WILLIAM KENNETH HAMBLIN Dept. Geology, the University of Kansas, Lawrence, Kans

WILLIAM KENNETH HAMBLIN Dept. Geology, The University of Kansas, Lawrence, Kans.

Paleogeographic Evolution of the

Lake Superior Region from Late Keweenawan

to Late Cambrian Time

Abstract: A combined study of regional stratigra- highland persisted as a positive area, and the shape phy, petrology, and paleocurrents was made of the and extent of the basin remained much the same. A Freda sandstone, Jacobsville sandstone, Bayfield lacustrine environment predominated in the central group, Dresbach formation, and Franconia forma- part of the basin, but much of the Jacobsville sand- tion. Data pertaining to the location and nature stone and Bayfield group undoubtedly represents of the source of the sediments were obtained prima- fan deposits which merged northward into sedi- rily from petrology and directional sedimentary ments of an alluvial plain. During Dresbachian time structures. Environmental reconstructions were the Northern Michigan highland remained as a based on patterns of lithologic variations, kinds of positive area, but shallow seas invaded the Lake sedimentary structures, and heavy minerals. Superior region from the northwest and central This information indicates that the Northern Wisconsin from the south. Most of the Dresbach Michigan highland extended through northern sediments accumulated in a beach environment, Wisconsin and northern Michigan and acted as a but in southern Wisconsin an offshore neritic source of sediment from Late Keweenawan through environment predominated. Prior to Franconian Dresbachian time. The Freda formation accumu- time there was a widespread regression of the seas, lated in the Keweenawan basin, which was north and most of the region was subjected to subaerial of the Northern Michigan highland approximately erosion. By Franconian time the Northern Michi- in the present site of Lake Superior but extended gan highland was reduced to a surface of low relief, considerably farther to the southwest. Deposition and the seas re-advanced across the entire area from took place in a flood-plain and lacustrine environ- the southwest. An appreciable amount of the Fran- ment. Prior to the deposition of the Jacobsville- conia sediments accumulated in an offshore environ- Bayfield sediments, the Keweenawan sequence was ment. deformed and eroded, but the Northern Michigan

CONTENTS Introduction 2 4. Diagrammatic section of Upper Cambrian de- Freda formation 2 posits in the Lake Superior region showing Jacobsville sandstone and Bayfield group 6 facies variations and nomenclature of rock Dresbach formation 8 units 9 Franconia formation 11 5. Isopach map and cross-stratification directions Paleogeography 13 of the Dresbach formation 10 Late Keweenawan time 13 6. Variations in sorting, mean grain diameter, and Jacobsville-Bayfield time 14 per cent garnet in the Miner's Castle mem- Dresbachian time 15 ber of the Munising formation 12 Franconian time 15 7. Isopach map and cross-stratification directions References cited 18 of the Franconia formation 13 8. Paleogeographic maps of the Lake Superior Figure region during Late Keweenawan, Jacobs- ville-Bayfield, Dresbachian, and Franconian 1. Geologic map of the Lake Superior region show- time 16, 17 ing area! distribution of formations studied 3 2. Cross-stratification directions in the Freda sand- Table stone 6 1. Heavy minerals in the Freda formation, Jacobs- 3. Cross-stratification directions in the Jacobsville ville-Bayfield sandstones, and Dresbach and sandstone and Bayfield group 8 Franconia formations 4

Geological Society of America Bulletin, v. 72, p. 1-18, 8 figs., January 1961 1

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 \V. K. HAMBLIN-PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

from surface and subsurface data collected by INTRODUCTION the writer and from sections published by Berg One of the most significant changes in the et al. (1956). paleogeography of the Lake Superior region occurred near the beginning of the Paleozoic, FREDA FORMATION when the Keweenawan basin ceased to be the STRATIGRAPHY AND PETROLOGY: The Freda site of accumulation of lava and coarse clastic sandstone constitutes what is generally con- sediments and the entire region became inun- sidered to be the youngest formation of the dated by shallow Paleozoic seas. This change is Keweenawan sequence. It is exposed along the recorded in the sandstones deposited during western coast of the Keweenaw Peninsula and this critical interval of geologic time. They are, can be traced southward into Wisconsin (Fig. in ascending order, the Freda sandstone (Upper 1) almost to the 46th parallel. Sandstones Keweenawan), the Jacobsville sandstone and identical to the Freda in gross lithology, heavy Bayfield group (Cambrian or Precambrian?), minerals, and sedimentary structures are ex- the Dresbach formation (lower St. Croixan), posed intermittently along the east coast of and the Franconia formation (middle St. Lake Superior from Gros Cap to Alona Bay. Croixan). Preliminary work by the writer on These rocks apparently occupy the same strati- the Jacobsville and Munising formations graphic position as the Freda and are therefore (Hamblin, 1958) indicated that although considered to be its equivalent. This correla- many of the stratigraphic problems of the tion is substantiated by paleomagnetic measure- "Lake Superior sandstones" could not easily ments made by P. M. Du Bois (Personal com- be solved, a regional study of stratigraphy, munication), who found that the paleomagnet- petrology, and paleocurrents of these rocks ism of the Freda equivalent of Canada is very would shed much light on the paleogeographic similar to that of the Freda of Michigan and evolution of the region. Thus in 1958 the work Wisconsin but dissimilar to that of the Jacobs- was extended to include the Freda sandstone ville sandstone and Bayfield group. of Michigan and the Franconia sandstone, The Freda formation lies conformably upon Dresbach sandstone, and Bayfield group of older Keweenawan sediments, but the upper Wisconsin and Minnesota. contact is not exposed, so its stratigraphic re- This study was supported by a research grant lationships with the overlying Bayfield group from the University of Kansas general research are still a matter of conjecture. Outcrops of the fund. The writer gratefully acknowledges the Freda are too scattered to permit accurate assistance extended by members of the Michi- measurements of its maximum thickness; how- gan Geological Survey in supplying a boat and ever, semicontinuous exposures along the access to unpublished data and expresses his Montreal River on the Michigan-Wisconsin appreciation to Mr. Harry Sorensen and Mr. boundary and along the coast south of the Por- Charles Bondurant for assistance in the field. cupine Mountains indicate that the Freda Dr. Louis I. Briggs and Dr. Ralph H. King could be more than 14,000 feet thick. In both kindly read the manuscript and suggested vari- areas detailed examination of the section re- ous improvements. veals a gradual change in grain size, type of METHODS: The writer examined essentially bedding, and sedimentary structures from bot- every outcrop of sandstone on the southern tom to top, indicating that there is probably coast of Lake Superior and selected exposures no important repetition due to faulting. inland in Michigan, Wisconsin, and Minnesota The Freda consists principally of alternating during this study. Cross-stratification directions layers of fine arkosic sandstone and red mica- were measured at each locality, and samples ceous silty shale. The texture in the lowest were collected for petrographic and heavy- 1500 feet of the section is generally coarser, and mineral analysis. Except in the Dresbach for- a conglomerate 15 to 150 feet thick is found in mation, all cross-stratification measurements many places several hundred feet above the were made on a plan view of the structure, and base of the formation. Pebbles from the con- at least three measurements were made on two glomerate average 3 inches in diameter, but stratigraphic horizons at each outcrop. Petro- several boulders exceed 1 foot in their longest graphic and heavy-mineral studies were made dimension. Most are composed of basalt and on selected samples in order to supplement quartzite, although an appreciable number published data. Isopach maps were constructed were derived from the Huronian Iron forma- for the Dresbach and Franconia formations tion. The upper part of the Freda is character-

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 FREDA FORMATION

EZJ St. Croixon Series

Jacobsville - Bayneld Sandstones

MINNESOTA IOWA Pre - Freda Rocks

Source of data: Geologic map of the United States, (932; Michigan Geological Survey Publication 51, 1958 , Field work by the writer, 1958 - 1959.

FIGURE \.—GEOLOGIC MAP OF THE LAKE SUPERIOR REGION SHOWING AREAL DISTRIBUTION OF FORMATIONS STUDIED

ized by a monotonous sequence of alternating lens of shale-pebble conglomerate can be traced beds of siltstone and shale 3 to 4 feet thick. laterally to the truncated edge of the shale bed Shale-pebble conglomerates are common at from which it was derived. This is especially many horizons throughout the Freda forma- true of the larger blocks, which were obviously tion but are especially abundant in the lower transported only a few feet. part of the section. The individual pebbles Although quantitatively negligible, the shale range from pellets a fraction of an inch in di- pebbles and blocks are significant in that they ameter to blocks at least 10 feet long. Most of represent numerous intraformational breaks in the shale pebbles are irregular and show little the stratigraphic record. Generally the indi- effect of transportation and abrasion, but the vidual deposits can be traced laterally only smaller pebbles are characteristically well several hundred feet, but their occurrence at rounded and discoidal. All larger blocks are hundreds of places in the section testifies to the tabular, the length and width being as much vast number of pauses in sedimentation, ac- as 15 times the thickness. companied by local erosion. Generally the shale-pebble conglomerate The heavy-mineral assemblage of the Freda forms thin beds or lenses from a fraction of an formation is distinct in that ilmenite, leucoxene, inch to several feet thick, but many isolated and epidote are the only abundant species. In "floating pebbles" are scattered throughout most samples ilmenite (in part altered to the section. The lithology of the pebbles is leucoxene) constitutes over 70 per cent of the identical to that of the siltstone and shale units total heavy minerals. Tyler et al. (1940) report of the Freda formation, and in many areas a up to 11 per cent epidote from the Freda at

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 4 W. K. HAMBLIN-PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

TABLE 1.—HEAVY MINERALS IN THE FREDA FORMATION, JACOBSVILLE-BAYFIELD SANDSTONES, AND DRESBACH AND FRANCONIA FORMATIONS Per cent by number of grains V ™ Location <« OJ ^ - '= .i § U '-1 = G. 'E. .0 1 < uT 1 1 = H £ £ s I'reda formation Gros Cap X X 71 4 2 X X Montreal River . . 20 62 18 Shore cliffs 9 miles north of Little Girl Pt. X 6 84 8 X X Shore cliffs 2 miles south of Gratiot River 10 64 23 X 2 Potato River Falls* 11 87 X X X 2 Montreal River Park* 2 35 49 4 X X 4 Tyler Falls* 1 15 i 78 9 X 2 Jacobsville-Bayfield sandstones He Parisienne 1 X 41 26 21 5 2 3 Au Sable Falls* 28 9 37 10 X 12 Grand Island* . . 2 9 54 23 3 4 5 Marquette* 3 33 62 2 X Sand Point 4 5 5 37 8 20 21 Stockton Island 9 1 3 3 5 36 22 5 Bear Island 1 7 3 X 64 9 11 Cornucopia* 1 X 50 15 X 12 19 Dresbach formation Tahquamenon Falls 3 3 4 32 19 9 8 27 Chapel Falls 25 7 9 18 40 Miner's Castle* 6 12 28 9 12 33 Au Train River 3 4 29 3 4 51 Laughing Whitefish Lake X 18 18 2 50 11 Dresbach** X 38 X 3 25 25 Stoddard** 3 3 35 48 Wood h ill" 3 3 18 3 35 35 Franconia formation Tahquamenon Falls X 42 6 14 11 7 18 Chapel Falls* 48 X 10 39 X Miner's Castle* 55 2 14 18 4 4 Laughing Whitefish Falls* 41 1 4 22 25 2 1 Core near Kiva X 42 3 2 21 II Core 9 miles east of Waucedali . . 96 X X 2 La Cross** 91 1 8 Tunnel City** 82 12 5 Black Earth** 92 6 2 Mazomanie** 92 6 2 * After Tyler et al. (1940) * After Driscoll (1959) ** After Wilgus (1933) ** After Pendand (1931) X Less than I per cent

Potato River Falls, Wisconsin, and an average the total sample and average 7.4 per cent of 35 per cent epidote from samples collected (Tyler et al, 1940). at Montreal River Park. Minor amounts of The heavy-mineral assemblage of the Freda apatite, garnet, rutile, tourmaline, and zircon strongly suggests that the main source rocks are also present (Table 1). Heavy minerals were the Keweenawan flows and the Huronian range from 1.4 to 21.2 per cent by weight of Iron formation. The large average percentage

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 FREDA FORMATION 5

of heavy minerals in each sample further indi- structure was formed from ripple marks. In cates that the sediments were subjected to in- many places along the coast of Lake Superior complete weathering and transported for short where wave action has formed large wave-cut distances. terraces, the plan view of literally thousands of A great variety of ripple marks is found at sets of micro-cross-laminae is well exposed on a numerous horizons throughout the Freda single horizon. Throughout such exposures the formation, especially in the upper part of the sets of cross-strata are strikingly parallel and section, where siltstone predominates. Current remarkably uniform in size. This suggests that ripple marks are most abundant, although they the micro-cross-lamination was not formed in vary a great deal in form and detail. The wave small channels but originated on wide flats length ranges from less than 0.25 inch to more where current energy was weak but relatively than 6 inches, and the crests commonly bi- constant. In numerous places, cuspate ripple furcate and die out, forming an anastomosing marks exhibiting micro-cross-stratification in pattern. A less common variety of current vertical sections are found on the upper surface ripple marks consists of irregular rows of of a coset of micro-cross-laminae. The form of crescent-shaped ridges and depressions which the cusp ripple mark is identical to that of a produce a cusplike pattern in plan view. In micro-cross-lamina. It seems likely therefore many places the cusp ripple grades laterally that micro-cross-lamination results from depo- into parallel ridges characteristic of normal cur- sition on the lee side of a cusp ripple in much rent ripples. Interference ripple marks result- the same way that cross-stratification develops ing from a complex current system and sym- in a barchan. metrical ripples are less abundant. The writer has observed recent cuspate Mud cracks, rill marks, and rain imprints are ripple marks and micro-cross-stratification in a found at many horizons but are not nearly so number of sand bars and flood-plain deposits common as ripple marks. All the sedimentary of mature to old-age streams. These structures structures in the Freda clearly indicate deposi- apparently are formed in fine-grained sedi- tion in shallow water and repeated exposure to ments by slow, sheetlike moving bodies of subaerial conditions. water. Such conditions could exist not only on CROSS-STRATIFICATION: Cross-stratification in flood-plain and wide-stream channels but also the Freda sandstone is predominantly the on tidal flats. trough type (McK.ee and Weir, 1953, p. 385), Throughout the Freda formation there but two distinct varieties are distinguished on seems to be a direct relationship between av- the basis of size and mode of origin. The most erage grain size and scale of cross-stratification. abundant and characteristic variety is ex- Micro-cross-lamination is restricted almost en- tremely small scale and in this paper is referred tirely to the fine-grained parts of the formation to as micro-cross-lamination. The second vari- and is most abundant in the upper part. Small- ety is somewhat larger scale, but few of the dip- scale cross-bedding (McKee and Weir, 1953) ping strata exceed 1 foot in length. The shape dominates in the lower part of the section, of the micro-cross-laminae does not differ sig- where the average grain is between one-eighth nificantly from the typical festoon described of a millimeter and 1 mm in diameter. The sets by Knight (1929), in which the individual of small-scale cross-bedding are on the order of cross-stratum has the shape of a quadrant of an 1.5 feet wide, 6 inches deep, and 3 feet long. In elongate ellipsoid. Each stratum, however, av- several places the small-scale cross-bedding lies erages only 0.5 inch in depth, 2 inches in within a channel structure and seemingly width, and 1.5 inches in length. Most sets of originated by current action within a stream. cross-strata are less than 1 foot long and are The azimuth of both the micro-cross-lamina- truncated in a down-current direction by tion and small-scale cross-bedding in the Freda erosion which preceded deposition of a younger formation was measured wherever exposures set. Throughout most of the upper part of the were accessible. All measurements were made Freda, cosets of micro-cross-laminae averaging from a plan view of the structure and indicate 3 feet in thickness are interbedded with hori- the plunge direction of the trough, which is zontally laminated siltstone and reflect a cyclic considered to be the true direction of sediment pattern in Freda sedimentation. transport. Where micro-cross-lamination was Although the origin of micro-cross-lamina- exposed over a large area, only one measure- tion in the Freda formation is not completely ment was made at each stratigraphic horizon, clear, field evidence strongly suggests that the inasmuch as all sets on each horizon are parallel

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 W. K. HAMBLIN-PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

and represent only one current direction. JACOBSVILLE SANDSTONE Throughout the entire outcrop area the direction of sediment transport as indicated by AND BAYFIELD GROUP micro-cross-lamination is essentially the same STRATIGRAPHY AND PETROLOGY: A sequence as that indicated by small-scale cross-bedding. of red sandstones and shales more than 2000 Figure 2 shows the results of a statistical feet thick separates the Freda formation from analysis of cross-stratification in the Freda the overlying Croixan series. In Michigan these formation. In almost every exposure the mean rocks are known as the Jacobsville sandstone,

EXPLANATION

Average cross - bedding direction

Outcrop belt of Fredo Sandstone

Distribution of cross - bedding

number of readings

FIGURE 2.—CROSS-STRATIFICATION DIRECTIONS IN THE FREDA SANDSTONE

direction of current flow was northwestward. and they extend along the southern coast of Notable exceptions, however, are found in the Lake Superior from the tip of the Keweenaw northernmost exposures on the Keweenaw Peninsula to Sault Ste. Marie (Hamblin, 1958). Peninsula near the mouth of Gratiot River In Wisconsin and Minnesota they are referred where the direction of sediment transport was to as the "red elastics" or Bayfield group southward. In Michigan and Wisconsin the (Thwaites, 1912) and are exposed predom- standard deviation in the paleocurrent direc- inately in the Bayfield Peninsula and Apostle tion is relatively small, and a dominant north- Islands, although they are recognized in the west trend persists throughout the entire section. sub-surface as far south as southern Minnesota Considerable variation in the azimuth of (Fig. 1). Although it cannot be demonstrated cross-stratification was observed in the few ex- conclusively that the Jacobsville formation and posures of the Freda in Canada, even in a very the Bayfield group are exactly contemporane- small stratigraphic interval. Although a north- ous, they occupy essentially the same strati- west trend predominates, more reversals were graphic position and are remarkably similar in observed than would normally be expected to gross lithology, heavy minerals, sedimentary result from a meandering-stream pattern. It is structures, and paleomagnetism. It is therefore reasonable to assume therefore that Freda quite probable that if the Jacobsville is not sediments in Canada were derived from a high- equivalent to the entire Bayfield group, it is land to the north, east, and south and that the at least equivalent to part of it. This correla- eastern margin of the basin of deposition was tion dates back at least to Irving (1883, p. 366) only a few miles east of Sault Ste. Marie. and has subsequently been followed by

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 JACOBSVILLE SANDSTONE AND BAYFIELD GROUP

Thwaites (1912, p. 102-105) and Tyler et al. group averages only 0.18 per cent. From 50 to (1940, p. 1481). 80 per cent of the heavy minerals are opaque, Sedimentary rocks similar to the Jacobsville although garnet, tourmaline, and zircon are have also been reported from the Owen Sound present in appreciable amounts in most places area on Georgian Bay, Lake Huron, by B. A. (Table 1). The high quartz content, good sort- Liberty (Personal communication). Inasmuch ing, and simple heavy-mineral suite indicate as these rocks are isolated and so far removed that the Jacobsville-Bayfield sediments were from the easternmost extent of the Jacobsville subjected to considerably more transportation at Sault Ste. Marie, correlation is difficult at and winnowing action than Freda sediments best. and represent, at least in part, a second-cycle The relationship between the Jacobsville sandstone. formation-Bayfield group and the Freda forma- Current and oscillation ripple marks, mud tion has been a matter of conjecture for many cracks, and shale-pebble conglomerate are gen- years. Thwaites (1912) maintained that the erally not prominent in the Jacobsville and Bayfield group grades downward into the Bayfield rocks but are abundant in units rela- Upper Keweenawan sediments, although the tively free from cross-bedding. The lithology actual contact was never observed. In Michi- of the shale pebbles is identical to that of the gan, east of the Keweenaw fault, the entire shale within the Jacobsville formation and Upper Keweenawan is absent, and the Jacobs- clearly indicates repeated interruptions in ville rests directly upon Middle Keweenawan sedimentation and penecontemporaneous ero- basalts and older rocks with a marked angular sion. Sandstone pebbles are also found im- discordance. Suggestions of an angular uncon- bedded in the basal part of the Jacobsville for- formity between the Freda and Jacobsville mation, especially along the coast from L'Anse formations are also found in several outcrops in to Marquette (Fig. 1). The lithology of the Whitefish Bay (Hamblin, 1958). In the writer's pebbles, however, is quite unlike the typical opinion, available evidence indicates a signifi- Jacobsville but very similar to the Freda. It cant break in sedimentation between the thus seems reasonable to conclude that the Upper Keweenawan sequence and the Jacobs- Jacobsville was derived in part from the Freda. ville formation and suggests that the Jacobs- Channel structures are abundant throughout ville-Bayfield sediments were deposited after much of the Jacobsville and Bayfield sand- Keweenawan but before St. Croixan time. stones and clearly indicate a fluvial environ- The petrology of the Jacobsville sandstone ment. It is evident, however, from the hori- and Bayfield group differs considerably from zontal bedding and oscillation ripple marks in that of the Freda formation. Whereas the Freda the massive-sandstone facies and in the Chequa- consists of arkoses, feldspathic sandstones, and megon sandstone that appreciable sedimenta- micaceous shales, the Jacobsville and Bayfield tion also occurred in a standing body of water. rocks are characteristically quartzose. The CROSS-STRATIFICATION: Medium- to large- average grain ranges between 0.25 and 0.5 mm scale trough cross-bedding is the most abund- in diameter, and except for minor conglomer- ant sedimentary structure in the Jacobsville- ate lenses, most of the samples studied are well Bayfield rocks. Each set of cross-strata forms a sorted and skewed toward the finer grains. Four wide, shallow, concave-upward channel which distinct facies are recognized in the Jacobsville is U-shaped in plan view and wedge- or lens- formation. Lenses of conglomerate occur near shaped in longitudinal section. These structures the basal contact and interfinger with the range from 3 to 6 feet in width, 1 to 2 feet in lenticular sandstone facies which constitutes depth, and 5 to 10 feet in length, but troughs the major part of the formation. In a number more than 50 feet wide were observed in the of outcrops, however, massive sandstones or red Apostle Islands. Small-scale cross-bedding is siltstones predominate and presumably inter- not abundant, and micro-cross-lamination is finger with or overlie the lenticular sandstone absent. In many exposures it is evident that facies. In detail the lenticular sandstone facies the trough cross-bedding was formed in stream of the Jacobsville is identical with the Orienta channels, as it is commonly preserved within sandstone of the Bayfield group, whereas the channel structures. massive sandstone facies closely resembles the Figure 3 shows the results of cross-bedding Chequamegon sandstone. measurements in the Jacobsville-Bayfield rocks. The total weight per cent of the heavy min- The arrows indicate the mean current direction erals in the Jacobsville sandstone and Bayfield at each locality, and rose diagrams represent

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 W. K. HAMBLIN—PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

the percentage of current azimuths measured Upper Mississippi valley, but eastward around in each 20-degree interval. the Wisconsin arch it is generally covered with It is apparent from Figure 3 that the regional glacial drift, so details of its northern extent drainage in the Lake Superior area during are vague and incomplete. The writer (1958, Jacobsville-Bayfield time was northward to- p. 114) has demonstrated, however, that the ward the Canadian Shield. Minor variations in Chapel Rock member of the Munising forma- the direction of sediment transport are evident tion is at least equivalent to part of the Dres- from east to west, however, and are significant bach, and that the outcrop belt extends in that they most likely reflect the configura- around the northwestern rim of the Michigan

: Outcrop belt of Jacobsville - Bayfield group 25 50 75 miles

Distribution of cross - bedding directions. Numerals represent number of readings FIGURE 3.—CROSS-STRATIFICATION DIRECTIONS IN THE JACOBSVILLE SANDSTONE AND BAYFIELD GROUP

tion of the sedimentary basin. The direction basin. Unless otherwise indicated, the term of sediment transport was to the northwest Dresbach in this report, includes, therefore, near the eastern margins of Lake Superior, due the Chapel Rock member of the Munising north from Grand Marais to Marquette, and formation. Sandstones at the base of the to the northeast from the tip of the Keweenaw Paleozoic section exposed at Limestone Moun- Peninsula to Duluth. Thus, except in the vi- tain are identical to the Dresbach in gross cinity of the Huron Mountains, the paleo- lithology, heavy minerals, and sedimentary currents converge toward the center of Lake structures and probably constitute the north- Superior. A second significant variation in the ernmost exposure of the formation. This indi- paleodrainage is the divergence of the direction cates that the original distribution of the Dres- of sediment transport away from the Huron bach was far beyond the present outcrop belt. Mountains area. The Huron Mountains evi- In southern Wisconsin the Dresbach forma- dently existed as a positive area, protruding tion is more than 1000 feet thick, but it thins into the basin of sedimentation and shedding progressively northward and pinches out com- debris to the northeast and northwest. The out- pletely over the Wisconsin arch a few miles line of the Jacobsville-Bayfield basin, was, in north of Iron Mountain, Michigan (Figs. 4, 5). all probability, very similar to the present Lake North of the Wisconsin arch the Dresbach re- Superior basin, although somewhat larger. appears and thickens to more than 75 feet in the vicinity ot Grand Island, Michigan. DRESBACH FORMATION Both surface and subsurface data indicate STRATIGRAPHY AND PETROLOGY: The DfCS- that the Dresbach formation is separated from bach formation is well exposed throughout the the older rocks by a widespread unconformity.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 DRESBACH FORMATION

Exposures in the Grand Island area show that members of the Dresbach in Minnesota and the Jacobsville was tilted and eroded prior to Wisconsin (Fig. 4). the first advancement of the Croixan seas. The heavy-mineral suite of the Dresbach These exposures indicate that the Jacobsville is formation is characterized by a dominance of not part of the Croixan series as was thought by zircon. Tourmaline, leucoxene, and ilmenite many previous workers. Elsewhere, at Taylors are common, and garnet is generally rare, al- Falls, Minnesota, and in the Baraboo area of though it occurs in appreciable amounts in the Wisconsin, where the pre-Croixan uncon- Eau Clair member (Tyler, 1936). Table 1 formity is exposed, the erosional surface is de- shows the relative percentage of the heavy-

s.w. N. E.

Sandstone - medium to coarse large - scale cross - bedding

Sandstone ~ fine to coarse EZJ small - scale cross - bedding FIGURE 4.—DIAGRAMMATIC SECTION OF UPPER CAMBRIAN DEPOSITS IN THE LAKE SUPERIOR REGION SHOWING FACIES VARIATIONS AND NOMENCLATURE OF ROCK UNITS (Wisconsin data in part from Berg et al., 1956.)

veloped upon Keweenawan basalts and older mineral species from selected samples of the rocks. Dresbach formation in Michigan, Minnesota, Berg et al. (1956) recognize three distinct and Wisconsin. lithic units within the Dresbach formation in The heavy-mineral suite of the Dresbach is Minnesota and Wisconsin: (1) medium- to thus significantly different from the other for- coarse-grained quartzose sandstone (Mt. Simon mations studied in this report and probably and Galesville members); (2) shaly siltstone reflects differences in both environment and and very fine-grained quartzose sandstone (Eau source area. Clair member); and (3) a red-shale phase of Ripple marks and mud cracks are present in the Eau Clair member identified in wells as the Dresbach formation but are generally re- Waconia and Mankota. Although these units stricted to sections relatively free from cross- cannot be traced with certainty into northern bedding. These structures are most abundant Michigan, the lower part of the Munising for- at Tahquamenon Falls and in the Eau Clair mation consists of a coarse orthoquartzitic member in Wisconsin and Minnesota. Both conglomerate 10 to 15 feet thick overlain by a oscillation and current ripple marks are present well-sorted medium-grained quartzose sand- and generally have a wave length between 2 stone identical in gross lithology to the Mt. and 4 inches. The ripple crests generally strike Simon and Galesville members. On the basis east-west or essentially perpendicular to the of lithologic similarity and stratigraphic posi- direction of sediment transport as indicated by tion, therefore, this quartzose sandstone, the cross-bedding. Chapel Rock member of the Munising forma- CROSS-STRATIFICATION: A distinctive large- tion, is considered to represent a shore facies scale trough cross-bedding characterizes the equivalent to the Mt. Simon and Galesville Dresbach formation and is the major type of

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 10 W. K. HAMBLIN—PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

stratification in most exposures. In plan view cut deep channels in the older units prior to the trough is commonly more than 50 feet the deposition of each succeeding set of cross- wide; several troughs more than 600 feet wide strata. This indicates that the structures were were observed along the coast approximately formed in an environment of strong currents 10 miles east of Munising. Most outcrops are capable of considerable sublevation or else in too small to expose these structures in three water sufficiently shallow to permit repeated dimensions, but along the Pictured Rocks cliffs subaerial erosion. It appears from the tre-

FlGLRM 5. IsOPACH MAP AND CROSS-STRATIFICATION DIRECTIONS OF TIIK DRHSBACII FoRMATIOM

in Michigan and in the Dells of Wisconsin, ex- mendous size and troughlike shape of these posures are adequate to give a true picture of structures that they were formed in embay - their exact form and size. In most outcrops, ments along a cuspate shore. The trough of the however, only a very small part of the trough cross-bedding would thus plunge seaward and is exposed. indicate the direction of regional slope. In many respects the large-scale cross-bed- In Minnesota and Wisconsin the scale of the ding of the Dresbach formation is similar to cross-bedding decreases significantly south- the large-scale eolian cross-bedding of the ward, and the average trough is only between Navajo sandstone of the Colorado Plateau, but 3 and 5 feet wide in the southernmost ex- stringers of large pebbles more than 3 inches in posures. diameter are incorporated in many places Figure 5 shows the mean dip direction of the within the Dresbach and testify to its aqueous large-scale cross-bedding at each locality where origin. The lower bounding surface of each set sufficient numbers of measurements could be of cross-strata shows that considerable erosion obtained to give a reliable average. Most out-

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 DRESBACH FORMATION 11

crops are too small to expose the entire trough hill member—cross-bedded, medium- to coarse- in three dimensions, so the plunge direction grained sandstone; the Birkmore member— could not be determined accurately at every glauconitic, fine-grained sandstone; the Tomah exposure. Dip directions of the limbs of the member—very fine-grained sandstone and in- trough were therefore measured as close to the terbedded shale; the Reno member—glaucon- axes as possible so that when averaged to- itic, fine-grained sandstone; and the Mazo- gether they would indicate the plunge direc- manie member—thin-bedded or cross-bedded tion and regional slope. In a few localities such sandstone that forms a nonglauconitic facies as Pictured Rocks in Michigan, the Dells of interfingering with the Reno member. The Wisconsin, and in several quarries, plan views glauconitic facies of the Reno and Birkmore of the structures are well exposed, and many members is thickest in the southwest, but it true bearings ot the plunge of the trough were interfingers with and passes laterally into the easily obtained. The effect of measuring the nonglauconitic Mazomanie member to the dip direction of the limbs is indicated in the northeast and is completely absent near the relatively large distribution of cross-bedding northernmost limit of the formation in Wiscon- azimuths shown in the rose diagram of measure- sin. In northern Michigan, fossils indicate that ments made in Minnesota and Wisconsin. The the upper part of the Miner's Castle member grand average for all dip directions is probably of the Munising formation is equivalent to the reasonably accurate, but the writer believes middle Franconia of Minnesota and Wisconsin that if only the plunge directions were meas- (Hamblin, 1958). The Miner's Castle member ured, the rose diagram would show a much of the Munising formation, however, consists greater central tendency and would be similar only of the nonglauconitic facies and is highly to what is indicated in Michigan where es- variable in grain size and degree of sorting. In sentially every measurement is the direction gross lithology and general characteristics it is of plunge. identical to the Mazomanie member (Fig. 4). The average dip of the cross-bedding in the On the basis of variations in sorting, mean Dresbach formation is to the south in Minne- gram diameter, and heavy minerals in samples sota and Wisconsin and to the north in northern collected in Alger County, Michigan, Driscoll Michigan. In each area the dip tends, to a re- (1959) questions this correlation and proposes markable degree, to be perpendicular to the that the Miner's Castle member of the Munis- isopach lines and points in the direction of ing formation is transitional between the trans- greatest thickness. This clearly indicates that a gressive Franconia sandstone and the regressive highland extended in an east-west direction Jordan sandstone. If Driscoll's interpretation is through northern Michigan and into Wisconsin correct, parameters of the Prosaukia subzone and that sediments were shed off both to the should be very similar to those of the lower north and south. Miner's Castle member in Alger County. The The location of the source area established writer found however that the mean grain by the measurements given is further sub- diameter, sorting, and heavy-mineral assem- stantiated by an increase in the average thick- blage in the Prosaukia subzone in northern ness of the cross-bedding units from south to Michigan are almost identical to those in the north in Minnesota and Wisconsin. According upper part of the Miner's Castle member and to Schwarzacher (1953), cross-bedding units are significantly different from those in the thicken toward the source of the sediment. lower part (Fig. 6). Inasmuch as the Prosaufya subzone is recognized in drill hole cores less FRANCONIA FORMATION than 12 miles west of sections studied by Dris- STRATIGRAPHY AND PETROLOGY: The Fran- coll, and because parameters, gross lithology, conia formation consists of a highly variable and sedimentary structures of the Prosaukia sequence of sandstones that range from 75 to subzone are strikingly similar to those of the 200 feet thick within the area covered by this upper part of the Miner's Castle member, it study. Berg (1954) recognizes five distinct seems very unlikely that the upper part of the lithic units within the Franconia in Wisconsin Miner's Castle member is equivalent to the and Minnesota which are distinguished pri- Jordan sandstone of Wisconsin. Although a re- marily on the basis of glauconite content, and gressive phase no doubt was once present in to a lesser extent on such features as grain size, northern Michigan, it was probably eroded bedding, and the presence of carbonate or away prior to deposition of the Au Train shale. They are, in ascending order, the Wood- formation.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 12 W. K. HAMBUN—PALF.OGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

Regional variations in thickness of the Fran- are exposed, it is reasonable to assume that they coma formation are substantially different from exist but are not discernible in many vertical those of the Dresbach formation (compare exposures. Both oscillation and current ripple Figs. 5 and 7). The Franconia thickens to the marks are present, but in many exposures the northeast in Minnesota and Wisconsin and ob- crests have been flattened by erosion, so it is viously once extended far beyond its present impossible to distinguish between the two eroded limits. In northern Michigan it thins forms in all places. Wherever ripple marks are over the central part of the peninsula but does found, their crests strike essentially perpen- not pinch out completely as does the Dresbach dicular to the current direction that is indi- formation. cated by cross-bedding.

L 50- 60- 60 - \ 70 - 80 - 90 - 2.0 2.2 .06 .07 .08 .09 1.0 1.3 1.3 1.4 1.5 40 50 60 70 90 90 100

FIGURE 6.—VARIATIONS IN SORTING, MEAN GRAIN DIAMETER, AND PER CENT GARNET IN THE MINER'S CASTLE. MEMBER OF THE MUNISING FORMATION Solid line represents average of four sections in Alger County, Michigan (Driscoll, 1959), broken line represents Prosaukfa sub/one in Menominee County, Michigan.

Basins in which the Franconia accumulated CROSS-STRATIFICATION: The Franconia for- to greatest thickness are located in west-central mation is characterized by small-scale trough Wisconsin and probably in the Lake Superior cross-bedding which is remarkably persistent and Michigan basins. Significant thinning oc- in size and shape throughout the entire area. curs in the Baraboo Mountains area and locally This structure constitutes the dominant type in northern Michigan where erosional remnants of stratification in the nonglauconitic facies but of Precambrian rocks produced islands in the is much less common where glauconite is Franconia sea. abundant. The troughs range from 3 to 8 inches The heavy-mineral assemblage of the Fran- in depth and are 20 to 40 inches long in a down- conia formation is characterized by a remark- current direction. ably high percentage of garnet and is, there- The environment in which the small-scale fore, readily distinguished from that of the cross-bedding originated is not completely Freda, Tacobsville, or Dresbach formations. understood, but several features suggest that Throughout the entire extent of the Fran- it originated offshore by currents moving down conia, garnet is present in amounts between the regional slope. The orientation of the cross- 40 and 96 per cent of the total heavy minerals. bedding is remarkably uniform throughout an Tourmaline and zircon are the only other area of more than 8000 square miles (Fig. 7), nonopaque heavy minerals consistently present and the direction of current flow is, by and (Table 1). large, perpendicular to the major isopach Ripple marks and mud cracks are abundant trends. In addition glauconite is concentrated in the Franconia formation in some localities, along the inclined strata in many places, sug- but in many vertical exposures they are absent. gesting that the structures were produced in an Inasmuch as these structures are readily ap- offshore environment. It therefore seems more parent only where large areas of bedding planes likely that the cross-bedding orientation was

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 FRANCONIA FORMATION 13

controlled by the regional slope than long- directions average 49°, which is somewhat shore currents. larger than that of the older formations con- Figure 7 shows the results of a statistical sidered in this report. This is probably due to analysis of the cross-bedding direction in the the fact that the Franconia cross-bedding re- Franconia formation. Each arrow indicates the sulted from marine currents that were not con- mean direction computed from an average of fined to the coast line or restricted to a drain- 25 measurements made at each locality. The age system.

FIGURE 7.—ISOPACH MAP AND CROSS-STRATIFICATION DIRECTIONS OF THE FRANCONIA FORMATION

two rose diagrams illustrate the degree of Excepting a small area in northern Michigan, central tendency of current direction in the paleocurrent directions are perpendicular northern Michigan and in the upper Mississippi to the isopach lines but do not converge to- Valley. ward the thickest area in western Wisconsin. From Figure 7 it is readily apparent that the The prominent southwest trend is maintained current directions during Franconian time throughout the entire area. were consistently toward the southwest and were only slightly affected by the Precambrian PALEOGEOGRAPHY highs such as those in northern Michigan and in the Baraboo region. Furthermore the paleo- Late Keiveenawan Time current direction remained remarkably stable During Early and Middle Keweenawan throughout all Franconian time, as there is no time more than 30,000 feet of basaltic flows significant variation in cross-bedding directions and medium- to coarse-grained clastic sedi- from the top to the bottom of the formation. ments were deposited in the Keweenaw basin The standard deviations of the cross-bedding which was located in approximately the present

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 14 W. K. HAMBLIN—PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

site of Lake Superior. Bent pipe amygdules west, suggesting a continuation of Keweena- and other primary structures indicate that the wan rocks in that direction. Sedimentary lava flowed from the center of the basin out structures in the westernmost exposures of the toward the margins, whereas cross-bedding and Freda show no indication of a change in the imbricate structure in conglomerate beds indi- direction of paleocurrents such as might herald cate that the clastic material came from a high- proximity to the end of the basin. It is there- land to the south. This apparent paradox was fore concluded that although the basin in resolved by White (1957), who suggested that which the Freda sediments accumulated may the floor of the basin was nearly flat and was not have extended throughout the complete continuously being downwarped by tectonic length of the gravity anomaly, it most likely movements. If the extrusion of the lava kept extended an appreciable distance southwest of pace with subsidence, the lava surface would be the present outcrop belt. essentially flat or would slope gently toward The Keweenawan basin continued to subside the margins so the streams coming from the during all of Late Keweenawan time. Sub- highland to the south could not extend out sidence was both rapid and pulsating, as is in- into the basin. When extrusion of lava was dicated by the abundance of feldspar through- interrupted for any period of time, continued out the formation and the alternating units of downwarping would then produce a topo- coarse and fine material. The abundance of graphic basin into which streams could flow feldspar, ilmenite, and epidote grains together and deposit the clastic sediments derived from with the low percentage of quartz indicates the south. Subsequent lava flows would flow that the Keweenawan basalts and the Huronian toward the lowest points in the basin and would Iron formation and granite were the dominant spread out toward the margins from there. rock types m the source area. The major elements of paleogeography did The Freda formation is considered to have not change significantly from Middle to Late accumulated in a terrestrial environment. Keweenawan time, except that extrusion of Micro-cross-lamination which is prominent lava gradually ceased. Freda sediments were throughout most of the section indicates rela- derived from a southern source area that ap- tively weak currents and is presumed to have parently extended through northern Michigan been formed on a flood plain. The uniform and into Canada (Fig. 8). This positive area, regional orientation of the micro-cross-lammae named the Northern Michigan highland by precludes the possibility of a tidal-flat environ- the writer (Hamblin, 1958), was probably ment, since in this environment cross-bedding formed near the beginning of Keweenawan forms in two opposing directions resulting from time and persisted as a highland up until middle currents associated with ebb and flood tides. Late Cambrian time. Remnants of these moun- Innumerable interruptions in sedimentation tains are found in the greatly deformed are indicated by the abundance of shale pebbles Iluronian rocks of the Huron Mountains. The and blocks scattered throughout the formation. eastern margin of the basin was located in the Persistent beds that can be traced laterally Batchawana Bay area of Canada, and although over several miles were apparently formed in the dominant direction of sediment transport standing bodies of water. in that area was to the northwest, sediments were also derived from the south and northeast. Jacobsville-Bayfield Time The northern boundary of the basin cannot Prior to Jacobsville sedimentation, the be definitely determined from available data, Keweenawan sequence was deformed, and a but cross-stratification in the northernmost period of weathering and erosion produced an exposures of the Freda indicates that the sedi- irregular surface and a zone of weathered debris ments in that locality were possibly derived upon the Keweenawan basalts and older Irom a northern source. Inasmuch as Freda granites and metamorphic rocks (Hamblin, rocks are not found on the northern shore of 1958). The Northern Michigan highland re- Lake Superior, it seems reasonable to assume mained as a positive area during Jacobsville- that the northern boundary of the basin was Bayfield time and supplied sediments to the probably very near the northern shore of the subsiding basin to the north (Fig. 8). The shape Lake. A considerable southwestern extension and extent of the basin apparently did not of the Keweenawan basin is indicated by a change significantly from Late Keweenawan gravity anomaly that correlates with the mafic time. Sediments were derived from the south- Keweenawan rocks (Thiel, 1956). This anomaly east in the Sault Ste. Marie area and probably extends without interruption far to the south- reflect the eastern limits of the basin. In the

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 PALEOGEOGRAPHY 15

environs of Munising the direction of sediment effective barrier to the encroachment of the transport was essentially due north, but a Cambrian seas onto the Canadian Shield from salient of the Northern Michigan highland the south. The first Paleozoic seas, therefore, extended northward in the Huron Mountains entered the Lake Superior area from the north- area and shed off debris to the north, northeast, west and probably never completely covered and northwest. Throughout the Apostle Islands the Northern Michigan highland (Fig. 8). and on up to the tip of the Keweenawan To the north the seas advanced rapidly from Peninsula, the dominant trend of sediment the northwest over a surface of low relief de- transport was to the northeast. This trend is veloped upon the slightly tilted and eroded not indicated by cross-stratification of Upper Jacobsville sediments. The orthoquartzitic Keweenawan rocks but could have been pres- conglomerate and large-scale cross-bedding ent during Freda time, as it is located west of indicate that the Dresbach in northern Michi- the present Freda exposures. This dominant gan accumulated in a near-shore environment northeast trend could, however, reflect a post- north of the Northern Michigan highland. The Freda uplift in the southwest. Sediments were large-scale cross-bedding was probably formed also derived from an area west of Duluth, indi- along a cuspate shore where wave action was cating that the western margin of the basin strong and capable of considerable erosion. was not far from the western coast of Lake In Wisconsin and Minnesota the Dresbach Superior. seas invaded the area from the south. The seas Fluvial sedimentation predominated near evidently transgressed over a surface of low re- the margins of the basin, but a lacustrine en- lief developed upon Laurentian granites vironment was present in the deepest parts and Huronian metamorphic rocks, and Keweena- became more prominent during late Jacobs- wan basalts. Local highs on the Precambrian ville-Bayfield time. Although a large part of surface such as Baraboo Mountains and Taylors the Jacobsville sandstone and Bayfield group Falls formed islands and peninsulas in the Dres- accumulated in a fluvial environment, large- bach sea, but their local relief was only several scale cross-bedding and numerous channels of hundred feet, and they exerted little effect on conglomerate indicate that the type of environ- regional sedimentation. ment was considerably different from that in A large part of the Dresbach in northern which the fine micro-cross-laminated sands of Wisconsin accumulated in a near-shore en- the Freda accumulated. Much of the Jacobs- vironment, as is indicated by the large-scale ville now exposed undoubtedly represents fan cross-bedding identical to that found in deposits which merged northward into sedi- northern Michigan. To the south, however, ments of an alluvial plain. an offshore, neritic environment predominated, Subsidence of the basin apparently did not and more than 1000 feet of sediment accumu- keep pace with sedimentation, and the North- lated in the areas of maximum subsidence. ern Michigan highland progressively became Most of the Dresbach sediments were de- buried in its own debris. Numerous interrup- rived directly from the Northern Michigan tions in sedimentation are attested by the highland which extended through northern abundance of shale pebbles scattered through- Michigan and northern Wisconsin and shed off out the formation. debris to the north and south. Jacobsville- The relatively simple heavy-mineral suite Bayfield sediments which at one time at least combined with the quartzose nature of the partly covered the Northern Michigan high- Jacobsville sandstone and Bayfield group land were subjected to another cycle of erosion strongly suggests a second-cycle sand. Sand- and acted as a major source for the Dresbach stone pebbles similar in lithology to the Freda sands. found embedded in the Jacobsville further support this conclusion and suggest that outer Franconian Time margins of the Freda were eroded and supplied The Northern Michigan highland which was appreciable amounts of material to the Jacobs- the major paleogeographic element from ville formation. Keweenawan to Dresbachian time exerted little influence upon the paleogeography during Dresbachian Time Franconian time. A widespread unconformity If the Northern Michigan highland was es- at the top of the Dresbach formation indicates sentially covered with Jacobsville-Bayfield a regression of the seas accompanied by sub- sediments, it was uplifted and partly exhumed aerial erosion and the removal of part of the prior to Dresbachian time and formed an Dresbach formation. Thus, if the Northern

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 16 \V. K. HAMBI.IN—PALKOGKOGRAPHIC INVOLUTION, LAKE SUPERIOR RliGION

Late Keweenawan Time

Jacobsville - Bayfield Time I'K;LTRI-: 8. — PAL[^)GEOGRAPHIC MAPS OF THE LAKE SUPERIOR REGION DURING LA' Arrows indicate direction

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 PALEOGEOGRAPHY 17

Dresbachian Time

Francor.ian Time ^EENAWAN, JACOBSVILLE-BAYFIELD, DRESBACHIAN, AND FRANCONIAN TIME tnent transport.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021 18 W. K. HAMBLIN—PALEOGEOGRAPHIC EVOLUTION, LAKE SUPERIOR REGION

Michigan highland was ever completely A large proportion of the Franconia sedi- covered with Dresbach sediments, it was again ments in southern Wisconsin accumulated in partly exhumed and existed only as a subdued an offshore environment, as is suggested by the crystalline landmass in small parts of northern distribution of glaucomte. However, ripple Wisconsin and northern Michigan. The Fran- marks and mud cracks indicate that the shallow conia seas transgressed across the entire area seas were occasionally exposed to desiccation from the southwest (Fig. 8). Numerous islands on tidal flats. Nonglauconitic small-scale cross- existed in the early Franconia seas in the bedded sandstones accumulated nearer the Baraboo area and where higher parts of the strand line, but somewhat removed from the Northern Michigan Highland existed in beach. The major source area was to the north- northern Michigan. These islands exerted only east, and undoubtedly the Franconia seas re- local influence on regional sedimentation and worked a large part of the Dresbach sediments, were apparently soon covered by the rapidly transgressing sea.

REFERENCES CITED Berg, R. R., 1954, Franconia formation of Minnesota and Wisconsin: Geol. Soc. America Bull, v. 65, p. 857-882 Berg, R. R., et al., 1956, Upper Cambrian rocks in southeast Minnesota: Geol. Soc. America Field Trip Guidebook to the Lower Paleozoic of the Upper Mississippi Valley, p. 1-23 Driscoll, E. G., 1959, Evidence of transgressive-regressive Cambrian sandstones bordering Lake Superior: Jour. Sed. Petrology, v. 29, p. 5-15 Hamblin, W. K., 1958, Cambrian sandstones of northern Michigan: State of Michigan, Dept. Conserva- tion, Geol. Survey Div., Publ. 51, 146 p. Irving, R. D., 1883, The copper-bearing rocks of Lake Superior: U. S. Geol. Survey Mon. 5, 366p. Knight, S. H., 1929, The Fountain and the Casper formations of the Laramie Basin: Univ. Wyoming Publ. Sci. Geology, v. 1, no. 1, 182 p. McKee, E. D., and Weir, G. W., 1953, Terminology for stratification and cross-stratification in sedimentary rocks: Geol. Soc. America Bull., v. 64, p. 381-390 Pentland, A., 1931, The heavy minerals of the Franconia and Mazomanie sandstones, Wisconsin: Jour. Sed. Petrology, v. 1, p. 23-36 Schwarzacher, W., 1953, Cross-bedding and grain size in the Lower Cretaceous sands of East Anglia: Geol. Mag., v. 90, p. 322-330 Thiel, E., 1956, Correlation of gravity anomalies with the Ke\veena\van geology of Wisconsin and Minne- sota: Geol. Soc. America Bull., v. 67, p. 1079-1100 Thwaites, F. T., 1912, Sandstones of the Wisconsin coast of Lake Superior: Wis. Geol. Nat. Hist. Survev, Bull. 25, 109 p. Tyler, S. A., 1936, Heavy minerals of the St. Peter sandstone in Wisconsin: four. Sed. Petrology, v. 6, p. 55-84 Tyler, S. A., et. al., 1940, Studies of the Lake Superior Precambrian by accessory-mineral methods: Geol. Soc. America Bull., v. 51, p. 1429-1537 White, W. S., 1957, Regional structural setting of the Michigan native copper district: Institute on Lake Superior Geology, Mich. College of Mining and Tech. Press Wilgus. W. L., 1933, Heavy minerals of the Dresbach sandstone of \vestern Wisconsin: Jour. Sed. Pe- trology, v. 3, p. 83-91 '

MANUSCRIPT RECEIVED BY THE SECRETARY OF THE SOCIETY, JUNE 8, 1959

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/72/1/1/3416832/i0016-7606-72-1-1.pdf by guest on 03 October 2021