TheMiddle Jurassic Summeruille Formation, northern New Mexico by0. J. Anderson,New Mexico Bureau of Minesand Mineral Resources, Socono, NM 87801; andSpencei G, Luns,New Mexico Museum of NaturalHistory, 1801 Mountain Road NW, Albuquerque, NM87104

Abstract name used in New York state at the time it was introduced in Colorado. Clearly, usage of the name Wanakah outside of New We recommend continued usage of the name Summerville For- York should be discontinued. mation as a stratigraphic unit in the upper San Rafael Group in In northeastem New Mexico, the thinly bedded stratathat overlie northern New Mexico. Our reasonsare as follows. Subsequentto the lower San Rafael Group have been designated as Bell Ranch recognition in 1947of the Entrada Sandstoneand overlying Todilto Formation since 1959. Based on similarity of lithologies, deposi- Formation (the lower San RafaelGroup) in the southern San Juan tional environments, and stratigraphic sequenceand position, it Basin of northwestern New Mexico, field-oriented geologistsgen- is now virtually certain that the Bell Ranch is equivalent to and erally agreed to regard the overlying rocks as the upper part of the westward correlateswith Summerville strata. We thus recommend San Rafael Group. Although fossil control was lacking, the strati- that the name Bell Ranchbe abandonedand the name Summerville graphic sequenceand position were right, and lithologies and fa- retained for upper San Rafael Group strata acrossnorthern New cies were similar. Hence, until 1986, Summerville was used for Mexico. This simplifies the nomenclatureand properly relatesMid- these upper San RafaelGroup strata in all geologic mapping, and dle Jurassicrocks of the High Plains to the Colorado Plateau se- the usage becameentrenched. quence. In 1986, the name Wanakah Formation was introduced for the Other questions regarding correlation and nomenclatureof Mid- strata previously designated and mapped as Summerville in the dle Jurasiic units--Curtis, Pony Expressbeds, Bilk Creek, Bluff, San Juan Basin of New Mexico. The Wanakah type area is in the Cow Springs, Junction Creek, Zuni. and Tidwell-are discussed Ouray dishict of southwestemColorado, where the strataso named as they relate to the Summerville Formation' had been recognizedas equivalent to the Summerville Formation; the usageof Wanakahwas local only. Sinceequivalency with Sum- Introduction merville can be demonstrated, no scientific purpose is served by replacing the more widespread name Summerville with the pro- This report discusses the Summerville Formation and associ- vincial term Wanakah. Of greater significance, however, is that ated strata of the San Rafael Group (Middle jurassic) in northern Summerville has precedence,and that Wanakahwas a stratigraphic New Mexico and adjacent areas (Fig. 1), with Particular respect

108" 106' 104"

Moab It o Monlrose

?iver co Nl n,l Beclablto

Horse Mesa \t

i

D/t\ 'lf;"tfitl l Church Hasan ul Rock '\, /i (t $ Basin \\in l J\ Y>,f\l" Zuni . )l aAlbuquerque Pueblof;\ r \\" t\ i# 110' 108" 106' FIGURE1-Index map of the study areashowing outcropsof Middle Jurassicrocks (J, shaded)and someareas of pre-Jurassic(pre-J) and Cretaceous (K) rocks.

Nan Merico Geology November 1992 to 1) lithologies, 2) depositionalenvironments, 3) regional cor- and Wanakah in the southwesternpart of the statewith no criteria relations, and 4) the concept of a climatic change marking the given for distinguishing the two. The only early U.S. Geological transition from Middle to Late furassic time. Survey publication to recognizeWanakah was Wood et al. (1953), The Summerville consistsof alternating fine-grained sandstone and curiously this was in northeastern, not northwestern New and gypsiferous siltstone or mudstone that overlies the Curtis Mexico. They were thus the first to recognizethe Wanakah/Sum- Formation and forms the uppermost unit of the San RafaelGroup merville strata on the High Plains. Griggs and Read (1959)pointed in the San RafaelSwell of Utah. To the south, near Moab and La out that since 1953,usage of the name Wanakah had been limited Sal function, the unit thins significantly acrossthe sill or divide to the type area in southwestern Colorado, and they Proceeded that separatedthe marine basin of Curtis deposition to the north- to introduce the name Bell Ranch Formation for strata overlying west from the salina-lakebasin of Todilto deposition to the south- the Todilto in northeastern New Mexico. east. Still further southward near Bluff, Utah, the Summerville More recently Condon and Peterson(1985) and Condon and thickens but retains much of the lithologic characterof the type Huffrnan (1988)"reinstated" the name WanakahFormation, which area, though it is somewhat more sandy. It includes an eolian- they asserted "replaces the name Summerville Formation" sandstone lithosome at Bluff-the bed at Butler Wash-that has throughout northwestern New Mexico. They alleged this change been regarded as a tongue of the Entrada Sandstone.Also at in nomenclafure was necessarybecause of shatigraphic reports Bluff, a prominent sandstoneunit overlies the Summerville, and (O'sullivan, 1980;O'Sullivan and Pierce, 1983)indicating that all is appropriately named Bluff Sandstone.The Bluff Sandstonehas the Summerville Fotmation was cut out by an unconformity in been assigned to both the San RafaelGroup and to the Morrison southeastem Utah. But, in actuality, the Summerville is not cut Formation (Gregory, 1938; Craig, 1955; O'Sullivan, 1980). This out, although when O'Sullivan (1980;1981) realized this he had conflict has arisen because the Bluff contains both eolian and already decided to use the name Wanakah for these strata; this fluvial facies, and becausean unconformity has been perceived is documented by personal correspondence(1986) as well as the at the baseof the Bluff by those who would have it in the Morrison regional cross sections (Fig. 2). In essence,the Summerville is Formation. In this report we treat all eolianites overlying the continuous throughout the area and can be recognized in north- Summerville as essentially contemporaneous dune-field migra- western New Mexico on the basis of lithology, sequence, and tion onto the arid, sandy plain formed during and following the stratigraphic position. Condon and Huffman (1988) also 1) re- northward regressionof the epicontinental Curtis sea.Thus, they duced the Todilto Formation to a member-rank unit at the base all have affinities to and are included in the San Rafael Group. of the newly establishedWanakah Formation, 2) introduced the No major unconformity, and in most instancesno demonstrable name Beclabito Member for the overlying Summerville, and 3) unconformity, exists between the eolianites or mixed eolian-flu- replaced the name Bluff Sandstone with Horse Mesa Member, vial sand bodies and the underlying Summerville. These eolian- which forms the upper part of their Wanakah Formation. They dominated units include strata that have been referred to as Bluff. then establisheda referencesection in New Mexico using member Horse Mesa Member of the Wanakah, Junction Creek Sandstone names not used in the type area, and without benefit of an ac- (Colorado), RecaptureMember of the Morrison, Zuni Sandstone, companying stratigraphic section from the tyPe area. and upper part of Cow Springs Sandstone.Overlying the eolian- ites, or in places typical Summerville, is the fluvially dominated sequence regarded by most as the Morrison Formation. These Statement of problem sediments indicate increasedclastic input during a Late Jurassic The Summerville Formation overlies the Curtis in the type area gradual change to more humid climates. and demonstrably overlies the Todilto in northern New Mexico. Southeastward from Bluff, Arizona and into New Mexico, the In both areas the Summerville has been described as a thinly Summerville is present and continuous acrossthe Four Corners bedded unit consistingof arid coastalplain or sabkha,supratidal platform, the San|uan Basin, and the northeastern portion of the or hypersaline marine, wave-dominated deltaic (McKnight, 1940), state. Throughout much of this areathe Summerville overlies and and lacustrine deposits. These sediments were thus regressive overlaps the Todilto Formation. We discussthe equivalent strata and prograded across facies deposited during a high sealevel in southwestern Colorade-the Wanakah Formation----onlv1) to stand that was synchronousin the Curtis and Todilto depositional first describe in detail a type section, 2) to demonstrate how the basins. In the southern San Juan Basin and adiacentnortheastern Wanakah Formation units, the Pony Expressand the marl beds, Arizona, the Summerville grades southward and southwestward can be correlatedwith the Todilto Formation and the Summerville into an eolian sandpile that has been called the Cow Springs Formation, respectively, and 3) to lend our support to previous Sandstone. workers who have correlated the overlying Junction Creek Sand- Our interpretation of this physical stratigraphic sequenceis that stone westward with the Bluff Sandstone. the onset of regressiveconditions in the Todilto salina basin and the Curtis Seawaywas synchronous and initiated by relative sea- level fall. Ensuing depositional environments in both basinswere Previous sfudies similar, with eolian processesdominating along the southwest An exhaustive review of previous work on furassic rocks of the margin of the integrated basin. It is therefore in the interest of Colorado Plateau area is not undertaken here, as it has recently regional correlations to perpetuate the name Summerville For- been reported on by Condon and Peterson(1986), Condon and mation for these strata in both areas. Huffman (1988),and in other works. It is of interestto note that, The Wanakah Formation is another name for Summerville and following the suggestionby Baker et aI. (1947)to extend the name associatedstrata in southwestern Colorado. Recently, Wanakah Wanakah into New Mexico for the Todilto and overlying strata, Formation has been proposed as a more appropriate name for few geologists working the Colorado Plateau or San Juan Basin Summerville strata in southeastem Utah, and BeclabitoMember have used that name. The list of authors ignoring the name Wa- of the Wanakah in northwestern New Mexico (Condon and Pe- nakah in favor of Todilto and Summerville is long and impressive terson, 1986; Condon and Huffman, 1988; Condon, 1989).The and includes Harshbargeret al. (1951;1957), Rapaport et al. (1952), reasonsgiven by thoseauthors were 1.)an earlierU.S. Geological Smith (1957;196L), Moench and Puffet (1963),Schlee and Moench Survey report that had (mistakenly) concluded that Summerville (1963), Moench Q9e), Maxwell (1976),and Green and Pierson strata were missing in southeastem Utah; 2) when the Summer- Q9n). A significantwork is Dane and Bachman(1955), who in ville error was discovered, the name Wanakah had already been their compilation of the New Mexico state geologic map used San installed for the missing strata and thus the name had to be Rafael Group with Todilto and Summerville Formations as com- recognized in all areas to the south and east; and 3) Baker et al. ponent units. A correlation chart for the state of Colorado com- (L947)had "rccognized" the Wanakah in New Mexico. piled by Pearl(1972) used TodilteSummerville in placeof Wanakah. We question the logic and procedure followed in the proposed The Colorado geologic map (Tweto, 7979)usedboth Summerville name change and consider the stated reasonshivial becausethey

November 7992 Nal Mexico Geology were based on an error in judgment, in this casea double error, mudstone, minor, thinly bedded, dense gray limestone, and one since introducing the name Wanakah was done prematurely and gypsum bed (Peterson,1988). Based on the presenceof marine before full disclosure of all the stratigraphic data in southeastern invertebrate fossils and glauconite, a marine origin is indicated; Utah. In reference to 3) above, the Baker et al. paper consisted lithology and depositional setting further identify this unit as of interim recommendations; Baker et al. merely stated that "it shallow marine, Iikely hypersaline marine toward the end of dep- seems desirable to extend into New Mexico" use of the name osition. The Curtis representsa highstandfacies tract associated Wanakah becausethe upper contact of the San RafaelGroup was with maximum inundation of the region during Middle furassic "not yet definitely known." This is not synonymous with rec- (Callovian) time. ognizing strata that correlate with Wanakah, particularly so be- Relatedto and indirectlv sustainedbv the Curtis Seawas the causeno detailed stratigraphic studiesor sectionswere presented. paralic salina in which th6 Todilto Formation accumulated. Hy- A further objection to the use of the name Wanakah for these persaline conditions are indicated by the scarcity and low diver- strata in New Mexico is that it relegatesthe Todilto to a member- sity of fossils in the limestone member and the presenceof a thick rank unit at the base of the Wanakah Formation. While we are gypsum faciesin the central basin. The possibility of brief periods not preoccupied with rank, we do recognize and accept the con- of direct marine influence cannot be discounted based on the cept of formation. The Todilto, which has two mappable facies, presenceof dasycladalgae in the limestone(Armstrong et al., in locally exceeds150 ft (45.7 m) in thickness, and extends over an press).This interpretationhas alsofound support in the work of area of 33,000mi'? (86,000 km'?), not only meets all the contem- Ridgley (1989),who postulated a connection between the Todilto porary criteria for a formation-rank unit, but has been recognized basin and the Curtis seaway. A connection is suggestedby the as such since the name was introduced by Gregory (1917). northward-extending tongue of the Todilto Formation reaching Our usage of Todilto Formation would extend into southwest- the latitude of Montrose,Colorado (Fig. 1). No detailof the strat- ern Colorado to include the Pony Expressbeds at the base of the igraphic relationship of the northem terminus of the Todilto (called Wanakah. The overlying strata of the Wanakah are herein as- Pony Express beds in southwestern Colorado) with marine |u- signed to the Summerville Formation, which has precedence;Wa- rassicrocks to the north was provided by Ridgley(1989), and the nakah is abandonedin recognition of the fact that it is an occupied evidence for a seaway link thus is inconclusive at this point. The name in the state of New York. sulfur-isotopedata presentedby Ridgley and Goldhaber(1983) The Zuni Sandstone,long an undefined or poorly defined term, were used to support their casefor a direct seawayconnection. is retained for use in New Mexico and applied to strata formerly We contend that the isotope data merely confirm the proximity designatedas Cow Springs Sandstone.In the Acoma-MesaGi- of marine watersto the Todiltobasin. These marine waterscould gante areaof the southeasternSan fuan Basinthe upper part of then move as shallow ground water into the Todilto basin with the Bluff Sandstonewith contrasting bedforms has been mapped little changein isotopiccomposition. The isotope data do more as the Zuni Sandstone(Maxwell, 1975).Inasmuch as this upper to strengthenthe casefor synchronousCurtiFTodilto deposition unit can be traced southward into the main Zuni, the entrenched than they do to support direct marine influence. local usage of Zuni is herein considered valid and is given the Without a direct seawayconnection the high evaporative rates name Acoma Tongue of the Zuni Sandstone. prevalent at the time would have necessitatedreplenishment via ground-watermovement. Continuous transferof marine waters through a divide perhaps as much as 25 mi (40 km) in width is Stratigraphy and depositional setting the most plausible explanation. Hydraulic conductivity of the At the type area in eastern Utah the Summerville Formation eoliansand (Entrada)that formed the shallowbrine aquiferspan- forms the uppermost unit of the SanRafael Group. The remainder ning the divide would have been sufficient and adequateto sup- of the group in descendingorder consistsof the Curtis Formation, ply the water. Runoff would have been another source of water the Entrada Sandstone,and the Carmel Formation. and dissolvedsolids in the system,although climateswere very In extreme southeasternUtah and northwestern New Mexico arid. It is further possiblethat eoliantransport of gypsum south- the Carmel is not present and thus the Entrada rests unconform- eastwardfrom the margins of the Curtis Seawayselectively en- ably (the ]-2 unconformity, Pipiringos and O'Sullivan, 1978)on riched the Todilto salina in sulfate. Paleowind studies by Poole Triassicrocks (Chinle Group) in those areas.The Curtis Formation (7962)do not, however, Iend great support to this interpretation; likewise pinches out to the southeast due to pre-Summerville the study did neverthelessindicate a strong northerly wind pat- erosion or nondeposition and thus from the Moab areasouthward tern during deposition of the Cow SpringsSandstone, which is to the Four Corners area the Summerville rests paraconformably in part correlativewith the Todilto. on the Entrada Sandstone.Southeastward into northern New Regardlessof how the Todilto salina basin and its carbonate- Mexico the Todilto Formation is present between the Entrada and evaporite sequence are interpreted, contemporaneity with the the Summerville. The Todilto is therefore homotaxial to, and as Curtis Seawayis well supportedby the physicalstratigraphy and we shall argue, correlativewith, the Curtis Formation.Our con- hydrologic conceptsin terms of the constrainingeffects of arid cept of the regional correlation of the San RafaelGroup is shown climates on runoff. Thus the fall of relative sealevel (or increased in Fig. 2. clastic supply) following the Curtis-Todilto highstand initiated In southeastemUtah, a low sill or divide on an arid landscape simultaneousprogradation in both basinsas basalSummerville of no relief other than dunes separatedthe basin of Curtis dep- stratawere deposited. osition from the salina basin to the southeastin which the Todilto was deposited.The Moab Tongue of the Entrada accumulated acrossa portion of this divide, and local eolian deposition thus Summerville Formation persisted throughout Todilto and well into Summerville time. Named by Gilluly and Reeside (7928) for exposures at Sum- Similar eolian depositionin post-Entradatime is representedby merville Point in the northern San RafaelSwell, the Summerville the bed at Butler Wash. This informal sandstoneunit within the Formation consists of reddish-brown, thinly bedded sandstone Summerville Formation is 30-70 ft (9-23 m) thick. It extends with interbedded gypsiferous siltstone, sandy siltstone, or mud- southward into the Four Comers area, and northward it merges stone. Lithologies may vary somewhat, specificallymudstones with the main body of the Entradanear Blanding. are not everywhere present, the sequenceis not everywhere gyp- Establishing the correlation of the Curtis and Todilto Forma- siferous, and coarser-grainedconglomerate facies are present near tions, which was doubted by Pipiringos and O'Sullivan (1978), the depositionalbasin margins on the south and southwest.Bed- is of importance in correlating the younger sediments that pro- ded gypsum is generallynot presentin the SanJuan Basin.The graded out acrosstheir respectivebasins to form the Summerville flat, thin bedding, however, is a characteristicfrom the type area strata.The Curtis Formation is composedlargely of grayish-green, to the southern SanJuan Basin and eastward into the High Plains glauconitic, flat and crossbeddedsandstone, with thin beds of of east-centralNew Mexico (Fig. 2B). 13

Nm Mexico Geology November 7992 trlW Acoma Mesa I Lupton, Zuni The scale Arizona Pueblo El Morro Narrows Pueblo Mesita Gigante (ft) I I I I I I I 1200

1100-

1000-

900-

800-

700 Kd

Jm 600

500 JO

400

300 -

200-

100

0-

C NNW scate Whitewash. Moabarea, Blandingarea, Bluffarea, Beclabito (tt) Utah Utah Butler#1 Butler#4 Dome eoo- i I I I I

800-

700-

600-

500-

400-

300-

Je 200 - JE

up lo 100- base not\ \ exposed 400' thick \ \ locally 0-

--&-oo"'"u

rco'r"z Jpqronso UT

FIGURE 2-Regional correlation of the San RafaelGroup. A, lines of stratigraphic cross sections shown in B-D; B, Lupton, Arizona oi,,\*"'"* ,/ Basin \ to northeastern New Mexico; C, Whitewash, Utah to Acoma Pueblo Mesa /.D"---" area in southeasternSan Juan Basin, New Mexico; D, Whitewash, i"..-rie"..l,,.o' L ntuuquerque Utah to Ouray County, Colorado showing relationshipof Sum- merville strata to the type area of the Wanikah Formatron.

November 1992 Nan Mexico Geology Hagan Ouay N.E. E Basin County NewMexico l" I I

Kd - DakotaFm. Kdg - DakotaGp. Kr - RomerovilleSs. Kp - PajaritoFm Kr Kmr - Mesa RicaSs. - lKp Kts TucumcariSh. Klt - Lytle Ss. Jm - MorrisonFm, Jmj - JackPileMbL, MorrisonFm. Jz - AcomaTongue,Zuni Ss Jza - Zuni Ss. Jb - BlutlSs. Js - SummeruilleFm. Jtl - limestonembr., Todilto Fm Jtg - gYPSummbr., TodiltoFm. Jex - ExeterMbr., Entrada Ss Je - EntradaSs.

SSE Kd - Dakota Fm. Horse Todilto Thoreau Acoma East - Pueblo Mesa Jm lvlorrisonFm. Mesa Park alea - Recaptu(eMbr., Morrison Fm. I Jmr - SaltWash MbL,Morrison Fm I I I Jms I I - Ss .? Jb Bluff Js - SummeruilleFm. Jsbw- bed at ButlerWash, SummeruilleFm. Jz - ZuniSs. Jza - AcomaTongue ol Zuni Ss. Jtl - limestonembr., Todilto Fm. Je - EntradaSs. Jem - MoabTongue, Entrada Ss. mn - manganesebed at base of MorrisonFm. ffisandstone [i'.]!il sre556g6 6 sd sandstone t:-,....:,t:I ffitir""ton" l-=l;1 silty sandstone

l.;-:-. I sandy slltstone exposed [-----J shale

OurayCo., E Colo. D Whitewash, Moab area SlickRock, BilkCreek, scate Colo. flype Wanakah) (tr) Utah Utah Colo. I I 600 - I I I

500 -

400 -

300 -

200 - oase nol exPosed

100- \'

0

- Jm MolrisonFm. Jt TodiltoFm. - Jms - SaltWash Mbr.,Morrison Fm. Je EntradaSs, - Jjc - JunctionCreek Ss. Jem N.4oabTongue, Entrada Ss Js - SummeruilleFm. - MorrisonFm Jsbc - BilkCreek Ss. bed,Summeruille Fm mn manganesebed at base01

New Merico GeologY Novembet 1992 Thickness is also similar from the type area southward into the inated, the shrinkage factor was low, and therefore San Basin, mudcracks Juan as both-areaswere viry shallow epicontinental were not formed. basins during the Middle Jurassic.Maiimum thick'nessmav be The generallack of fluvial faciesin the fine-grained Summerville as much as 300 ft (91 m) in Utah but is generallyless and iri the suggeststhat a significant portion of the sediment was introduced jype areais.163 ft (50 m) (Gilluly and Reeside,ieZAy.In the San as eolian material. Fluvial deposits in the form of coarser-grained Juan Basinthicknesses varyfrom 90 to 150ft (22.Sto45 m) in the lenticular sandstone and conglomeratic sandstone exisiwithin southern portion to more than 200 ft (61 park mj at Todilto on the Acoma Sag in the extreme southern San Basin south of the Arizona-New Mexico state line (Aarshbarger fuan et al., l95T). Acoma Pueblo (Maxwell, 1979).Further to the northwest, along Thicknessesare vastly reduced in the areathat fbrmed the divide the solthwestem margin of the basin of Summerville deposition] between Todilto and Curtis deposition. Here the Summerville fluvial facies consisting of pebbly conglomeratic-sandstonebeds may_beas thin as 4-5 ft (1.2-1.5 (O,Sullivan, m) 19g0)and rests containing clay clasts have been noted by Peterson (1988),who ttt" Moab Tongue of the 9l Entrada. Even in those aieas where assigned them to a new unit called the Romana Sandstone.The it is thinnest,the thinly beddedaspect identifies it as summerville usage of Romana is restricted strata. to the Kaiparowits basin where it is considered to be correlative with the Summerville. Bedform Intraformational deformation and slumping are additional and lithology are sufficiently different in the Romana Sandstone characteristics of the summerville. In northeastern Arizona this to warrant recognition as a separate stratigraphic unit, albeit a slumping was thought by Harshbarger et al. (1957) to,"pr"r".i member or tongue of the Summerville would have served. soft-sedimentdeformation of water-safuratedstrata on a eentlv sloping sea floor. Inasmuch as the Summerville is nonmarlne ii Wanakah Formation thi.9are.1, we do not agree entirely with their interpretation. The The name Wanakahwas first applied to a lithostratigraphic unit soft-sediment deformation is uncontested;however, the deform_ in the contiguous United Statesby Grabau(1917,p. 338,footnote), ing mechanism probably relatesto dewatering and local variation who introduced the name "Wanakah shales" fol the lower part in confining pressure.In the san Basinin additionar Juan feature of the Devonian Hamilton Formation in western New York. eoo- of the Summerville is the sandstone pipes up to 100 ft (30 m; or per (1930,p. 225) subsequently designated a type section ("the more in diameter. It has been suggesfea (Moench urrd -deposition S.hi"", Wanakahand lakewood Beachcliffs, Lake Erie") of what he termed 1967)that the pipes formed durinf of the uppermosi the "Wanakah Member" of the Ludlowville Formation of the part of the formation-bythe founderin! of sandinto spriig vents Hamilton Group. The Middle Devonian Wanakah Member is a as compaction and dewatering proceeded. Locally tire rEmoval well recognized shatigraphic unit in western New York. For ex- of underlying gyRgur-nmay haie been a factor in pipe formation; on the "Geologic Map of New York" published by the however, many of the pipes do not extend downward 9TRt", to the New York StateMuseum of Science/NewYork GeologicalSurvey, Todilto and thus are unrelated to dissolution and removal ol sum. Wp_ the Wanakah Member is recognized on the Finger Lakes Sheilt (Rickard and Fisher, L970a)andNiagara Sheet (Riclard and Fisher, In east-centralNew Mexico strataequivalent to the Summerville 1970b).Burbank's (1930)use of the name Wanakah thus is pre- have been recognized since the 195ds.These strata overlie and occupied by Grabau's (1917). According to the acceptedrules of ovellap the Todilto Formation and range in thickness from 0 to stratigraphic procedure (NACSN, 1983,p. 852-853,Article 7(b)), 66 ft (0-20 m) (Lucas et al., 1985).th6 thinly bedded, ribbed_ the name Wanakah should not be used for a lithostrati- outcrop aspect is well displayed furassic and lithologies are similar to graphic unit in Colorado since it is being used in New York for lithologies of the Summervitte in the type are"ao., the Colorado a Devonian unit named 13years earlier (Lucas al., 1985; P.lateau_.. et Lucas, 9tlg-gt and Read (1959)recogriized the equivalence of less). these.High Plains strata to the Wanaka'hFormation, a name they Burbank (1930,p. 172)inhoduced the name Wanakah Member statedwas used for Summerville stratain southwestern Coloradd. of the Morrison Formation for strataexposed in the Ouray mining However, since the name Wanakah had not been used outside district of southwestern Colorado named for its exposure in the of Colorado since 1953, Griggs and Read (1959)introduced the Wanakah mine near Ouray. Eckel (1949)subsequently raised Wa- name Bell Ranch Formation-fbr these strata in east-cenkal New nakah to formation rank, and it currently is recognized as a for- Mexico. We agree with Griggs and Read in the correlation Sum_ mation in Colorado, New Mexico, and iJtah by"rarorkersof the merville : Wanakah : Bell Ranch. Thus in the interest of clar_ U.S. GeologicalSurvey (e.g. Condon and Peterson,1986; Con- ification of regionalshatigraphy and correlationsthe nomenclature don, 1989)despite the fact that no type section of the Wanakah ought to be simplified. We-th-erefore recommend that the name Formation has been established or described. It is of interest to Bell Ranch Formation be abandoned in easternNew Mexico and note that Goldman and Spencer (1941)recognized the Summer- instead the name Summerville be used for these strata. The name ville and Wanakah as "being equivalent to each other." Summerville thus is applied across the entire northern half of The only surface section of the Wanakah described that ap- New Mexico, recognizing the_correlation plains of these High proximates a type section is one measured "on the east side of strata with the type Summerville. the Uncompahgre Valley 6 miles north of Ouray" (Burbank, 1930, A.nonmarine origin for the Summerville is indicated by the p. V!. We measured a section of Burbank's (1930) Wanakah nearly complete lack of fossils, the gypsiferous character of the Member approimately 6 miles north-northwest of Ouray, just sediments, and also by the assumption that the dewaterinS phe- east of the Uncompaghre River in the NW1/4 NW1/4 sdc.'35, nomena noted above could only occur subaerially. Nonetlieiess, T4SN, R8W Ouray County, Colorado (Table1). We consider this some type of quiet-water or intermittent quiet-water deposition section (Fig. 3) to be the type section (principal referencesection is strongly indicated by the great lateral co-ntinuity of iniividual or lectostratotype of some authors) of the Wanakah (Formation). strqt-a,lle localized presence of ripple marks, arid the nodular At its type section the Wanakah Formation is 85 ft (25.7 m) and.bedded-gypsum. Thesefeatures along with the near absence thick, includirig a basal L2 ft (3.6 m) thick interval of limestone of channel-form structures indicate depos-itionin broad, shallow and limestone breccia (Fig. a). This interval, named the "Pony temporary lakes and tidal flats on an arid coastal plain of verv Express Beds" by Burbank (1930),rests directly on the Entrada low relief. High water tables related to the nearbv ieawav chai- Sandstoneand clearly is the Todilto Limestone of Gregory (19L6). acterized the entire area and evaporative pumping contiibuted We thus follow earlier authors and abandon the superfluous term to the gypsum content_ofthe sedimints. Durirg pioloiged drought Pony ExpressBeds. eolian-sand sheetsand dune fields encroachedupon tXe lake bJds Above the limestone breccia north of Ouray are 43.5 ft (13.2 to be later worked into thin strata by wave actidn, as brackish to q) of silfy, gypsiferous sandstonethat are equivalent to the Bilk saline lakes formed and waned. Much of the finer material and CreekSandstone of Goldmanand Spencer(1941). The remainder salt was perhaps deflated by the prevailing winds. Sand predom- of the type Wanakah section is 29.4 ft (8.9 m) of gypsiferous

November 7982 Nao Mexico Geology TABLE 1-Measured tvpe section of the Wanakah Formation, NW1/n NW1A sec. 35 T45N RSW Ouray County, Colorado.

Thickness Unit Lithology (m)

Morrison Formatron: Morrison 12 Sandstone; grayish-orange pink (10 R 6/2); fine- grained, subrounded, well-sorted, quartzose,tal- iareous; trough-crossbedded;forms a ledgy cliff. not measured

11 Mudstone; brownish-gray (5 YR 4/1) and grayish red-purple (5 RP A2); silty and smectitic; deeply Junction weaihered, partially covered slope. 72.7 CreekSs, 9 I unction CreekSands tone : n 10 Sandstone;yellowish-gray (5Y 712and 5 Y 8/1); very fine to medium-grained, subrounded, mod- 7 erately sorted, quartzose, very calcareous;lami- 0t nar, in flaggy weatheringbeds 0.3-0.6 m thick. 3.3 o," -] 9 Sandstone;pale-red (10 R 6/2)and very pale orange (10 R/YR 812); fine- to medium-grained, sub- rounded, moderately sorted, quartzose, slightly calcareous;laminar bedding; forms a prominent ledge. 3.0 BilkCreek Ss. "WanakahF ormation" (type section) : 8 Muddy and sandy siltstone;Srayish-orange (10 YR 714) and moderate yellowish-brown (10 YR 5/4); slightly calcareous;massive; laterally gives _Ponyl;; way to massivemedium-gray (N5), kerogenic, mi- tsxpress-' critic limestone. 0.8 I

7 Siltstone; same color and lithologies as unit 6 but better cemented and thus forms a more resistant EntradaSs. cliff than 6. 5.1

6 Siltstone and sandy siltstone; siltstone is pale red (5 R 6/2) to grayish red (10 R tV2), slightly calcar- eousand gypsiferous;sandy siltstonecontains very fine grained quartzose sand and is pale yellowish brown (10 YR 612),very calcareousand gypsifer- ous; unit contains fist-sized SyPsum rosettes that are mottled pale yellowish orange (10YR 8/6),very pale orange (10 YR 8/2) and moderate reddish FIGURE3-Measured section at thetype Wanakah location, NW'/+NW orange (10 R 6i6); bedding is laminar and small- r/qsec. 35 T45N R8W OurayCounty, Colorado; strata dip 23'N30'W. scale-troughs. 3.0

mostly light olive-gray (5 Y 6/1) 5 Silty sandstone; limestone near their top overlain grayish-orange(70 YR 7l\; very fine siltstoneswith thin interbedded but mottled Creek Sandstone grained, subangular,well-sorted, gypsiferous,cal- bv ouartzose,laminar sandstone,the Junction iareous; small trough crossbeds and much bio- oi Cttd-un and Spencer(1941). Strata included in the Morrison turbation; partly covered by slope debris; this is Formation (Salt Wish Member) overlie the Junction Creek Sand- the "Bilk Creek Sandstone"of Goldman and Spen- stone north of OuraY. cer. l3'2 Eolianites overlying Summerville Formation "Pony ExpressBeds": 4 Limestone breccia in gypsiferous matrix; angular Creek Sandstone limestone fragments of unit 3 lithology and color Junction up to 2 cm in diameter in light olive-gray (5 Y The function Creek Sandstoneis a massive,whitish, cross- 5/1) very calcareousgypsiferous matrix; forms a bedded sandstone that has a lithotogy and bedform similar to ledge. 2-8 that of the much older Entrada Sanditone (Figs' 2D and 4F)' It is a dominantly eolian unit that attains thicknessesof 300-500ft (10YR 22) and 3 Limestone;dusky yellowish-brown The Creek is correlative with the Bluff Sand- (N4); laminar and wary (g7-152m). function medium-dark gray micrite; into Utah (Goldman and Spencer, 1941; laminar (crinkly). 0.8 rtor," to,rih*estward Craig, L955;Poole,7962;Ekren and Houser, 1965)'The Bluff, also dorriinantly eolian, conformably overlies the Summerville in the EntradaSandstone: Utah (Harshbargeret al', L957;Ctaig, paleyellowish-brown (10 YR 6i2);very Wpe area of southeastern 2 Sandstone; overlies the Wanakah For- fine grained,subrounded, well-sorted, quartzose, iSSS);tn" function Creek conformably very-calcareous,trace micas, laminar. 0.4 mation (now Summerville) in southwestern Colorado' The significanceof the function Creek Sand.stoneis that Gold- 1 Sandstone;dusky yellowish-brown(10 YR Z2); very *utt u.rdsp"ncer (1941),who named the unit, recognized.itas quartzose, also realized the fine grained, subrounded, well-sorted, the upper irart of the San Rafael Group and calca"reous,trough-crossbedded. not measured reglondl correlations. They specificallydiscussed the correlations

Nan Mexico Geology Novembet 1992 ado. A, Overview of type section, showing Entrada Sandstonefle) overlain 0) named these strata the "Pony Expressbeds" at the baseof the Wanakah. se the major Portion of the type Wanakah, here considered strata of the type Wanakah and is in turn overlain by the Morrison Formation 0m). B, re Todilto Formation. C. Limestone breccia in a gypsum matrix (unit 4) at sandstone (unit 5) at type Wanakah section, a unit of the Summerville nits 5-8) of Summerville Formation (upper beds of type Wanakah) overlain ;tone (Jjc).

November 1992 New Mexico Geology with the Utah section, the Four Corners area, and even correla- ognized the equivalency of their lower function Creek and the tions of the Wanakah or post-Entradaunits into the Dry Cimarron Horse Mesa Member of the Wanakah (our Bluff Sandstone).They River valley in the northeastern corner of New Mexico. They did include the Horse Mesa Member (Bluff) in the San Rafael noted the similarities of the Pony Expressbeds to the Todilto, the Group, and we thus think their lower Junction Creek would by Wanakah units-Bilk Creek Sandstoneand the marl beds-to the inference also be included in that group. This supports our case Summerville, and, as stated, the function Creek-Bluff Sandstone because we regard the entire |unction Creek as one formation correlation. Griggs and Read (1959)were aware of the eastward and include it as the uppermost unit of the San Rafael Group regional extent of the Entrada and Summerville on the basis of (with perhaps even more Summerville above it locally; see Gold- physical shatigraphy.They, however, choseto introduce the name man and Spencer,I94L, p.1766.). Thus the function Creek,Horse Bell Ranch Formation for Summerville strata in east-centralNew Mesa Member, and Bluff Sandstoneall are homotaxial sand bodies Mexico. No eolianite such as the Junction Creek Sandstone is and are likely correlative,but this lacksa definite proof. We regard systematically present at the top of the San RafaelGroup (above the name Horse Mesa as superfluous. the Bell Ranch Formation-now Summerville) in these eastern The following quote from Goldman and Spencer (1941)sum- locations on the High Plains. However, an upper member of the marizes our concept of Junction Creek Sandstone: Entrada Sandstoneis locally well developed and apparently oc- Any correlationof the bedsoverlying the Entradasandstone in cupies the stratigraphic interval of the Todilto Formation and a southwestemColorado with thosein Utah must, in our opinion, part of the overlying Summerville. This upper eolianite, called be basedon recognitionof the fact that the functionCreek sand- Exeter Member of the Entrada Sandstone, is analogous to and stone,by its purity, its eoliantype of crossbedding,the uniform homotaxial with the Moab Tongue of the Entrada; both occupy sizeand relativeroundness of its grains,and thegreat thicknesses stratigraphic levels below the Bluff Sandstoneand are therefore it attains,is of thetype of the pre-Morrisonsandstones of the east- older than that unit. central Utah section, the Wingate, Navajo and Entrada, and not Source area for the Creek Sandstone lay to the west of the type of the sandstones in the Morrison Formation. If this Junction plausible and southwest based on paleowind studies by Poole (1962)who fact is accepted a very conelation can be made, of the and the Creek sandstone, with reported eastward and northeastward transport beds betwe'en the Entrada function direction. Eolian- the pre-Morrison beds of east-central Utah. sand bodies tend to thicken downwind, being thickest at their downwind limit (Poole, 1962).The |unction Creek is at its thickest in the La Plata mining district in southwestern Colorado, reported Bluff Sandstone by Eckel (1949)to be 200-500 ft (61-152m) thick. The eolianites Overlying the Summerville Formation through most of the San thus are locally well developed, well defined units of sufficient juan Basin is a cliff-forming unit of similar lithology but with more extent and thickness to merit separatestratigraphic names. Con- sandstone and thicker beds than the underlying section. It is sequently, we recognize|unction Creek Sandstoneas one of sev- largely eolian, but fluvial faciesare locally common and Predom- eral homotaxial eolian units that overlie a regionally extensive inate. This unit, named the Bluff SandstoneMember of the Mor- Summerville. rison Formation by Gregory (1938)and subsequentlyassigned to Since the earlier work pertaining to upper San Rafael Group the San RafaelGroup by Craig (1955),consists of a well-cemented lithologies and correlation(Goldman and Spencer,1941; Eckel, sandstoneof variable thickness, best developed near Bluff, Utah, 1949;Griggs and Read, 1959)and the relationship to the overlying from which it takes its name. The thickness near Bluff is as much Morrison Formation (Craig, 1955),there has been a great deal of as 300 ft, but decreasesto approximately 30 ft in the Horse Mesa effort on the part of U. S. Geological Survey investigators to area of northwestem New Mexico. The unit extends southward disprove those original findings and provisional correlations,most in the subsurfaceand eventually merges with the Bluff and Zuni of which we feel were valid. Specific attempts have been made Sandstonesin the southwesternand southernSan ]uan Basin.In to find a way to get the eolianites into the Morrison Formation. the southeastern San fuan Basin the Bluff is delineated from the A "horizontal parting surface" within the function Creek Sand- Summerville on the basis of bed form and bed thickness, and is stone was noted by Ekren and Houser (1965),who very obser- shown with an overlying tongue of the Zuni Sandstone(Fig. 2C). vantly placed the entire formation within and at the top of the The Bluff has been recognized as a medium- to thick-bedded San RafaelGroup. Condon and Huffman (1988),however, have sandstoneof eolian and fluvial origin conformably overlying dis- elevated the status of this "horizontal parting surface" to that of tinctive, thinly bedded Summerville strata. An upper part of the the basal contact of the Morrison Formation and by inference to Bluff, commonly lighter in color and with thick sets of high-angle a regional unconformity. Lithologies aboveand below the parting crossbeddedstrata, has been mapped (Maxwell, 7976,1979) and surface are similar, as is bedform. We contend that there is little, recognized as a northward-extending tongue of the Zuni. We if any, scientific basis for the stratigraphic division or assignment show it as a separateunit in our crosssections (Fig. 28, C) and proposed by Condon and Huffman. Rather it is a contrivance to refer to it as the Acoma Tongue of the Zuni. Southward and reconcile a problem created when O'Sullivan (1980)reassigned southwestward the Bluff and Summerville lose their identity and the Bluff Sandstoneto the Morrison Formation, and in doing so become one eolian-sand sequencethat is called Zuni Sandstone, envisioned an unconformity (the J-5) at the base. There had been the preferred name in New Mexico. The Zuni is well exposed at a prior general agreementbased on observablecriteria and strat- Zuni Pueblo and thus a local name is justified; however, it is the igraphic position that the Bluff and Junction Creek were correl- sameunit that Harshbargeret al. (1957)and Condon and Peterson ative. Thus, with the Bluff included in the Morrison a means to (1986)called the Cow Springs Sandstone. reassign the function Creek Sandstone,or part of it, to the Mor- The Bluff-Summerville contact is not ever)'where apparent or rison becamenecessary, and the "horizontal parting surface" was readily picked. Harshbarger et al. (1957)stated that this contact selected for that purpose. Placementof the basal Morrison For- is arbitrary in areas,suggesting that the transition from arid coastal- mation contactwithin the ]unction Creek Sandstonecreated other plain sabkha and eolian deposition to dominantly eolian and flu- minor problemsfor Condonand Huffman (1988).The upperJunc- viatile deposition was unbroken and gradual. The contact is es- tion Creek was Morrison, the lower Junction Creek was "pre- sentially based on bed form; the thinly bedded Summerville Morrison," and they concluded that this rendered the function contrastswith the much thicker bed setsand large-scalecrossbeds Creek both Middle and Late furassic in age. All these age as- that typically occur in the overlying BIuff Sandstone. signments were without benefit of any fossil control and with no radiometric dates. The result was that function Creek could no longer be included in the San RafaelGroup, which is entirely of Bluff-Zuni Sandston+Morrison Formation relationships Middle furassicage. We do not afford as much significance to the bedform break However, in their correlations of the function Creek southward between the Bluff Sandstoneand the overlying tongue of the Zuni into New Mexico, Condon and Huffman (1988)tentativelv rec-

Nm Mexico Geology November 7992 in the southern SanJuan Basinas did Condon and Peterson(1986) nized as the Entrada Sandstone.Only the upper 200 ft (61 m) are or Condon (1989). The former favored an interpretation of the assigned to the Zuni. The reason given by Anderson (1983)for relationship between the two units as "an unrecognized time the initial inclusion of the Entrada is that the break, or "Todilto boundary" and that the lower part was the San Rafael Group, notch," between the upper and lower parts is not everywhere whereas the upper, more crossbeddedportion was the Morrison present; where it is not present no basis exists for subdividing Formation. Their only reason was that other workers had re- the sandpile. While this remains true, we now prefer to call the ported the upper, crossbeddedsandstone to intertongue with the undivided section Zuni-Entrada Sandstoneand reservethe name Morrison Formation, and thus it had to be the Morrison. We Zuni only for the upper part, the post-Todilto portion above the regard the crossbeddedsandstone as a faciesof the San Rafael notch. Group. Condon (1989)was more definitive. Although conceding The Zuni Sandstoneis well exposed in the Zuni Pueblo area, that in most placesthe break between the upper and lower sand- which is in the type area of the "Zuni sandstones" of Dutton stones was gradational and basedon the presenceof eolian cross- (1885).The Cow Springs Sandstone,named by Harshbargeret bedding, Condon nevertheless saw it as a formational contact aI. (1957)for exposures on the north side of the Black Mesa in between his Wanakah Formation and an "eolian facies of the Arizona, was projected into western New Mexico by those au- Recapture Member of the Morrison Formation." This "contact" thors. We, however, object to usage of Cow Springs in New was of importance to Condon becauseit had been described as Mexico, because Peterson (1988)has indicated that in the type an unconformity (the I-5) by O'Sullivan (1980) in areas to the area the Cow Springs is a member of the Entrada Sandstone. northwest of the San Juan Basin; all strata above the |-5 by def- Peterson(1988) correlated the Cow SpringsSandstone eastward inition belonged in the Morrison Formation. The subdivision of as a pre-Todilto unit beneath the J-3 unconformity of Pipiringos the Bluff and repositioning of the baseof the Morrison Formation and O'sullivan (1978).This indicates that the type Cow Springs is analogousto what was done with the function Creek Sandstone is older than the Cow Springs strata projected into New Mexico. described above. We contend that a formation contact ought to Although it is doubtful that the pre-Todilto ageof the Cow Springs be physically observableat the point where it is picked and not can ever be demonstrated, we accept it because so defined it based on interpretative information such as "unrecognized time should not be applied to the Zuni Sandstonethat is demonstrably boundaries" projected in from severalhundred miles away. Con- post-Todilto. Also a local name is highly appropriate for this unit; don (1989)was perhaps describing his Wanakah-Morrison con- it is well exposed at Zuni Pueblo and usage of the name dates tact as a correlative conformity, correlative with what was perceived from 1885.In compiling the geologic map of New Mexico, Carle as the J-5 unconformity at the baseof the Morrison at Bluff, Utah. H. Dane (1956memo) and Dane and Bachman(1965) very clearly Our contention is that lithologies and bedform of the "eolian expressed their preference for the name Zuni. facies of the Morrison" suggestit more properly belongs with the The Acoma Tongue of the Zuni is recognizedas an eolian facies Bluff Sandstone, or a tongue of the Zuni Sandstone. Moreover, extending northeastward from the main Zuni (Fig. 2B). The main by definition, a formation is a lithostratigraphic unit, not a chron- body of the Zuni Sandstone is developed as the Summerville ostratigraphic unit. Accordingly, we name and map lithostrati- Formation and Bluff Sandstone Iose their identities southward graphic units, not chronostratigraphic units. With these and grade into a thick, crossbeddedsandstone sequenceof un- shatigraphic concepts in mind, we can readily see that Condon doubted eolian origin. This relationship can be traced out in the has tried to project in and recognize chronostratigraphic units. southern San Juan Basin (Maxwell, t976, 1979;Moorc, 1990)and As the Bluff supposedly intertongues with the Morrison, it is thus lends validity to the extensive use of the name Zuni Sand- obvious that the base of the Morrison is not everywhere of the stone in the area by geologic maPPersand explorationists during same age or, conversely, that eolian deposition persisted locally the past 20 years. into the Late Jurassic. The type area of the Acoma Tongue is the east side of Seama The basalMorrison Formation contact in the Mesa Gigante area Mesa in sec. 19 T9N R6W and in secs.24, 25, 34, and 35 T9N is more properly placed at the base of the fluvial section that RZW on the Acoma Pueblo quadrangle. The Acoma Tongue con- overlies the eolian facies. This fits much better conceptually as sists of as much as 3fi) ft (91 m) of fine-to medium-grained, well- well, becauseMorrison climateswere decidedly more humid than sorted quartzosesandstone. Large-scale eolian crossbeddingalong those that prevailed during San Rafael Group deposition. The with the light color are distinguishing characteristics.It is sharply climatic change, in part related to continued northward drift of overlain by the coarser-grained,darker-colored Dakota Sandstone the North American plate, was gradual, however, and this is at the K-l unconformity. We, however, defer to the local inves- shown in the sedimentary units westward from Mesa Gigante. tigator the decision to recognize, or not recognize, the Acoma In the Thoreau-Church Rock area, interbedding of fluvial and Tongue as a separateunit on the basis of bedform alone. In his eolian facies creates even greater difficulty in placing the basal map of the Acoma Pueblo quadrangle, Maxwell (1976)used a Morrison contact.We agreewith Robertson(1990), nonetheless, dashed line everywhere for the Bluff-Zuni contact. in being reluctant to place the stratigraphically highest eolian sandstone in the RecaptureMember of the Morrison. Paleotransport direction during deposition of the Bluff deter- Correlation and discussion mined from eolian faciesin the Horse Mesa areawas to the south- During the 1980s,two arguments were presented by workers east and northeast (Condon, 1989).Thus, the sourcearea may at the U.S. GeologicalSurvey to suPPortuse of the nameWanakah well have been in part the vast plain left exposed by the retreat Formation in place of Summerville in southeastern Utah and of the CurtirSummerville sea. An additional sand sourcemust northwestern New Mexico. First, O'Sullivan (1980)claimed that have been present to the southwest, however, supplying sedi- the Summerville Formation pinched out beneath the J-5 uncon- ment to the Cow Springs eolian system in Arizona. Poole (1962) formity near Moab, Utah, and was not Present southeast of the reported generally eastward transport direction in the type area pinchout. Subbequently,O'Sullivan (1981)himself corrected this of the Bluff. This is consistent with transport direction observed ionclusion and recognizedSummerville strata southeastof Moab in the Junction Creek Sandstonewith which the BIuff has been as the "Wanakah Formation." Nomenclature aside, O'Sullivan correlated. demonstrated that although the Summerville thins significantly southeast of Moab, it can be traced throughout the Four Corners Zuni-Cow Springs Sandstonerelationship area and southeastward into the San Juan Basin. The name Zuni Sandstoneis retained for use in New Meico. We further note that identification of the Tidwell Member of Our usage of Zuni differs slightly from that suggestedin a pre- the Morrison Formation by Peterson (1988)removed strata long liminary report by Anderson (1983)by excluding the lower part and, we believe, correctly included in the Summerville Formation of the Zuni in the type area at Zuni Pueblo. This lower part, (e.g., McKnight, L94O)from that unit. This contributed to the totaling some 280 ft (85 m) of crossbeddedsandstone, is recog- apparent thinning of the Summerville locally in southeastem Utah.

November 79f2 New Mexico Geology The second arBument focused on correlation of the Wanakah, the Zuni Sandstone(previously Cow Springs). 3) Northward Summerville, and Curtis Formations. In the Blanding area of withdrawal of the Curtis Sea and the related disappearanceof southeastern Utah, O'Sullivan (1980)applied the informal name the Todilto salina initiated progradation of arid coastalplain-sab- "Bed at Black Steer Knoll" to a locally persistent sandstoneledge kha depositional systemsthat laid down the bulk of Summerville in the upper part of the Wanakah (Summerville) Formation. He strata. 4) Eolianite depocenters, thickest in their downwind di- then identified this bed as far east as Slick Rock, Colorado and rection, developed as penecontemporaneoussandpiles on the as far north as White Wash, northwest of Moab (Fig. 2C). At the Summerville strata as the San Rafael Group deposition came to White Wash locality, O'Sullivan (1980)claimed that the "Bed at an end. The cessationof eolianite,sabkha, and salina{akedep- Black Steer Knoll" was correlative with the upper Curtis For- osition was the direct result of progressivelymore humid climates mation and thus beneath the Summerville Formation. Condon during latestMiddle and earliestLate Jurassictimes. The humid and Peterson(1986, p.15) thus concludedthat "a markerbed [the climatesinitiated fluvial deposition as representedby the lower Bed at Black SteerKnolll in the upper part of the Wanakah is also members of the Morrison Formation culminating in the Brushy present in the upper part of the Curtis Formation of the San Basin Member. Rafael Swell area which indicates that the Wanakah is older than Sedimentary rocks are now being described in terms of se- the Summerville." quence stratigraphy, which defines a sequenceas an unconfor- We, however, question O'Sullivan's (1980)identification of the mity-boundedsuccession of geneticallyrelated strata. As a result "Bed at Black Steer Knoll" at localities such as White Wash, Utah, of the nearly complete lack of fossils, unconformities have become and Slick Rock, Colorado, (Fig. 5) that are a long distance from an increasinglypopular tool with Middle Jurassicstratigraphers. Black Steer Knoll west of Blanding and between which the San A study of the Jurassicunconformities by Pipiringosand O'Sul- Rafael Group strata cannot be directly traced. Furthermore, at livan (1978)resulted in the identificationof six regionallysignif- these localities the lithology, thickness, and exact stratigraphic icant unconformities labeled J-0 through J-5. position of the "Bed at Black Steer Knoll" differ (Fig. 6), thus The oldest unconformity pertaining to the San Rafael Group making it difficult to acceptO'Sullivan's correlations. In the Black strata is the j-2. It representsa hiatus between the Enhada (or SteerKnoll area the "Bed at Black SteerKnoll" is very gypsiferous the Carmel where present) and underlying strata. This uncon- and massive (Fig. 5C), and is in the middle of the Summerville formity is widespread, coextensive in fact with the San Rafael when the Tidwell Member is, as it indeed should be, included in Group, and is easily recognized. The J-3 is recognizedat the base the Summerville. At White Wash, the unit identified by O'Sul- of the Curtis Formationand may representsome minor reworking livan as the "Bed at Black Steer Knoll" is a calcareousripple- of sedimentassociated with the transgressionof the shallowCur- Iaminated sandstone near the top of the Summerville Formation tis Seaway.It is also recognizedat the base of the correlative (Fig. 5A). It is not at the top of the Curtis Formation;no Curtis Todilto Formation to the southeast(Maxwell , 7976;1982;7986). Formation strata are present at White Wash (McKnight, 1940). The J-4 is of limited applicability here, being present in the north- Finally, at Slick Rock, Colorado, gypsiferous crossbeddedsand- ern San RafaelSwell where it representsa break between the stone at the top of the Summerville Formation (Fig. 5D) was upper and lower Curtis or locally betweenthe Summervilleand identified by O'Sullivanas the "Bedat BlackSteer Knoll." Clearly, the Redwater ShaleMember of the SundanceFormation (Pipirin- O'Sullivan'i identification of different lithologies at different strai- gos and O'Sullivan, 1978). igraphic levels over a relatively large geographicarea as the "Bed Forming the top of the San Rafael Group is the J-5 unconfor- at Black Steer Knoll" is difficult to accept. mity, which is generallyregarded as regionalin scope.The prob- Since the Summerville Formation is so readily correlatedon the lem is that not all workers have recognizedit nor placedit at the basis of lithology and physical stratigraphy, we are puzzled by same horizon. No attempt is made here to resolve this problem recent attempts to restrict usage of the name. Given that the except to state that conceptually the contact between the San ultimate goals of detailed stratigraphic studies are regional cor- RafaelGroup and the Morrison Formation should be placed at relations,paleogeographic reconstructions, and conceptualdep- the break between the quartzose eolianites or thinly bedded sab- ositional models, it is paramount that stratigraphers strive to kha-type gypsiferoussandstones and the overlying coarser-grained recognize regionally interrelated genetic units. fluvial sandstonesof more lithic composition.As suchit may not Formation contacts cannot always be drawn at boundaries be- be an isochronoussurface and perhapswould not fit the definition tween genetic units, nor does definition demand this. Contacts of a sequenceboundary. It is mentionedhere briefly only because must, nevertheless,be placedat lithostratigraphicbreaks, not at in the more northwestern extent the Morrison directly overlies "unrecognizedtime boundaries" or chronostratigraphicbreaks; the SummervilleFormation. this is simply in accordancewith the basic definition of a for- mation. For purposesof discussionwe canrecognize and correlate Correlation of the Bluff Sandstone chronostratigraphic units, conceivably worldwide, but by defi- Condon and Huffman (1988)questioned correlation of the BIuff nition we map formations. Sandstonefrom the type area at Bluff, Utah (Fig. 5E) to Horse Sufficient stratigraphic information exists to permit accurate Mesa, New Mexico (Fig. 2C). This is the direct result of strati- regional correlations of the San Rafael Group strata (Fig. 2B-D) graphicreassignment by O'Sullivan(1980) of the type Bluff Sand- and to relate component formations and facies to depositional stone to the basal member of the Morrison Formation. Because settings. Paleotransport directions and paleohydrologic consid- the 30-ft-thick(l0-m-thick) eolian section at HorseMesa has been erationsallow an accuratepaleogeographic model of the Middle recognizedas part of the San RafaelGroup and as correlative furassicin the Southwest.In post-Carmeltime it consistedin turn with the Bluff Sandstone,reassignment of the type Bluff neces- of 1)an extensiveeolian erg represented by the EntradaSandstone sitated a new name-the Horse Mesa Member of the Wanakah that contains indicators of southward paleotransport of sand in Formation-for the sectionat Horse Mesa. To support this new migrating dunes. 2) Encroachmentinto southernUtah of a shal- name, Condon and Huffman (1988)stated that some contorted low epicontinental seawayin which the fine-grained, glauconitic beddingand fold structuresin the "HorseMesa Member" at Tsitah Curtis Formation was deposited.An associatedhighstand dep- Wash, on the west side of the Carrizo Mountains, Arizona, are ositional systemdeveloped as a salinalake or a paralicsalina to truncated by the younger BIuff Sandstone. We believe the con- the southeastin what is now northern New Mexico;in this salina torted bedding in the "Horse Mesa Member" is due to soft-sed- the Todilto Formation accumulated.Peripheral to the water masses iment deformation, dewatering or piping in the underlying were an arid coastal plain-sabkha and shallow temporary lacus- Summerville,or some large-scale,primary-bedding feature. The trine environments in which wave-dominated deltas prevailed deformation thus is syndepositional and not related to tectonic and in which the basalSummerville stratawere laid down. Farther deformation; therefore, little basis for an unconformity exists. away from the water masses,especially to the south and south- Horse Mesa and Tsitah Wash are 23 mi (37 km) apart, and the west, eoliandeposition persisted and resultedin a sandpilecalled

Natt Mexico Ceology November 1992 FIGURE 5-A, Part of the Jurassic section exposed at White Wash, NE1/+ NWr/+ sec. 14 T23S R178, Grand County, Utah, showing lower (fsl) and upper (fsu) Summerville Formation overlain by Morrison Formation 0m). B, Close-up of sandstone interval at White Wash identified by O'Sullivan (1980) as the "Bed at Black Steer Knoll." C, Close-up of "Bed at Black Steer Knoll" near Black Steer Knoll, SW1A NWr/r sec. 21 T35S R21E, San fuan County, Utah. Note gypsiferous nodules weathering out, especially around the rock hammer. D, Close-up of the sandstone interval near Slick Rock, Colorado (NEr/r NEl/r sec. 36 T44N R19W, San Miguel County) identified by O"Sullivan as the "Bed at Black Steer Knoll." Thickness of outcrop in photo is 10 ft. E, Summerville Formation (fs) overlain by Bluff Sandstone (fb) near Bluff, Utah, sec. 20 T40S R21E, San Juan County. F, Entrada Sandstone (Je) overlain by type Bell Ranch Formation (fs) overlain in turn by Morrison Formation (fm) at Carpenter Mesa, SW1/r NEIA sec 22 T16N R30E, Quay County, New Mexico. The uppermost sandstone bed in the Bell Ranch is an eolianite homotaxial to the Bluff Sandstone on the Colorado Plateau.

November 7982 Nau Mexico Geology WhiteWash, Utah scoured surface. Ekren and Houser (1965)reported that in south- (Dip2-3'N20'E) western Colorado fluvial sandstoneof the Salt Wash Member has NE+NE+sec.14T23S R17E been deposited in channelsscoured into the function Creek Sand- the unconformity. GrandCounty, Utah stone. This is J-5 Summary MorrisonFm. ,6 Our correlations of Middle Jurassicstrata are summarized by the cross sections (Fig. 2B-D). The thinly bedded strata overlying the Curtis Formation in the San RafaelSwell of Utah were given the name Summerville Formation by Gilluly and Reeside(1928). 'l The Summerville can be physically traced into southeasternUtah and correlated into adiacent southwestern Colorado and New 4mJ Mexico on the basis of lithology, bedform, and stratigraphic po- sition. Thus the name Wanakah Formation applied to these strata in southwesternColorado by Burbank (1930),but never widely used, is abandonedin favor of the widely used name Summer- ville. The Pony Express(limestone and gypsum) beds at the base SlickRock, Colorado of the Wanakah are correlative with the Todilto Formation, and (Dip50" E) thus the name Pony Expressbeds is also abandoned' NE+NE+sec.36T44N R19W Attempts to rejuvenate and expand the use of the name Wa- SanMiguel County, Colorado nakah Formation in New Mexico in place of the entrenched name Summerville Formation (Condon and Peterson, 1986; Condon and Huffman, 1988;and Condon, 1'989)are without merit and tend to stifle or inhibit regional correlations. Most importantly, SummervilleFm Summerville has precedence,and the name Wanakahis occupied in New York State. Therefore, the name Wanakah should not be used to designatestrata in Colorado. In northeastem New Mexico the strata overlying and overlaP- ping the Todilto Formation have been since 1959referred to as the gell Ranch Formation (Fig. 5F). The Bell Ranch is demon- strably equivalent to the Summerville strata to the west, and thus is abandoned in favor of Summerville. The gypsiferous, thinly bedded Tidwell Member of the Mor- rison Formation is not recognLed. Genetically and lithologically, these strata are properly included with the Summerville strata. Eolianites overlying the Summerville and/or forming the top of the San Rafael Group are homotaxial and very likely correlative. Theseinclude the function CreekSandstone, the Bluff Sandstone, EntradaSs, EntradaSs and the Acoma Tongue of the Zuni Sandstonein the southeastern SanJuan Basin. The HorseMesa Member of the Wanakah,named for exposures in extreme northwestern New Mexico, does not demonstrably differ from the Bluff Sandstonein age or lithology. FIGURE6-Stratigraphic sectionsat Whitewash,Utah and SlickRock, Horse Mesa thus is a superfluousname. Coloradoshowing position and thicknesscontrasts of the "bed at Black The Bluff Sandstoneand SummervilleFormation grade south- SteerKnoll." g, gypsiferous;mn, manganesebed. ward and southwestward into an undifferentiated eolian sand- stone. In west-central New Mexico and in the southern San fuan Basinthis eolianunit hasbeen called the Cow SpringsSandstone, Bluff Sandstonemerely thickens and developsan upper facies named for exposuresnear Cow Springsat the north end of Black westward into Arizona, where it exhibits internal deformation Mesa, Arizona. The Cow Springs Sandstone,however, is con- featurescommon in gypsiferoussediments. sidered by Peterson(1988) to be a pre-Todiltounit in Arizona, The "unconformity" or "truncation" describedby Condon and and as suchis merely the upper part of the Enhada.We note that Huffman is chronologicallyinsignificant. It is apparentlyanother the unit called Cow Springs in west-centralNew Mexico is a post- contrivance and argument put forth to defend placing the Bluff Todilto unit and consequently,we use the local name Zuni Sand- in the Morrison Formation. Regardlessof whether or not the stone, for exposuresnear Zuni Pueblo,in placeof Cow Springs. Horse Mesa Member can eventuallybe establishedas older than Usageof the name Zunilor thesestrata dates from Dutton (1885). the fype Bluff, they are here both considered to be part of the In the Laguna, Acoma, and Mesa Gigante areasa northerly San RafaelGroup. On the basisof lithology, bedform, and strat- extendinguppermost unit of the main Zuni Sandstonewas Pre- igraphicposition they appearcorrelative. We thereforeregard the viously called the upper Bluff or assignedto the RecaptureMem- name Horse Mesa Member as superfluous.Should an age differ- ber of the Morrison Formation. This highly crossbeddedunit, enceever be established,the option would be to regardthe thinned recognized by Maxwell (1976)Iatgely on the basis of bedform, is Horse Mesa section as a tongue of the Bluff. herein designatedthe Acoma Tongueof the Zuni Sandstonefor The Bluff Sandstonewas observedby O'Sullivan(1980) to pinch those who wish to differentiate it from the underlying Bluff. out into the Morrison Formation.We havenot visitedthe location AcrNowr-BocMENTS-The manuscript was reviewed and im- where the pinchout occurs,but if it indeed does so, it is simply proved by Stan R. Hafenfeld, Frank E. Kottlowski, CharlesH. an exampleof formationsintertonguing. A preconceivedJ-5 un- Maxwell, and Jiri Zidek. We are grateful for their comments, conformity should not be sought at the San RafaelGroupMor- corrections,and additions. Any remainingerrors are the full re- rison contact. This contact must be based on lithology, which sponsibilityof the authors. We also wish to acknowledgeLynne then dictatesplacement of the Bluff Sandstonein the San Rafael Hemenway who typed the manuscript and RebeccaTitus who Group (Craig,1955; Ekren and Houser,1965; Goldman and Spen- drafted most of the illushations. The New Mexico Bureauof Mines cet, l94l). The Salt Wash Member of the Morrison overlies the and Mineral Resourcesand New Medco Museum of Natural His- San RafaelGroup at a commonly,but not universally,channel- tory supported this research. rj

New Mexico Geolog3r November 7992 References Maxwell, C. H , 1976, Geologic map of the Acoma Pueblo quadrangle, Valencia Anderson, O. J.,7983, Preliminary report on redefinition of Zuni Sandstone,west- County, New Mexrco: U S Geological Survey, Geologic Quadrangle Map, GQ- central New Mexico: New Mexico Geology, v. 5, no. 3, pp. 56-59. 1298.scale 7:24.0ffi. Armstrong, A. K., in press,The JurassicTodiJto Limestone Member, Grants uranium Maxwell, C H., 1979,Geologic map of the EastMesa quadrangle,Valencia County, district, New Mexico-facies, diagenesis,and mineralogy: New Mexico Bureau New Mexico: U S GeologicalSuroey, Geologic Quadrangle Map, GQ-1522, scale of Mines and Mineral Resources,Bulletin 147. l:24,000 Maxwell, C H., 7982, Mesozoic stratigraphy of the Laguna-Grants region: New Baker,A. A., Dane,C H., and Reeside,J.B.,lr.,1947,Revised conelation of Jurassic formations of parts of Utah, Arizona, New Mexico and Colorado: American As- Mexrco GeologicalSociety, Guidebook to 33rd Field Conference,pp 26"1-266 sociation of Petroleum Geologists,Bulletin, v 31, no. 9, pp.7664-1668. McKnight, E T., 1,940,Geology of area between Creen and Colorado River, Grand Burbank, W. S., 1930,Revision of geologic structure and straligraphy in the Ouray and San Juan Counties, Utah: U.S GeologicalSuruey, Bulletin 908, pp 85-102 district of Colorado and its bearing on ore deposition: Colorado-scientificSociety Moench, R. H., 1954,Geology of the Dough Mountain quadrangle, New Mexico: Proceedings,v. 12, no. 6,231,pp. U.S. GeologicalSurvey, Geologic Quadrangle Map, GQ-354, scale1:24,000. Moench, R. and Puffet, W. P.,1953, map of the Mesa Gigante quad- Condon. S. M., 1989,Revisions of Middle |urassicnomenclature in the southeastern H., Geologic San Juan Basin, New Mexico: U.S. GeologicalSuruey, Bulletin 1808-E, 15 pp. rangle, New Mexico: U S Geological Survey, Geologic Quadrangle Map, GQ- Condon, S. M., and Huffman, A. C., Jr., 1988, Revision in nomenclature oi the 272, scale7:24,N0 Middle JurassicWanakah Formation, northwestern New Mexicoand northeastern Moench,R. H., and Schlee,l.5.,7967, Geologyand uraniumdeposits of the Laguna Ari'ona. U.S. GeologicalSurvey, Bulletin 7633-A,72pp. district, New Mexico: U S GeologicalSurvey, ProfessionalPaper 579,774pp Moore, S. L 1990,Geologic map of the Blue Mesaquadrangle, Cibola County, New Condon, S. M., and Peterson,F., 1986,Stratigraphy of Middle and Upper Jurassic , rocks of the San fuan Basin: historical perspective,current ideas, and remaining Mexico: U S. Geological Survey, Gmlogic Quadrangle Map, GQ-1652, scale1:24,000 problems: American Associationof PetroleumGeologists, Studies in Geology no. North American Commissionon ShatigraphicNomenclature, 1983, North American 22, pp.7-26. Stratigraphic Code: American Association of Petroleum Geologists, Bulletin, v Cooper, G. A., 7930, Stratigraphy of the Hamilton Group of New York, Part II: 67, no 5, pp. &41-875 American Journal of Science,5th series, vol. 19, pp.274-236. O'Sullivan, R. 8., 1980,Shatigraphic sectionsof Middle furassic San RafaelGroup Craig, L. C., 1955,Stratigraphy of the Morrison and related formations, Colorado from Wilson Arch to Bluff in southeasternUtah: U S GeologicalSurvey, Oil and Plateau region, a preliminary report: U.S GeologicalSurvey, Bulletin 1009-E,43 Gas lnvestigationChart 102. PP, O'Sullivan, R. B., 1981,Shatigraphic sections of somefurassic rocks from near Moab, Dane, C. H., and Bachman,G. O., 1965,Geologic map of New Mexico:U.S. Ge- Utah to Slick Rock, Colorado: U S. GeologicalSurvey, Oil and Gas Investigation ological Suruey and New Mexico Bureau of Mines and Mineral Resources,scale Chart 107 1:500,000. O'Sullivan, R B., and Pierce,F W., 1983,Stratigraphic diagram of Middle furassic Dane, C. H., 7956,Unpublished internal memo to GeologicNames Committee, U S San Rafael Group and associatedformations from the San Rafael Swell to Bluff Geological Survey. in southeastern Utah: U S Geological Survey, Oil and Gas Investigation Chart Dutton, C. 8., 1885, Mount Taylor and the Zuni Plateau: U.S Geological Survey 119. 6th Annual Report, 198pp. Pearl, R. H.,7972, Colorado stratigraphic nomenclaturechart: Colorado Geological Suruey Peterson,F , 1988,Shatigraphy and nomenclatureof Middle Jurassicrocks, western Colorado Plateau, Utah and Arizona: U S Geological Suruey, Bulletin 1533-8, pp 1.7-56 Pipiringos, G N., and O'Sul.livan,R. 8., 1978,Principal unconformities in Triassic and Jurassicrocks, Westem Interior, U.S.; a preliminary survey: U,S Geological Survey, ProfessionalPaper 1035-A, 29 pp. Goldman, M. I , and Spencer,A. C ,1941, Conelation of Cross' La PlataSandstone, Poole, F G ,7962, Wind direction in late Paleozoicto middle Mesozoictime on the southwestern Colorado: American Associationof Peholeum Geologists,Bulletin, Colorado Plateau:U S. GeologicalSuruey, Professional Paper 450-D, pp 147-755 v. 25, no. 9, pp.1745-7767. Rapaport,L, Hadfield, J P, and Olsen, R H., 1952,Jurassic rocks of the Zuni Grabau, A. W., 1977, Age and shatigraphic relations of the Olentangy Shale of uplift, New Mexico: U S Atomic Energy Commission, RMO-542, 49 pp. central Ohio, with remarks on the Prout Limestoneand so-calledOlentingy shales Rickard, D V, and Fisher, D. W.,1970a, Geologicmap of New York, Finger Lakes of northern Ohio: Journal of Geology, v.25, pp.337-343. sheet: New York State Museum and ScienceSeruice Map and Chart series, no Green, M. W., and Pierson, C. T., 7977,A summary of the shati$aphy and dep- 15. scale1:250.000. Rickard, D V, and Fisher,D. W., 1970b,Geologic map of New York, Niagara sheet: ositional environmentsofJurassic and relatedrocks in the SanJuan basin, Arizonl, Colorado, and New Mexico: New Mexico GeologicalSociety, Guidebook 28, pp New York StateMuseum and ScienceSewice Map and Chart series,no 15, scale 147-1.52. 1:250,000. Gregory H. 8.., 191,6,The Navaio country-a geographicand hydrographic recon- Ridgley, J. L., 1989, Trace fossils and mollusks from the upper members of the naissanceof parts of Ar2ona, New Mexico and Utah: U.S. Geologiial Suruey, WanakahFormation, Chama basin, New Mexiceevidence for a lacustrineorigin: Water-supply Paper 380, 219 pp. U.S Geological Suruey, Bulletin 1808-C, pp C1-C16 Gregory H. E., 7977,Geology of the Navajo country: U.S. GeologicalSurvey, Pro- Ridgley, l. L., and Goldhaber, M., 1983,Isotopic evidence for a marine origin of fessional paper 93, 161 pp. the Todilto Limestone, north-central New Mexico: GeologicalSociety of America, Abshacts v. 15, no. 4, p. 41.4. Gregory, H. 8., 1938,The San Juan country, a geographicand geologic reconnais- with Programs, sanceof southeastem Utah: U.S. GeologicalSuruey, ProfessionalPaper 188, 123 Robertson,I F.,1991J,Geologic map of the Thoreau quadrangle,McKinley County, New Mexico: GeologicalSuwey, Geologic Map, GQ-1575, scale PP' U.S. Quadrangle Griggs, R. L., and Read, C 8., 1959,Revisions in shatigraphic nomenclature in 1:24,000. Tucumcari-Sabinosoarea, northeastern New Mexico: American Association of Schlee,J. S., and Moench, R H., 1953, Ceologic map of the Mesita quadrangle, Petroleum Geologists, Bulletin, v. 43, no. 8, pp.20032007. New Mexico: U.S. GeologicalSurvey, Geologic Quadrangle Map, GQ-210, scale 7:24,4N Harshbarger,J. W., Repenning,C. A., and lrwin, I. H.,7957, Stratigraphyof the the Zuni Mountains, Valenciaand McKinley Counties, uppermost Triassicand the Jurassicrocks of the Navaio country: U.S. Geological Smith, C T ,1957, Ceologyof Survey, ProfessionalPaper 297,77 pp. New Mexico: Four Comers GeologicalSociety Guidebook, pp 53-61 Smith, C T 1961,Triassic and rocks of the Albuquerque area;New Mexico Harshbarger,J. W, Repenning,C. A., andJackson, R. L., 1951,Jurassic stratigraphy , furassic of the Navaio country: New Mexico GeologicalSociety, Guidebook 2, pp.9S-fS. Geological Society,Guidebook to 12th Field Conference,pp 127-128 Lucas, S. G., 1988,Todilto as a formation, not a member of the WanakahFormation, Tweto, O , 1979,Geologic map of Colorado: U S GeologicalSuruey in cooperation with the GeologicalSuruey of Colorado, scale1:500,000 Middle Jurassicof northern New Meico (abs ): New Mexico Geology, v. 10, no. 2, p. 4[. Wood,G H, Northrop,S A, andGriggs,R L,7953, Geologyandshatigraphy of KoeNer and Mount Laughlin quadrangles,eastern Colfax County, New Mexico: Lucas,S. G., Kietzke,K. K., and Hunt, A. P.,1985,The JurassicSystem in east- central New Mexico: New Mexico Geological Society, Guidebook to 36th Field U.S. GeologicalSurvey, Oil and Gas Investigations, OM-141 n Conference, pp. 273-242.

Continue to read New MexicoGeology . . .

All subscriptionsexpire with this issue.Renew your subscriptionby sending$5.00 in the renewal envelopethat has been insertedin this issue.

92 November 7992 New Merico Geology