windermere Walker (1926, page 14) describes the Horsethief Creek formation as "largely grey, green, and purplish slate with several lenticular beds of coarse quartzite and pebble conglomerate and numerous thin interbeds of blue-grey, crystalline, and mostly non-magnesian limestone, which occur at different horizons but form a relatively small part of the whole formation". The formation conformably overlies the Toby Conglomerate and is almost 4000 feet thick. Nelson Area East Northeast of Kootenay Lake the Horsethief 'Creek conforms closely to Walkers description of the Horsethief of Windermere map-area. Rice (1941, page 18) gives detailed section measured west from Rose Pass. The argillite is mostly dark grey to black. Beds and lenses of blue-grey, 55 crystalline, essentially non-magnesian limestones are conspicuous. Beds of conglomerate, very much like parts of the Toby, occur all through the formation. Salmo • Walker (1934, page 7) described the Horsethief Creek Series as "a heterogeneous assemblage of sheared and schistose, argillaceous rocks with some beds of limestone and grit, with at the base a well defined bed of boulder conglomerate, and bounded at the top by the massive grits of the overlying Three Sisters formation." Metaline Park and Cannon (1943, page 11) estimate the thickness of the Monk Formation to be 3800 feet. The base of the Monk formation has been drawn arbitrarily at the base of a conglomerate layer or, when the conglomerate is absent; at the top of the Leola volcanics. This is the same conglomerate horizon which occurs at the base of the Horsethief Creek formation in the Salmo map-area. The top contact is placed so that dominantly quartzitic sediments are put in the Gypsy quartzite, whereas sediments that contain mostly beds of phyllites are referred to the Monk formation. The fine-grained phyllites that predominate in the Monk formation contain numerous intercalations of carbonate rocks, quartzite, and grit. In the extreme northern part of the Metaline map-area the lower half of the Monk formation contains numerous beds of carbonate rocls, 56

very few beds of quartzite, and none of grit. In the upper part quartzites and grits increase in abundance toward the top but relatively few beds of carbonate rocks are present.

HAMILL GROUP Nelson Area East Rice (1941, page 20) measured 10,600:. feet of mainly siliceous quartzite overlain by a succession of phyllites, schists, and quartzites all of which he assigned to the Hamhjill Series. These rocks conformably overly the Horsethief Creek formation. The boundary between the two formations is placed between the grey, rather gritty, arkosic quartzite and pebble conglomerate, which is characteristic of the Horsethief, and the very siliceous, relatively fine-grained, light-coloured quartzite. Quartzite is the most striking and typical component of the Hamill formation. The first 5000 feet are largely composed of this rock. Most of the quartzite is siliceous, fine-grained, white to rose red and grey and green in colour. In the basal members it is usually in thick, massive beds with narrow partings of argillite. In the upper members of the Ham^ill group schist and phyllite are common. These rocks are similar to parts of the overlying Lardeau group and in some map-areas the boundary between these two groups is difficult to define. However, in the Nelson map-area an excellent horizon of 57 of limestone known as the Badshot formation marks the upper "boundary of the Hamill group. Salroo In the Salmo map-area Walker made three subdivisions of the great.succession of grits, quartzites, and phyllites. They are from bottom to top; Three Sisters, Quartzite Range,and Reno formations. Walker (1934, page 7) found the Three Sisters formation to be about 5400 feet thick. The lower 2000 feet is composed of massive, greenish grey grit, or fine conglomerate. These are succeeded by 450 feet of alternating beds of grit and white quartzite overlain by 1700 feet of grey» gritty quartzites. A greenish grey boulder conglomerate, averaging 100 feet in thickness succeeds the gritty quartzites and forms a distintive member. About 1150 feet of gritty quartzites overlie the conglomerate and form the uppermost member of the formation. Rice (1941 page 19) states: "As the Three Sisters beds are traced northeasterly from the Salmo area they lose those features that distinguish them from the Horsethief Creek beds, and in fact seem to merge along the strike into normal Horsethief Creek strata." For this reason the Three Sisters formation is included in the Horsethief formation In the Nelson map area. The Quartzite Range formation is 4400 feet thick as measured by Walker (1934, page 8). The lower 1600 58

feet Is essentially massive, white quartzite in beds up to 4 feet in thickness. Ripple marks are present in these beds. A well-defined horizon of argillaceous to slaty rocks up to 200 feet in thickness succeeds the massive quartzites. About ]100 feet of white, crumbly quartzites and white-massJs/ e hard quartzite overlie the argillaceous member and they are in-turn overlain by about 1500 feet of white, faintly tinged green or grey quartzite with a few interbeds of slate. Walker (1934, page 9) measured nearly 3500 feet of beds which he assigned to the Reno formation. A condensed statement of the section is included in his report. The lower part consists of about 1100 feet of argillites, phyllites and quartzites. The lower part grades upwards to' about 500 feet of calcareous schist, grey limestone and argillaceous schist. The Upper part of the Reno formation is essentially, quartzite resembling the Quartzite Range formation. Metaline Park and Cannon (1943, page 13) measured 8500 feet of Gypsy Quartzite beds on a ridge on the east side of the Pend Oreille River. On the west side of the river the thickness is only 5300 feet. This thinning is due to the fact that the 4525 feet of grits and conglomerate present at the base of the formation east of the Pend Oreille River are represented by only 1260 feet west of the river. The quartzites above the basal grits and conglomerates are surprisingly similar in the two sections. 59

The lower member of the Gypsy formation consists of 4525 feet of conglomerate, quartzite, and alternating beds of grits. Park and Cannon recognized the same conglomerate horizon as Walker (1934, page 8) reported in his description of the Three Sisters formation. The conglomerate thins southward and probably averages less than 100 feet thick in the Metaline Quadrangle. A band of dark green schist 20 to 195 feet thick overlies the lower membe.r of grits and conglomerate. Walker (1934, page 8) reported a well-defined horizon of argillaceous rock up to 200 feet thick when he was describing the Quartzite Range formation. They are probably the same horizon. The schist is overlain by 800 to 1625 feet of massive white to pinkish cliff-forming quartzite. Cross- bedding occurs in this rock. The massive quartzite is succeeded by 2200 to 3000 feet of thin platy quartzite, in places with shaly layers and a few intercalated limy beds. Near the top phyllite layers are more numerous and are similar to those in the overlying Maitlen phyllite. The upper contact of the Gypsy formation is' placed at the top of a band, 50 to 300 feet thick, of alternating beds of quartzite and. phyllite in nearly equal parts. These beds are characterized by fucoidal cylinders called "burrows" which are of organic origin. The "burrows" are approximately circular in cross section, 60

about a half an inch in diameter, and generally 2 to 3 inches long. The cylinders are commonly oriented normal to the bedding and gently tapered. The "burrows" occur usually in the more shaly layers. In the Metaline quadrangle these "burrows" are found throughout the transition zone between the Gypsy Quartzite and the Maitlen phyllite.

BADSHOT LIMESTONE Nelson A^ea East It has been pointed out that the Badshot limestone is a very useful horizon marker in the Nelson map-area. This formation divides the Lardeau and Hamill groups and is defined by Walker, Bancroft, and Gunning, (1929) in the Lardeau area as the lowest, stratigraphically, of three prominent beds of limestone occurring at the base of the Lardeau group. The formation is very conspicuous on the east side of Kootenay lake where it maintains a width of 100 to several hundred feet throughout its length. However on the west side of the lake it is not nearly so conspicuous. In some places it may even be missing and other places it is only 50 to 100 feet thick. The Badshot consists of grey to cream-coloured, in places siliceous, or magnesian limestone, which weathers a light buff colour. The numerous limestone beds in the Lardeau group may be confused with the Badshot formation. 61

Salmo Walker (1934, page 9) made no attempt to correlate the limy rocks of the Reno formation with the Badshot limestone. In connection with this Rice (1941, page 20) states: "As no limestone has been seen in the Hamill, the limit of that series is tentatively set at the first bed of limestone in the Reno formation, and this limestone is considered equivalent to the Badshot. The Hamill is thus correlated with the Quartzite Range and that part of the Reno lying below the first bed of limestone."

LARDEAU GROUP Nelson Area East The Lardeau group conformably overlies the Badshot formation. The thickness of the Lardeau has been variously estimated at 10,000 to 15,000 feet thick. In the Nelson area Rice (1941, page 21) believes the thickness is probably near the smaller figure; however, the Lardeau has been subjected to severe metamorphism, shearing and folding so that an accurate measurement is impossible. The Lardeau groups consists of slate, schist, gneiss, quartzite, and several prominent bands of ; magnesian limestone. A dark grey to black slate constitutes a large part of the formation. A muscovite-biotite schist much of which contains garnet, andalusite, and sullimanite, forms £he^ part of the Lardeau group. Where the metamorphism is more intense the rocks become gneissic. 62

Quartzite occurs in beds from 1 to over 100 feet thick. Calcareous quartzite is most common near the base of the formation. Conspicuous belts of creamy white to light grey buff weathering, in places magnesian, limestone occur all through the Lardeau group. They constitute a relatively small proportion of the group and some of the members have even been given formational rank. They are similar to the Badshot limestone which is really the lowest, stratigraph- ically, of these limestone belts. Salmo Area The Pend Oreille group conformably overlies the Reno formation in the Salmo area. Walker (1934, page 9) believed it to occupy the same position in the Windermere succession as the Lardeau group. However, if the correlation of the Badshot limestone with the calcareous member of the Reno is correct then the Upper part of the Reno also belongs with the Lardeau group. The Pend Oreille group is composed chiefly of dark, grey to black phyllites. In the lower part the phyllites grade into bed's of dark grey, almost black, quartzites, and four well-defined horizons of limestone are present. The phyllites are in places highly carbonaceous. Walker (1941, page 10) gives an account of the section which indicates the heterogenous character of the group. 63

Metaline Park and Gannon (1943, page 15) define three formations which overlie the Gypsy quartzite. These formations are equivalent to at least the lower part of the Pend Orielle group of the Salmo area. The formations are from bottom to top: Maitlen phyllite, Metaline limestone,and the Ledbetter slate; The Maitlen phyllite is about 5500 feet thick. The base of the phyllite is placed at the top of a bed of Gypsy quartzite which contains abundant "burrows". This horizon lies about a 100 feet below a gray white limestone band 200 feet thick. At the top of the Maitlen phyllite the beds become limy and grade into the Metaline limestone. The upper contact is placed where the phyllite predominates over limestone. The most common rock type is a gray greenish, fine-grained, and conspicuously banded phyllite. Quartzite beds are present near the base of the formation. Limestone layers are common in addition to the 200 foot bed near the base. The Metaline limestone appears to be about 3000 feet thick and four members have been distinguished (Park and Ca>nnon (1943, page 18). The lower member of interbedded limestones and limy shales is 1200 feet thick. The next member is also 1200 feet thick and consists of fine-grained cream-coloured dolomite. This member is overlain by 450 feet of mottled dense gray limestone with many cherty nodules. 64

The top member is also a mottled ,dense gray limestone but contains only a few chert nodules and measures 150 feet thick. On paieototolGgical evidence the Metaline limestone is middle Cambrian and the Ledbetter slate is Ordovician j therefore, a disconformity must exist between the two formations. The thickness of the Ledbetter slate appears to be about 2500 feet. It is a black, fine-grained, generally homogeneous-appearing rock. In the upper part the slate becomes limy and bedding is more conspicuous. 65

GENERAL SUMMARY AND THEORETICAL CONSIDERATIONS OF THE WINDERMERE SERIES

At the close of Purcell times the strata were deformed into broad open fold mountains. Present day structures consists of folds plunging both north and south with axial planes in general striking north-south. The original mountains consisted of north-south trending cuestas. Dips are somewhat steeper on the west side of the Purcell series as presently exposed; therefore, originally the western escarpments were steeper than those on the eastern side of the former Purcell land mass. They were probably hogback ridges. Toby-Shedroof Conglomerate The obvious derivation of the boulders from the immediately underlying Purcell Series, and the size and shape of the boulders indicate the materials were not transported very far. The surface upon which the conglomerate rests is irregular and all facts point to rapid erosion. Walker (1926) states "The marked variation in thickness of conglomerate from 50 - 2000 feet is strongly suggestive of fan structure and the rock may well be called a fanglomerate,u Walker reports an angular discordance of as much as 45° between the Toby and the underlying Purcell strata. Rice (1941, page 14) observed no such discordance in the Nelson area. He states, "Nevertheless, it is 66 evident from the fact that the Toby lies on the Mount Nelson in certain places and on the Dutch Creek in others that a considerable unconformity must be present". In the Metaline Quadrangle an unconformity has not been unequivocally demonstrated. Wherever the base of the Shedroof conglomerate has been examined the bedding in the Priest River rocks is apparently parallel to the surface on which the conglomerate was deposited. The Priest River group is probably the southern extension of the Upper Purcell group. An unconformity at the base of the Shedroof conglomerate is indicated by pebbles with• in the latter derived from the Priest River group. The writer is inclined to agree with Walkers interpretation of the origin of the Toby conglomerate. It is further suggested that the fanglomerates were located on the steep western flank of the newly formed Purcell Mountains. The conglomerate is somewhat thicker near the international boundary than to the north. It is possible that the mountains adjacent to the thicker conglomerate were steeper and ^ss> higher. As the mountains became worn down the fanglomerate would extend landward or to the east in this case. Thus, the Toby-Shedroof conglomerate may be interpreted as a continental deposit. Irene-Leola Volcanics Greenstone, up to 5000 feet in thickness, is exposed at the International boundary in the south-east corner of the Salmo map-area. This rock is called the Irene formation. It thins to the north and has not been observed more than 20 miles north of the 49th parallel. To the south as the Leola formation, it extends well into the Metaline Quadrangle. Field evidence points to the volcanic origin of most of the Irene-Leola formation. The base is marked by a zone of transition, several hundred feet thick where conglomerate and greenstone are interbedded. Rice, (1941,. page 15) mentions one prominent bed of conglomerate well up in the formation. The Irene-Leola volcanics appear to have been extruded rather locally following and during the rapid erosion which produced the Toby-Shedroof conglomerate. Rice .(1941, page 23) in view of the marked thinning to the north suggests a possible source of volcanic rocks to the south of the International boundary. He also states, "The exceptional thickness of the underlying Toby conglomerate suggests that both it and the volcanic rocks were deposited in a basin in the land surface of that time. Park and Cannon (1943, page 10) mention vestigual pillow structures. If these can be interpreted as indicating a submarine environment then the sea must have extended efest as far as the Irene volcanics at this time. Horsethief Creek Monk Walker (1926, page 17) states: "The Horsethief formation was deposited in standing water in a subsiding 68

basin or valley bordering the area of high relief along the margin of which the Toby conglomerate was deposited." He further suggests that sedimentation took place in the basin contemporaneously with the formation of the conglomerate and that the latter was reworked to some extent. Park and Cannon (1943, page 13) state, "The great diversity of sedimentary materials in the Monk formation indicates extreme fluctuations and a wide range of conditions of sedimentation." They suggest a source of sedimentary debris to the north based upon the fact that the Monk formation is uniform from east to west but from north to south the amount of grit decreases and carbonate beds are proportionately more abundant. The writer believes that the Purcell mountains existed at this time and formed the source.of the Winder• mere sediments. Park and Cannon (1943, page 13) noted that the grit pebbles are quartzites rather than single quartz grains. This would be in accord with the Purcell Series of rocks forming a source of sedimentary material. It is further postulated that marine conditions existed on the western flank of the Purcell mountains. Fluctuations of sea level might account for the diverse nature of sediments. Parts' of the Toby conglomerate was reworked and finer material carried out to sea. Erosion of the Kitchener-Wallace formation of the Purcell Series where then exposed would yield limy material and result in 69

limestone beds in the Horsethief Creek formation. Age of the Hamill and Lardeau Groups The main source of evidence for the age' of the Hamill group comes from its equivalent the Gypsy quartzite in the Metaline Quadrangle. The "Burrows" of the top member of the Gypsy quartzite are interpreted as Scolithus and are thus believed to be of Lower Cambrian age. Also a broken piece of quartzite believed to have been derived from the Gypsy quartzite was found to contain several poorly preserved fragments of trilobites of Cambrian age. The Pend Oreille group in the area west of the Metaline Quadrangle was found to contain pleospongia and is of Lower Cambrian age. The overlying Metaline lime• stone is quite fossiliferous and definitely contains a fauna of middle Cambrian age. Because the Maitlen phyllite conformably underlies the Metaline limestone it is believed to be of late Lower Cambrian age. The conformable underlying Gypsy would then be early Lower Cambrian. The writer thus believes that in the Salmo- Metaline area the Hamill group is essentially of Early Waucobian age and the Lardeau group below the Metaline limestone, is late Waucobian in age. Correlation of the Hamill Group The earlier Lower Cambrian strata of the Dogtooth Mountains described by C.S. Evans (1932, page 119) consists of two massive quartzite formations separated 70

by a formation that is dominantly argillaceous. The formations are from bottom to top: Port Mountain, Lake Louise, and St. Piran. They disconformably overlie the Horsethief Creek formation. On the basis of similar lithology and age the Port Mountain, Lake Louise, and St. Piran formations may be correlated with the Hamill group. In the Glacier Area Okulitch (1948) recognized the Hamill and Lardeau groups with the intervening Badshot limestone. After a careful search for fossils it was concluded that the formations are here unfossiliferous. Since the same formations contain fossils in the Dogtooth area it was concluded that diachronous situation existed. If attention is turned to the correlation chart (see pocket) it will be observed that the formations become younger from Glacier to the Dogtooth Mountains. This fact is best interpreted as indicating deposition in a transgressing sea which moved slowly from west to east. In the Cranbrook area the Eager formation is known to be Upper Lower Cambrian in age on the evidence of trilobites found by Schofield (1922) near St. Eugene Mission. The Cranbrook quartzite is about 800 feet thick and underlies the Eager formation. The following evidence presented by Rice (1937, page 21) points to a lower Cambrian age: 1. It overlies the Purcell series in places with a marked unconformity. 71

2. It is different lithologically from any of the Beltian formations. 3. The annelid-like markings and the "punctate" forms, particularly the resemblance of the latter to those found in the Lower Donald suggest that the Cranbrook formation is Palaeozoic rather than Precambrian in age. Similar structures are found in the Nelson area (Rice 1941, page 29). 4. Schofield has correlated the basal quartzites of the Cranbrook formation with certain quartzites and conglomerates conformably underlying the lower Cambrian Burton formation at Ram creek and Elko. In the Nelson area Rice (1941, page 29) observed the Cranbrook formation to lie upon Siyeh, Kitchener, and Creston rocks. Clearly recognizable Creston cobbles occur in the basal conglomerate. Thus, there is little doubt of an important unconformity between the Purcell Series and the Cranbrook formation. Rice (1937, page 18) describes the Cranbrook formation as follows: "They are massive, coarse-grained, siliceous rocks, white, rose-red, green, and grey in colour. They occur in beds up to 4 feet thick separated by thin partings of argillite. Small beds of conglomerate, with quartz pebbles up to an inch across, occur locally. The basal members in places contain small fragments of the underlying rocks. Ripple-marks and cross-bedding occur locally." Beds of magnesite and conglomerate each up to 72

100 feet thick occur within the Cranbrook formation, The Cranbrook formation and the Hamill group both consist mainly of coarse clastic sediments. Since they are also of the same age the writer believes that the Cranbrook formation is the eastern equivalent of the Hamill group. The lithology of the two formations is not markedly similar; however, this in part may be due to a difference in the degree of metamorphism. Rice (1941, page 30) presents a problem which exists in the Nelson area. On Goat River the Cambrian Cranbrook formation in places rests upon the Creston with no evidence of marked Precambrian folding. Only 13 miles to the west, the lower Purcell group is overlain by the Upper Purcell, which is, in turn overlain by the Windermere Series. It then appears that 35000 feet of sediments are missing in the Goat River section. However, if the Hamill group is the western equivalent of the Cranbrook formation then this figure is greatly reduced and the problem dissolves itself. After the "Purcell Uplift" erosion proceeded rapidly on the western flank of the Purcell mountains. In places the Purcell Series was eroded down to the Aldridge formation. This sedimentary debris went to form the Toby Conglomerate, Horsethief Creek, and part of the Hamill group. Then the early Waucobian sea spread east and deposited the lower Cambrian Cranbrook formation on the eroded Purcell Group. Thus, the base of the Cranbrook quartzite marks a great 73

unconformity which is the eastern extension of the base of the Toby Conglomerate. It has been pointed out that the Hamill group of the Glacier-Dogtooth area was probably deposited in a sea which transgressed the Purcell land mass as it moved slowly east. The same situation apparently existed in the Cranbrook area. Here, the main basin of deposition was located in the present position of Kootenay Lake. As the Purcell mountains were worn down to supply sediments to the main basin the shore line extended east and a shallow sea covered the Cranbrook area. The magnesite in all probability was derived from the erosion of the Kitchener formation under favourable conditions. It was in these somewhat shallow marine to continental conditions with deltas and a fluctuating shore line that the Cranbrook formation was deposited. Thusduring early Waucobian times, the Cranbrook formation is believed to be a near- shore equivalent of the Hamill group. Plate 3 shows the writers conception of the distribution of land and sea during early Waucobian times. It will be noted that the geosynclinal basin trends north- south from Metaline to the Field-Golden area. Location of the Olenellus Zone The Olenellus zone of lower Cambrian is believed to pass through the limy member of the Reno formation. To the south in the Metaline Quadrangle the ©lenellus zone has not been recognized but probably lies within the Gypsy Pa I eo g e o g v-o p In ic Map

o-f

E airly Waucobian T~\ m e S

Plate I 1 1 75

quartzite. To the north of the Salmo area the limy horizon of the Reno becomes the Badshot limestone. In the Dogtooth area Evans (1932, page 122A2) found the Qlenellus zone at the base of the Donald formation. This formation consists mainly of alternating impure limestone, sandstone and slate. They are believed equivalent to the Badshot limestone and the base of the Lardeau group. In the Cranbrook area a formation similar to the Badshot does not occur. The Qlenellus occurs within the Eager formation. On the correlation chart (see pocket) it has been placed at the base of the Eager formation. Precambrian-Palaeozoic Boundary By definition the Qlenellus zone should mark the Precambrian-Palaeozoic boundary. In some areas the Qlenellus zone lies within the Hamill quartzite. "Scolithua" of the Gypsy quartzites is considered to be of Lower Cambrian age. The sea that deposited the Hamill formation has been shown to have moved slowly eastward. Since the Hamill in the Glacier area is Precambrian in age and is of Lower Cambrian age in the Dogtooth mountains then clearly the Hamill group spans the Precambrian-Palaeozoic boundary. On a formational basis the writer proposes that the top of the Horsethief Creek be taken as the Precambrian Palaeozoic boundary. The Horsethief Creek and all its equivalents are considered to be definitely of Proterozoic age. 76

Correlation of the Lardeau Group The three formations of the Metaline Quadrangle are represented in the Salmo area by the Fend Oreille group. The Lardeau is equivalent to the Pend Oreille plus the top of the Reno formation. In the Cranbrook area the Eager formation is lower Cambrian; therefore, rock equivalent in age to the Metaline limestone and Ledbetter slate in the Cranbrook area has been either eroded away or was never deposited. In the Dogtooth area the Donald and Canyon Creek formations are of the same age as the Maitlen phyllite and Metaline limestone. In the Glacier area the Badshot and Lardeau were recognized all of which are apparently unfossiliferous. Conditions of Sedimentation of the Lardeau Group The rocks of the Lardeau group are limy and argillaceous in contrast to the quartzitic character of the Hamill group. This change may in part be ascribed to the decreasing relief of the Purcell mountains by the processes of erosion. It is believed that the shoreline of the Waucobian sea moved slowly from west to east transgressing the Purcell land mass. This may be due to a rise in sea level in addition to the tendency for base leveling and crustal disturbance. Deeper water conditions obtained in the basin -deposition- led to the deposition of limy and argillaceous sediments. 77

The Montana Island No mention has yet been made of the important geographic element of the Cordilleran trough of Cambrian, Ordovician and Silurian periods known as the Montana Island. The moat northern locality in south western United States from which Lower Cambrian fossils have been reported is on the north side of Great Salt Lake. The most southern Waucobian fossil locality on the north side of the Montana Island is 6 miles northeast of Cranbrook. To the west Lower Cambrian fossils have been recognized at Colville, Washington. Deiss (1940, page 788) states: "Similar lower Cambrian trilobites do occur in the Canadian Rockies and in Utah, and parts of the Qlenellus fauna reached as far south as north-eastern Washington, but all other field evidence suggests that two separate invasions occurred in the Cordilleran geosyncline in Waucobian time; one from the south through southern California, and one from the north through Alaska." The writer believes this intervening positive area was the land mass produced by the Purcell Uplift and represents the Purcell Mountains which formed the source of the Windermere sediments. Further it is proposed that this land mass, better known as the Montana Island, had a considerable extension north and east. The regional structures of the Purcell mountains trend north. Along the International boundary the Purcell Series outcrops as far east as Waterton Park. It is probable that the Purcell 78

Series continue^ beneath the foothills of Alberta at least sa far as Calgary. All of this area would have constituted the Montana Island. Prom a study of the sections exposed on Mounts Bosworth and Assiniboine in the Canadian Rockes near Banff, Deiss (1940, page 789) concludes: uThe transgression was halted on the north shore of the Montana Island where sands and gravels were being deposited during late Waucobian times while limestones and shales were being deposited in the north." The late Waucobian shore line can be extended to Cranbrook and thence around the north west side of the Montana Island to Colville, Washington. The writer considers the lower part of the Lardeau to be of late Waucobian age so that it would also be deposited by this sea. A paleogeographic map of late Waucobian time (see Plate 4) shows the distribution of the sea in relation to the Montana Island. The Metaline limestone is of middle Cambrian age and probably indicate^deeper water conditions in the * Metaline Quadrangle during this time. The geesynclinal basin of Waucobian time still existed and in the Dogtooth area the Canyon Creek slates and shales were laid down. They were presumably deposited nearer to shore than the Metaline limestone. Upper Cambrian strata are absent from the Metaline Quadrangle. This indicates emergence or subsequent erosion prior to the deposition of the Ledbetter 79

slate. However, in the Dogtooth area the Ottertail and ^ Goodsir formations of shales and limestones were deposited. During Ordovician times the Metaline area was again depressed and the Ledbetter slate deposited, thus completing the Lardeau group. 80

PaWooe Map

Of

Late Wa u cob va n T\rne

Plate A- 81

BIBLIOGRAPHY Anderson, A.L. (1930) ' Geology and Ore Deposits of the Clark Fork District, Idaho; Idaho Bur. Mines and Geol. Bull 12. Anderson, A.L. (1940) Geology and Metalliferous Deposits of Kootenai County, Idaho; Idaho Bureau of Mines and Geology, Pamphlet No. 53. Calkins, F.C. (1913) Geology and Ore Deposits of the Philipsburg Quadrangle, Montana; U.S. Geol. Surv. Prof. Paper 78. Clapp, C.H., and Deiss CF. (1930) Correlation of Montana Algonkian Formations; Bull, of Geol. Soc. of America, Vol. 42. Daly, R.A. (1912)_ North American Cordillora, Forty- ninth Parallel; Geol. Surv., Canada, Mem.38, A Deiss, C. (1940) Lower and Middle Cambrian Stratigraphy of Southwestern Alberta and Southeastern British Columbia; Bull, of Geol. Soc. of America, Vol. 51. Evans, C.S. (1932) Brisco-Dogtooth Map-area, British Columbia; Geol... Sur. of Canada, Summary Report, Part A 2. Gibson, R., Jenks, W.F., and Campbell I. (1941) Stratigraphy of the Belt Series in Llbby and Trout Creek Quadrangles, Northwestern Montana and Northern Idaho; Bull, of Geol. Soc. of America, Vol. 52 - ? *o-*-p._«,A. Kindle, E.M. (1917) Recent and Fossil Ripple Marks; Geol. Surv., Canada, Mus. Bull. 25. Kirkham, R.D., and Ellis, E.W. (1926) Geology and Ore Deposits of Boundary Co., Idaho, Idaho Bur. Mines and Geol., Bull. 10. Park, C.F., and Cannon, R.S. (1943) Geology and Ore Deposits of the Metaline Quadrangle, Washington; U.S. Geol. Sur. Prof. Paper 202 Ransome, P.L. and Calkins P.C. (1908) Geology and Ore Deposits of the Coeur d' Alene District, Idaho, Prof. Paper 62, U.S.G.S. Rice, H.M.A. (1937) Cranbrook Map-area, British Columbia; Geol. Surv., Canada, Mem. 207 82

Rice, H.M.A. (1941) Nelson Map-area, East Half, British Columbia, Geol. Surv. Canada, Mem.228 Schofield, S.J. (1915) Geology of the Cranbrook Map-area British Columbia; Geol. Surv. Canada Memo. 76. Schofield, S.J. (1922) Relationships of the Precambrian (Beltian) Terrain to the Lower Cambrian Strata of Southeastern British Columbia; Geol. Surv. Canada, Mus. Bull. 35 Geol Sur.42 Trask, P.D. and Wu, CC. (1930) Free sulphur in recent sediments; Bull. Geol. Soc. Am. Vol. 41. Twenhofel, W.H. (1939) Principles of Sedimentation; McGraw-Hill Book Company. Walcott, CD. (1906) Algonkian Formations of Northwestern Montana; Bull. Geol. Soc. Am. Vol. 17. Walcott, CD. (1910) Smith. Misc. Coll; Vol. 57 Walker, J.F. (1926) Geology and Mineral Deposits of Windermere Map-area, British Columbia; Geol. Surv. Canada, Mem. 148 Walker, J.F., Bancroft M.F. and Gunning, H.C (1929) Lardeau Map-area, British Columbia, Geol. Surv. Canada, Mem. 161. Walker, J.F. (1934) Geology and Mineral Deposits of Salmo Map-area, British Columbia; Geol. Surv. Canada, Mem. 172. AMERICAN MAP-AREAS

5 Metal me Quadrangle (\9 37) : Prof. Paper 202, U.S.G.S

Park and Cannon ; Scale 1 mch = \.5 miles.

6 Boundary County(l926V Bull. 10, Idaho Bur. of

Mines, Kirkham and Ellis; Scale 1 mch - 2 miles.

7 Geologic Mapof Montana(l944)'- Oil and Gas Inves•

tigation t U.S. Dept of \nf., Geol. Surv., Scale! wli* 8 miles.

\ Salmo Map-Area (\93l) • Map 299A , Mem. 172, 8 Clark Fork Dl strict (I 9 29) '• Bui 1. 12 , Idaho Bur. of

Geol. Surv. of Canada; J. F. Walker; Scale 1 "= i mile. Mines, A.L.Anderson ; Scale \ mch-2 miles.

2 Nelson Area - East (\938): Map 603 A ,Mem.228, 9 Trout Creek Quadrangle (,^41). Bui I. cf Ge

Amer \ ca.Vol 52 .Gibson, Jenks and Campbell, Scole 1"^ 3 m\ les. Geol. Surv. of Canada , H. M.A. R > ce ; Scale l"-4 miles.

3 Cranbrook Map Area-north (l 935) Map 396 A , O Kootenai County (1940) : Parn. Mo. 53, Idaho Bur. of

Mem.207, Geol. Surv. of Canada ; H.M.A. R \ce ; Scale \%%* 1 mile Mines ; A.L. Anderson ; Scole \ inch - 2. miles.

: 4- Cranbrook Map Area-South (\9I5) • Map 147 A 30N 11 Coeur O^Alene Di Strict (1907) Prof. Poper 62 ,

U.S.G.S , F.C Calkms , Scale mch mile. Mem.76, Geol. Surv. of Canoda , S. J. Scbof \e\d .Scale f=4 miles 1 m'\

H.I WINDERMERE SERIES

Me"Va l\ne Glacier Dogtooth and Wmdevmere Nelson Area E.as"T .a Irno

JTVMalker ParK and Cannon OKul itch C.S.Evans J.F. WolKev H.M. A Rice H.M.A. Rice

194-S I 932 \935 938 931 1937 (V J OHul.tcVi) M eTa \me 3ooo' SOUTH NORTH Creggan Creek 1500', v. Reeves -

Mc Donald x 'Ledbelfer Pend 3Soo' x\ Skfte . Ord ovician 250b' S» I ur i on Beoverfoot Driscuix X x x > D' Oreille 760' F Me1b\ine Wonoh QuavtziTe Top LirnesTo'-ie M»dd\< 1ST- not 3.000 Gle.rioqle SKolc F e»po&ed _ Group j Solmo River Ordov ician 2l5o" slale, schist, gneiss, Good TOOO sha /t. and limetdonn. (juarl^i fe and MoiTleo ambr \an Lav-deou Several bonds- ofL Upper urey rrtagnesjpn, —' PhvlliTe Ottertail 1 • v dork grey mayne>ian f^es-lo-'' I m-,e s If ne Gvoup 5000 Lower Z7ci-2ooo Eoger , Fusty org 1 dar.-- grey sld/es ar*.i //ilk. Combnan Middle tQ^on Creek feooo loooo - ""2.000' Shofy tirriC b^-tncj => 3500 Donald I mpurcj OlenelluS Zone Lowe*- Badshot loo" / tm es lone- /1 n-i e jione arid pmh'ish tftsarTxttfi CranbrooK Gypsy I - 800^^— s. of Lower Cambr\an Q uo rtzite \ Si" Pi rain,' and w-or argtWfe QuarTziTe maSSttrc qua rl JLi Ic. J .2360 x PorceW 53oo-85oo| 1 I \ Ronge f-n.nor- ot-y, I'I'1 'G- Lake Louise Qrgiilite and pr>y H'fe- ^ Qr-Gy grc-er, and Series \ N Harn.H 4400 ,/ /looo S pc-phsh auarTz:/ /e £J/->c/ while, sihc'tovs / 1 Fov-i" Mountain Group rlzife ite / >-l 2,000 Se«c, tit argi Ilaceorj-* ciuorlzi/e IOfc.00 quarT z Lardeau y Cc.*3r-sc ants / "Three Mor>K Horselbief Sisters . 380O Creek 5400 4000 Pv-oterozoic / C.Onq lon-re r**7L / / "Toby Conql. Leolo / So-2.0 OO 'ath*fil}tTe and / x v * *,y l.m e sTo n e green argil/aceous HorseTViief] Vo\conics x Hor-seTh.ef A Tblc S^-hi^T and Badshot Puree 1 I quartz.! te ( blue g*~ i Creek 5000+ Serves \G-eeG W 5000 lim&fTcne. 1 arkase COngJcxnC-raJe / \ pet

mas±ir£ and 'Sr-edroot Prater ozoic Trerie Vol - sheared •aneJc^li''^ Irene / - X VcJcanic- I-0C- Sheared jrr-enflhne Corg\omef| Toby x Conq\. ^ H©m i 5000

Puv-cell Ser\es Pr\esi" P»ver I ver>e Cong lorn era '

PURCELL-BELT SERIES

Trout Creek Coeur 0 Alene Cranbrook Area Nelson Area Easi" boundary Couniy Clark Fork District

Gibson , Je'nks, Campbell F.C. Calkins RM.A.R\ce H.M.A.R\ce K\rham and ElliS A.L.Anderson 1908 1937 194-1 1926 1928 \94\

Mount laminated argil-'te Nelson tnagnesian /'mestone SOUTH NORTH 3£oo

OutcVi lom/n a fed Org,

magn esian /i/r?&s7c"->e

Quarry, fe

purple . red and g r-e '1~tk K\TcViener-| Van C.0I0 ured Co/c«reouS ar^lU'te. and Peak Sandstbr>es and shales Wallace magnes/ar-i limestone 4000 S"Tr v p ed Ca/rarcous argi /life Wallace Variously colodnsd^ b>v£f S"\ yeri 4-500 and arqr/l'te PeoK weo I her my Co/careauS 7500 8000 ^Vooo Kvtckenev Calcareous qua r-t z /- ^ Re*<*tt T 6000 quard~z.i1e Reveii / • Ravall 1 Some cfuardz.ltb BurUe Protero Z.OIG

qree nt purple and jjnsy: 50OO SiUeouS shale 3ur-Ke . - 2ooo / purple. f green ond gr*y CresTor. dry if/a c&o us quar fx • t&, / / Crestor* massive cj r-Txife id green shales 6000 5o/r?a argd!'-&. €>3oo N / Pn chard St. Regis Wollace / ©000' g^een one! purple massive green ,sl> <*nW IOS00 Ot-g'i./life and qUorTx'te. A\dr.dge - 7500 purplish quarT^ile. qUQrt~^.t1& and ^ \ Pv-iclnovd argillaceous tfuarT~2.ilo \ T~r-a n s i~f~i'or-i pink quar-fxiTe. . IO OOO \ platy at-gi IhTes red shaly parTinyS- fimeslone and Cfua r-Txi Te and dotcrn'Te orai I la ceo (us QuarV^zite - / / pinKish rna*-K ngS /-thin bedded bkiish gray vvilh shaly par-Tings qr&y , rusty • weathering( But-Ke , Sandy, argillite^' Qnjy greenish f -fine grained Aldridge N ahyi Ilaccoas ua^rTzi te 35oo greemsh gray shales Sever-a I cJiorifc si/Is I6000i S^indstb r;a 7r majsivu heds s / and at-gillite Or as ttim beds *sdf> a>-y,ll'7& ThicK beddea ouar-Tz'/es \ 7000 yun,te quarTziiTe darh J,lc,sh argil>fifes and shales

grayish sandstone, an*

quarTZ.1b inT*ri*dd*°> rH ' if fa and Sanely WilrT arg-l'- te ty brovan dark y-us Sha/c i/UTe »chard PricViard grey argi 20000 970o

J im e To n e

do/ctm'i'Tic. argilfrfe

Fort Sfeele dark grey argi/life and TOOO grey quarTziTe

rr>a3Sive while quar-Tz-ife

base r\o"f exposed

PWilipsburg Belt Mountains Mission Range Mi SSOula Kootenai County Coeur 0 Alene

ColkvnS Woloott Wilson Walcott Clapp ond De\SS A. L. Anderson F. C. Calkins 1898 \924 \900 1930 \9\3 1940 1908

WES

Apri\ \949

C. G. Cheriton