University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

USGS Staff -- Published Research US Geological Survey

7-1948

Mississippian-Pennsylvanian Boundary Problems In The Rocky Mountain Region

James Steele Williams United States Geological Survey

Follow this and additional works at: https://digitalcommons.unl.edu/usgsstaffpub

Part of the Earth Sciences Commons

Williams, James Steele, "Mississippian-Pennsylvanian Boundary Problems In The Rocky Mountain Region" (1948). USGS Staff -- Published Research. 502. https://digitalcommons.unl.edu/usgsstaffpub/502

This Article is brought to you for free and open access by the US Geological Survey at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in USGS Staff -- Published Research by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. The Journal of Geology, Vol. 56, No. 4, Jul., 1948

MISSISSIPPIAN-PENNSYLVANIAN BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION'

JAMES STEELE WILLIAMS United States Geological Survey

ABSTRACT A variety of paleontologic and stratigraphic problems are presented by rocks near the Mississippian- Pennsylvanian boundary in the central and northern Rocky Mountains. Stratigraphic sections of these rocks show diverse interpretations of fundamental concepts of stratigraphy and paleontology. In many places where Upper Mississippian rocks directly underlie Pennsylvanian rocks it is difficult to determine the precise location of the boundary between these units. Formations that straddle the boundary are very useful and satisfactory over large areas. Most geologists use various types of lithologic criteria to distinguish forma- tions, but some appear to rely mainly on faunal data, unconformities, or attempts to trace prominent beds. More uniformity in criteria than now exists for the delimitation of formations is desirable. Surface and sub- surface formations should conform to the same definition. Critical paleontologic studies of several common species and genera, if based on a large number of specimens, might help solve the boundary problem. More correlations based on several lines of paleontologic evidence and less reliance on a few index fossils wouldalso help. Larger and more varied collections of well-preserved fossils stratigraphically located are needed from critical areas. Additional stratigraphic work in this region should be of a detailed nature and should prefer- ably be done in connection with detailed mapping. Ecologic and paleogeographic factors merit more atten- tion. The age significance of unconformities has perhaps been overestimated generally.

INTRODUCTION comes well informed on a single small As in many other parts of the United problemor spends considerabletime on a States, the Mississippian and Pennsyl- large problem. There is, however, also vanian rocks of the Rocky Mountain re- roomfor broadlyinterpretive work. Con- gion present many unsolved problems. sidering the vast area of the Rocky These problemsrelate to all stratigraphic Mountains underlain by Carboniferous zones from the base of the Mississippian rocks and the difficultyof access of many to the top of the Pennsylvanian.A group of the exposures, a very creditable of problemsthat involve beds at or near amount of knowledgeof the stratigraphy the Mississippian-Pennsylvanianbound- and paleontology has existed for a long ary are especially interesting because time; but not all of it is published, and they not only show places at which the much that is published is in papers con- geological data are sadly deficient but cerned also with general and economic also involve interpretations and differ- geology,with which papersmany stratig- ences in viewpoints on fundamentalprin- raphers appear to be unfamiliar. This knowledgemust be consideredby anyone ciples of paleontology and stratigraphy. All students of Carboniferousprob- starting work in the Rocky Mountains. lems, especially those who have them- Problemsin the Rocky Mountain Car- selves worked in the Rocky Mountains, boniferous(not all of which will be solved will agree that much geologicwork needs or even reviewed in this paper!) range to be done there. The type of work most from the need for more and better fossils, carefully needed, in the writer's opinion, is not, collected with respect to their however, reconnaissancework but de- geographic locations and stratigraphic tailed work, wherebythe investigatorbe- horizons, to the need for reviews, and perhaps reappraisals,of some of the fun- ' Published by permission of the Director, U.S. damental hypotheses and Geological Survey. Manuscript received February definitions 24, 1948. used in stratigraphy and paleontology. 327 328 JAMES STEELE WILLIAMS

Among these last-namedare such things an importanttime and time-rockbound- as definitions of various rock and time ary, as distinct from a lithologic bound- units and the applicationsof these defini- ary. It may (and does) happen to coin- tions in the field; hypotheses of, and fac- cide with distinct lithologic changes in tors in, the correlationof strata; and the- some places but not with important ories of speciesdefinition in paleontology. lithologic changes in others. It coincides Despite the two hundred and thirty or with an unconformity in some regions more years of the existence of the science and not with a recognizableunconform- of stratigraphyand stratigraphicpaleon- ity in others. It is a practical boundary tology, many disagreementsexist in the for mappingin some places, and in others application, if not in the definition, of it is not. Nevertheless, this boundary is many of the fundamentalor near-funda- one of the more important ones in the mental concepts upon which the daily United States. work of the stratigrapher and strati- graphic paleontologist is based. POSITION OF MISSISSIPPIAN-PENN- Whether one considers the Mississip- SYLVANIAN BOUNDARY pian-Pennsylvanian boundary a sys- All who are familiar with the general temic, subsystemic, or series boundary geology of the central and northern depends on the definitionsof a system, a Rocky Mountains know that, broadly subsystem, and a seriesto which one sub- speaking, the Mississippian rocks there scribes and on the applications(or inter- constitute a sequence mainly of lime- pretations) of these definitions in par- stones, whereasthe Pennsylvanianrocks ticular regions and with particular se- constitute a dominantly sandstone or quences of rocks; also involved are the "quartzite"sequence. Between the domi- uses or underlyingpurposes that one has nantly limestone sequence of the Missis- in mind for each of the units, the general sippian and the dominantly sandstone usage throughout the world, the degree sequence of the Pennsylvanianthere lies of relianceand degreeof finenessof inter- a series of thin and in many places alter- continental correlationsof the particular nating beds of sandstones, shales, thin units of rocks under consideration,and limestones, cherts, and other kinds of the breadth of experience one has with rock. In many places this series of rocks the rocksinvolved. All these are variable, contains red or purple beds, material and there is certainly adequate room for from which stains associated beds and at justified disagreements in the weights many exposures the whole series has a and interpretations given each of the reddishtinge. In many places the Missis- above factors and for disagreementin the sippian-Pennsylvanianboundary is with- rank assignedto the units called "Missis- in this series of rocks, some of the beds sippian" and "Pennsylvanian."A defi- being Mississippianand others Pennsyl- nite agreement is not necessary, and it vanian. In other places, however, the would be outside the scope of this paper Mississippian-Pennsylvanianboundary, to present argumentsfor or against any as determinedby fossils, appearsto coin- specific conclusion.The writer considers cide with a lithologic boundary. The that the Mississippian-Pennsylvanian Mississippian-Pennsylvanian boundary boundary is an important boundary in is placed within a series of alternating the United States (more so in some re- thin-beddedrocks-a nonresistantseries gions than in others) and believes it to be -not only in the area here discussedbut BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 329 in a far wider area in the western part of Mississippian fossils from beds that the United States. might be considered Lower Morgan. The name "Morgan" has been extended FORMATIONS INVOLVED into the Cottonwood-American Fork re- Early practice.-The variable beds be- gion of the Wasatch Mountains and to tween the Mississippian limestones and other areas in this part of the Wasatch the Pennsylvanian sandstones or quartz- Mountains and has also been used at sev- ites have, in the area under discussion, eral places in the Uinta Mountains. been placed in different formations in dif- In central-western Utah, the nonre- ferent parts of the area. In western and sistant unit of alternating shales, lime- central Montana and in northwestern stones, and sandstones in which the Mis- Wyoming they were generally assigned sissippian-Pennsylvanian boundary oc- to the lower part of the Quadrant forma- curs has little red material. At this place tion and widely, but not universally, con- the unit was called the Manning Canyon sidered Mississippian in age. In west- shale by Gilluly (1932, pp. 31-34). The central and central-northern Wyoming name Manning Canyon has been used and in parts of Montana contiguous to also for units of approximately the same northern Wyoming, they were placed in age in eastern Nevada and at other places the Amsden formation, which has from in central Utah, including at least one 1906 (Darton, 1906, p. 5), two years after area in the Wasatch Mountains, near the time of the proposal of the name Provo (Baker, 1947). "Amsden" for the beds, been generally The above paragraphs give the general considered to be of both Mississippian usage as of about 1930 (fig. i). This shows and Pennsylvanian age. In southeastern that in most places the Mississippian- and eastern Idaho and contiguous parts Pennsylvanian boundary was frankly of western Wyoming and Utah, the lower acknowledged not to be a practical map- beds of the sequence were for a long time ping (i.e., formational) boundary but to placed in the upper part of the Brazer lie within a formation. Exceptions to this limestone and the upper beds in the lower were found, however, in the southeastern part of the Wells formation. In mapping Idaho area, where attempts (abandoned begun in 1931 in the Afton quadrangle, by W. W. Rubey and the writer) were southeastern Idaho and southern Wyo- made to map the Mississippian-Pennsyl- ming, but as yet unpublished, W. W. vanian contact as the Brazer-Wells con- Rubey and the writer grouped the beds tact and, to a degree, in the area of the together in a single mapping unit, to typical Morgan formation, where all of which a field name has been applied the Morgan was considered to be of pending decision as to which of the avail- Pennsylvanian age. able names to use. In north-central Utah Recent work.-Much work has been the beds were put in the Morgan forma- done in the central and northern Rocky tion, which was considered by its namer, Mountains since 1930. The United States Eliot Blackwelder (1910, p. 530), to be Geological Survey has had many field Pennsylvanian in age, but which may parties working in various parts of the contain Mississippian beds in its lower region. Renewed interest in stratigraphic part. The writer has collected Pennsyl- problems has been shown by some of the vanian fossils from the type section and state geological organizations and espe- other exposures of the Morgan, and cially by geologists on the faculties of col- 330 JAMES STEELE WILLIAMS leges and universities in the area. Very some of the identificationsare as yet pro- important contributionshave been made visional. Most of the sections measured by the faculties and student bodies of the by GeologicalSurvey parties have been many summercamps maintained by mid- published. Unfortunately, wartime and western and easternuniversities in differ- other duties have prevented the writer ent parts of the Rocky Mountains. from compiling and publishing many of

NORTH WYOMING SW. WYOMING NORTH MONTANA AND AND SOUTHERN W. CENTRAL AND NORTHERN CENTRAL NW. WYOMING MONTANA WYOMING S.E. IDAHO UTAH UTAH

PENN. QUADRANT TENSLEEP TENSLEEP WELLS WEBER OQUIRRH

MORGAN MANNING AMSDEN AMSDEN CANYON C GREAT BLUE S.G. BRAZER BRAZER

HUMBUG M MISS. DESERET

0 MADISON MADISON MADISON MADISON MADISON MADISON

SPRE- MADISON? E. STROMBER6

FIG. i.-Correlation chart showing widespread usage as of about 1930. The stratigraphic usage shown for several of the areas on the chart is still the preferred and most satisfactory, but it has changed in other areas. The diagonal lines at the base and top of the cross-lined areas represent an attempt to show that both the top of the Madison limestone and the base of the Amsden formation probably are of different ages in different places, or are thought to be by different geologists. The abbreviations are for the well-known Mis- sissippi Valley subdivisions of the Mississippian.

The writer has aided nearly every the sections that he has measured inde- United States Geological Survey party pendently and from studying carefully that has worked in this area since all the collections of fossils in his hands. 1930 with its problems in Carboniferous rocks and, in addition, while engaged in SELECTED STRATIGRAPHIC SECTIONS stratigraphic projects of his own, has The total stratigraphicwork done on measured many sections in areas not the Mississippian-Pennsylvanianbound- worked in detail by the Survey mapping ary problem by all the geologists who parties. Fossils have been systematically, have workedin this regionis so great and though not always adequately, collected the numberof stratigraphicsections is so from nearly all these sections, and most large that only a small proportion of of these have been identified, although them can be discussedin this paper. Con- BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 331

sequently, a few sections have been se- (4) a basal unit of yellow sandstone, lected to show the trends in each of sev- sandy shale, and gypsum, I50-200 feet eral areas. It is hoped that these sections thick. Girty, who examinedthe fossil col- will reveal general tendencies in proce- lections made by Reeves, identified fos- dure that can be evaluated later and will sils from the middle and lower part of the show some of the specific deficienciesof upper limestone unit that, though not knowledge, disagreements in philoso- definitely diagnostic, were consideredto phies, and other problems that exist in be upper Mississippian in age (either connection with the stratigraphy and Ste. Genevieve or Chester or both), but paleontology of the Mississippian-Penn- he considereda collection from near the sylvanian boundaryin the Rocky Moun- top of this limestone unit at another tains. The sections given here were se- place to be clearly Pottsville (Pennsyl- lected because they are more or less typi- vanian). The writer has collections from cal of the areas or becausethey show fos- this limestone, but he has not yet been sil occurrences,lithologic features, or ter- able to study them or to re-examine minology that the writer believes are of Girty's collections. interest. The writer has personally vis- The main sandstone or quartzite, to ited the area of each section cited from which the name "Quadrant"is now gen- the various publications. Several of the erally restricted and which is probably sections have been examined in the field younger than any of the beds in Reeves's with the men who measuredthem. section, is representedonly by a few out- liers, if at all, in the Big Snowy area, but SECTION IN THE BIG SNOWY MOUNTAINS considerable thicknesses of this sand- CENTRAL MONTANA stone occur in other parts of Montana. The section in the Big Snowy Moun- In many places Jurassic beds rest on tains of central Montana (sec. i, fig. 2) is Reeves's upper thin-beddedlimestone of condensed from one published by Scott the Quadrant formation. In 1935, Scott in 1935 (p. 1024). It is given to indicate described some fifteen or more strati- changes in usage that have gained wide graphic sections in Montana, including acceptancein a part of the areawhere the the one here cited in the Big Snowy nonresistant alternating beds were for a Mountains. Like Reeves, he divided the long time placed in the Quadrantforma- Quadrant into four units in the Big tion. The locality is "on north flank of Snowies, but his units did not coincide Big Snowy Mountains, sec. 6, T. 12 N., precisely with those of Reeves. For his R. 20 E." two lower units he brought in the names Reeves in 1931 (p. 140) recognized four Kibbey and Otter, names used by units in the Quadrant formation which Weed for members of the Quadrant in he described as follows: (i) an upper the Little Belt Mountains. To a unit thin-beddedfossiliferous limestone inter- composed in large part of "black petro- bedded with red shale, ioo-200 feet liferous shales and sandstones"immedi- thick; (2) a unit of red, brown, and black ately below the upper limestone unit of shales and cross-beddedsandstones, 300- Reeves, he gave the name Heath for- 400 feet thick; (3) a 5oo-foot unit of mation. For the upper limestone unit he variegated calcareousshale, with a few brought in Darton's (1904, p. 396) Ams- thin limestones and including a pre- den from central-northern Wyoming. dominantly green shale near the middle; Scott consideredthe Kibbey, Otter, and boundary

Mississippian-Pennsylvanian the crossing sections

stratigraphic

2.-Selected

FIG. BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 333

Heath to be of formationalrank and to den to be both upper Mississippianand belong to a group to which he applied the lower Pennsylvanian. The overlap, ac- name Big Snowy group. Evidence from cording to them, would then have oc- several classes of fossils publishedby dif- curred in upper Mississippian time. ferent investigators has confirmed the Thom and others have mentioned some Upper Mississippian age of the Big evidence of physical unconformityat the Snowy group and has suggested correla- base of the Amsden. Perry and Sloss also tions of specific units with other units in describe an unconformityat the base of the Mid-Continentregion. The age of the the Big Snowy group, with sandstonesof Amsden of Scott of this region is not, the Kibbey formation filling channels however, firmly established. As stated and solution cavities some 300 feet be- earlier, Girty identified some of the low the top of the Madison limestone. faunules from it as upper Mississippian On the other hand, Scott (1935), Perry and others as Pennsylvanian. Scott in (1937), Pardee (I937), and others suggest 1935 (p. 1032) considered the unit that an unconformityabove the Amsden for- he called Amsden in the Big Snowy mation in Montana. Mountains to be of Mississippian age. The three units that make up Scott's Later in 1945 (1945a, p. 1195; 1945b, p. Big Snowy group have been mapped as 1196), he assigned the Amsden of central formationsin several places in Montana Montana to the Pennsylvaniannot only and have been recognizedin subsurface because, presumably,of a big overlap at as far east as the Dakotas and northward its base that he had recognized in 1935 into Canada.The variations in thickness but also because,mainly, of the presence and paleogeography of these units are of certain speciesof Millerellain it. Other shown on maps by Perry and Sloss (i943). microfossils occur in the Amsden, but The Charles formation, proposed by those so far identified are not definitely Seager as part of the Big Snowy group diagnostic. Though Millerella was for a below the Kibbey, has not yet gained time thought to be only of lowest Penn- wide recognition in surface work. The sylvanian age, it now is known to have a validity of Scott's use of the Kibbey, long range, which includes Mississippian Heath, and Otter as formations would as well as late Pennsylvanian, and some seem to the writerto be establishedeither paleontologists maintain that species of on the surface or in subsurfacein those Millerellaare not yet safe zone markers. areas where each has sufficientthickness Perry and Sloss in 1943 (p. 1293) also and distinction to qualify it as a map- described an overlap at the base of the pable lithologic unit on the scales or- Amsden, but they considered the Ams- dinarily used for topographicquadrangle

Explanation of Figure 2 The sections of figure 2 were selected to show interesting fossil occurrences, lithologic features, and strati- graphic terminology. Many of them are condensed in order that they may be shown together on one figure. The references are given in connection with the discussions of the sections on pp. 331-340. Only small parts of the lowest and uppermost formations in each of the sections are shown. The following are the names for which the abbreviations in the respective sections stand: I: B.S., Big Snowy group; Q?, Quadrant?; T?, Tensleep?; A, Amsden; H, Heath; O, Otter; K, Kibbey; M, Madison. 2: Q, Quadrant; A, Amsden; M, Madison. 3: T, Tensleep; A, Amsden; M, Madison. 4: T, Tensleep; A, Amsden; B, Brazer. 5: T, Tensleep; Ch, Chester; S, Sacajawea; M, Madison. 6: We, Wells; Un, unnamed beds; B, Brazer. 7: W, Weber; U. Mo., Upper Morgan; L. Mo, Lower Morgan; M.B.S., Mississippian black shale; H.B., Humbug. 8: W., Weber; U. Mo., Upper Morgan; L. Mo., Lower Morgan; M.B.S., Mississippian black shale; M, Mississippian. 9: Oq., Oquirrh; M.C., Manning Canyon; G.B., Great Blue. 334 JAMES STEELE WILLIAMS work in the area in which it occurs. In The location is near Sappington, south- other places these units would, if recog- western Montana, in Sappington Can- nizable, be reduced to member or bed yon, in Section 25, T. i N., R. 2 W., status. In this last alternative the ques- about 14 miles southwest of Three Forks tion arises as to what formation they and about 35-40 miles southwest of the would be referred as members-an en- Big Snowy Mountains. The formations larged Amsden formation, a Big Snowy shown are the Madison,the Amsden,and formation, or an entirely new formation. the Quadrant.A limestone brecciaat the Terminology used in sections measured base of the Amsden formation suggests by Gardner, Hendricks, Hadley, and an unconformity.L. L. Sloss has identi- Rogers (1945) of the United States Geo- fied Mississippianfossils, including Dia- logical Survey, suggests that the Kibbey, phragmus,from the lowest thick lime- Otter, and Heath lose their identity, at stone unit shown in the Amsden of the least as formations, in short distances. section and Pennsylvanianfossils from a This seems so because in some areas that series of siltstones, 50-100 feet above the are near the stratigraphic sections in top of the limestone that bears Mississip- which these authorsrecognize Scott's for- pian fossils. The top of the Amsden for- mations-Kibbey, Otter, and Heath- mation is here drawn at the top of the the same authors place all the beds be- nonresistant beds, and a considerable tween the Madison and the Quadrantin thickness of alternating sandstones and the Amsden. The Amsden as thus used dolomites is included in the Amsden. may or may not contain equivalents of The uniformityin lithology of the Quad- the Kibbey, Otter, and Heath. Some rant seems to have been the decidingfac- would maintain that these three units tor in delineatingit as a formationand in have pinched out and been overlapped distinguishingit from the Amsden,which by the Amsden. Another interpretation is recognizedby its heterogeneityin com- would be that the Kibbey, Otter, and position and nonresistantcharacter. Heath in some areas contain lenticular beds that change laterally so that the SECTION IN THE BIG HORN formations soon lose their identity in MOUNTAINS, WYOMING these areas. The apparent increase in Section 3 of figure 2, from the Big thickness of the Amsden in places where Horn Mountains, is condensedfrom one the formations of the Big Snowy group measuredby Darton (1906, p. 5) in the are not recognizableand the Amsden is canyon of Little Tongue River, Dayton the only formationbetween the Madison quadrangle,Wyoming. This section was and the Quadrant would favor the last selected because it is only about 6 miles interpretation. Unfortunately, fossils so from the type locality of the Amsdenfor- far obtained are inadequate to solve this mation. The type locality is given by problem. Darton (1904, p. 396), who defined the formations as the Amsden Branch of SECTION IN SOUTHWESTERN MONTANA Tongue River, about 5 miles southwestof Section 2 of figure 2 is a section taken Dayton, in the Dayton quadrangle,Wy- from those published by Gardner, oming. As shown in the figure, some of Hendricks, Hadley, Rogers, and Sloss the thicknesses are slightly exaggerated. (1946), mentioned above. The section is The total thicknessof the Amsden is 190 condensed for graphic representation. feet. The terminology, which is that of BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 335

Darton, is widely used in this region at fossils have been reported from the Ams- the present time. Judging from C. C. den of the Big Horn Mountains from Branson's(1939, pp. 1202-1213) usage in several localities. The writer has col- sections in the Big Horn Mountains, lected them from other localities in the probablysome of the lower beds of Dar- Big Horns. The presence of fossils of ton's section, possibly those up to the probable Carboniferous age from the base of the 12-foot sandstone, would be upper part of the Amsden was mentioned included by him in his Sacajawea, or in by Darton (1904) in his original descrip- other beds referredby him to the Missis- tion of the formation. Pennsylvanian fos- sippian. The typical area of Branson's sils from the Amsden have since been Sacajawea formation, discussed later in listed in many United States Geological this paper, is in the Wind River Moun- Survey papers and in other publications. tains. To the writer's knowledge, Bran- Some of these fossil lists were brought son has not cited faunal evidence for ex- together by C. C. Branson in 1939. Fos- tending it into the Big Horn Mountains, sils are not abundant in the lower part of so presumably the extension was based the Amsden of the Big Horn Mountains; mainly on lithology. No specificlithologic but, in 1906, Darton (p. 5) mentioned the features,however, have been given to ex- occurrence of a coral identified by Girty plain the extension or to furnish criteria as Menophyllum excavatum Girty and for the lithologic differentiation of the fragments of a Spirifer and a "Zaphren- Sacajaweafrom overlying beds. tis" from a limestone bed in the lower In the Big Horn Mountains, the Ams- part of the Amsden near Soldier Creek, den is a variableformation, both in thick- some 15-20 miles southwest of the type ness and in lithology. Because of the locality. Girty believed that these sug- thickeningand thinning of many beds in gested Mississippian age for the lower short distances, one suspects that at Amsden; and the Amsden in this general some stratigraphiczones they are lentic- area has for a long time been generally ular. In several places the basal bed is a considered to contain both Mississippian sandstone,which may attain a thickness and Pennsylvanian beds. The writer does of as much as 1oo feet or more, but sand- not consider these fossils of themselves stones occur at several stratigraphicpo- adequate to establish a Mississippian age sitions in the Amsdenof most areas. In a for the lower part of the Amsden; but the section measuredby the writer in 1920, absence of definite Pennsylvanian fossils along Little Goose Creek,about 25 miles near the base of the Amsden in the Big southeast of the type locality, a thin- Horn Mountains and the presence of bedded sandy limestone that is dense to Mississippian fossils in the alternating finely crystalline and has a purplishcast series of beds between the Madison lime- was considered the basal bed of the stone and the Quadrant (Tensleep) for- Amsden. mation north and west of the Big Horns The contact with the underlyingMad- in Montana and south and west of them ison limestone in this region in many in central Wyoming strengthen consid- places is irregular,and shales from the erably the meager fossil evidence for a Amsden fill depressions some of which Mississippian age of the lower Amsden in are probably sinkholes in the Madison. the Big Horns. If definite stratigraphic The writerknows of no fossils from the tracing could show that the lowest beds type locality of the Amsden, but in print of the Amsden of the Big Horns were 336 JAMES STEELE WILLIAMS younger than the youngest near-by beds lithologic criteria to those used elsewhere with Mississippian fossils, it would be an for the base of the Amsden, as there are argument for the absence of Mississip- several feet of nonresistant alternating pian beds in the Big Horn Mountains, shaly beds below the Darwin, or so-called but this does not appear to be possible "Darwin," at many places. The Amsden because of the areas of younger rocks be- in many different places in the Big Horn tween the mountain ranges. From time Mountains and Big Horn Basin and at to time, verbal reports are made of the other places in Wyoming has been de- discovery of Pennsylvanian fossils in the scribed and mapped by many geologists. basal Amsden of the Big Horns, but the Most geologists have considered it in writer does not know of any published part Mississippian and in part Pennsyl- record. Collections from the Amsden vanian, but in local areas it has been re- made in conjunction with field work in ferred entirely to the Pennsylvanian, and the Big Horns and mountains west of the in a few areas, notably in the Wind River Big Horn Basin by W. G. Pierce and Mountains, it has been referred by a few D. A. Andrews and the writer have not geologists entirely to the Mississippian. yet been carefully studied. Those who have placed the Amsden In the Big Horn Mountains the sand- wholly in either the Mississippian or the stone near the base of the Amsden has Pennsylvanian have not cited definite been identified by some geologists as the fossil evidence from critical beds. Darwin sandstone member, whose type locality is in the Gros Ventre Moun- SECTION IN THE GROS VENTRE MOUNTAINS, tains. This identification appears to be WYOMING based on lithologic similarity and similar The section selected from the Gros position in the stratigraphic succession. Ventre area (sec. 4 in fig. 2) was meas- Some geologists would place the Missis- ured by Wanless and Bachrach and pub- sippian-Pennsylvanian boundary at the lished in 1947 by Helen L. Foster (p. base of the Darwin sandstone, not only 1557). The locality is north of Sheep in the Gros Ventres, but in other areas Creek in the N.E. a, Section 10, T. 42 N., where they believe the Darwin can be R. 115 W., Teton County, Wyoming. recognized. Such an interpretation would This locality is near the type area of the not give a boundary closely tied to faunal Darwin sandstone. Miss Foster desig- data, as diagnostic fossils are not known nates the beds above the Darwin as from beds near this boundary. The prac- upper Amsden; those below it as lower tical advantages in mapping a unit Amsden. The Darwin sandstone in this boundary at the base of a prominent section is given as 97 feet thick, but its sandstone where that sandstone can be thickness varies in the surrounding re- definitely identified are easily under- gion. The Amsden formation was ex- stood; but to assume without faunal data amined briefly by the writer last summer that such a boundary is a period or epoch at one locality in the Gros Ventres. The boundary seems to the writer to be unjus- beds below the Darwin, in the lower part tified, even though an unconformity of the Amsden, consist of red shales, might be present. To use the base of the gray, pink, and lavender fine-grained Darwin sandstone unit as the base of the limestones, and beds of chert. No fossil Amsden, even if that sandstone can be names were listed by Miss Foster, but definitely identified from section to sec- she states that her "lower Amsden" con- tion, would add still another type of tains fossils resembling both Mississip- BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 337 plan and Pennsylvanian types. She be- minology is taken from a 1939 publica- lieves that the Mississippian-Pennsyl- tion (pp. 1209-1210) showing the same vanian boundary is near the base of the beds. The locality is Bull Lake Creek. In Darwin. In connection with studies in his 1937 paper, Branson proposed to split this area by Eliot Blackwelder, C. W. the sequence that had previously been Tomlinson, and other geologists of the generally called Amsden into three units. United States Geological Survey in 1911 The lowest of his three units consisted of and in near-by areas in subsequent years, 43 feet of cherty limestone underlain by Dr. G. H. Girty examined many collec- 2- I feet of "red and buff sandstone and tions of fossils from beds referred to the shaly sandstone, breccia in places, shale Amsden. Lists from two different strati- cave filling in places." The 43-foot lime- graphic zones are published in Black- stone is the zone of the invertebrate welder's paper (1913, p. 176). The list for fauna described by E. B. Branson and the uppermost zone, which is said to be D. K. Greger in 1918 (pp. 309-326) as of from a thin group of limestone beds a Ste. Genevieve age. For this lower unit of little below the middle of the formation, limestone and underlying red rocks Carl shows the zone clearly to be Pennsyl- Branson proposed formation rank and vanian. The fossils from a zone 60 feet gave them the name "Sacajawea forma- lower are not so diagnostic but appear to tion." One of the reasons for giving these the writer to be also Pennsylvanian. beds formation rank appears to have Girty, however, reserves the possibility been that they were Mississippian, that they may be Mississippian. There whereas the upper beds of the Amsden has not been an opportunity for the were Pennsylvanian. C. C. Branson con- writer to re-examine the actual collec- sidered the Sacajawea formation to range tions or to study Blackwelder's field possibly from Salem to Ste. Genevieve notes. Neither collection is located strat- age. He thus considered it pre-Chester. igraphically with respect to the Darwin Beds that Branson probably would refer sandstone, but it is probable that both to the Sacajawea have been examined at came from beds above it. This cannot, several localities by the writer, but they however, be definitely stated. The United contain very few fossils or none at all. States Geological Survey has had a field Aside from the faunas described by E. B. party working in the Gros Ventres dur- Branson and Greger and by C. C. Bran- ing the past season, and several univer- son in his 1937 paper, ostracodes have sities have had students and faculty been described by Morey (1935, pp. 474- members working there in recent sum- 482) and by Croneis and Funkhauser mers. It is to be hoped that the examina- (1938, pp. 331-360). These last two con- tion of fossils obtained by these investi- sider the ostracodes examined by them to gators will soon supply some definite be Chester. faunal data on the age of the Darwin Above the Sacajawea formation is a sandstone member of the Amsden. series of limestones about 60 feet thick that C. C. Branson believes is Mississip- SECTION IN THE WIND RIVER pian, probably Chester, but few fossils MOUNTAINS, WYOMING have been found in it or, if found, have Section 5, figure 2, shows relationships not been reported in publications. Above of the strata in the Wind River Moun- the limestone beds is a sandstone 80 feet tains. The section was described by C. C. thick that has been identified by some Branson in 1937 (p. 651), but the ter- geologists as the Darwin sandstone. The 338 JAMES STEELE WILLIAMS

Mississippian-Pennsylvanian boundary and the Wells formations.The many dif- is drawn at the base of this sandstoneby ficulties accompanying the use of this some investigators; but, as elsewhere, boundaryas a mapping boundaryin the thereis little directpaleontologic evidence field led Rubey to erect a new mapping for its precise location at the base of the unit that includes the nonresistant and Darwin. Pennsylvanianfossils have been alternating thin beds that occur in both collected from a limestone some 80 feet the lowerpart of the Wells and the upper above the top of the sandstone. Other part of the Brazer. A silty sandstone in geologists have rejected Branson's pro- the lower part of this unit may be the posed Sacajawea formation on the Darwin sandstone, or a lenticular sand- grounds that in actual field practice it is stone other than the Darwin. Mississip- not a mappableunit over an appreciable pian fossils, including Diaphragmus area. They have mapped the entire se- elegans (Norwood and Pratten) n. var., quence of variable nonresistantbeds be- Camarophoriacf. C. explanata(McChes- tween the Madison and Tensleep forma- ney), and Linoproductusovatus (Hall), tions as the Amsden formationand have were collected from a zone in a massive not ventured to indicate the preciseloca- limestone about 400 feet above the base tion of the Mississippian-Pennsylvanian of the nonresistant unit, and Chaetetes boundary within the Amsden. Some of milleporaceusMilne-Edwards and Haime these geologistshave recognizedthe Dar- and other Pennsylvanian fossils, includ- win as a member of the Amsden. C. C. ing a new species of Orthotetinathat ap- Branson, in 1939, proposed dropping the pears to be representedin the western name Amsden. He would include in United States only in Pennsylvanianor the Tensleep formation all those beds Permian rocks, from the beds immedi- called. Amsden by other investigators ately above the massive limestones. that are above the limestone beds re- Specimensof a species of Lithostrotionella ferred by him to Chester? age. Geolo- occur at the same zone as the D. elegans gists generally have not followed this mentionedabove. The lists of fossilsiden- suggestion, but it is used in figure I to tified in i931 and 1932 from this section illustrate various types of usage. The will be publishedin full in W. W. Rubey's lower beds of the Amsden, as identified report. by most geologists, would be referredby Branson, as before stated, either to his MORGAN LIMESTONE AREA Sacajaweaformation or to the unnamed The Morganformation is typically ex- unit that he believes may be of Chester posed in WeberCanyon near the town of age. Morgan, Utah, where, in 1935, the writer measured a section including the Mor- SECTION IN SOUTHWESTERN WYOMING gan, Brazer, and Weber formations Section 6, figure 2, measuredby W. W. (1936). The section has not been pub- Rubey and the writer along the Covey lished because it seemed preferable to cutoff trail in the Salt River Mountains, await detailed mapping to determinethe near Afton, Wyoming, has been selected extent and nature of the probablefault- to show features present in that part of ing. Girty had measured a section there southeastern Idaho and southwestern some years before. The fossils collected Wyomingwhere the Mississippian-Penn- during both investigations are being sylvanian boundary has been drawn studied, but the studies have not been along the boundary between the Brazer completed.A casual inspectionof the col- BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 339 lections definitely confirms the state- upper and a lower unit. The contact of ment of Blackwelderthat the Morganis, the formationwith the overlying Weber in part, of Pennsylvanianage. It has not is drawn at the place where the sand- been definitely determinedwhether Mis- stones cease to be dominantly red. This sissippianbeds occur in the lower part of coincides generally, but not precisely, the Morgan.If the lowerboundary of the with the beginning of the massive, Morgan is drawn on the basis of fossils coarser sandstones and the termination and its age stated as Pennsylvanian, of soft silty and shaly sandstones. Red then, of course,Mississippian beds would and purplesandstones and shales and red be excluded from it. If, however, it is cherts occur in the dominantly nonre- drawn on a lithologic basis that includes sistant lower part of the Morgan. The all beds in the nonresistant unit, then Mississippian-Pennsylvanian boundary Mississippianbeds might be present. is placed at the top of a black shale from Since the writer's section was meas- which the writer has identifiedfossils as ured, A. J. Eardley (i944) has mappedin Mississippian. Fossils from the lower this area, and J. Stewart Williams (1943, part of the Morgan, above the black p. 607) has publisheda descriptionof the shale, that have so far been studied by Morganat the type area, wherehe found the writer are either too incompleteor of it to be i,o60 feet thick. types too generalizedand long rangingto The Morgan formation has been ex- indicate a definiteage determination,but tended south and east from the type they suggest Pennsylvanian age. area. Calkins and Butler (1943, p. 28) re- ferredbeds in the Cottonwood-American SECTION NEAR VERNAL, UTAH Fork area, Utah, to the Morgan(?). Section 8, figure 2, near Vernal, Utah, Others who have recognizedthe Morgan is condensedand generalizedfrom a sec- outside the type area include J. Stewart tion, mostly along Ashley Creek, meas- Williams (i943); Thomas, McCann, and ured by D. M. Kinney and J. F. Ro- Raman (1945); Huddle and McCann minger (1947) in the Whiterocks River- (1947a); McCann, Raman, and Henbest Ashley Creek area on the south flank of (1946); K. G. Brill, Jr. (1944); and Kin- the Uinta Mountains, Utah. ney and Rominger (1947). As in the Tabiona regionfarther west, The beds identifiedas Morgan by the the Mississippian-Pennsylvanianbound- above geologists vary considerably in ary is placed at the top of a black shale, lithology, and one might well ask whether but the black shale is not so well exposed it is advisable to carry the name Mor- as in the Tabiona region. The Morgan is gan so far afield. Thompson (1945, p. divided into two parts, both of which are 31) has applied a new name, Hells tentatively consideredto be of Pennsyl- Canyon formation,to beds in the Uintas vanian age. The collections need to be that may be of Morganage. Possibly his carefully studied, however, as they con- Younghall formation is equivalent to tain many forms that are generalized. some part of the Morgan. The lower part of the Morgan is mostly limestone and is more resistant than at SECTION NEAR TABIONA, UTAH other localities where shales and sand- Section 7, figure 2, is taken from a stones are intercalatedat short intervals. paper by Huddle and McCann (1947b). The upper part consists of three sub- At this locality the Morgan formationis divisions, the lowest of which is mainly divided on lithologic grounds into an soft red shale, sandstone, or sandy shale. 340 JAMES STEELE WILLIAMS

It probably correspondsapproximately about 350 feet higher. These are listed in to the upper unit of the Morgan in the Gilluly's report. area near Tabiona. The middle subdivi- The Manning Canyon shale has been sion of the upper part of the Morgan is recognized at various places in north- mainly hard buff to red sandstones, and central Utah. A section showing the the upper subdivision is tan limy sand- Manning Canyon shale in the Wasatch stones and gray cherty limestones.It ap- Mountains area near Provo is given by pears to the writer that the upper and Baker (1947). Nolan (1935, P. 31) de- middle subdivisionsprobably correspond scribedthe Manning Canyon of the Gold to at least part of the beds consideredto Hill area, central western Utah. Bissell be Weber in the Tabiona region. The and Hansen in 1935 (p. 163) discussed upper contact of the Morganis drawn at briefly the gradational character of the the top of the highest limestone bed. Mississippian-Pennsylvaniancontact in Stratigraphic sections here and in Spring Creek Canyon, east of Spring- near-by parts of the Uintas have been ville, Utah. made at or near the same localities as those already discussedby several geolo- PROBLEMS IN PALEONTOLOGY gists. The informationin these separate NEED FOR MORE PALEONTOLOGICALDATA investigations needs to be better inte- The need for more paleontologicdata grated than it is at present, and work is is definitely shown in the precedingdis- now in progressto that end. cussions of the few selected stratigraphic sections. Not only is the need for addi- SECTION IN STOCKTON-FAIRFIELD tional and largercollections from certain AREA, UTAH zones indicated,but there is also shown a Section 9 of figure 2, condensed from need for more studies and better integra- one given by Gilluly (1932, p. 31) from tion of collections already made. The Soldier Canyon near Stockton, was se- problem of additional collections from lected to show features existing in an specific zones in critical areas is not ev- area where the Mississippian-Pennsyl- erywhere easily solved. In many moun- vanian boundary is within the Manning tain ranges the Carboniferousrocks are Canyon shale. exposed mainly in areas of high altitude Both the upper and the lower contacts that are difficultof access. Furthermore, of the Manning Canyon shale are grada- many of the beds are either unfossilifer- tional. Gilluly places the upper contact, ous or contain very few fossils. Some with the Oquirrhformation, at the point beds that have fossils do not yield them where the limestones start to become readily, and it is difficultto obtain speci- more abundant than shales. The lower mens that are well enough preserved to contact is drawn where shales become permit definite identifications. When more abundant than limestones. The closely related genera and species differ Manning Canyon, in contrast to other from one another only in some small in- formationsfarther north that include the ternal character or in some particular Mississippian-Pennsylvanianboundary, type of ornamentation,as many do, large is almost devoid of red beds. Mississip- collections are frequently required to pian (Chester) fossils occur about 500 provide specimens to show adequately feet above the base of the Manning these characters. Canyon, and Pennsylvanianfossils occur If one could everywheredetermine the BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 341 age of a formation or parts of a formation ous, certain individuals of the Pennsyl- by making collections from a few locali- vanian Spirifer occidentalis-rockymon- ties known to provide good material, the tanus group so closely resemble individ- problem would be simplified. Collections uals of the Mississippian S. increbescens- definitely located geographically and keokuk group, as identified in the West, stratigraphically from many widespread that one is forced to the conclusion that, areas are, however, needed to determine if these species have been correctly iden- the age ranges of formations. The deter- tified, they either overlap or are not dis- mination of the stratigraphic positions of tinct species. Studies of western repre- collections in the Rocky Mountain region sentatives of this group based on large, is frequently a problem because of fault- definitely located collections need to be ing and folding. Especially are so-called made. The nomenclature should reflect "bedding-plane faults" likely to be as nearly as possible the actual relation- missed and duplications or omissions of ships, and if in any instance it is desirable strata unnoticed. Stratigraphic sections to combine two or more groups consid- and the fossil collections from these sec- ered species into one, then that combina- tions, in the Rocky Mountain and Great tion should be made. This would serve Basin areas especially, are, as a rule, on the cause of stratigraphic paleontology much firmer ground if the stratigraphic more than maintaining fictional differ- work is done in connection with rather ences would. Progress in paleontology, it detailed mapping. is true, is made by finer and finer sub- divisions, if they are truly useful, but SPECIES DIFFERENTIATION progress is also made by combining so- Studies in systematic paleontology called "species." It is frequently true that deal with the relationships of the that, when only a few specimens are species of several genera need to be known, several apparent species can be made. These should be made by paleon- distinguished, whereas in larger collec- tologists who have available large num- tions species lines disappear and fewer bers of specimens, stratigraphically and species are recognizable. geographically well located. Only by the Similar investigations need to be made study of large collections can differences of Mississippian and Pennsylvanian spe- between two individuals be correctly cies of the genera Composita, Chonetes, evaluated, that is, whether they are in- Lithostrotionella, Reticulariina, Linopro- dividual, varietal, specific, subgeneric, or ductus, and others. generic. Extensive experience in applying USE OF FOSSILS IN CORRELATIONS zo6logical concepts in classification to closely related forms may substitute in One of the most important problems in part for the lack of large collections. The the use of fossils for correlation in the mere presence of some differences from a Carboniferous of the Rocky Mountain type specimen of a species does not, as region has just been discussed. More all paleontologists know, constitute rea- needs to be known about the ranges in sons for specific, or even varietal, differ- variation within certain common species entiation. The ranges of variation within and about effective criteria for distin- several species of the Rocky Mountain guishing these species, if they are dis- area need to be determined. tinct. Another problem that has also been In the Rocky Mountain Carbonifer- mentioned lies in the difficulty of obtain- 342 JAMES STEELE WILLIAMS

ing large collections of well-preserved writer's experience,include Lithostrotion materialsat criticallocalities and at criti- (or Lithostrotionella),Caninia, Leptaena cal stratigraphic zones. Criticism has analoga (Phillips), Leiorhynchus car- been made of lists of fossils that contain boniferum Girty, Reticulariina spinosa many question marks and provisional (Norwood and Pratten), and Millerella. identifications, but species cannot be Few paleontologistshave ever consid- positively identifiedwhen the diagnostic ered correlationsin the Rocky Mountain characters are not preserved. Correla- Carboniferousthat were based solely on tions can rarely be positively made if one or two index fossils to be more than species or genera cannot be definitely temporary,nor have they postulated the identified. presenceor absenceof beds or zones on a Not all species in the Mississippian similar basis. In using index fossils, the and Pennsylvanian rocks in the Rocky possibility always exists that the ranges Mountain regionare long rangingor gen- of these fossilsmay be extended, and it is eralized. Some are distinct and are of preferableto use several types of fossil definite stratigraphic value, along with evidence, where possible, for every cor- other evidence, in the correlationof beds relation. Whole faunas are decidedly and zones. They have been called "index more useful in making correlationsthan fossils,"2but the writer hesitates to use are conclusions based on several index this term because too many fossils are fossils or on fossils of any one class. index fossils only as long as relatively The province of this paper does not little is known about them. Since the permit an exhaustive discussion of the writerfirst began paleontologicalwork in varioususes of fossils in correlationor the the West, the range of many so-called theories behind these uses. Paleontolo- "index fossils" has been extended so that gists working on late Paleozoic rocks in they can no longer be used in the re- the Rocky Mountains are generally stricted sense of the term. One of these is aware of the many complex problemsin Archimedes.For many years it was con- any correlation and are usually on the sidered an index of certain zones in the alert for influences of less obvious fac- Mississippianin this country;but in 1927 tors. For instance, in parts of the Rocky Girty and Gilluly (1932) discovered it Mountain region certain types or classes above Chonetes mesolobus (Mesolobus of fossils are relatively common in spe- mesolobus of recent authors) in the cific types of lithology and nearly absent Oquirrh Mountains, Utah; and it has in others-to cite one example, the large since been discovered in many places in horn corals of the crystalline limestones Pennsylvanian rocks in the West and in of the Brazer. The possibility is always the Permian in Russia. Other fossils, to kept in mind that these corals are absent mention a few, that have lost their early at certain stratigraphiczones merely be- meaning as index fossils, within the cause of ecological conditions (to be dis- 2 An "index fossil" is generally defined as one cussed more fully later). Another ex- that is characteristic of a specified time unit, more ample: certain types or classes of fossils, or less irrespective of facies, whereas a "facies fossil" is defined as being characteristic of a certain facies such as crinoids and cephalopods, are and crossing one or more time boundaries. In the relatively rare, and it is possible that larger aspect, of course, all fossils are index fossils their assumedexcellence as zone markers and all fossils are probably facies fossils, but most geologists define and use the terms in the more is enhanced by their rareness. Still an- restricted sense. other example: several Mid-Continent BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 343 species have been identified in the Rocky 433) published a notice of a fauna from Mountains, and it is possible that some beds that he assigned to the Spergen of these species migrated into the Rocky limestone, which was then considered Mountains so slowly that they occur part of the St. Louis limestone, and that there at a later time than in the Mid- are probably now included in the Madi- Continent region. The possibility is kept son limestone in Montana. The presence in mind that the absences of species from of a zone of St. Louis or Meramec age in certain zones may mean only inadequate the Mississippian rocks of the Rocky collecting; also that classes of fossils only Mountains has been mentioned in print recently recognized and as yet little by Willis (pp. 316, 324) in 1902 (St. Louis known may yield results that appear to fossils identified by Stuart Weller in the be more definite than if they were better Yakinikak limestone of Willis in north- known. The value of relative abundance western Montana); by Girty (1927, p. of specific forms in making local correla- 71), both published and unpublished; by tions has for a long time been realized Sloss and Laird in 1945; and by several and utilized in the Western states. others. The writer and others have recog- Not only do paleontologists use com- nized it in the field at several places parisons of total faunas and combina- where its presence has not been men- tions of index fossils for correlations, but tioned in published accounts. Other zones the reported first appearances of new in the Mississippian that correspond to forms and the reported extinctions of old pre-Chester Mississippian zones in the forms are sometimes given special con- Mid-Continent area have been recog- sideration. Certain evolutionary stages nized in local areas. in specific species and especially genera, In 1927, in a general summary of the if adequately tested, and certain trends Mississippian rocks of southeastern Ida- in evolution, if firmly established, are ho, Girty (p. 71) recognized affinities of also used. An example of this last is the different faunules from the Brazer with tendency for coarsely plicated Spirifers of those from the Spergen, St. Louis, and the rockymontanustype to develop prom- Chester strata. He realized, however, inently in the middle Mississippian and that the Spergen fauna may be a facies to die out at the top of the lower Penn- fauna rather than one that can be defi- sylvanian (Des Moines equivalent). It is nitely correlated with the typical Sper- realized that whole faunas may trans- gen. The writer in 1935 and also in later gress time lines, but such transgressions years measured and collected from many are relatively unimportant in correla- stratigraphic sections in an effort to fix tions within local areas. these and possibly other zones in the Brazer, as well as to determine possible ZONATION zones in the Wells and Phosphoria for- For many years rocks of Chester age mations. J. Stewart Williams and J. S. have been generally separated from rocks Yolton in 1945 recognized five zones in of older Mississippian age throughout the the Brazer limestone as exposed at Dry northern and central Rocky Mountain Lake, Utah. The lowest zone was cor- area, and certain Mid-Continent zones related by them with the Warsaw forma- have been recognized within the pre- tion and the next highest zone with Chester Mississippian rocks in several higher beds in the Meramec group, pos- local areas. As early as 1873, Meek (p. sibly with the St. Louis limestone. Their 344 JAMES STEELE WILLIAMS third highest zone is considered by them Girty in several written opinions, and he to be transitional between the Meramec has identified, more or less provisionally, and Chester groups, and the two upper beds of Pottsville and post-Pottsville zones are Chester. Most of the Chester age. The writer does not believe that forms cited by them are generalized these correlations, when based on in- forms that have not only a considerable vertebrate megafossils, were very re- stratigraphic range in the Chester but liable as to detail, and Girty appears not occur or are represented by very closely to have considered them very reliable. In related forms both in younger and in 1934, C. B. Read described a Pottsville older rocks. One exception, however, is flora, probably of middle Pottsville age, Diaphragmus elegans (Norwood and from beds exposed near Leadville, Colo- Pratten). The proposed correlation of rado, from which Girty had described a Branson's Sacajawea formation (lower macrofauna that he believed represented Amsden of most authors) with Mid- a zone that was "very early in Pennsyl- Continent beds and of the units of the vanian time and probably older than any Big Snowy group with specific units in beds of the Kansas and Nebraska sec- the Chester of the Mid-Continent has tion." Read regarded this flora as older been discussed before. than the Glen Eyrie flora. On the Pennsylvanian side of the Mis- During the last five or six years, beds sissippian-Pennsylvanian boundary, a thought to be immediately above the zone of post-Morrow-pre-Kansas City Mississippian-Pennsylvanian boundary age has been recognized widely for a long have been referred to a zone that is con- time. The zone was first recognized and sidered the equivalent of the Morrow of identified widely by megafaunas (Wil- the Mid-Continent. Most of these refer- liams, in Moore, 1944); but fusulinids at- ences have been based mainly on fusuli- testing to its existence were also identi- nid evidence. In general, the larger in- fied by Girty prior to 1927. During recent vertebrate fossils, where they are pres- years fusulinids from the Quadrant for- ent, do not give adequate evidence-if, mation of Montana, the upper part of in fact, any positive evidence at all-for the Amsden and Tensleep of Wyoming, distinguishing the so-called "Morrow the Wells formation of Idaho and Utah, zone" from the zones bearing both Des the Oquirrh of Utah, and other forma- Moines-Lampasas macrofossils and tions in neighboring states have supple- microfossils. This does not necessarily mented and confirmed the evidence from mean, however, that larger collections the larger invertebrates. Although some may not provide material whereby such efforts have been made to divide this a zone can be recognized by larger fossils. zone into two-a lower Lampasas and an L. G. Henbest (1946) seems to have upper Des Moines-it has not proved been the first actually to designate beds feasible on the basis of the megafossils. in the area under discussion as of Mor- Some fusulinid workers also have not row age, but the presence of very early found such a division practicable. Pennsylvanian beds in near-by areas had Correlation of Lower Pennsylvanian been noted before by C. B. Read, G. H. beds with the Pennsylvanian formations Girty, the writer (1944, P. 700), and of the Appalachian region rather than others, and M. L. Thompson (1936) had directly with the formations of the Mid- previously suggested the presence of beds Continent areas was attempted by G. H. of Morrow age in the Black Hills. Paleon- BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 345 tologists had also identifieda zone prin- termination on Millerella, with supple- cipally in the subsurface that they called mentary evidence from other microscopic "Morrow"in other parts of the general forms. Rocky Mountain area, namely, New Despite the recognition of beds at the Mexico and Colorado.M. L. Thompson places mentioned above and at other (1945) identified a zone correlated by places in the northern and central Rocky him with the Morrowover a wide area in Mountains, correlated with faunal zones the Uinta Mountains;Williams and Yol- in the Mid-Continent, zonation studies ton (1945, p. 1152) referred rocks in cen- useful over wide areas in the rocks near tral-northern Utah to the Morrow; the Mississippian-Pennsylvanian bound- Thomas, McCann, and Raman (1945) re- ary have not been published. One reason ferred rocks in northwestern Colorado for this deficiency is the paucity of fossils and northeastern Utah to the Morrow; and the poor preservation in many beds. and H. W. Scott (1945, p. 1195) referred Large collections of good fossils from rocksin the sequencewhich he had iden- many zones and localities are needed if tified as Amsden in central Montana to zonation is to be based on trustworthy the Morrow. grounds. Another reason is the great In assigning rocks to the Morrow, number of relatively generalized long- Henbest appears to have laid consider- ranging species. Still another reason is able stress on the occurrencein them of that faunal zones do not coincide in many the fusulinid genus Millerella,which was instances with lithologic divisions, as once thought to be restricted to rocks of they more or less do in some places. This Morrowage; but the range of Millerella reduces the general use and recognition is now generally considered to extend of faunal zones in areas where the natural from late Mississippianto late Pennsyl- mapping units are not also faunal units. vanian time. Thompson recognizes the In the Rocky Mountain area there are long range of the genus Millerella in great thicknesses of rather uniform li- Pennsylvanian rocks and the difficulty thology, especially in the Mississippian of using species of Millerella to denote rocks. This contrasts with the situation different stratigraphiczones. He identi- in the Mississippi Valley, where com- fies his zone of Millerella, which zone paratively frequent alternations in depo- name he applies to rocks of Morrowage, sition have occurred and where, conse- by the predominance of specimens of quently, the typical formations are thin- Millerella and the absence of the more ner lithologic units. The Mississippian highly developed fusulinids with which faunas in the Mississippi Valley were Millerella is commonly associated in thus subjected to more frequent changes Pennsylvanian rocks of post-Morrow of conditions than were those in the age. Williams and Yolton base their as- Rocky Mountain area. Both the Missis- signment of beds to a Morrow age mainly sippian and the Pennsylvanian rocks of on microscopicbryozoans and ostracodes the Mid-Continent were deposited in a identified by ChalmerL. Cooper. They different basin (though there were con- also list larger fossils, but nearly all the nections) than the one in which the rocks larger fossils are forms which, if identi- of the same approximate ages in the fiable, have been collected elsewhere in Rocky Mountain region were deposited. beds associatedwith fusulinidsof younger Although some species have been identi- age than Morrow.Scott bases his age de- fied as Mid-Continent species, the faunas 346 JAMES STEELE WILLIAMS in general are quite different. Most cor- been the determiningfactor in the sepa- relations of zones in the Rocky Moun- ration of one formation from another, tains with zones in the Mid-Continent, and the lithologic change at the bound- when made in detail, have appeared ary only secondary.Few geologists defi- based upon inadequate evidence. The nitely state that the basis for their for- writer has believed for a long time that mation boundariesis faunal, but those either the southeasternIdaho section or who refuse to put Mississippianbeds in one of the sections in central Utah in an the same formationwith Pennsylvanian area that has been mapped in detail and beds, simply because of this differencein from which a great many collections age, are neverthelessusing faunaldata as have been made should be selected as a their main criterion. Rocky Mountain standardand that cor- Even those who have used lithologic relationsin this regionshould be made in composition as the main factor in the terms of this standard rather than in definition of their formations have not termsof Mid-Continentstandards. There agreedon the type of lithologic criteriato seems no reason to suppose that in this use. It is probablyunlikely that absolute differentbasin the rangesof fossils would uniformityin usage will ever be attained be precisely the same as in the Mid-Con- over wide areas or in differentsections of tinent basin or basins. Broad correlations the country; and it probably is not de- can be made, but only broadones are jus- sirable.Geologists as a rule will use what- tified at the present. Several instances ever they find in the particular region are on recordof rangesof Mid-Continent that provides mappable units, and the species, and even of genera, that are dif- choice will dependnot only on individual ferent from rangesin the Mid-Continent. preferencesbut also on the characterof the topography and climate of the par- STRATIGRAPHIC PROBLEMS ticular area, the lithology of the rocks, and other factors. CRITERIA FOR FORMATIONS In work in the Rocky Mountain re- A study of the selected stratigraphic gion, the writerhas been impressedwith sections discussed on pages 331-340 shows the usefulness of lithologic formations that a considerablevariety of criteriahas based upon featuresof the total lithologic been used for the definitionand delimita- compositionof the formation more than tion of formations near the Mississip- that of formationsbased on the presence pian-Pennsylvanian boundary in the of some specific color; or on the highest northern and central Rocky Mountain or lowest occurrenceof some particular region. These criteria have given quite type of lithology, such as the highest different results concerning the content limestone or the lowest phosphate bed; and thickness of a formationin the same or on the tracing of some conspicuous or in near-by regions. Most geologists bed, such as a sandstone;or on the trac- have used lithologic criteria, but many ing of unconformities,either by faunal have combined lithologic criteria with data or by matchingunconformities seen structuralrelations, such as unconformi- in separateoutcrops. Nearly all these cri- ties, karst surfaces, or brecciation, or teria have been, or are being, used in the with faunal data. In some instances the Rocky Mountain area, and local condi- location of an unconformity or the tions might make any one of them more boundaryof a faunal zone seems to have desirable than the others. The first ap- BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 347 pearance of a changed type of lithology tions as close as are desired can be recog- especially has cogent arguments in its nized by some sort of designation. As in- support. dicated in the remarks explaining fea- Far more important than the adoption tures of the code, formations may include of any one particular type of criterion is parts of two systems or even major un- the desirability of uniform usage, at least conformities where there is a sequence of in local areas, to which, after all, for- similar beds and where in practical work mation names only apply. it is not useful to make a division. During recent years detailed strati- Much has been written about the ne- graphic work and subsurface work have cessity of distinguishing between rock greatly increased in the Rocky Mountain and time (or so-called "time") units in area. This work has added much needed stratigraphic nomenclature, and interest detailed knowledge regarding the indi- in this question has been revived lately vidual beds that make up the formations because of many examples of widespread and has resulted, in some areas, in the confusion and because a third category- breaking-up of larger formations and the time-rock units-has been proposed. It is giving of formational rank to units that not appropriate to discuss this question constituted subdivisions or merely unrec- here, but a satisfactory understanding of ognized parts of the larger formations. these differences, if agreed upon, would In the writer's opinion, new and thinner aid in achieving more uniformity in usage formations are desirable if, and wherever, in some parts of the Rocky Mountains. they conform to the generally accepted criterion that a formation "shall . . . UNCONFORMITIES meet the practical and scientific needs of Unconformities exist between and the users of geologic maps" (Ashley et al., within several formations that occur 1933, p. 431). As stated in the remarks ac- near the Mississippian-Pennsylvanian companying the stratigraphic code just boundary in the northern and central quoted, "practicability of mapping is Rocky Mountains. Many of these are usually an essential feature" of a forma- discussed in connection with the different tion. It is only "under exceptional condi- stratigraphic sections, to which the tions" that this criterion should be reader is referred, but additional data on waived. The test of the validity of new many of the unconformities are in publi- formations proposed for use in rocks of cations from which no sections were se- Carboniferous age in the Rocky Moun- lected or in the writer's files and notes. tain area, whether surface or subsurface Williams and Yolton (1945, p. II50) formations, would, in the writer's opin- show a widespread unconformity at the ion, lie in the decision as to whether or base of the Meramec division. Uncon- not they could reasonably be supposed to formities in local areas at this strati- form practical mapping units, under the graphic zone have been mentioned by conditions existing in the region at the others. A widespread unconformity oc- present time, if exposed over a consider- curs at the top of the Madison limestone able area. The validity of any surface for- over a wide area. The writer believes that mation can be tested by mapping. Ample this is mainly a pre-Chester or pre-Ste. terminology exists for the recognition Genevieve unconformity. In some places and designation of lithologic units below in Montana and Wyoming, beds that the rank of formation, so that distinc- have been considered as young as St. 348 JAMES STEELE WILLIAMS

Louis in age have been included in the quartzites suggests the wide range of Madison; but in other places the upper ecological conditions that existed during part of the Madisonlimestone is of Osage late Mississippian and early Pennsyl- age. The disagreementregarding the age vanian time. These also indicate the of the Madison is beyond the scope of the variety of ecological problems. present paper. Some of the disagreement Existing data on the paleogeography may be due to a differencein interpreta- need to be consolidated and published tion of the boundariesof formationsand and new data added. Geosynclinal and groups in the Mississippi Valley, with shelf areas have been indicated in parts which correlationshave been attempted. of the area. The location of major land An unconformity and overlap between masses is known in a general way, but the Heath and the Amsden formations details of the paleogeography,including that may be within the Chesterhas been preciselocations and extent of majorfea- described from Montana, but the age of tures, are not known. Relationships to the unconformitydepends on the age as- other areas, such as the Mid-Continent signed to the Amsden there. An uncon- and Pacific Coast areas, need to be more formity occursbeneath the Darwin sand- definitely worked out and sea connec- stone unit at some places where it has tions more preciselylocated. Much addi- been identifiedin Wyoming. Unconform- tional data are needed on sourcesof sedi- ities have been placed at the Mississip- ments. Additional isopach maps need to pian-Pennsylvanianboundary as delim- be constructed; but these, to be widely ited by fossils mainly because it appears understood, either will have to await that in some places fossil zones of the unification of some of the stratigraphic Mississippi Valley Basin are not repre- terminology or will have to be accom- sented here. Lithologic criteria that sug- panied by considerablestratigraphy, ex- gest terrestrial beds and other types of plaining the author's usages of forma- lithologic criteria indicate unconformi- tions. ties at various places in the Amsden for- Of quite a number of recent articles mation, some of which are probablyvery containing maps or discussions of paleo- local in extent. geographic features, papers by Nolan (1943) and Eardley (1947) are especially PROBLEMS IN ECOLOGY AND comprehensiveand significant and con- PALEOGEOGRAPHY tain valuable references to other litera- Problems in ecology and paleogeogra- ture. phy that are as challenging as those in paleontology and stratigraphy occur in CONCLUSIONS the Carboniferousrocks of the northern i. As elsewhere in the United States, and central Rocky Mountains. Mention many unsolved problems are presented may be made of a few. The occurrencein by rocks at and near the Mississippian- these rocks of black shales, dwarffaunas, Pennsylvanianboundary in the northern oilitic limestones, great thicknesses of and central Rocky Mountains. crystallinelimestones containingfew fos- 2. The problems cover various phases sils other than corals, alternating ter- of stratigraphy,systematic paleontology, restrial and marine beds, considerable stratigraphic paleontology, ecology, and thicknesses of red beds, finely laminated paleogeography. limestones, gypsums, and cross-bedded 3. Adequate collections of well-pre- BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 349 served fossils are needed from critical with zones in the Mid-Continent region stratigraphic zones at critical localities. based on evidence from larger inverte- Collectinghas beenlimited in some places brate fossils have in several instances by the absence of marine invertebrates been made on data that are inadequate due to unfavorableconditions either dur- in both places. ing the deposition of the beds or for the 8. Detailed correlations within the preservation as fossils. At some other Rocky Mountain area should be made places it is limited by the difficulty of mainly with typical sections within that separating fossils from the enclosing area. There is little reasonto believe that rock. At many places, however, the small the Rocky Mountain and Mid-Continent collections made have shown that more areas, being distinct, though in places time and more experience in collecting connected, basins, had the same geologic would have yielded the larger and more history. Nor did they necessarily have varied collections needed. the same faunal zones, if one uses the 4. The considerable amount of paleon- term "zone"in a restrictedsense, or pre- tologic data already obtained should be cisely the same ranges of species. The better organizedfor effective use. gross differences in the faunas suggest 5. Criticalpaleontologic studies of sev- that the rangesof species may, in fact, be eral common species are essential. These quite different. should be based on enough specimens to 9. The use of fossils in determining the show the range of variation within each relative ages of beds in the Carboniferous species. Species that are too narrow be- of the Rocky Mountain area has been ex- cause of the lack of knowledge of the tensive and, despite the deficiencies of range of individual variation encourage the data at some places, appear to have erroneous and unsubstantiated correla- been as successful,though perhapsnot so tions. Species that are too broad contrib- detailed, as in most other areas. Many ute to the difficultyof establishingrecog- hundreds of collections have been stud- nizable paleontologic zones. The fact ied, and the age conclusionsbased upon that certain species definitely are long these studies have been given to field rangingand do not contributeeffectively men for furtherwork. Many hundredsof to detailed zonation studies should be squaremiles involving Mississippianand recognized and the nomenclature ad- Pennsylvanian rocks have been mapped justed accordingly. in detail, and some areas have been re- 6. Paleontologic correlations should be mapped on different scales. Many areas based on several lines of evidence. Index have been complexly folded and faulted, fossils are very useful in combination and other areas have been highly meta- with one another and with other faunal morphosed. Yet detailed remapping of data, but changes in the stratigraphic some areas and mapping of neighboring ranges of some so-called "index fossils" areas have changed few of the paleon- argue for caution in the use of all of tological age determinations. them. Correlationdata from one class of o10.Formations have been defined and fossils need to be checked against those their limits set by many different cri- from other classes whereverpossible. No teria. More uniformity than now exists, one class is of itself totally sufficient. as to both criteriaand terminology,is de- 7. Correlations of restricted paleon- sirable. A big step toward uniformity tologic zones in the Rocky Mountains would result from general agreement to 350 JAMES STEELE WILLIAMS define and actually employ formations, of various formations, lateral gradation, whether surface or subsurface, as litho- and other points would remain to be logic units which could be effectively solved. mapped, or could logically be supposed to I1. Additional work on the Carbonif- be mappable, on the ordinary scales of erous rocks and their faunas in the Rocky topographic quadrangle mapping used in Mountain area should be very thorough the area where they occur. Smaller units and detailed, or frankly interpretive in could be given varying degrees of sub- localities for which the data are inade- formational rank to allow as fine dis- quate. All types of previous knowledge crimination as desired. Such a definition should be utilized. Reconnaissance work would not prohibit formations from con- has been done in nearly all areas. taining unconformities, parts of two geo- 12. The combination of detailed strati- logic systems, or several paleontologic graphic and paleontologic work with de- zones or from varying somewhat in age tailed mapping offers the best possibility from place to place. Even with such a of obtaining adequate and reliable data general agreement, problems regarding needed from the northern and central the details of the lithologic composition Rocky Mountain region.

REFERENCES CITED

ASHLEY, G. H., et al (1933) Classification and CRONEIS, C. G., and FUNKHAUSER, H. J. (1938) nomenclature of rock units: Geol. Soc. America New ostracodes from the Clore formation [11.]: Bull. 44, PP. 423-459. Denison Univ. Bull. 38, no. o10(Sci. Lab. Jour., BAKER,A. A. (1947) Stratigraphy of the Wasatch vol. 33, art. 7), pp. 331-360. Mountains in the vicinity of Provo, Utah: DARTON,N. H. (1904) Comparison of the stratig- U.S. Geol. Survey Oil and Gas Inv. Preliminary raphy of the Black Hills, Big Horn Mountains Chart 3o. and Rocky Mountain front range: Geol. Soc. BISSELL, H. J., and HANSEN, G. H. (1935) The America Bull. 15, pp. 379-448. Mississippian-Pennsylvanian contact in the --- (1906) Description of the Bald Mountain central Wasatch Mountains, Utah: Utah Acad. and Dayton quadrangles, Wyoming: U.S. Geol. Sci. Proc., vol. I2, p. 163. Survey Geol. Atlas, Bald Mountain-Dayton BLACKWELDER, ELIOT (1910) New light on the Folio 141. geology of the Wasatch Mountains, Utah: Geol. EARDLEY, A. J. (1944) Geology of the north-central Soc. America Bull. 21, pp. 517-542. Wasatch Mountains, Utah: Geol. Soc. America ---- (1913) New or little known Paleozoic faunas Bull. 55, pp. 819-894. from Wyoming and Idaho: Am. Jour. Sci., --- (947) Paleozoic Cordilleran geosyncline 4th ser., vol. 36, pp. 174-179. and related orogeny: Jour. Geology, vol. 55, BRANSON, C. C. (I937) Stratigraphy and fauna of pp. 309-342. the Sacajawea formation, Mississippian, of Wyo- FOSTER, H. L. (1947) Paleozoic and Mesozoic ming: Jour. Paleontology, vol. ii, pp. 650-660. stratigraphy of northern Gros Ventre Mountains --- (1939) Pennsylvanian formations of central and Mount Leidy Highlands, Teton County, Wyoming: Geol. Soc. America Bull. 50, pp. Wyoming: Am. Assoc. Petroleum Geologists 1199-1225. Bull. 31, pp. 1537-1593. BRANSON,E. B., and GREGER,D. K. (1918) Ams- GARDNER,L. S.; HENDRICKS,T. A.; HADLEY, H. D.; den formation of the east slope of the Wind River and ROGERS, C. P., JR. (1945) Columnar sec- Mountains of Wyoming and its fauna: Geol. Soc. tions of Mesozoic and Paleozoic rocks in the America Bull. 29, pp. 309-326. mountains of south-central Montana: U.S. BRILL, K. G., JR. (I944) Late Paleozoic stratig- Geol. Survey Oil and Gas Inv. Preliminary raphy, west-central and northwestern Colorado: Chart 18 (also published as Stratigraphic sections Geol. Soc. America Bull. 55, pp. 621-655. of Upper Paleozoic and Mesozoic rocks in south- CALKINS, F. C., and BUTLER, B. S. (1943) Geology central Montana, with description of fauna of and ore deposits of the Cottonwood-American Amsden and Heath formations by L. L. Sloss: Fork area, Utah: U.S. Geol. Survey Prof. Paper Montana Bur. Mines and Geology Mem. 24, 2CI. 1946). BOUNDARY PROBLEMS IN THE ROCKY MOUNTAIN REGION 351

GILLULY,JAMES (1932) Geology and ore deposits REEVES, FRANK (1931) Geology of the Big Snowy of the Stockton and Fairfield quadrangles, Mountains, Montana: U.S. Geol. Survey Prof. Utah: U.S. Geol. Survey Prof. Paper 173. Paper 165, pp. 135-149. GIRTY, G. H. (1927) In MANSFIELD, G. R., Geog- SCOTT,H. W. (1935) Some Carboniferous stratig- raphy, geology, and mineral resources of part of raphy in Montana and northwestern Wyoming: southeastern Idaho: U.S. Geol. Survey Prof. Jour. Geology, vol. 43, pp. 1011-1032. Paper 152. --- (1945a) Age of the Amsden formation HENBEST,L. G. (1946) Correlation of the marine (abstr.): Geol. Soc. America Bull. 56, p. 1196. Pennsylvanian rocks of northern New Mexico -(945b) Pennsylvanian stratigraphy in and western Colorado (abstr.): Washington Montana and northern Wyoming (abstr.): ibid., Acad. Sci. Jour., vol. 36, p. 134. p. 1196. HUDDLE,J. W., and MCCANN,F. T. (i947a) Geo- SLOSS, L. L. and LAIRD, W. M. (1945) Mis- logic map of Duchesne River area, Wasatch and sissippian and Devonian stratigraphy of north- Duchesne counties, Utah: U.S. Geol. Survey Oil western Montana: U.S. Geol. Survey Oil and and Gas Inv. Preliminary Map 75. Gas Inv. Preliminary Chart 15. ------(I947b) Late Paleozoic rocks THOMAS, C. R.; MCCANN, F. T.; and RAMAN, N. D. exposed in the Duchesne River area, Duchesne (1945) Mesozoic and Paleozoic stratigraphy County, Utah: U.S. Geol. Survey Circ. 16. in northwestern Colorado and northeastern KINNEY, D. M., and ROMINGER, J. F. (1947) Utah: U.S. Geol. Survey Oil and Gas Inv. Geology of the Whiterocks River-Ashley Creek Preliminary Chart 16. area, Uintah County, Utah: U.S. Geol. Survey THOMPSON, M. L. (1936) Fusulinids from the Oil and Gas Inv. Preliminary Map 82. Black Hills and adjacent areas in Wyoming: MCCANN, F. T.; RAMAN, N. D.; and HENBEST, Jour. Paleontology, vol. o10,pp. 95-113. L. G. (1946) Section of Morgan formation, --- (1945) Pennsylvanian rocks and fusulinids Pennsylvanian, at Split Mountain in Dinosaur of east Utah and northeast Colorado correlated National Monument, Uintah County, Utah: with Kansas section: Kansas Univ. Geol. Survey U.S. Geol. Survey Mimeographed Rept., 1946. Bull. 60. MEEK, F. B. (1873) Preliminary paleontological WILLIAMS,JAMES STEELE (1936) Stratigraphic report, consisting of lists and descriptions of section and faunules of some western Carbonifer- fossils, with remarks on the ages of the rocks in ous formations at or near the type localities which they were found: U.S. Geol. and Geog. (abstr.): Geol. Soc. America Proc. 1935, pp. Survey Terr. 6th Ann. Rept., pp. 429-518. 118-119. MOREY, P. S. (1935) Ostracoda from the Amsden --- (1944) In MOORE,R. C., et al., Correla- formation of Wyoming: Jour. Paleontology, tion of Pennsylvanian formations of North vol. 9, pp. 474-482. America: Geol. Soc. America Bull. 55. NOLAN, T. B. (1935) The Gold Hill mining district, WILLIAMS, JAMES STEWART (1943) Carboniferous Utah: U.S. Geol. Survey Prof. Paper 177. formations of the Uinta and northern Wasatch ---(943) The Basin and Range Province in Mountains, Utah: Geol. Soc. America Bull. 54, Utah, Nevada, and California: U.S. Geol. pp. 591-624. Survey Prof. Paper I97-D, pp. 141-196. --- and YOLTON,J. S. (I945) Brazer (Mis- PERRY, E. S., and SLoss, L. L. (1943) Big Snowy sisippian) and lower Wells (Pennsylvanian) group, lithology and correlation in the northern section at Dry Lake, Logan quadrangle, Utah: Great Plains: Am. Assoc. Petroleum Geolo- Am. Assoc. Petroleum Geologists Bull. 29, pp. gists Bull. 27, pp. 1287-1304. 1143-1155. READ, C. B. (I934) A flora of Pottsville age from WILLIS,BAILEY (1902) Stratigraphy and structure, the Mosquito Range, Colorado: U.S. Geol. Lewis and Livingston Ranges, Montana: Geol. Survey Prof. Paper I85-D, pp. 79-96. Soc. America Bull. 13, pp. 305-352.