AMERICAN MUSEUM

Norntates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2840, pp. 1-68, figs. 1-1 3, tables 1-11 February 28, 1986

Fossil Mammals from the "Mesaverde" Formation (Late Cretaceous, Judithian) of the Bighom and Wind River Basins, Wyoming, with Definitions of Late Cretaceous North American Land-Mammal "Ages"

JASON A. LILLEGRAVEN1 AND MALCOLM C. McKENNA2

ABSTRACT Mammalian faunas are documented for the first oming for the first time, and new records involve time from the "Mesaverde" Formation (Late Cre- both geographic and geologic range extensions. The taceous) ofWyoming. Nonmarine fossils from the temporal record ofthe dryolestid appears relictual, Bighorn and Wind River basins indicate a Ju- beingpreyiously unknown from post-Jurassic strata dithian (revised definition) "age" for the assem- in North America. Tax6nomic comparisons sug- blages through comparisons with approximately gest that the Judithian mammalian fauna of what contemporaneous faunas of the Judith River was then coastal parts of the western interior was (Montana) and Oldman (Alberta) formations. essentially homogeneous geographically, at least Three previously unknown genera are recognized, from southern Alberta to central Wyoming. although not named herein, from the "Mesa- Nonmarine mammalian assemblages from the verde" Formation (Multituberculata, new genus Oldman, Judith River, and "Mesaverde" forma- and species, unidentified; Dryolestidae, new genus tions correlate temporally with more easterly ma- and species, unidentified; and Falepetrus barwini). rine rock units that lie within (or within four zones Three new species of previously described genera above) the Baculites gregoryensis (cephalopod) (Alphadon sahnii, A. attaragos, and Paranyctoides Zone, part ofthe standard zonation ofUpper Cre- megakeros) are named. All species-level taxa (16 taceous rocks of the North American western in- total) except Alphadon lulli are reported from Wy- terior. The upper part ofthe Red Bird Silty Mem-

' Professor, Departments of Geology and Geophysics/Zoology and Physiology, The University of Wyoming, Lar- amie, Wyoming 82071. 2 Frick Curator of Fossil Mammals, Department of Vertebrate Paleontology, The American Museum of Natural History; Professor, Department of Geological Sciences, Columbia University.

Copyright © American Museum of Natural History 1986 ISSN 0003-0082 / Price $5.50 2 AMERICAN MUSEUM NOVITATES NO. 2840 ber ofthe Pierre Shale at Redbird, Wyoming, holds Aquilan (oldest), Judithian, and Lancian the largely endemic invertebrate macrofaunal as- (youngest) provincial North American land-mam- semblage characteristic ofthe B. gregoryensisZone mal "ages" are redefined for the Late Cretaceous as well as a newly described planktonic forami- nonmarine sequence of the western interior from niferal assemblage. The microfossils allow corre- an older stage concept; the "ages," based upon lation to the upper Taylorian and/or lower Na- species-level mammalian assemblages, are mod- varroan foraminiferal stages of the Gulf Coast. eled after the system used successfully for non- This, in turn, correlates approximately to the marine Cenozoic faunas of North America. An Campanian-Maastrichtian stage boundary at Edmontonian "age" (previously used as a stage Gubbio, Italy, and European stratotypic sections. term, chronologically intermediate between the Judithian mammal faunas of the Rocky Moun- Judithian and Lancian) is probably identifiable as tains, therefore, must be younger in age (i.e., late a discrete interval of geologic time, but is not yet Campanian and/or early Maastrichtian) in terms defensible on the basis of mammalian assem- of the European stages than usually is considered blages. Therefore, it is not redefined as a land- on the basis of molluscan zonations within the mammal "age." North American western interior (e.g., well within We concur with the interpretation that the Dja- the Campanian). Judithian mammals from the dokhta Formation of southern Mongolia and the Rockies probably lived about 74-76 million years Judith River and Oldman faunas of North Amer- ago during the late part of geomagnetic Polarity ica are essentially ofthe same age. This compresses Chron 33 or the early part of Polarity Chron 32, the faunas of the Barun Goyot and Nemegt for- during the regressive phases of the Claggett cyclo- mations of southern Mongolia, plus that of the them as recognized for the western shoreline of poorly known, nearby Bugeen Tsav locality, into the Western Interior Seaway. They correlate strati- an interval of time equivalent to the Judithian graphically with the lower part of the Aquilapol- and/or Lancian North American land mammal lenites quadrilobus palynomorph Interval Zone of "ages" and, most probably, to the Maastrichtian the northern Rockies. stage as typified in western Europe.

INTRODUCTION This is the first systematic account ofmam- Cifelli and J. G. Eaton, personal commun.), malian remains from the Late Cretaceous the Kirtland and Fruitland formations ofNew "Mesaverde" Formation ofthe North Amer- Mexico (Clemens et al., 1979; Flynn, in press), ican western interior. As discussed below (see the El Gallo Formation (Clemens, 1980) of Geological Framework), the name Mesa- Baja California, Mexico, Eutaw Formation verde Formation is used incorrectly in Wyo- (Emry et al., 1981) of Mississippi, or Mount ming; thus we consistently place the name Laurel Formation (Krause and Baird, 1979) within quotation marks to distinguish its use of New Jersey. in Wyoming from its proper usage nearer the New faunas from two general areas (see type section in southwestern Colorado. figs. 1-3; Locality Data, below) within Wy- The fossils described herein are of special oming are described. The first is from mul- biogeographical importance because they tiple localities in the north-central part ofthe represent the most southerly, well-docu- state in the Bighorn Basin, with most sites mented record of Judithian (revised defi- being northeast of the town ofWorland. The nition, see under Biostratigraphy) mammals second is from two, closely adjacent localities known from the continent. Although prelim- at the center of the state in the Wind River inary, we considered the development ofthis Basin, in the area of the Rattlesnake Hills paper worthwhile because a new collecting Anticline (Barwin, 1959, 196 la, 196 lb; Os- effort in the "Mesaverde" Formation has been trom, 1965; Shapurji, 1978). Virtually all of mounted by Lillegraven. It will be some time the reported specimens occur along with before the results ofthat could be made avail- abundant nonmammalian remains within able in published form. Because of the lim- yellow channel sandstones, with considerable ited number of specimens available, we are reworking of the fossils. Most mammalian not considering elements ofmammalian fau- specimens are represented by isolated teeth nas of Late Cretaceous age from Utah (R. L. or edentulousjaw fragments as shown through 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 3

the various collecting techniques of surface- deposits in older Cretaceous rocks of Wyo- crawling, quarrying, underwater screen- ming and Utah, including the "Mesaverde" washing, and dry-screening. No indication of Formation (Barwin, 1959, 1961a, 1961b). the existence ofarticulated mammalian skel- Only one locality, however, the future site of etal remains has yet been observed. Barwin Quarry (see fig. 2 and Locality Data, below), central Wyoming, was judged prom- ising. ABBREVIATIONS Prospecting at this site yielded the first fos- INSTITUTIONAL sil mammals from the "Mesaverde" For- mation. Searches in nearby areas of outcrop AMNH, Department of Vertebrate Paleontology, west of the minimally productive Barwin The American Museum of Natural History, New Quarry were carried out in 1965 with the York of new and NMC, National Museum of Canada, Ottawa hope finding better sites. Results PMAA-P, Provincial Museum and Archives of were negative, however, and a decision was Alberta, Paleontological Collections, Drum- made to work Barwin Quarry. The richness heller ofthe site was minimal, yielding several teeth UA, Collection ofFossil Vertebrates, Department per ton of rock processed. A full-scale wash- ofGeology, The University ofAlberta, Edmon- ing program was continued at Barwin Quarry ton in 1966. UCMP, Museum of Paleontology, University of After extensive overburden was removed California, Berkeley by blasting and bulldozing, approximately USGS, United States Geological Survey 100 identifiable mammalian specimens and UW, The Geological Museum, The University of a Wyoming, Laramie large number of other vertebrate remains eventually were obtained from Barwin Quar- STANDARD DENTAL MEASUREMENTS ry by screening about 80 tons of the fossil- iferous sandstone. Unfortunately, total pro- AP, Anteroposterior length ductivity was even less than the minimal value ANW, Anterior width (=width oftrigonid oflower anticipated, so excavation was terminated at molariform teeth; =width of tooth on teeth in the end of the 1966 season. However, com- which only one width measurement was taken) paratively minor excavations were made by POW, Posterior width (=width oftalonid oflower the AMNH crew molariform teeth) in 1970. The collections LTRI, Length of trigonid made in 1965, 1966, and 1970 were cata- logued by Thomas H. Rich, but no research MISCELLANEOUS was undertaken until Lillegraven rejuvenated the present project in 1981. ACSN, American Commission on Stratigraphic Surface collecting, including dry-screening, Nomenclature was done by J. Howard Hutchison and Mi- IUGS, International Union ofGeological Sciences chael T. Greenwald (University ofCalifornia, MPR, Mongolian People's Republic Berkeley) at a newly discovered extension to NACSN, North American Commission on Strati- the east ofthe fossil-bearing level represented graphic Nomenclature at NALMA, North American land-mammal "age" Barwin Quarry in August of 1976. Two P4, ml, Tooth designations: capital letters, upper days were spent at the site, resulting in the jaw; lower case letters, lower jaw UCMP specimens used in the present study. The new locality (UCMP V-81101, Fales Rocks 1) is the same as UW V-81006, Fales HISTORY OF COLLECTING Rocks. WIND RIVER BASIN Lillegraven visited UW V-81006 for one day in June of 1981, collecting a few teeth Earlier successes in fossil recovery by through dry-screening. He revisited the site screen-washing of large volumes of sedi- in November of 1982, collecting about half ments in the type Lance Formation ofeastern a ton of rock for screen-washing. A sample Wyoming prompted explorations by Mc- of roughly 25 tons was taken by Lillegraven Kenna in 1961 to similar point bar and pond from the Fales Rocks locality in August of 4 AMERICAN MUSEUM NOVITATES NO. 2840

Lancian "Edmontonian" Judithian Aquilan

0 500 km

FIG. 1. Reference map of western North America showing general locations (numbers 5 and 6) of mammalian assemblages from "Mesaverde" Formation described in present paper (see fig. 2 for greater detail). Numbers refer to important Late Cretaceous mammal-bearing areas listed in table 10. See text conceming nature of "Edmontonian."

1983, but remains found there are not in- weeks in the area discovered by Case in late cluded in the present paper. July and early August of 1981. They pros- pected widely, searching with only modest success for new and richer mammal-bearing BIGHORN BASIN localities. Specimens recovered by surface The first mammals from the "Mesaverde" prospecting, dry-screening, and test screen- Formation of the Bighorn Basin were found washing are described in the present paper. in 1978 by Gerard R. Case. The specimens One site in the southwestern Bighorn Basin were recovered from many sites (see fig. 2 (UW V-81016, Late for Lunch locality, see and Locality Data, below) to the northeast of fig. 2) was discovered by Kenneth E. Jackson Worland and were given to McKenna for of Lillegraven's field party in July of 1981; it study and inclusion within the AMNH col- was revisited briefly by Lillegraven's crew in lections; Case was searching primarily for re- August of 1982. The site appears highly mains of fossil sharks. Case returned to the promising for recovery of additional speci- area briefly in 1979. mens ofhigh quality, but the relative richness Lillegraven and his party spent roughly two was not appreciated until the 1982 visit. A 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 5

0 Pow

PARK CO. i ~~~RIVER ARIN CH '1 Cody S

I _ HARTVILLE I I < UPLIFT

I, WYOMING 0 I O 10 1 BIGHORN CO. 0- ____ _ km- __ __ _ I_ _L____ 9\\ A

"Case s

HOT SPRINGS CO. Lucerne Thermopolis '111-.N w

*Shoshoni FREMONT CO. 0 Riverton

"0 HuIdsonl .0

Laonder

0 50 km Alcova 0 - FIG. 2. Reference maps of Wyoming showing semi-detailed positio;ns of mammalian assemblages from "Mesaverde" Formation. A, Inset, showing principal tectonic features surrounding Bighorn, Wind River, and Powder River basins. B, Outcrop pattern (solid black; traced from Love and Christiansen, 1983) of "Mesaverde" Formation of Bighorn and Wind River basins showing approximate areas for the Case sites, Late for Lunch locality, and Barwin Quarry-Fales Rocks (see section on locality data for more detail). 6 AMERICAN MUSEUM NOVITATES NO. 2840

0N

, ans6& 6\"\ ,, NE a SW c1 s':bb\ .

FIG. 3. Cross section of lower part of "Mesaverde" Formation in southeastern Wind River Basin in area of mammal-bearing localities (see fig. 2 and section on locality data). Stratigraphic terminology follows Barwin (1961a; see fig. 4). Section is presented normal to strike (strike 42-47°W, dip 31-33°N), and shows local topography relative to horizontal (heavy dashed line) and prominent sandstone outcrops (dotted pattern). Base of section is top of main body of Cody Shale. Top of section is unexposed and within "Mesaverde" Formation at top of level (best seen to NW of measured section) rich in petrified wood. Numbers in parentheses are thicknesses of individual rock units. Solid triangle indicates top of Wallace Creek Tongue of Cody Shale as delimited by Barwin (1961 a); we place contact lower because of presence of nonmarine fossil vertebrates within contested part of section. Fales Rocks (spelling as used on USGS Garfield Park Quadrangle, 1959, 7.5 min topographic map) is a series ofgray, oil-stained sandstones that serves as a prominent local landmark. sample of about half a ton was secured in verde" Formation (northwestern and central June of 1983, but fossils recovered then are Wyoming). All fossils are from nonmarine not included in the present report. strata, and a tacit assumption is that the var- ious rock units are contemporaneous at a coarse level of resolution. Such an assump- GEOLOGIC AGE FRAMEWORK tion is important, because it implies that most BASED ON ZONATION BY MOLLUSCS differences observed within species collected from the various areas are a result of indi- INTRODUCTION vidual or populational variation rather than Most comparisons made within the Sys- evolutionary change expressed through a sig- tematic Paleontology section are made among nificant interval of geologic time. Is the as- fossils recovered from the upper part of the sumption of contemporaneous existence jus- Oldman Formation (southern Alberta; Ju- tifiable? dith River Formation, Oldman beds of The primary means for determining tem- McLean, 1971), the Judith River Formation poral correlation for strata ofLate Cretaceous (north-central Montana), and the "Mesa- age in the North American western interior 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 7

I B I G H O R N / BIGHORN BASIN

" OWL CREEK '- WIND RIVER 0, ss\\ BASIN

ND t-s_, "4 G \I MOU

100__~~km

SE WIND RIVER BASIN CASPER ARCH SW POWDER RIVER A BASIN 0% # _ A B C D E F

^-'.,.OMEETEETSE"j-*-M T N K: -6WA per tongu of LEWIS SHALE LEWIS SHALE < .,._ .~I~IiiI~ltongue of LEWIS SHALE Teapot Sands.ton...M" ; unnamed middle member __ unnamed tongue of STEELE SHALE A "'' Parkmanrs I_-- w--~ W Nace--WalCreek Tongue of CODY SH, STEELE SHALE ,::.:,,,,.-m,bhae; P yesnr. .... -. , C O DY SH ALE fluvial, lagoonal, and coastal A Barwin Quarry and Fales Rocks Locality swamp deposits littoral and nearshore marine deposits Z J]offshore marine deposits FIG. 4. Index map (above) ofpart of Wyoming (refer to fig. 2) and interpretive geologic cross section (below; modified from Barwin, 1961 a, pl. IV, not drawn to scale) to show local relationships ofimportant stratigraphic units. Approximate position ofmammal-bearing localities (Barwin Quarry and Fales Rocks locality) is shown by solid triangle. Cretaceous strata above Cody Shale are lost to erosion across uplifted Casper Arch in chosen section. is by the distribution ofmarine invertebrates. developed has been presented by Waage A history of how the basic framework of the (1975). We are concerned only with the Late Cretaceous System in the western interior was Cretaceous (Senonian). Detailed zonations 8 AMERICAN MUSEUM NOVITATES NO. 2840 exist, based primarily on marine molluscs, The Foremost Formation is underlain by that are generally assumed to be contempo- the Pakowki Formation (Dowling, 1915; rep- raneous geographically (see Kauffman, 1970, resenting a westerly equivalent of the upper 1975; Obradovich and Cobban, 1975). The part of the marine Lea Shale); the Pakowki following section reviews specifically how the Formation bears the distinctive Baculites ob- Oldman, Judith River, and "Mesaverde" for- tusus fauna, characterized by Jeletzky (1968, mations relate to marine invertebrate zona- p. 45). The B. obtusus fauna of southern Al- tions. The review is geographically based, berta correlates in age in the United States progressing southward. with rocks of the Sharon Springs Member of the Pierre Shale and its more westerly equiv- OLDMAN FORMATION alents such as the Claggett Shale (Hatcher and Stanton, 1903; Hatcher, 1904; Stanton and The Oldman Formation (Russell and Hatcher, 1905; see also Gill and Cobban, Landes, 1940) overlies and partly interdigi- 1965, fig. 3, 1966a, table 2, 1973, fig. 12). tates with the Foremost Formation (Oldman The Oldman Formation is overlain by the and Foremost members of the Belly River Bearpaw Formation (Hatcher and Stanton, Formation as discussed by Williams and 1903; Stanton and Hatcher, 1905; Forester Burk, 1964, p. 175). McLean (1971) suggest- et al., 1977), bearing at its base (in Alberta) ed extension of the name Judith River For- the Baculites compressus Zone, also recog- mation (type section in Montana) into the nized widely to the south (Cobban and Ree- plains of Canada, with relegation ofthe Old- side, 1952; Obradovich and Cobban, 1975; man and Foremost formations to the status Riccardi, 1983). Thus, the vertebrate assem- of beds. Although such terminology is now blages of the Oldman Formation are reason- in general use among Canadian geologists ably constrained in terms of their positions (e.g., Forester et al., 1977), for purposes of within the North American western interior convenience in designation of fossil beds we marine invertebrate zonation. retain the older terms (the extensive need for qualification as "Judith River Formation of JUDITH RIVER FORMATION Alberta" is thereby obviated). The Foremost Formation (Dowling, 1915) The type Judith River Formation (Hayden, is composed of fine clastic sediments, occa- 1871; see also Stanton and Hatcher, 1903; sional coals, and abundant strata bearing a Hatcher, 1904; Sahni, 1972; Gill and Cob- brackish-water fauna. The Oldman and Fore- ban, 1973) is biostratigraphically bracketed most formations were considered by Jeletzky in a similar fashion to the Oldman-Foremost (1968, p. 46) to represent the Baculites greg- complex of southern Alberta. That is, the oryensis (cephalopod) Zone of the United nonmarine strata are underlain by the Clag- States (see Cobban and Reeside, 1952, chart gett Shale, bearing the Baculites asperiformis 1Ob; Obradovich and Cobban, 1975, table 1). Zone (ofCobban and Reeside, 1952, p. 1020; The upper reaches ofthe Oldman Formation two zones higher than the B. obtusus Zone), in Alberta, however, probably involve lateral and overlain by the Bearpaw Shale, bearing equivalents of as many as four cephalopod the B. compressus Zone (ibid.). The Judith zones higher (approximately 2 million years River Formation grades eastward into the B. younger) than the B. gregoryensis Zone (i.e., gregoryensis Zone within the Pierre Shale progressing upward, B. scotti through B. ru- (ibid.). Thus the biostratigraphic correlation gosus zones; see Forester et al., 1977, fig. 2). between nonmarine strata ofthe Oldman and The principal assemblages offossil mammals Judith River formations is reasonably close from Alberta's Oldman Formation occur high and well controlled, in addition to corre- in the section. Dr. Dale A. Russell suggested sponding to the same regression of the sea. to Lillegraven (letter dated November 19, 1984) that the mammalian assemblages of "MESAVERDE" FORMATION the upper Oldman Formation may be as GENERAL: As discussed by Reeside (1924), young as the B. rugosus (=Exiteloceras jen- Weimer (1960), Fisher et al. (1960), and Mo- neyi) Zone. lenaar (1983), the strata in Wyoming referred 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 9 to in current literature as the Mesaverde For- Although no marine fossils have been re- mation or Group (Holmes, 1877: "Mesa covered from the unnamed middle member Verde Group") and Lewis Shale (Cross et al., in the immediate vicinity ofthe mammal lo- 1899) are significantly younger than the Me- calities, the presence of Inoceramus subcom- saverde and Lewis Shale at their type sections pressus was reported by Keefer and Rich in the San Juan Basin in southwestern Col- (1957, p. 73) from strata equivalent to the orado. Not only are the more northerly rock basal part ofthe member in the southeastern units younger, they also represent clastic de- corner of the Wind River Basin (see Barwin, bris from quite different transgressive-re- 1961 a, p. 34). Inoceramus subcompressus oc- gressive sequences of the Western Interior curs elsewhere in zones just below the Bac- Seaway (see Weimer, 1960, fig. 7). Moreover, ulites gregoryensis Zone (see Kauffman, 1975, a widespread unconformity exists within the fig. 4). "Mesaverde" of Wyoming at the base of the Underlying the local unnamed middle Teapot Sandstone Member (Gill and Cob- member of the "Mesaverde" Formation is ban, 1966b). Thus use of the names Mesa- the Wallace Creek Tongue (ofBarwin, 1961 a, verde Formation and Lewis Shale for strata p. 14) ofthe Cody Shale. The Wallace Creek in Wyoming represents a long-standing, al- Tongue has yielded a variety of marine in- though generally understood, error. vertebrates (see Barwin, 196 la, pp. 17-19), BIGHORN BASIN: Nearly all mammalian including Inoceramus subcompressus and specimens from the Bighorn Basin available Baculites sp. aff. B. haresi (identifications by for the present study come from the middle W. A. Cobban in written commun. to Bar- to upper parts ofthe "Mesaverde" Formation win, 1957). Elsewhere, Baculites haresi is also in the southeastern part of the basin. The characteristic of the zones just below the B. localities are a few kilometers northeast of gregoryensis Zone (see Cobban and Reeside, the No Water Creek section (number 5) of 1952, chart 10B). Gill and Cobban (1966b, fig. 1). According No marine invertebrates have yet been de- to their interpretation, the part ofthe section scribed from the upper part of the unnamed bearing the mammalian fossils would overlie middle member of the "Mesaverde" For- the Baculitesperplexus Zone (see Obradovich mation in the southeastern Wind River Ba- and Cobban, 1975, table 1) and would cor- sin. However, according to Barwin's (1961) relate to the east with: (1) the lower marine interpretation of the correlation of the "Me- part of the Parkman Sandstone Member of saverde" sections between the southeastern the "Mesaverde" Formation; and (2) the B. Wind River Basin and the southwestern gregoryensis Zone as seen in the Pierre Shale Powder River Basin (see fig. 4), the mammal- (see Gill and Cobban, 1966a). At least three bearing localities within the unnamed middle species of Baculites that represent zones member correlate with the Parkman Member slightly lower than B. gregoryensis underlie of the "Mesaverde" Formation. Cobban the "Mesaverde" Formation in the southern (1958, p. 116) reported the presence of Bac- Bighorn Basin within the Cody Shale. An ero- ulites gregoryensis within the Parkman Mem- sional unconformity separates the upper part ber in the southwestern Powder River Basin. ofthe mammal-bearing section from the sig- Following the above review, it seems rea- nificantly younger Teapot Sandstone Mem- sonable to accept the assumption that the ber ofthe "Mesaverde" Formation (Gill and mammalian assemblages of the Oldman, Ju- Cobban, 1966b, fig. 1). dith River, and "Mesaverde" formations do, WIND RIVER BASIN: All mammalian spec- indeed, represent essentially the same ceph- imens known from the Wind River Basin alopod-based biostratigraphic unit, correlat- come from the Barwin Quarry and Fales ing approximately with the marine Baculites Rocks sites (see figs. 3 and 4; Locality Data gregoryensis Zone (of Cobban and Reeside, below). These sites are within the lower part 1952); as discussed above, the material from of the mainly nonmarine unnamed middle the upper Oldman Formation may be as much member of the "Mesaverde" Formation as four zones younger. Because there is little (Rich, 1958) as described by Barwin (1959, evidence to suggest that the limits ofthe ma- 196 la, p.29, 1961b; Zapp and Cobban, 1962). rine invertebrate zonation are significantly 10 AMERICAN MUSEUM NOVITATES NO. 2840 time-transgressive from one geographic area ranges in conjunction with their known in- to another, we can further assume that the tervals ofgreatest abundance (Pessagno, 1967) mammalian faunas from Alberta, Montana, suggests that the foraminiferal fauna of the and Wyoming are essentially contempora- Red Bird Silty Member represents the cor- neous. relative of the Globotruncana elevata Sub- zone and/or the next higher Rugotruncana BASED ON ZONATION BY PLANKTONIC subeircumnodifer Subzone of the Gulf Coast FORAMINIFERA (fig. 5). In the GulfCoast, Rugotruncana sub- circumnodifer first appears as a rare element The Red Bird Silty Member of the Pierre of the upper half of the G. elevata Subzone, Shale (see Gill and Cobban, 1966a) as seen but becomes a common species only in the in east-central Wyoming at Redbird is ofpar- R. subeircumnodifer Subzone. Although un- ticular importance because it provides a link known from the section of the Pierre Shale between zonations based on ammonites and at Redbird, R. subeircumnodifer is well doc- those based on planktonic foraminifera. umented nearby in upper parts of the Nio- The Red Bird Silty Member contains a brara Formation in rocks unquestionably macrofauna characteristic of the Baculites older than the Red Bird Silty Member (Fre- gregoryensis (cephalopod) Zone (see Gill and richs and Dring, 1981, p. 65; Bergstresser, Cobban, 1966a, table 2) which, as discussed 1981, p. 31). The planktonic foraminiferal above, is a principal correlative ofthe mam- fauna of the Red Bird Silty Member thus al- mal-bearing strata of the nonmarine Old- lows correlation with the upper Taylorian man, Judith River, and "Mesaverde" for- and/or lower Navarroan stages of the Gulf mations. Unfortunately, the Late Cretaceous with a Navarroan cephalopod zones of the North American Coast (see Pessagno, 1969), western interior are represented by highly en- assignment being probable. demic species (see Kennedy and Cobban, BASED ON ZONATION BY BENTHONIC 1976; Young, 1963) that have not proven FORAMINIFERA useful for temporal correlation with warmer- water areas to the south. Moreover, the gas- As discussed by Caldwell and North (1984), tropods at Redbird are of little use in high- assemblages ofbenthonic foraminiferans may resolution correlation with areas outside the be diachronous, and species usually have ex- western interior (Sohl, 1967). tensive stratigraphic ranges that span several Bergstresser (1981, fig. 7) reported, among cephalopod zones. Nevertheless, a zonation other taxa, the following species ofplanktonic of marine Cretaceous strata of the North foraminifera from the Red Bird Silty Mem- American western interior has been estab- ber: Archaeoglobigerina cretacea, A. sp. cf. A. lished on the basis of benthonic species blowi, Globigerinelloides multispina, G. (Caldwell et al., 1978). The Eoeponidella linki prairiehillensis, G. volutus, and Heterohelix Zone (ibid., p. 551; same as "Zone XI" of globulosa. All of these widely distributed Caldwell and North, 1975, p. 323) encom- species constitute parts of the foraminiferal passes the lowest part of the Bearpaw For- zonation of Upper Cretaceous rocks in the mation plus underlying marine strata later- North American Gulf Coast (see Pessagno, ally equivalent to the Judith River Formation. 1967, 1969). Comparison of foraminiferal The E. linki Zone involves the Baculites species' ranges between the two areas suggests gregoryensis through Didymoceras stevensoni correlation of the Red Bird Silty Member of cephalopod zones (see Caldwell et al., 1978, the Pierre Shale with Pessagno's (1967, text- table 1, p. 503). figs. 3-5) Globotruncanafornicata-G. stuart- Benthonic foraminiferans from the Red iformis Assemblage Zone of the Gulf Coast. Bird Silty Member ofthe Pierre Shale at Red- Species of planktonic foraminifera of the bird, Wyoming, were listed by Bergstresser Pierre Shale are temporally long-ranging, and (1981, p. 86); this member is represented by thus are of limited utility to detailed bio- the Baculites gregoryensis (cephalopod) Zone. stratigraphic zonation. Nevertheless, com- Although Bergstresser (1981) did not discuss parison of species' concurrent stratigraphic the section at Redbird in terms of Caldwell 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS I1I

STANDARD NORTH ASSEMBLAGE SBOE OUE AMERICAN ZONES SUBZONES ZONULES STAGES

Abathompha/us t3 maryaroensis CZ CZ z ..Z. Q 4ZI 0 .z I' Globotruncana gansserl cc zK~~~~~~~~~~~~~~9 Rugotruncano Rugotruncana subpenny/i subcircumnodifer Globotruncana lapparenti s.s. i.4 (a) G/obotruncanc Globotruncona ca/carata elevato Pseudotextularla elegans Planog/obul/na Archoeog/obigerlna glabrata b/owi Dlctyomitra multicosPata

FIG. 5. Correlation chart for Upper Cretaceous sequences ofwestern GulfCoastal Plain and Caribbean Basin (abbreviated from Pessagno, 1969, pl. 2). et al.'s (1978) zonation, the Red Bird Silty from the E. linki Zone (Nodosaria proboscid- Member is a temporal equivalent of part of ea) is reputed unique to the Taylorian, and the Eoeponidella linki zone, originally de- it is a rare form in the Gulf Coast, originally fined in Saskatchewan. described from Europe (Cushman, 1946, p. We compared lists of species of benthonic 72). Another species (Glomospira gordialis) foraminiferans from Upper Cretaceous strata is restricted to sub-Navarroan strata in the in Canada (Caldwell et al., 1978), the Pierre Gulf Coast, but is identified from rocks in Shale at Redbird (Bergstresser, 1981), and at Canada known to be above the E. linki Zone the Gulf Coast (Cushman, 1946, pp. 9-13). (Caldwell et al., 1978, pp. 557-558). As Although the percentage of species endemic pointed out by Cushman (1946, p. 18), how- to the western interior is high, many were in ever, identification of G. gordialis from the common with the Gulf Coast. GulfCoast is in doubt, and the species is very Of the species shared between the Eoepo- similar to Recent forms. Thus, the presumed nidella linki Zone of Canada and the Gulf restriction of Nodosaria proboscidea or Glo- Coast, most in the latter area are found both mospira gordialis to sub-Navarroan strata in in Taylorian and Navarroan foraminiferal the North American GulfCoast inspires little stages (see Pessagno, 1969). Only one species confidence. 12 AMERICAN MUSEUM NOVITATES NO. 2840

Contrariwise, two species occur in the Eo- the Baculites gregoryensis Zone occurs at 76.1 eponidella linki Zone that, in the Gulf Coast, Ma. Large potential error, however, should are linked specifically with the Navarroan be anticipated for these crude and admittedly Stage; Reophax texanus is an index fossil for oversimplified calculations. Kennedy and the Navarroan (Cushman, 1946, p. 16) and Odin (1982, p. 590) estimated the Campa- Spiroplectammina semicomplanata, though nian-Maastrichtian boundary to be about 72 also known in the upper Taylorian, is char- Ma. Harland et al. (1982) placed it at 73 Ma. acteristic of the Navarroan Stage (ibid., p. Both of these estimates, however, are based 28). on rocks occurring higher in the section than Data from comparative stratigraphic ranges the B. gregoryensis Zone. of benthonic foraminiferans of the Eoeponi- della linki Zone cannot be used with total confidence in correlation to the Upper Cre- BASED ON EUROPEAN STAGES taceous sequence of the Gulf Coast. Never- GENERAL INFORMATION theless, the most parsimonious interpretation is consistent with that derived from plank- Attempts to correlate the nonmarine mam- tonic species, as discussed above. Using ter- mal-bearing Oldman, Judith River, and minology from the Gulf, correlation is most "Mesaverde" formations to the stages used probable with upper Taylorian and/or lower for the Late Cretaceous of Europe presently Navarroan stages, with correlation to the Na- depends most importantly on planktonic fo- varroan being probable. raminifera collected in laterally equivalent marine strata. As stated above, the cepha- lopod zonation of the North American tem- BASED ON RADIOISOTOPIC DATING perate western interior is too endemic to al- TECHNIQUES low satisfactory links with the more tropical Radioisotopic dates are not yet directly Tethyan realm. At the species level, gastro- available for mammal-bearing strata of the pods give similar results (Sohl, 1967, p. 9), Oldman, Judith River, or "Mesaverde" for- although genera from the Pierre Shale are mations nor for marine equivalents close to common to the Campanian and Maastrich- the Baculites gregoryensis (cephalopod) Zone tian of the Gulf Coast. Stratigraphic range in the North American western interior. data for the species of marine bivalves from However, Obradovich and Cobban (1975) the western interior discussed by Kauffman summarized data from potassium-argon (especially 1970, 1973, 1975, 1979) are sum- techniques applied to bentonite deposits marized by a method that does not allow found associated with bracketing cephalopod detailed interbasinal correlation. Further- zones. The weighted mean for several sites more, Kauffman (1968) based zonations for was 72.2 Ma (74.0 using corrected constants Caribbean inoceramids on European stages of Steiger and Jager, 1977; see Dalrymple, as determined through ammonites for the 1979) as determined from the Didymoceras North American western interior. nebrascense Zone, two zones higher than the Assemblages of radiolarians, marine dia- Baculites gregoryensis Zone. Obradovich and toms, and calcareous nannoplankton are not Cobban (1975) reported a weighted mean of sufficiently known from the Late Cretaceous 77.9 Ma (79.9 using new IUGS constants) for sequence of the western interior to inspire the Baculites obtusus Zone, some seven zones confidence in long-distance correlations. below the B. gregoryensis Zone. Following Palynomorphs, because of marked biotic the questionable procedure ofassigning equal provincialism within the western interior, durations to each of the nine cephalopod cannot be used for purposes of correlation zones intervening between the two available beyond that area. The mammals, of course, K-Ar age assemblages, we calculated an age cannot be linked directly to the European of 76.6 Ma (using new IUGS constants) for stratotypes, all of which are represented by the Baculites gregoryensis Zone. If the same marine rocks. Finally, the greatly scattered assumption is made for the entire suite of radioisotopic ages available from the western zones in Obradovich and Cobban's table 1, interior are difficult to tie with confidence to 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 13 stratigraphic sections in other parts of the In terms of Tethyan planktonic foraminif- world. era, Premoli Silva (1977, fig. 2) recognized Despite the above-cited practical restric- the base of the Globotruncana elevata Zone tions to correlation with the Old World as the base of the Campanian (note that Van stratotypes, the Oldman, Judith River, and Hinte, 1976 carried the G. elevata Zone into "Mesaverde" formations have been corre- the late Santonian, below the Placenticeras lated, with seeming confidence, to the Cam- bidorsatum Zone) and the top of the G. cal- panian stage in commonly used reference carata Zone as the Campanian-Maastrich- sources (see, for example, McGookey et al., tian boundary (Kennedy and Odin, 1982, ta- 1972; Williams and Burk, 1964; Kauffman, ble 7; see Marks, 1984). Comparison of 1977, 1979); the basis for all ofthese is ceph- species lists from Premoli Silva (1977, fig. 1) alopod and/or bivalve zonation. and Pessagno (1969, pls. 3-5) shows that both Rawson et al. (1978; also see Harland et defined the base of the G. elevata Zone (a al., 1982) summarized the histories of the Subzone in Pessagno's usage) using essen- concepts of the type Campanian (district of tially the same range zone criteria. Pessagno, Charente, west-central France) and Maas- however, considered the base ofthe Archaeo- trichtian (near the town of Maastricht in the globigerina blowi Zone (next below the G. Netherlands on the Belgian border; see Ken- elevata Zone) ofthe GulfCoast to be the base nedy, 1984) and the way these stages are rec- of the Campanian; this appears to be late ognized in practice today. The base of the Santonian in the usual European sense (Van Campanian is presently drawn at the base of Hinte, 1976, fig. 2) and as applied to the Gub- the Placenticeras (Diplamoceras) bidorsatum bio section by Premoli Silva. Although the (cephalopod) Zone (see Van Hinte, 1976, fig. Santonian-Campanian boundary problem as 2; Kennedy and Odin, 1982, table 6). The recognized in North America is of interest Campanian-Maastrichtian boundary is usu- (see Frerichs, 1980 for discussion), it is ofless ally recognized (see Birkelund et al., 1984; importance than the position of the Cam- Schulz et al., 1984; Surlyk, 1984) at the base panian-Maastrichtian boundary in reference of the Belemnella lanceolata (temperate be- to the mammalian faunas that constitute the lemnite) Zone (Acanthoscaphites tridens basis for the present paper. Tethyan Zone; Van Hinte, 1976, fig. 2). As summarized by Lanphere and Jones As summarized by Alvarez et al. (1977), (1978), general disagreement exists on the the Late Cretaceous marine section in the placement of the Campanian-Maastrichtian Umbrian Apennines of east-central penin- boundary within the fossil-rich marine se- sular Italy exposed at Gubbio holds a key to quence of the North American western in- worldwide correlation of biological/geologi- terior. Because of the proven utility of range cal events ofthat interval oftime. Abundant zones of planktonic foraminifera to distant planktonic foraminifera and well-defined correlation, we emphasize that procedure in magnetic polarity zones developed through a the discussion that follows. roughly 300 m thick Late Cretaceous section Comparison of species range zones from allow close temporal comparisons with ma- Premoli Silva (1977) for the Gubbio section rine rock units in distant oceanic basins. and Pessagno (1969) for the GulfCoast shows Premoli Silva (1977) used the Tethyan ver- that they used identical criteria for recogni- sion of the cephalopod zonation scheme dis- tion ofthe Campanian-Maastrichtian bound- cussed above to define the base of the Cam- ary. Both recognized the top of the Globo- panian and the Campanian-Maastrichtian truncana calcarata Zone as the boundary (fig. boundary at Gubbio. She recognized that the 5). The G. calcarata Zone is a total-range Gubbio section was well within the warm- zone for that species and contains the first water Tethyan realm during Late Cretaceous appearance of G. subcircumnodifer. The same time, thus allowing direct comparisons of sequence was described by Olsson (1964) for planktonic foraminiferal zonations with the the North American Atlantic seaboard in New Caribbean Basin and North American Gulf Jersey and Delaware, and he too considered Coast (see Kauffman, 1977 for reconstruction the top of the G. calcarata Zone as the Cam- of distribution of marine climatic zones). panian-Maastrichtian boundary. The same 14 AMERICAN MUSEUM NOVITATES NO. 2840 system was used by Sissingh (1977, 1978) for suggested for the Campanian-Maastrichtian calcareous nannoplankton zonation and by boundary by Berggren et al. (in press, fig. 1 Barrier (1980) for specific application to the and appendix III). Gulf Coast. As discussed in previous sections, fora- CAVEAT miniferans from the Red Bird Silty Member ofthe Pierre Shale (representing the Baculites In terms of application of the above infor- gregoryensis (cephalopod) Zone and the lat- mation to mammalian faunas discussed in eral equivalent of the mammalian assem- the present paper, one important lesson is blages in question) probably correlate with learned. Because ofthe proximity ofthe var- the late Taylorian and/or early Navarroan ious Oldman, Judith River, and "Mesa- stages of the North American Gulf Coast. verde" mammalian assemblages to the Cam- The Taylorian-Navarroan boundary is set at panian-Maastrichtian boundary as now the top of the Globotruncana calcarata (fo- recognized, we can no longer blithely consid- raminiferal) Zone, recognized at Gubbio, the er them to be Campanian in age. They could North American Atlantic seaboard, and the represent the late Campanian, early Maas- Gulf Coast as the Campanian-Maastrichtian trichtian, or both. boundary. Thus, because of a downward re- A provincial zonation is therefore needed location ofthe boundary, the mammal-bear- for the nonmarine faunas ofthe North Amer- ing strata of the Oldman, Judith River, and ican Late Cretaceous to avoid the use of du- "Mesaverde" formations that have so con- biously applicable interpretive names such as fidently been considered Campanian in age Campanian or Maastrichtian. Such a need (see Clemens et al., 1979) lie precariously close was recognized decades ago by Professor Lo- to the relocated Campanian-Maastrichtian ris S. Russell (1964, 1975), who developed a boundary, and possibly within limits of the local system of terminology. The provincial early Maastrichtian since that stage has now terms Aquilan, Judithian, and Lancian will been expanded. be used within the Systematic Paleontology Such a conclusion is incompatible with the section below, but in somewhat different in- concept of the Campanian-Maastrichtian terpretations from those originally intended boundary as it generally has been identified by Russell. The differences are specified in within the western interior on the basis of the section below entitled Biostratigraphy. molluscan faunas; the boundary on the basis of planktonic foraminifera would be roughly BASED ON MAGNETOSTRATIGRAPHY nine or ten cephalopod zones lower than the If we assume that the correlations dis- level selected, for example, by Jeletzky (in cussed above are correct, the Baculites greg- Cobban and Reeside, 1952) and by Mc- oryensis Zone and mammal-bearing Old- Gookey et al. (1972, fig. 9) and six or seven man, Judith River, and "Mesaverde" cephalopod zones below the level suggested formations should be close to the late part of by Obradovich and Cobban (1975). The ra- geomagnetic Polarity Chron 33 or the early dioisotopic dates (mean of 74.0 Ma, correct- part ofPolarity Chron 32 (see Palmer, 1983). ed) determined by Obradovich and Cobban Unfortunately, however, this part of the Up- (1975) for the Didymoceras nebrascense per Cretaceous section ofthe western interior (cephalopod) Zone ofthe western interior (near lacks a magnetostratigraphic zonation. the Campanian-Maastrichtian boundary as determined by Pessagno, 1969, and Olsson, 1964) are somewhat earlier than the date rec- IN RELATION TO PULSES OF ognized for the Campanian-Maastrichtian SEA LEVEL CHANGE boundary (72 Ma) by Lowrie and Alvarez Vail et al. (1977) summarized the impor- (1977, fig. 3) in their revised magnetic-po- tance of global cycles of changes in sea level larity time scale for the Late Cretaceous. Ra- (recognized in three orders of magnitude) to dioisotopic age estimates are close, however, interpretations of earth history, though they between the Obradovich and Cobban (1975) did not release data specific for Cretaceous date and the estimate of 74.5 Ma recently time. Ryer (1983) recognized a still smaller, 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 15 fourth-order, transgressive-regressive cycle gested that because of biotic provincialism, as pertaining specifically to the Cretaceous this palynomorph interval zone may be dif- sequence of Utah. Matsumoto (1980) dis- ficult to recognize in Canada. cussed sea level changes during Cretaceous time from a global perspective. SUMMARY OF GEOLOGIC SETTING Weimer (1960) recognized four important transgressive-regressive sequences of the The mammalian fossils from the Oldman, strandline of the Western Interior Seaway Judith River, and "Mesaverde" formations within Upper Cretaceous rocks ofthe Rocky were deposited essentially contemporane- Mountains; these correspond in a general ously in nonmarine sediments near the west- sense to third-order cycles of Vail et al. (1977). ern shoreline ofthe Western Interior Seaway The mammalian faunas of the Oldman, Ju- during the regressive phase of the Claggett dith River, and "Mesaverde" formations are cyclothem. The mammal-bearing sedimen- found within Weimer's third regressive (R3) tary sequences were essentially contempo- pulse. raneous in deposition with more easterly, Kauffman (1977) and Hancock and Kauff- marine sediments bearing the Baculites greg- man (1979) recognized a sequence similar to oryensis (cephalopod) Zone (or as many as Weimer's for the Upper Cretaceous of the four zones younger) of the North American western interior, but began their numbering western interior. We correlate the B. grego- system with the base of the Cretaceous. The ryensis Zone, on the basis ofstratigraphically mammalian faunas described in the present associated foraminiferans, with the upper relate to Taylorian and/or lower Navarroan forami- paper the regressive phase of the niferal stages of the North American Gulf Claggett cyclothem (R8) of Hancock and Coast; an early Navarroan assignment is Kauffman (1979, table 3). slightly favored. When linked by planktonic foraminifera to equivalents of the European BASED ON PALYNOMORPH BIOZONES stratotypes, the correlation suggests time of deposition of the mammalian assemblages Nichols et al. (1982) devised a zonation for near the end of the Campanian and/or the the Upper Cretaceous of the northern and beginning ofthe Maastrichtian, with possible central Rocky Mountains ofthe United States overlap. The mammals in on the basis of pollen and spores. Their bio- question probably zones have the unique advantages of strati- lived about 74 to 76 million years ago, near graphic utility within nonmarine rocks in the end ofgeomagnetic Polarity Chron 33 or subsurface exploration and, to some extent, the beginning ofPolarity Chron 32 and occur in recognizability within contemporary ma- stratigraphically within the lower part of the rine strata. The Judithian mammalian faunas Aquilapollenites quadrilobus palynomorph discussed herein fall within the Aquilapollen- Interval Zone of the northern Rockies. ites quadrilobus Interval Zone, the base of which is below the upper part of the type LOCALITY DATA Judith River Formation of north-central Montana, the part yielding the mammalian GENERAL INFORMATION fauna described by Sahni (1972). Further, the Locality data presented below are of sev- mammals are within Nichols et al.'s (1982) eral degrees of reliability (see also fig. 2). Po- informally recognized Siberiapollis monta- sitions of both localities in the Wind River nensis Subzone, the lower of the two sub- Basin and all UW localities in the Bighorn zones of the A. quadrilobus Interval Zone. Basin were plotted in the field with pertinent The Aquilapollenites quadrilobus Interval topographic maps in hand. Written com- Zone is extensive temporally, involving the munication from Dr. Donald Baird (Novem- early form of Baculites perplexus (below) ber 4 and December 29, 1981) to Lillegraven, through B. grandis (above) cephalopod zones on the other hand, suggested that positions for the western interior (see Obradovich and of the various Case Sites were not plotted in Cobban, 1975); this includes roughly 16 the field, but rather from memory some time cephalopod zones. Nichols et al. (1982) sug- after the collections were made. With one AMERICAN MUSEUM NOVITATES NO. 2840

exception, geographic coordinates ofthe Case 506). NE¼A of sec 27, T. 48 N, R. 91 W. Sites in all probability should not be consid- Collected by G. R. Case in 1978. ered accurate to less than a quarter-section. AMNH 109413, 109425, 109483 The exception is Case Site 5, which is essen- Jerry Case Five (UW V-81032 = AMNH un- tially the same as UW V-81032, as docu- numbered locality, Case Site 5; Gerard R. mented by photographs. Case field numbers MV496, 499, 516-517). Also included below are listings ofall avail- NE corner ofNW/4 ofNW/4 of sec 27 plus able mammalian specimen numbers from SE corner of SW¼A of SW¼A of sec 22, T. each locality. Many of the specimens, how- 48 N, R. 91 W. Collected by G. R. Case ever, were unidentifiable (edentulous jaw in 1978 and J. A. Lillegraven in 1981. fragments, fragmented teeth, isolated incisors AMNH 109418-109419, 109426- and canines, etc.) and do not appear in the 109435,109437-109446, 109448-109450, species discussions in the Systematic Paleon- 109462,109476,109485, 109487-109489; tology sections. UW 17071-17073 Topographic quadrangle maps involved Case Site 6 (unnumbered AMNH locality; are: Bighorn Basin-Blue Mesa (1914, 15 Gerard R. Case field number MV521). min; Late for Lunch locality); and, for the NW¼A of SW¼A of sec 22, T. 48 N, R. 91 remainder, Broom Draw, McDermotts Butte, W. Collected by G. R. Case in 1978. and Worland SE (all 1967, 7.5 min); Wind AMNH 109477 River Basin-Garfield Peak (1959, 7.5 min; Case Site 8 (unnumbered AMNH locality; Barwin Quarry and Fales Rocks). Gerard R. Case field number MV510). Pre- sumably south-center of sec 16, T. 48 N, Bighorn Basin (Washakie County, R. 91 W. Collected by G. R. Case in 1978. except UW V-81016) AMNH 109420 Case Site 8:30 (unnumbered AMNH locali- The following localities, with the exception ty; no field number). Presumably center of of UW V-81016, have similar lithologies of sec 16, T. 48 N, R. 91 W. Collected by G. friable, yellow channel sandstones with oc- R. Case in 1979. casional stringers ofclay galls and clay clasts. AMNH 109453-109455 The sandstones show local cementations that No locality name (unnumbered AMNH lo- weather into hoodoos and large cannonball cality; no field number), "12 mi NE of concretions. Worland." Collected by G. R. Case and R. Case Site 1 (unnumbered AMNH locality; Steiner in 1978. Gerard R. Case field number MV500). AMNH 109456-109460 NW¼ of sec 35, T. 48 N, R. 91 W. Col- Late for Lunch Locality (UW V-81016). lected by G. R. Case in 1978. South-central part of SW¼A of NW¼A of AMNH 109415 NW¼A of sec 25, T. 45 N, R. 97 W, Hot Case Site 2 (unnumbered AMNH locality; Springs County. On basal flanks of SE part Gerard R. Case field numbers MV497, 501- of tallest local hill below massive bands of 505, 507-509). SE/4 of sec 27, T. 48 N, R. white sandstone that make up the hill. Fos- 91 W. Collected by G. R. Case in 1978. sils come from a moderately cemented, AMNH 109414, 109416-109417, plant-rich channel sandstone set between 109424,109452,109461,109464-109466, low-grade lignite layers. The fossiliferous 109468,109475,109479,109481-109482, layer is only 10-15 cm thick and extends 109484, 109486 laterally less than 10 m. Collected by J. A. Case Site 3 (unnumbered AMNH locality; Lillegraven in 1981 and 1982. Gerard R. Case field numbers MV511- UW 17068-17070, 17087 515). Eastern center of sec 27, T. 48 N, R. Old Number One (UW V-81036). Near NE 91 W. Collected by G. R. Case in 1978. corner of NE/4 of SE/4 of NE/4 of SW/4 of AMNH 109421-109423, 109478, sec 16, T. 48 N, R. 91 W. Collected by J. 109480 A. Lillegraven in 1981. Case Site 4 (unnumbered AMNH locality; UW 17084 Gerard R. Case field numbers MV498, Old Number Four (UW V-8 1038). Center of 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 17

N1/2 of NW¼A of NW¼A of NE¼A of sec 21, UCMP 125336-125343, 125346- T. 48 N, R. 91 W. Collected by J. A. Lil- 125350;UW 15515-15518, 15520, 15535, legraven in 1981. 15538-15540, 15581, 17041-17062 UW 15529-15530, 17099-17101 Barwin Quarry (unnamed AMNH locality; Old Number Five (UW V-81040). SW¼A of no field numbers). 37 m northwest ofFales SE¼A of SE¼A of NE¼A of sec 21, T. 48 N, Rocks (UW V-81006) locality, at same R. 91 W. Collected by J. A. Lillegraven in stratigraphic horizon. Fossils occur in an 1981. identical geologic situation between the two UW 15531 named localities and fossils are present in Meteors (UW V-81066). Center of WI/2 of abundance between the two sites. Collected NE/4 of SE¼ of NW/4 of sec 27, T. 48 N, by M. C. McKenna and party in 1961, 1966, R. 91 W. Collected by J. A. Lillegraven in and 1970. 1981. AMNH 59695, 59697-59699, 80001- UW 15532-15533, 17095 80006, 80009-80022, 86301-86302, No Wind (UW V-81073). South-central part 86304-86319, 86321-86326, 86328- of NW/4 of NE/4 of SE¼ of sec 27, T. 48 86337, 86339-86343, 86345-86346, N, R. 91 W. Collected by J. A. Lillegraven 86348-86354, 86356-86365, 88482- in 1981. 88495, 88497-88499, 108676, 108678- UW 17096 108688 Sunset to Dawn Two (UW V-81076). NE¼A of SW/4 of SE¼ of sec 27, T. 48 N, R. 91 SYSTEMATIC PALEONTOLOGY W. Middle of outcrop area. Collected by J. A. Lillegraven in 1981. CLASS MAMMALIA LINNAEUS, 1758 UW 15522-15528, 17075-17083 SUBCLASS ALLOTHERIA (MARSH, 1880) Sunset to Dawn Three (UW V-81077). NE¼A of SW/4 of SE¼ of sec 27, T. 48 N, R. 91 ORDER MULTITUBERCULATA COPE, 1884 W. Low on outcrop area. Collected by J. SUBORDER PTILODONTOIDEA A. Lillegraven in 1981. SLOAN AND VAN VALEN, 1965 UW 17097-17098 FAMILY NEOPLAGIAULACIDAE AMEGHINO, 1890 (Natrona County) Wind River Basin GENUS MESODMA JEPSEN, 1940 Fales Rocks (UW V-81006 = UCMP Mesodma primaeva (Lambe, 1902) V-81101 and within general area ofUCMP 1 V-5833). Below top of southwestern slope Figure 6A-E; table ofhill 6450 (on Garfield Peak Quadrangle, HOLOTYPE: NMC 1890, right mandible 1959, 7.5 min topographic map) in NE¼A with p4ml. of SW/4 of SW/4 of SEI/4 of SW/4 of sec 4, TYPE LOCALITY: Oldman Formation in T. 33 N, R. 87 W. Vertebrate fossils occur valley ofRed Deer River near Steveville, Al- at multiple levels through at least 10 m of berta. yellowish to golden channel sandstone (see LOCALITIES REPRESENTED IN PRESENT figs. 3 and 4). The richest fossil-bearing STUDY AND REFERRED SPECIMENS: WIND level is within a basal ironstone-cemented RIVER BASIN: p4, AMNH 86305, 86323, pebble conglomerate and throughout the UW 15539 (fig. 6A, B); P4, AMNH 59697, overlying 2 m of more friable, yellowish 86304 (fig. 6C-E), 86356. sandstone. Fossils are seen both as isolated KNOWN DISTRIBUTION: Oldman Forma- occurrences in massively cross-bedded tion (Judithian), Alberta; Judith River For- sandstones and within concentrations of mation (Judithian), Montana; and "Mesa- clay-gall stringers alongthin bedding planes. verde" Formation (Judithian), Wind River A few thin lignite beds occur within sand- Basin, Wyoming. stone bedding planes. Collected by J. H. DESCRIPTION AND COMPARISONS: p4. The Hutchison and M. T. Greenwald in 1976 three p4s referred to Mesodmaprimaeva agree and J. A. Lillegraven in 1981. closely in structure with the holotype (Lambe, 18 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 1 TABLE 2 Measurements of Teeth Referable to Measurements of Teeth Referable to Mesodma primaeva Mesodma sp. AP ANW AP ANW p4 AMNH 86305 3.74 1.71 ml Wind River Basin UW 15539a 3.48 1.34 UW 17044a 2.11 1.01 P4 AMNH 59697 3.39 1.77 Ml Bighorn Basin 86304 3.60 1.89 UW 17083 2.36 1.41 aFrom Fales Rocks; unmarked teeth from Barwin Wind River Basin Quarry. UW 15538a 2.43 1.44 M2 Bighorn Basin UW 15528 1.61 1.41 1902, pl. 15, figs. 13, 14) and with AMNH Wind River Basin 77121, referred by Sahni (1972, p. 367) to AMNH 80003 2.24 1.96 this species. Anteroposterior lengths of spec- 86364 1.42 1.59 imens from the Wind River Basin (see table 108683 1.97 1.61 1) are below the range of variation reported a From Fales Rocks; unmarked teeth from Wind River by Novacek and Clemens (1977, table 1, p. Basin are from Barwin Quarry. 704) for specimens of M. primaeva from the Judith River Formation. Both complete specimens from the Wind River Basin 15538 (fig. 6F); M2, AMNH 80003, 86364, have ser- 108683. (AMNH 86305 and UW 15539) 11 DESCRIPTION AND COMPARISONS: Seven rations. isolated molars referable to the genus Me- P4. Three teeth from the Wind River Basin sodma but of uncertain species have been are referable to P4s ofMesodma primaeva on recovered. The largest among them (AMNH the basis of size and general morphology. 80003) may represent M. primaeva, but the However, several differences from the Mon- remainder probably belong to one or more tana specimens are noted (Sahni, 1972, p. undescribed smaller species. Measurements 369). First, the teeth from Wyoming are are provided in table 2. roughly a half millimeter shorter (table 1), A possibility exists as well that some of though they match well in length with teeth these teeth are referable to a species of Kim- from the same locality identified as p4s of betohia. Clemens (1963, p. 43) identified two this species. Secondly, the cusp count of the p4s from the type Lance Formation as ?Me- medial row in both complete specimens from sodma sp. Clemens later (1973a, p. 79) iden- Wyoming is seven, rather than six as seen in tified these teeth as Kimbetohia campi Simp- Montana. In both cases, however, the extra son, 1936 on the basis of a suggestion cusp is weakly developed and the main crest (personal commun.) from Robert E. Sloan. effectively has only five cusps; AMNH 59697 Sloan (1981, p. 150) defended that assertion has a minuscule first (anterior) cusp and by pointing out that the two isolated p4s from AMNH 86304 has the tallest and last cusp Wyoming are, as is true in P4s from New ofthe main crest incompletely bifurcated. Fi- Mexico presumed to be from K. campi, ".... nally the anterolabial and posterolingual rows intermediate in morphology between Mesod- have only single, strong cusps in contrast to ma thompsoni and those of various species the two seen in each case in specimens from of Ptilodus." The identification as K. campi the Judith River Formation. of isolated p4s from the Lance Formation of Wyoming is based on: (1) presumed inter- Mesodma sp. mediacy in p4 morphology (i.e., size, orien- Figure 6F; table 2 tation of labial wrinkles, lateral profile, INCLUDED SPECIMENS: BIGHORN BASIN: bifurcation of labial striations) between Me- ml, UW 17044; Ml, UW 17083; M2, UW sodma and Ptilodus (the holotype ofK. campi 15528. WIND RIVER BASIN: Ml, UW is an upper jaw); and (2) the interpretation 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 19

lpl.,.Prl, 1mm

1 F

E D

FIG. 6. Judithian Mesodma teeth from the "Mesaverde" Formation, Wyoming. A, UW 15539, right p4, Mesodma primaeva, labial view. B, Same, lingual view. C, AMNH 86304, right P4, Mesodma primaeva, labial view. D, Same, occlusal view. E, Same, lingual view. F, UW 15538, left M1, Mesodma sp., occlusal view. All x 10. that Kimbetohia was a phylogenetic inter- M2. The four available M2s follow Clem- mediate between Mesodma and Ptilodus. ens' (1963, p. 48) description of M2s of Me- Until physical evidence becomes more con- sodma, but show the lesser cusp formula vari- vincing, we do not consider documentation ation of 1:3:3-4; only one specimen (AMNH adequate for the former existence ofK. campi 86364) has four internal cusps. in Wyoming. ml. UW 17044 follows in all respects the SUBORDER INCERTAE SEDIS description provided by Clemens (1963, p. 45) for Mesodma sp. from the type Lance FAMILY CIMOLOMYIDAE (MARSH, 1888) Formation. SLOAN AND VAN VALEN, 1965 Ml. UW 15538 and 17083 follow most GENUS CIMOLOMYS MARSH, 1889 aspects ofClemens' (1963, p. 47) description clarki 1972 of MIs of Mesodma, but have a lower cusp Cimolomys Sahni, count (as also noted by Fox, 1971a, p. 920, Figure 7A-G; tables 3, 4 for a specimen of Campanian age from the HOLOTYPE: AMNH 77179, isolated right upper part of the Milk River Formation). p4. Cusp counts in the "Mesaverde" specimens TYPE LOCALITY: Clayball Hill locality, Ju- are 5:6:3-6. The internal cusp row of UW dith River Formation, Montana. 15538 differs from the usual situation in Me- LOCALITIES REPRESENTED IN PRESENT sodma by continuing all the way to the base STUDY AND REFERRED SPECIMENS: BIG- of the first cusp on the middle row. The an- HORN BASIN: ml, AMNH 109422, teriormost cusp on the internal row is mi- 1094235 109466, 109485; m2, AMNH nuscule and anteroposteriorly elongated, but 109444, 109457 (fig. 7B); P4, AMNH 109419, the last five cusps are all well formed and UW 17071 (fig. 7E-G); Ml, AMNH 109424, conical. 109425, 109438, 109456, 109464, 109479, 20 AMERICAN MUSEUM NOVITATES NO. 2840

A 7wxB - - I

1mm

., G ;\ F ..Fsk as!s

.~-F= c - - s 1 s .3|D~~~~~.Ak } FIG. 7. Judithian Cimolomys, Meniscoessus, and Paracimexomys teeth from the "Mesaverde" For- mation, Wyoming. A, UW 15535, right ml, Cimolomys clarki, occlusal view. B, AMNH 109457, left m2, Cimolomys clarki, occlusal view. C, AMNH 88485, left Ml, Cimolomys clarki, occlusal view. D, UCMP 125336, right M2, Cimolomys clarki, occlusal view. E, UW 17071, right P4, Cimolomys clarki, labial view. F, Same, occlusal view. G, Same, lingual view. H, AMNH 86346, right M2, Meniscoessus intermedius, occlusal view. I, AMNH 88484, right Ml, Paracimexomys priscus, occlusal view. All x 10.

UW 15532, 17072. WIND RIVER BASIN: in table 9 (p. 93). Roughly one-third of the ml, AMNH 86363, UW 15535 (fig. 7A); P4, dental measurements presented in table 3 of AMNH 108679; Ml, AMNH 59695, 59698, the present paper fall below the range ofvari- 59699, 80002, 86306, 88485 (fig. 7C); M2, ation documented by Clemens (lengths: ml, AMNH 86307, 88482, 108678, UCMP 4.3-5.5; m2, 3.2-4.2; P4, 2.9-3.1; Ml, 4.3- 125336 (fig. 7D). 6.0; and M2, 3.1-3.8; widths: ml, 1.8-2.5; KNoWN DISTRIBUTION: Judith River For- m2, 2.1-2.9; P4, 1.3-1.4; Ml, 2.1-3.1; and mation (Judithian), Montana; and "Mesa- M2, 2.7-3.0). The diagnosis provided by verde" Formation (Judithian), Bighom and Sahni (1972, p. 371) for C. clarki reads: Wind River basins, Wyoming. "Cimolomys clarki is smaller than, but sim- DESCRIPTION, COMPARISONS, AND Dis- ilar to, its Maestrichtian descendant, C. grac- CUSSION: Identification of isolated teeth of i/is." Unfortunately, Sahni did not present Cimolomys from the "Mesaverde" Forma- full measurement data, but estimates of den- tion to the level of species is most difficult tal size from his illustrated specimens suggest because of the scanty pre-Lancian record of that the type material of C. clarki corre- the genus available for comparative purpos- sponds to or is slightly smaller than speci- es. Diagnostic features useful in differentiat- mens reported in table 3 ofthe present paper. ing the various named species are few. Mea- ?Cimolomys sp. B from the Milk River surements are provided in table 3. Formation (Fox, 1971 a, p. 932) has teeth sig- The largest existing samples of teeth from nificantly smaller than those from the "Me- Cimolomys were described by Clemens (1 963) saverde" Formation, and Cimolomys sp. A, from the type Lance Formation (for C. gra- also from the Milk River Formation (Fox, cilis), with measurements provided by him 1971 a, p. 930), has morphological differences 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 21

TABLE 3 (discussed below) from the specimens pres- Measurements of Teeth Referable to ently under study. Thus, from the criterion Cimolomys clarki of size, the "Mesaverde" fossils probably are AP ANW best identified as C. clarki. Unfortunately, cusp counts of the cheek ml Bighorn Basin teeth provide little immediate help in iden- AMNH 109422 4.28 2.22 tification, as can be seen by study of table 4. 109423 - 1.91 Specimens from the "Mesaverde" Formation 109466 - 1.88 have mis that are intermediate in external 109485 4.35 2.06 Wind River Basin row number between Cimolomys gracilis and UW 15535a 4.24 1.76 C. clarki or ?C. sp. A, but more like C. clarki and ?C. sp. A in internal row number. m2s m2 Bighom Basin AMNH 109457 2.92 2.11 from the "Mesaverde" Formation are more advanced in external row count than C. clar- P4 Bighom Basin ki, and more like C. gracilis. P4s show basic UW 17071 3.99 1.56 similarity of counts between C. gracilis and Ml Bighorn Basin C. clarki. Mls from the "Mesaverde" For- AMNH 109424 4.20 2.51 mation are closer in cusp counts to C. gracilis 109438 - 2.52 than to C. clarki. Finally, cusp counts ofM2s 109456 4.25 2.46 from the "Mesaverde" Formation appear 109479 - 2.37 est. Wind River Basin lower than either C. gracilis or C. clarki. Thus, AMNH 59698 - 2.57 although little weight can be placed on the 59699 - 2.53 interpretation of cusp counts at present, the 86306 - 2.73 specimens from the "Mesaverde" Formation 88485 4.36 2.34 are basically in agreement with those of C. M2 Wind River Basin gracilis, with certain lower (more primitive?) AMNH 86307 3.00 2.88 numbers of cusps. 108678 3.30 2.90 Almost all other morphological features of 125336a 2.74 2.60 the teeth from the "Mesaverde" Formation a From Fales Rocks; unmarked teeth from Wind River fit the descriptions by Clemens (1963) for Basin are from Barwin Quarry. Cimolomys gracilis; because few differences

TABLE 4 Cusp Counts (Not Including Tiny Cuspules) of Cheek Teeth of Various Species of Cimolomys See text for discussion. C. clarki C. gracilis C. clarki ?C. sp. A Present paper, Clemens, 1963, type Lance Sahni, 1972, Fox, 197 la, "Mesaverde" Fm.; Archibald, 1982, Judith Milk River Fm. Hell Creek Fm. River Fm. Fm. ml 6-7:4 7-8 5-7 5-6:4 6:4 mode 7 mode 5 m2 5:2 4-6 2-3 4:2 ? mode 5 mode 2 P4 2:6:1 1-2 5-6 : 1-3 2:5-6:2 1:5 mode 1 mode 5 M1 6-7:7:4-5 5-8 7-10 : 4-6 5:6:6 ? mode 7 mode 8 M2 1:3:3-4 2-3 : 3-4 4-7 2:3:4 ? mode 2 mode 3 mode 4 22 AMERICAN MUSEUM NOVITATES NO. 2840 were noted between C. clarki and C. gracilis Formation (see Fox, 1971 a, fig. 6D, p. 931). by Sahni (1972), however, the descriptions ?Cimolomys sp. A also has an anterolingual also hold well for C. clarki. Nevertheless, a lobe on the crown (Fox, 1971a, p. 931) that few minor dental differences from C. gracilis has not been observed on C. gracilis, C. clarki, were noted during the course of the present or specimens from the "Mesaverde" For- study. For example, cusps on ml and cusps mation. Thus, by a process of elimination on the medial row of Ml on specimens from and with great equivocation, we suggest that the "Mesaverde" Formation are slightly less the specimens referred to Cimolomys from crescentic than the usual condition of C. the "Mesaverde" Formation show overall gracilis; the more pyramidal shape is pre- greater similarity to C. clarki than they do to sumed the more primitive condition. Sec- C. gracilis, and we thereby use the former ondly, AMNH 109457, an m2 from the "Me- name for these temporally roughly equivalent saverde" Formation, has a ridge, though not fossil remains. However, we do not feel that a strong one, that connects the anterior in- the differences from C. gracilis are constant ternal and external cusps; such a ridge is lack- enough to justify the development of a re- ing in C. gracilis (see Clemens, 1963, p. 78) vised diagnosis for C. clarki. and usually present in C. clarki (see Sahni, 1972, p. 371). Thirdly, adequately complete specimens from the "Mesaverde" Formation GENUS MENISCOESSUS COPE, 1882 have small accessory roots on Ml on either side of the interradicular crest as described Meniscoessus intermedius Fox, 1976b for C. clarki by Sahni (1972, p. 373); this is Figure 7H in contrast to the condition in C. gracilis (see HOLOTYPE: UA 12081, right mandible with Clemens, 1963, p. 79) in which usually it is m 1-2. only the lingual accessory root that is devel- TYPE LOCALITY: Lowermost Oldman For- oped. mation along South Saskatchewan River, Differences also exist in structure ofP4 be- about 48.3 km north of Medicine Hat and tween specimens from the "Mesaverde" For- 24.1 km west of Hilda, Alberta. mation and Cimolomys gracilis. As illustrat- REFERRED SPECIMEN FROM PRESENT STUDY: ed by Sahni (1972, fig. 111, p. 372), the P4 Isolated right M2, AMNH 86346 (fig. 7H) is proportionately longer and narrower than from Barwin Quarry, Wind River Basin. in the more rectangular tooth in C. gracilis KNOwN DISTRIBUTION: Lowermost Old- illustrated by Clemens (1963, fig. 35, p. 78). man Formation (Judithian) and possibly UW 17071 (fig. 7E-G), a long, narrow tooth, Foremost Formation (Judithian; see Fox, is more similar in proportions to that of C. 1976b, p. 1217), Alberta; and "Mesaverde" clarki than to C. gracilis. The relative length Formation (Judithian), Wind River Basin, ofthe shearing blade to total tooth length (as Wyoming. measured to the apical crest of the blade) is DESCRIPTION: AMNH 86346 has a cusp comparable in C. gracilis (blade length 61% formula of 2:3:5 and, except for its large size oftotal length; measured from Clemens, 1963, (AP, 3.79; W, 3.41) and higher cusp count, fig. 36, p. 78), C. clarki (59%; Sahni, 1972, differs little from M2s referred to Cimolomys fig. 111, p. 372), and UW 17071 from the clarki. Complex flutings adorn both sides of "Mesaverde" Formation (67%; from speci- the cusps ofthe middle row, the lingual sides men). of the external cusps, and the labial sides of the internal cusps; the labial and lingual sur- Specimens of ml from the "Mesaverde" faces of the crown are smooth. All cusps in Formation (e.g., UW 15535) possess a short each row are sharply separated from one posterolabial cingulum in common with another by precipitous valleys, and the rows Cimolomys gracilis (see Clemens, 1963, p. are even more severely separated by deep, 77); no mention of such a structure in C. straight valleys. The anterior cusp ofthe mid- clarki was made by Sahni (1972), and the dle row is connected by a strong, anteriorly structure does not appear in the described convex crest to the anterior cusp of the ex- specimens of ?C. sp. A from the Milk River ternal row. A similarly oriented but much 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 23 lower and weaker crestjoins the anterior cusps appears to exceed that ofAMNH 86346, and ofthe middle and internal rows. A strong and probably should be written as 2-3?:4:5-6?, precipitously sloping crest joins the posterior with the queries reflecting uncertainty due to cusp of the middle row to the labial base of heavy cusp wear. The approximate measure- the external row's last cusp. The valley be- ments of AMNH 77261 (M2 illustrated by tween the middle and internal rows is vir- Sahni, 1972, fig. 12G, p. 375) are AP, 4.4 tually open posteriorly, being closed by only mm and W, 4.2 mm, significantly larger than a low and delicate ridge. Cusp wear renders for the specimen from the "Mesaverde" For- apical shape difficult to interpret, but the fol- mation. Measurements ofm2 ofthe holotype lowing appears correct: internal row-cusp 1 of M. major (see L. S. Russell, 1937, p. 76) conical, cusps 2-5 subcrescentic; middle are AP, 4.7 mm and W, 3.2 mm, again rep- row-all subcrescentic; and external row- resenting a significantly larger tooth. cusp 1 conical, cusp 2 pyramidal. Identification of AMNH 86346 as Menis- A strong appression facet from contact with coessus conquistus (see Sloan and Russell, Ml covers the dorsal half of the enamel in- 1974, p. 6) or M. robustus (see Clemens, 1963, cluding all of the anterior surface of the an- table 10, p.93 and Archibald, 1982, table 13, terior cusp of the middle row and adjacent p. 86) can be eliminated on the basis of size bases ofthe anterior cusps ofthe external and alone, although numerous morphological dif- internal rows. The forward root, as viewed ferences exist as well. anteriorly, forms a low isosceles triangle with its base at the edge of the enamel; in side view it is only slightly greater in anteropos- terior girth than the posterior root. The back SUBORDER AND FAMILY INCERTAE SEDIS root, as viewed posteriorly, is more columnar GENUS PARACIMEXOMYS ARCHIBALD, 1982 in shape, has its origin dorsal to overhanging enamel, and is much narrower than the for- Paracimexomys priscus ward root. No accessory roots are present. (Lillegraven, 1969) COMPARISONS AND DISCUSSION: Identifi- Figure 7I cation of AMNH 86346 as Meniscoessus in- HOLOTYPE: UA 3231, isolated right M 1. termedius is based in large part upon com- TYPE LOCALITY: University of Kansas lo- parisons with the holotype (m 1-2) and cality KUA- 1, Scollard Formation, Alberta. PMAA-P72.14.1, an isolated Ml referred to REFERRED SPECIMEN FROM PRESENT STUDY: M. intermedius by Fox (1976b, see fig. 3, p. Isolated Ml, AMNH 88484 from Barwin 1220 of that paper). The morphology de- Quarry, Wind River Basin. scribed above for AMNH 86346 would be KNOWN DISTRIBUTION: Trochu local fauna, expected for an M2 (as yet unknown) of the Scollard Formation (Lancian), Alberta; Hell type material; all are remarkably Cimolomys- Creek Formation, Montana (Lancian); and like, presumably primitive for the family. The "Mesaverde" Formation (Judithian), Wind size of AMNH 86346 also agrees well with River Basin, Wyoming. that of the type material, since Fox (1976b, DESCRIPTION, COMPARISONS, AND DISCUS- table 1, p. 1220) reported the length of m2 SION: AMNH 88484 (fig. 7I) differs in four in the holotype as 4.1 mm (roughly 0.3 mm respects from the holotype of "Cimexomys" longer than the M2 described here). priscus (transferred to Paracimexomys by Ar- Although the existence of an M2 of Me- chibald, 1982, p. 111). The specimen from niscoessus ferox from the somewhat older Wyoming: (1) shows less "waisting" of the Milk River Formation was mentioned by Fox crown as seen in occlusal view; (2) has but a (1976b, p. 1220), the specimen remains un- single cusp (rather than two) on the internal described. Direct comparison of AMNH cusp row; (3) has its internal cusp row ter- 86346 with M. ferox (see Fox, 1971 a, p. 933; minating at the middle (rather than at the known in published form only by P4) is im- anterior end) of the third cusp of the middle possible. row; and (4) is smaller (AP, 2.24 estimated; The cusp formula of M2 for Meniscoessus W, 1.40). major (see Sahni, 1972, p. 376 and fig. 12G) Fox (197 la, p. 922) defined a new species, 24 AMERICAN MUSEUM NOVITATES NO. 2840

Cimexomys magister, from the upper part of dition in P. priscus to involve a fundamen- the Milk River Formation. Archibald's de- tally four-cusped external row. fining criteria for Paracimexomys include Although AMNH 88484 is nearly a mil- features seen in "Cimexomys" magister, thus limeter shorter than the holotype of Paraci- requiring use of the name P. magister. Di- mexomys priscus (and of P. magister), the agnostic differences cited by Fox that are use- specimen does fall within the smaller end of ful in distinguishing teeth from those of"C." the range documented by Archibald (1982, priscus were few, and the greater age of the table 17, p. 114) for P. priscus from UCMP enclosing rocks probably weighed in the es- locality V-73087 in the Hell Creek Forma- tablishment of the trivial name. As might be tion of Montana. Thus the only significant expected in a fauna of intermediate age, a difference observed between AMNH 88484 dilemma is seen in attempting to identify and P. priscus is the presence of but a single AMNH 88484 to the species level; the spec- cusp in the internal row, presumably a prim- imen shares features with Paracimexomys itive feature shared with P. magister. For that magister (single cusp in the internal row) on reason, and until better evidence to the con- one hand and with P. priscus (four well-de- trary is forthcoming, we favor identification fined cusps in the external row) on the other. of AMNH 88484 as P. priscus, thereby ex- The cusp count on the external row requires tending the geologic range of that morpho- more elaboration. Lillegraven, in his original species from the Lancian into the Judithian. definition of "C." priscus, interpreted the worn holotype as having four cusps in the external row. Fox (1971a, p. 922), upon reex- Multituberculata, new genus and amining the specimen, suggested that wear species, unnamed had probably obliterated a small posterior- SPECIMENS: p4 fragments, AMNH 86301, most external cusp, thereby indicating a five- 86302. cusp row. Fox's prime evidence for the ex- LOCALITY: Wind River Basin, "Mesa- istence of the fifth cusp in the holotype was verde" Formation (Judithian), Barwin Quar- the presence ofa transverse lingual valley that ry. terminates at the wear facet; a discrete fifth DESCRIPTION AND DISCUSSION: The two p4s cusp does occur on specimens from the Milk (AMNH 86301, anterior half of tooth but River Formation. lacking labial lobe; AMNH 86302, fragment Whether the worn holotype ofParacimex- ofmid-dorsal part ofblade) are too fragmen- omys priscus had a fifth cusp or not probably tary for adequate description or even tenta- never will be known certainly, but new spec- tive identification, but stand out nevertheless imens of P. priscus collected and described because of their enormous size relative to by Archibald (1982) provide pertinent infor- other known Late Cretaceous multitubercu- mation. UCMP 117033 (Archibald, 1982, fig. lates. We estimate that the total tooth length 39a) and AMNH 88484 show identical struc- of AMNH 86301 prior to breakage would ture on the posteroexternal-most cusp. In have been at least 11 mm, much larger than both, the anterior part of the cusp is a low any previously described Mesozoic multitu- cone that constricts posteriorly to a broad berculate. The tooth is a fully developed ptil- ridge that curves posteromediad to terminate odontoid structure (see Clemens and Kielan- at the labial base ofthe last cusp ofthe middle Jaworowska, 1979, p. 142), not showing row. Admittedly, there is a hint of a trans- trends toward reduction (or plesiomorphi- verse valley that partly separates the cone- cally small size) typical of Meniscoessus ro- from the ridge-like parts of the cusp, but bustus (see Clemens, 1963, p. 88). neither Archibald nor we consider the sepa- AMNH 86301 has a minimum of 11 stria- ration great enough to warrant recognition of tions, both on the lingual and labial sides. At two cusps; there is no significant break in the least the first three striations on both sides of cusp height when seen in side view. Thus, in the tooth run to the apices of serrations. The contrast to the situation in P. magister, we first six striations on both sides begin sepa- consider AMNH 88484 and the usual con- rately from one another, run completely par- 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 25

lmm I .1

A B C D FIG. 8. Judithian dryolestid, genus and species unidentified, AMNH 109462, fragmentary right lower molar, from the "Mesaverde" Formation, Wyoming. A, Labial view. B, Occlusal view. C, Posterior view. D, Anterior view. All x 15. allel, and do not branch. From the seventh M2 striation posteriorly, their ventral ends pro- AMNH 109465 gressively diverge more ventrally and ter- UW 17045 minate well away from the edge ofthe enam- Unidentifiable molar fragments el. AMNH 86302 provides little information AMNH 80001, 80004, 88486 other than the presence of well-developed, Edentulous mandible though not unusually strong, serrations at UW 17087 midblade, with a one-to-one relationship of striation (gently convex anteriorly) to serra- SUBCLASS THERIA tion. PARKER AND HASWELL, 1897 Unidentified Specimens of SUBLEGION DRYOLESTOIDEA (BUTLER, 1939) Multituberculates FAMILY DRYOLESTIDAE MARSH, 1879 In addition to the specimens discussed Dryolestidae, genus and species above, a number ofisolated teeth remain un- unidentified identified and are listed below. Though most Figure 8A-D are fragmentary or heavily worn, some are excellently preserved and should be identi- SPECIMEN: AMNH 109462 (fig. 8A-D), fiable once larger and better com- fragmentary right lower molar. samples LOCALITY: Case Site 5, Bighorn Basin, parative series become available. "Mesaverde" Formation (Judithian). Isolated incisors, lower and upper DESCRIPTION, COMPARISONS, AND DISCUS- AMNH 86318, 86325, 86345, 86351, SION: AMNH 109462 is roughly the labial 86352,108677,109468,109470,109475 two-thirds of a right lower molar, split UCMP 125344, 125345 anteroposteriorly. Present are the entire pro- UW 15523, 17100 toconid, labial half of the talonid, and frac- p4 tured base of the metaconid; no parts of the AMNH 86324, 86342, 88483 paraconid or roots are preserved. The frag- UCMP 125335 ment follows in most respects the description ml of lower molars of Dryolestes priscus by AMNH 109429 Simpson (1929, p. 60), and is not greatly dis- Anterior upper premolars similar from the illustration of teeth of the AMNH 88491, 108684, 108687 same species by Prothero (1981, fig. 5, p. 294). UW 17041, 17070 The size ofAMNH 109462 is comparable to P4 known specimens of D. priscus. UW 17042, 17043 The tooth is moderately worn, with ob- Ml vious wear on the apex ofthe broadly basined AMNH 80013, 86360, 108676, 109437, protoconid, on the pre- and postprotocris- 109446 tids, and on a postvallid facet (extending from 26 AMERICAN MUSEUM NOVITATES NO. 2840

I ,0" .. _ WAN_ .. .> ..B C

1mm

I-q

F

E

A

FIG. 9. Judithian Alphadon russelli from the "Mesaverde" Formation, Wyoming. A, AMNH 80022, left m3, occlusal view. B, UW 17098, right P1, labial view. C, Same, occlusal view. D, Same, lingual view. E, AMNH 109483, right P2, labial view. F, Same, occlusal view. G, Same, lingual view. H, AMNH 86359, right P3, labial view. I, Same, occlusal view. J, Same, lingual view. All x 10.

the posterior surface of the apex of the pro- preserved on the transversely ridged talonid. toconid ventrolingually to just above the top A weak, transversely aligned swelling (with a ofthe talonid, then dorsolingually up the pos- spot worn through to the dentine) is seen on terior wall of the metaconid). Oval appres- the lingual termination ofthe preprotocristid. sion facets are present just ventral to the pro- The protoconid shows no sign of a labial cin- toconid-paraconid notch and on the posterior gulum. surface of the talonid. Light, dorsoventrally Identification of AMNH 109462 as a dry- aligned rugosities of the enamel are present olestid is based on similarities with teeth of on the basal postvallid surface; the enamel is Dryolestes and Laolestes (see Prothero, 1981, otherwise essentially smooth. No cusps are pp. 293-296). Identification to a lower taxo- 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 27 nomic level is impossible until more com- plete materials are discovered. The specimen is exciting, however, in suggesting that dry- olestoids, until now thought extinct in North America since the Late Jurassic (see Kraus, 1979), survived as rare animals until nearly the end of the Mesozoic. Thus they join the triconodonts and symmetrodonts as rare Late Cretaceous relicts (see Fox, 1976a, 1984b) 1mm from earlier times of greater diversity.

INFRACLASS, ORDER, AND FAMILY V- INCERTAE SEDIS Falepetrus barwini Clemens and Lillegraven, in press HOLOTYPE: AMNH 86316, isolated right upper molariform tooth. TYPE LOCALITY: Barwin Quarry, "Mesa- C verde" Formation, Wind River Basin, Wy- oming. FIG. 10. Judithian Alphadon russelli from the REFERRED SPECIMEN: Isolated left upper "Mesaverde" Formation, Wyoming (continued). molariform tooth, UCMP 118602, UCMP A,UW 17073, right MI. B, UW 15516, right M2 locality V-77083, Judith River (reversed to appear as a left M2). C, UW 17053, Formation, left M3. D, AMNH 108680, left M4. All occlusal Montana. views, xlO. KNowN DISTRIBUTION: Judith River For- mation (Judithian), Montana; and "Mesa- verde" Formation (Judithian), Wind River TYPE LOCALITY: Uppermost part of Old- Basin, Wyoming. man Formation, near Irvine, Alberta. COMMENTS: Falepetrus barwini is de- LOCALITIES REPRESENTED IN PRESENT scribed elsewhere by Clemens and Lillegrav- STUDY AND REFERRED SPECIMENS: BIG- en (in press), and is included here only to HORN BASIN: mx, AMNH 109430, complete the listing of mammalian taxa 109435, 109452, 109474, 109480; right P1, known from the "Mesaverde" Formation of UW 17098 (fig. 9B-D); right P2, AMNH 109483 9E-G); AMNH Wyoming. The two teeth represent an ad- (fig. P3, 109427, vanced (i.e., protocone-bearing) therian 109432, 109441, 109473, 109489, UW mammal that shares specializations with nei- 17080; right Ml, UW 17073 (fig. 10A); M2, AMNH 109461. WIND RIVER BASIN: ther the marsupials nor eutherians. The m3, AMNH 80022 (fig. AMNH species can be added to the peculiar, and 9A); P1, 108688; clearly paraphyletic, assemblage discussed by P2, AMNH 80011; right P3, AMNH 86359 (fig. 9H-J), UCMP 125338, UW Kielan-Jaworowska and others (1979) as 17049; M2, "therians of metatherian-eutherian grade." AMNH 80012, 86308, UW 15516 (fig. lOB), 17054; M3, UW 17049; M2, AMNH 80012, INFRACLASS METATHERIA 86308, UW 15516, 17054; left M3, UW HUXLEY, 1880 17053 (fig. 10C); left M4, AMNH 108680 ORDER MARSUPIALIA ILLIGER, 1811 (fig. 10D); Mx, AMNH 80017, 86336. FAMILY DIDELPHIDAE KNoWN DISTRIBUTION: Oldman Forma- GRAY, 1821 tion (Judithian), Alberta; and "Mesaverde" GENUS ALPHADON SIMPSON, 1927b Formation (Judithian), Bighorn and Wind Alphadon russelli Fox, 1979a River basins, Wyoming. DESCRIPTION, COMPARISONS, AND DIS- Figures 9A-J, 1OA-D; table 5 CUSSION: Lower premolars. Because ofthe dif- HOLOTYPE: UA 14805, right maxillary ficulty of differentiating isolated lower pre- fragment with P3MI-4. molars of Alphadon russelli from those of 28 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 5 Measurements of Teeth Referable to Alphadon russelli AP LTRI ANW POW m3 Wind River Basin AMNH 80022 2.96 1.26 1.81 1.53 P1 Bighom Basin UW 17098 2.09 1.17 Wind River Basin AMNH 108688 1.12 P2 Bighorn Basin AMNH 109483 2.41 1.21 Wind River Basin AMNH 80011 1.24 P3 Bighorn Basin AMNH 109427 2.48 1.47 109432 2.62 1.67 109441 2.48 1.73 109473 2.61 1.76 109489 1.75 UW 17080 2.52 1.74 Wind River Basin AMNH 86359 2.67 1.89 UCMP 125338a 2.41 1.53 UW 17049a 2.74 1.74 MI Bighorn Basin UW 17073 2.72 M2 Bighorn Basin AMNH 109461 2.70 2.94 3.08 Wind River Basin AMNH 80012 2.41 2.83 2.96 86308 2.77 2.84 3.00 UW 15516a 2.87 2.92 3.18 17054a 2.48 3.01 3.09 M3 Wind River Basin UW 17053a 2.41 3.05 3.17 M4 Wind River Basin AMNH 108680 1.92 est. 2.37 a From Fales Rocks; unmarked teeth from Wind River Basin are from Barwin Quarry. other species ofAlphadon, none here is iden- new information on lower molar structure for tified. Almost certainly, however, lower pre- the species. molars of this species are represented within P1-3. The Pl-3s here referred to Alphadon the following list of teeth identified as "Al- russelli follow in most respects descriptions phadon sp. px." for P3 provided by Fox (1 979a) for that m3. AMNH 80022 (fig. 9A; table 5) is of species (p. 98) and for its close relative, A. the appropriate size and follows the descrip- praesagus (p. 93); the P1 (fig. 9B-D) and P2 tions of m3s of Alphadon russelli by Fox (fig. 9E-G), previously unknown, are recog- (1 979a, table 2, p. 96 and p. 99) from Alberta. nized on the basis of their small size (table Further description is deemed unnecessary. 5), narrowness, and more delicate features. mx. All teeth identified as "mx" ofAlpha- Fox (1979a, p. 98) stressed the basic mor- don russelli are fragmentary, and provide no phological similarity of upper premolars be- 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 29 tween Alphadon praesagus and A. russelli; the Oldman Formation. Secondly, stylar cusps those from the latter species are smaller and C and D in UW 17053 are approximately porportionately narrower, but Fox noted no equal in size. This is of modest interest be- differences in the Alberta material in regard cause one ofthe diagnostic features ofA. rus- to cingula or furrows in the crown. A few selli from A. praesagus cited by Fox (1 979a, minor differences are observable in P3 be- p. 98) is that the former usually has stylar tween specimens ofA. russelli from the Old- cusp C larger than D on M3. man and "Mesaverde" formations. For ex- M4. AMNH 108680 (fig. 10D), here re- ample, the lingual cingulum in three of the ferred to Alphadon russelli, is a nearly com- nine available specimens from Wyoming plete M4, lacking only the roots and the labial continues weakly (though uninterruptedly) part of the anterolabial expansion of the sty- forward across the anterior halfof the crown lar shelf. We provide detail on the tooth in to be continuous with the anterior cingulum; addition to that given for specimens from the the lingual cingulum in the Alberta speci- Oldman Formation by Fox (1979a, p. 98). mens stops short of the anterior root. Sec- The protocone is intermediate in height be- ondly, the anterior cingulum of the speci- tween the para- and metacones and is sharply mens from Alberta are ". . . more prominent recumbent anteriorly. A weak anterior lin- lingually than labially . . ." (Fox, 1979a, p. gual cingulum is present on the base of the 93). No such consistency is noted in the spec- protocone dorsal to the protoconule. No hint imens from the "Mesaverde" Formation; the of a posterior lingual cingulum exists. The anterior cingulum is either equally developed protoconule is strong, forming a broad pre- on both sides (the most common situation), protoconular shelf that continues labiad, ap- or is stronger on the labial or lingual side in parently to the (broken away) stylar cusp A. equal proportions. A third difference is the The postprotoconular wing, however, is weak, vertical anterolingual furrow of enamel on terminating at the rounded base of the para- the main cusp. As with P3s from the Oldman cone. The metaconule is strong, with the pre- Formation, the furrow is usually present in metaconular wing terminating just dorsal to the Wyoming specimens (5 of8); in two, how- the paracone-metacone notch; the post- ever, the furrow is double. metaconular wing ends immediately poste- Orthal shear facets are observable on sev- rior to the lingual rounded base of the meta- eral specimens on the anterior and posterior cone. The protocone is markedly basined crests and on the labial and lingual sides of between the conules. the main cusp. These are in addition to the The metacone is a round cone, apparently dominant, horizontally oriented grinding lacking a postmetacrista. It has a weak pre- wear surfaces. metacrista that narrows to a shallow notch at Ml. Remaining parts of UW 17073 (fig. the base ofthe postprotocrista. Only a narrow 1 OA) agree in morphology with Ml s de- stylar shelf exists opposite the metacone, scribed for Alphadon russelli from Alberta by ending at the posterior base of the strong, Fox (1979a, p. 98). No further description is rounded stylar cusp C (cusp "D" of Fox, necessary. 1979a, p. 98). The towering paracone has a M2. The M2s referred to Alphadon russelli strong, posteriorly directed postparacrista and from the "Mesaverde" Formation (fig. 10B) an even stronger preparacrista. The latter follow the descriptions by Fox (1979a, p. 98) projects straight labiad toward the (broken in all but trivial aspects. The Wyoming spec- away) styloconal area. Just anterolabial to imens, however, show no development of stylar cusp C is a minuscule teardrop-shaped protoconal cingula. cusp (cusp "C" of Fox, 1979a, p. 98) with a M3. Only one specimen (UW 17053; fig. tapered crest that projects a short distance 1OC) ofM3 referable to Alphadon russelli has anterolingually. Despite Fox's (1979a, p. 98) been recovered from the "Mesaverde" For- assertions, we see no didactic or functional mation. It differs in two respects from com- reason to assume that the cusp seen at the parable teeth of that species from Alberta. depth ofthe ectoflexus, set between the bases First, the Wyoming specimen lacks proto- ofthe paracone and metacone, should not be conal cingula, variably developed in M3s from referred to as cusp "C." Fox's cusp "C" seems 30 AMERICAN MUSEUM NOVITATES NO. 2840

justification, should not be interpreted as car- rying an implication of homology. Wear is difficult to interpret on AMNH 108680 because of surface breakages, but it is heavy on the apex ofthe metaconule while being nonexistent on the apex of the proto- conule. Wear is also heavy on the pre- and postparacrista and on the anterior surface of A the preprotoconular wing. Wear elsewhere is either light or impossible to determine. Alphadon halleyi Sahni, 1972 Figure 1 1A-E; table 6 HOLOTYPE: AMNH 77367, isolated left ml.

-- -.., 7.r. TYPE LOCALITY: Clambank Hollow local- ity, Judith River Formation, Montana. ap;.l BJ A LoCALITIES REPRESENTED IN PRESENT Lb"r 6 M.. - STUDY AND REFERRED SPECIMENS: BIG- m i. HORN BASIN: ml, AMNH 109428 (fig. llD), UW 17081; m4, UW 15524. WIND RIVER BASIN: p3, AMNH 86343 (fig. 1 lA- C); m2, AMNH 80021, UCMP 125347 (fig. llE). KNOwN DISTRIBUTION: Oldman Forma- tion (Judithian), Alberta; Judith River For- mation (Judithian), Montana; and "Mesa- verde" Formation (Judithian), Bighorn and Wind River basins, Wyoming. DESCRIPTION, COMPARISONS, AND DISCUS- SION: Alphadon halleyi represents the middle- C size version offour small species ofAlphadon (all smaller than A. russelli) known from the "Mesaverde" Formation of Wyoming. Slightly larger forms are A. sahnii and A. lulli and a tiny form is A. attaragos. As discussed below, samples ofA. halleyi identified by Fox (1979a) from the Oldman Formation and Sahni (1972) from the Judith River Forma- D E tion both appear to contain teeth referable to A. sahnii. FIG. 11. Judithian Alphadon halleyi from the 1 "Mesaverde" Formation, Wyoming. A-C, AMNH p3. AMNH 86343 (fig. lA-C) is identified 86343, left lower jaw with p3 and roots of ml: as Alphadon halleyi on the basis of size; the labial, occlusal, and lingual views, respectively. D, structure of p3 ofthis species was previously AMNH 109428, left ml, occlusal view. E, UCMP undescribed. It is a strongly double-rooted 125347, left m2, occlusal view. All x 10. tooth that lacks an anterior accessory cusp, an anterior keel on the main cusp, and basal to be nothing more than a supernumerary cingula. The main cusp is strongly keeled pos- cusp, of variable development, on an ex- teriorly, with a simple, anteriorly keeled pos- panded region of the anterior stylar shelf. In terior accessory (talonid) cusp. A moderately any case, identification ofstylar cusps as "A," developed ridge runs anterolingually, then "B," "C," and so forth is based entirely on straight anteriorly from the apex of the tal- topographic position and, without additional onid cusp to terminate on the posterolingual 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS surface of the main cusp, about halfway up TABLE 6 its height. Obvious wear on the tooth is seen Measurements of Teeth Referable to only on the posterior keel of the main cusp, Alphadon halleyi, New Species anterior ridge of the talonid cusp, and as a AP LTRI ANW POW sharply troughed vertical groove on the lat- eral side ofthe tooth at the junction between p3 Bighom Basin the main and talonid cusps. As is usual for AMNH 86343 1.60 0.79 Alphadon, the posterior mental foramen is ml Bighom Basin placed below the posterior root of ml. AMNH 109428 1.98 1.00 1.02 0.98 ml. AMNH 109428 (fig. 11D) and UW UW 17081 2.03 0.89 1.04 1.06 17081 match the size and morphology ofthe m2 Wind River Basin holotype almost perfectly, and although more AMNH 80021 1.91 0.82 - - heavily worn than the type, follow in all rec- UCMP 125347a 1.86 0.74 1.00 1.02 ognizable features the description provided m4 Bighom Basin by Sahni (1972, p. 381). No further descrip- UW 15524 1.74 - - - tion is necessary. Measurements of the ho- ml cast of holotype lotype taken by Lillegraven (table 6), how- AMNH 77367 1.98 0.93 0.97 1.06 ever, differ significantly from those reported by Sahni (p. 381). AMNH 77368 (ml) and a From Fales Rocks; unmarked teeth from Wind River 77369 (m2) from the Judith River Formation Basin are from Barwin Quarry. were identified by Sahni as A. halleyi, but not specifically cited in his text; they are referable STUDY AND REFERRED SPECIMENS: BIG- to A. sahnii. Similarly, UA 1736 and UA HORN BASIN: ml, AMNH 109414, 6987, respectively identified by Fox (1979a, 109421, UW 17075 (fig. 12A); m2, AMNH p. 100, fig. 3a, c) as m3 and MI ofA. halleyi, 109426, 109431, 109442, 109482, 109484 are also identifiable as A. sahnii. (fig. 12B), UW 15530, 15531; m3, AMNH m2. UCMP 125347 (fig. lE) and AMNH 109478, 109481; P2, UW 15522, 15526 (fig. 80021 (labial halfoftooth only) have features 12E-G); P3, AMNH 109445, UW 17082 (fig. characteristic of m2s of Alphadon and agree 12H-J); Ml, AMNH 109439; M2, AMNH in size and all other morphological features 109417; M4, AMNH 109443. WIND RIV- with the holotype of A. halleyi. ER BASIN: ml, UCMP 125350; m2, AMNH m4. The crown of UW 15524 is heavily 86310, 86330, 86362, 88488, UW 17057, damaged, but it has the greatly narrowed tal- 17059; m3, AMNH 86309, 86328, 86331, onid and posteriorly elongated hypoconulid UW 15520 (fig. 12C), 17056, 17061; m4, area characteristic of m4. Although the size AMNH 86332 (fig. 12D),UW 15540,17055; is appropriate for reference to Alphadon hal- mx, AMNH 80016, 86329; P3, AMNH leyi, no other diagnostic features are recog- 86341, 86357; Ml, AMNH 86337; M2, nizable. AMNH 86312, 86313, 86321, 86333, 108686 The few scattered isolated teeth here iden- (fig. 13B); M3, UW 17052 (fig. 13C); left M4, tified as Alphadon halleyi provide no phy- AMNH 86311, UW 17050 (fig. 13D). logenetic information beyond that discussed KNOWN DISTRIBUTION: Probably Oldman by Fox (1979a, p. 101) in terms of possible Formation (Judithian), Alberta; probably Ju- ancestry to A. lulli. dith River Formation (Judithian), Montana; and "Mesaverde" Formation (Judithian), Alphadon sahnii, new species Bighorn and Wind River basins, Wyoming. Figures 12A-J, 13A-D; table 7 ETYMOLOGY: Species named in honor of HOLOTYPE: UCMP 125337 (fig. 13A), iso- Dr. Ashok Sahni for his primary research on lated right Ml. the vertebrate fauna ofthe Judith River For- TYPE LOCALITY: Fales Rocks locality, mation. UCMP V-8 1101, "Mesaverde" Formation, DIAGNOSIS: Small species ofAlphadon, in- Wind River Basin, Wyoming. termediate in size between A. wilsoni and A. LOCALITIES REPRESENTED IN PRESENT marshi; molar cusps slightly inflated; P2-3 32 AMERICAN MUSEUM NOVITATES NO. 2840

lower (fig. 12A-D), are slightly inflated in _e general construction; the cusps lack the del- icacy characteristic of Alphadon marshi and A; 1mB A. wilsoni. The protoconid is markedly the tallest cusp, .~~ with the paraconid slightly lower than the ; metaconid. The metaconid is set posterior to ..g..f the protoconid, but not distantly so. The an- ''{.£ X~~~~~~~~~~~~~ terior cingulid begins labial to the apex ofthe paraconid and terminates on the anterolabial *,_a surface of the protoconid. The posterior cin- gulid begins just below the apex of the hy- lm poconulid and terminates on the posterola- bial surface of the hypoconid; a labial basal cingulid is lacking. The paraconid is usually vertically keeled anteriorly. The posterior face of the paraconid and the anterior face of the metaconid are usually rounded, but some- E Ær ~~F times each is lightly ridged vertically. The metaconid usually has a vertical posterior keel that runs posteriorly to a slight notch at the junction with an anterior cristid from the ~1 entoconid. The hypoconid is the tallest of the talonid cusps and the hypoconulid the lowest. The entoconid is set slightly posterior to the level of the hypoconid and, on m4, markedly so. The entoconid is laterally compressed and, although usually having a rounded posterior edge, sometimes is posteriorly keeled. An- FIG. 12. Judithian Alphadon sahnii, new teriorly, the entoconid has a sharply descend- species, from the "Mesaverde" Formation, Wy- ing cristid, occasionally with an accessory oming. A, UW 17075, left ml, occlusal view. B, AMNH 109484, right m2, occlusal view. C, UW cuspule. The entoconid is set anterolingually 15520, left m3, occlusal view. D, AMNH 86332, from the hypoconulid. The talonid is deeply right m4, occlusal view. E-G, UW 15526, left P2, basined, with the deepest part just labial to labial, occlusal, and lingual views, respectively. H- the notch between the posterior metaconid J, UW 17082, left P3, labial, occlusal, and lingual ridge and the anterior entoconid cristid. The views, respectively. All x 10. cristid obliqua is either straight or slightly convex labially, and commonly has a short, vertical anterior continuation that climbs high usually lacking anterolingual parts of basal up the posterior wall of the talonid. cingulum; P2-3 with anterior border of main The hypoconulid is compressed, poste- cusp unkeeled and with strong wear on pos- riorly recumbent, and with the long axis run- terior crest of main cusp. ning posterolingually to anterolabially. The DESCRIPTION: Lower premolars. Because of hypoconulid and hypoconid are connected by the difficulty of differentiating isolated lower a continuous crest, but the posthypoconid premolars of Alphadon sahnii from those of cristid courses more directly labiad. Al- other species ofAlphadon, none is identified though usually absent, occasionally a weak, here. Almost certainly, however, lower pre- posterolingually convex cristid at the extreme molars of this species are represented below posterolingual corner of the crown connects under List of Unidentified Specimens of the bases of the hypoconulid and entoconid. Therians; px. Wear is heaviest across all main crests of m 1-4. All molars ofthis species, upper and the trigonid, the length ofthe cristid obliqua, 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 33

CI

lmm

4 3J F H G 'H J FIG. 13. Judithian Alphadon sahnii, new species (continued), Alphadon lulli, Alphadon attaragos, new species, Paranyctoides megakeros, new species, and Gypsonictops lewisi from the "Mesaverde" Formation, Wyoming. A, UCMP 125337, Alphadon sahnii, new species, holotype, right Ml, occlusal view. B, AMNH 108686, Alphadon sahnii, right M2, occlusal view. C, UW 17052, Alphadon sahnii, left M3, occlusal view. D, UW 17050, Alphadon sahnii, left M4, occlusal view. E, UCMP 125349, Alphadon lulli, labial half of right M3 in fragment of maxilla, occlusal view. F, UW 17079, Alphadon attaragos, new species, holotype, left ml, occlusal view. G, AMNH 109420, Paranyctoides megakeros, new species, holotype, left ml, occlusal view. H-J, UW 17047, Gypsonictops lewisi, right p4 (p3 of traditional designation), labial, occlusal, and lingual views, respectively. All x 10. the apex of the entoconid, and the crests that accessory cusp is enclosed by a cingulum that join the apices of the hypoconid and hypo- runs from the apex of the posterior accessory conulid. Lighter wear is also observed on the cusp to the labial midbase of the main cusp. anterior and posterior cingulids, on the pre- The widest part of the tooth is just posterior and postvallid shear surface of the trigonid, to the level of the apex of the main cusp and and on the posterior shear surface of the hy- the lingual cingulum is strongest at that point. poflexid. The central parts ofthe trigonid and Wear is strong on the posterior crest of the talonid basins remain unworn even into ad- main cusp and on the conjoined lingual sur- vanced stages of wear. Similarly, the apices face of the main cusp-posterior accessory ofthe entoconid and hypoconulid retain their cusp. Obvious abrasion facets do not exist distinctness even into heavy stages of wear. elsewhere, even under conditions of heavy P2. UW 15522 and 15526 (fig. 12E-G), wear. clearly referable to Alphadon, are identified P3. The P3 (fig. 12H-J) is merely a larger as P2s of A. sahnii on the basis of size. The variation on the structure of P2, with a pro- teeth have strong, double, divergent roots. portionately larger posterior accessory cusp The main cusp is only weakly keeled ante- and taller main cusp. There is some tendency riorly, but has a sharply developed posterior for the cingulum across the anterolabial quar- crest. There is no anterior accessory cusp per ter of the main cusp to approach weak com- se, but a sharply constructed cingulum begins pletion. The wear pattern is identical to that at the anterolabial corner ofthe crown, wraps of P2. around the anterior edge of the main cusp, Ml. Stylar cusps A, B, C, and D (fig. 1 3A) and continues in an undulating course along are well separated from one another and the lingual base of the tooth to the apex of strongly developed on Ml; A is the lowest. the moderately developed posterior acces- Stylar cusp B is the largest and is conical sory cusp. A small basin between the postero- except for a short, forwardly projecting spur labial base ofthe main cusp and the posterior originating at the anterolabial corner of the 34 AMERICAN MUSEUM NOVITATES NO. 2840 cuspule's apex. Cusp C is a sharply pointed this was the usual situation in M3s of this cone, intermediate in height between stylar species. cusps B and D. Stylar cusp D is anteropos- M4. The paracone of M4 (fig. 13D) is by teriorly elongated. The paracone and meta- far the tallest cusp, with the metacone and cone are roughly equal in height and general protocone ofthe same height and the conules development; the lingual surfaces ofboth are the lowest. The preparacrista is a sharp, thin, broadly rounded. The preparacrista makes an bladelike crest that runs anterolabiad but ter- anteriorly convex arc between the apices of minates abruptly well short of the anterola- the paracone and stylar cusp B. The postpara- bial extreme of the stylar shelf. The labial crista is straight, joining directly with an end of the preparacrista hooks forward with equally straight premetacrista. The post- a weak connection to the anterior cingulum, metacrista makes a gentle posteriorly convex a continuation of the preprotoconule crista. arc to terminate at the posterolabial extreme The lingual surface ofthe paracone is round- of the stylar shelf. ed. The labial surface is flattened, and almost The protoconule and metaconule are mod- concave as the sharp paraconal apex curves erately scaled and about equal in develop- ventrolingually. The postparacrista is short, ment. The preprotoconule crista continues as straight, and sharply keeled, connecting with a cingulum along the base ofthe crown to the the base of a still shorter and sharper pre- apex of stylar cusp A. The postprotoconule metacrista. crista is a short, rounded crest that terminates The posterior wall of the metacone is flat, at the lingual base of the paracone. The pre- with a crest running from its apex in a pos- metaconule crista is a short, rounded crest teriorly convex arc to become continuous with terminating at the anterolingual base of the the labial margin of the stylar shelf. Stylar metacone. The postmetaconule crista ter- cusp C is weakly developed and anteropos- minates at a level posterior to the apex ofthe teriorly elongated. Minor undulations of the metacone at the base of the vertical wall edge ofthe stylar shelfare seen between stylar formed by the postmetacrista. The protocone cusps C and B. The latter is low and placed is broadly based and lacks all but hints of just posterolabial to the termination of the lingual cingula. preparacrista. A weak notch separates stylar Heaviest wear is seen on the apices ofstylar cusp B from A, the tallest of the stylar shelf. cusps A-C, para- and metaconal crests, pro- The protoconule has a strong anterior cris- toconal and conular crests, and in the general ta that continues labially as a basal cingulum vicinity ofthe anterior faces ofthe paracone- to the apex of stylar cusp A. The postpro- preparacrista and adjacent cingulum plus sty- toconule crista is short, terminating at the lar cusp A. lingual base ofthe paracone. The metaconule M2. Except for the usual differences in co- lacks an anterior crista. The postmetaconule ronal proportions, the M2 (fig. 13B) differs crista is strong, terminating at the lingual base in only a few ways from the description given of the metacone. A pronounced indentation, above for Ml. Prime among these are: (1) as seen in occlusal view, exists on the pos- stylar cusps C and D in M2 are either sub- terior wall ofthe tooth between the metacone equally developed (the usual case) or D slight- and metaconule. ly exceeds the height ofC; and (2) the internal The protocone is anteroposteriorly com- crests of the conules tend to be sharp-edged pressed with strong pre- and postprotocon- rather than rounded. AMNH 109417 shows ular crests that terminate at the central bases weak lingual cingula on the protocone, but of the protoconule and metaconule, respec- the usual case (as in Ml) is to lack such cin- tively. There are no hints of lingual cingula. gula. The wear pattern is as in M 1 except that Specimens adequately worn to provide in- stylar cusp D comes into active wear sooner. formation on wear patterns are unknown. M3. UW 17052 (fig. 13C), heavily dam- RELATIONSHIPS: The teeth of Alphadon aged, is the only tooth referable to M3 of sahnii are most closely comparable to those Alphadon sahnii. Stylar cusp C is nearly as of A. wilsoni and A. marshi, species known strong as stylar cusp B; it is unknown whether from rocks of Lancian age. Alphadon sahnii 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 35 is intermediate in size between those two E per se, only the posterolabial extension of species (compare data in table 7 ofthe present a rather low postmetacrista. The wear pattern paper with figs. 16 and 17 of Lillegraven, on UCMP 125349 is as described by Clemens 1969), however, and differs as well in the (1966, p. 10). UW 17078 shows the conules minor features cited above in the diagnosis. to be strongly developed and fully winged. It is possible, though certainly not proven, There is no hint of protoconal cingula. that A. sahnii was ancestral either to A. wil- If properly identified, these two specimens soni or A. marshi, or to both. could be the first known pre-Lancian repre- As discussed in the section dealing with sentatives ofAlphadon lulli. They provide no Alphadon halleyi, it is highly probable that new information concerning phylogenetic re- AMNH 77368 (ml) and 77369 (m2) from lationships of the species. the Judith River Formation and UA 1736 (m3) and 6987 (Ml) from the Oldman For- mation also represent teeth ofA. sahnii. Alphadon attaragos, new species Figure 13F Clemens, 1966 Alphadon lulli HOLOTYPE AND ONLY KNOwN SPECIMEN: Figure 1 3E UW 17079 (fig. 13F), isolated left ml. HOLOTYPE: UCMP 47047, fragmentary left TYPE AND ONLY KNOWN LOCALITY: Sunset maxilla with M 1-2, alveoli of M3, and parts to Dawn Two locality, UW V-81076, "Me- of alveoli of P3 and M4. saverde" Formation, Bighorn Basin, Wyo- TYPE LOCALITY: Lull 2 quarry, UCMP ming. V-5620, Lance Formation, Powder River Ba- KNOWN DISTRIBUTION: "Mesaverde" For- sin, Wyoming. mation (Judithian), Bighorn Basin, Wyo- LOCALITIES REPRESENTED IN PRESENT ming. STUDY AND REFERRED SPECIMENS: BIG- ETYMOLOGY: Gr. attaragos, meaning mor- HORN BASIN: M3, UW 17078. WIND sel or bit, in reference to minuscule size. RIVER BASIN: fragmentary right M3, DIAGNOSIS: Minuscule size; metaconid of UCMP 125349 (fig. 13E). ml set far posterior to protoconid; hypocon- KNOWN DISTRIBUTION: "Mesaverde" For- ulid essentially conical; molar construction mation (Judithian), Bighorn and Wind River uninflated. basins, Wyoming; Hell Creek Formation DESCRIPTION: UW 17079 (fig. 13F) is a left (Lancian), Montana; and Lance Formation ml of minuscule size (AP, 1.68; LTRI, 0.74; (Lancian), Wyoming. ANW, 0.76; POW, 0.79) and uninflated con- DESCRIPTION AND COMPARISONS: The two struction. The metaconid is set far posterior fragmentary M3s from the "Mesaverde" For- to the protoconid, giving the trigonid the ap- mation agree well in size and general mor- pearance of an equilateral triangle (base at phology with homologous teeth ofAlphadon lingual side) in occlusal view. Although lulli from the younger Hell Creek Formation somewhat damaged, the hypoconulid ap- (Archibald, 1982, p. 128) and Lance For- pears to have been conical, rather than ridge- mation (Clemens, 1966, p. 8). Available den- like as in Alphadon sahnii. The tooth other- tal measurements ofthe Wyoming specimens wise follows the basic construction typical of are: UCMP 125349, AP, 2.06; and UW A. sahnii. 17078, POW, 2.70. Stylar cusp C is lacking RELATIONSHIPS: The hypodigm of a single in UCMP 125349 and apparently weak isolated lower molar does not allow adequate (enamel abraded away from labial edge of interpretation of phylogenetic relationships. stylar shelf) on UW 17078. The ectoflexus is The basic construction of ml, however, is rather broadly developed on both specimens. close to that of Alphadon sahnii, and A. at- UCMP 125349 shows stylar cusp B to be tall taragos may represent a counterpart of tiny and conical, A much lower, and D weak and body size. Alphadon attaragos is the smallest anterolingually elongated along the posterior species of the genus yet described; continued rim of the ectoflexus; there is no stylar cusp collecting with fine-meshed washing screens 36 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 7 Measurements of Teeth Referable to Alphadon sahnii, New Species AP LTRI ANW POW ml Bighom Basin AMNH 109414 2.32 1.06 1.18 1.21 109421 2.22 1.03 1.15 1.17 UW 17075 2.32 1.08 1.23 1.24 Wind River Basin UCMP 125350a 2.14 0.96 0.94 1.00 m2 Bighom Basin AMNH 109431 2.41 1.00 1.45 1.37 109442 2.49 1.05 1.42 1.50 109482 2.38 1.00 1.37 1.42 109484 2.16 0.87 1.29 1.30 UW 15530 2.27 0.96 1.40 1.34 15531 2.16 0.88 1.33 1.31 Wind River Basin AMNH 86310 2.33 0.95 1.46 1.41 86330 2.14 0.94 1.37 1.40 86362 2.28 0.91 1.31 est. 1.31 88488 2.32 0.99 1.44 1.38 UW 17057a 2.04 0.81 1.36 1.34 17059a 2.42 0.99 1.32 1.32 m3 Bighorn Basin AMNH 109481 2.42 1.06 1.62 1.44 Wind River Basin AMNH 86309 2.53 1.08 1.69 1.45 86328 - 0.98 1.47 - 86331 2.66 0.99 1.68 1.41 UW 15520a 2.60 1.07 1.58 1.43 17056a 2.19 0.91 1.44 1.24 17061a 2.15 0.84 1.37 1.26 m4 Wind River Basin AMNH 86332 2.41 0.94 1.52 1.17 UW 15540a 2.50 1.00 1.51 1.35 17055a 2.77 1.14 1.73 1.41 P2 Bighom Basin UW 15522 1.57 0.91 15526 1.52 0.95 P3 Bighorn Basin AMNH 109445 - 1.03 UW 17082 1.73 1.05 Wind River Basin AMNH 86341 - 1.09 86357 - 1.08 MI Bighom Basin AMNH 109439 2.31 2.47 2.62 Wind River Basin UCMP 125337a 2.40 2.25 2.43 M2 Bighorn Basin AMNH 109417 - 2.39 - Wind River Basin AMNH 86312 2.22 2.31 2.48 86313 2.49 2.57 2.84 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 37

TABLE 7-(Continued) AP LTRI ANW POW 86321 2.01 2.21 2.34 86333 2.32 est. - - 108686 2.14 2.23 2.48 M3 Wind River Basin UW 17052a 2.03 est. 2.55 2.75 est. M4 Wind River Basin UW 17050a 2.06 3.11 2.08 a From Fales Rocks; unmarked teeth from Wind River Basin are from Barwin Quarry. will be necessary for improving knowledge of SUPERFAMILY ERINACEOIDEA these interesting animals. (FISCHER DE WALDHEIM, 1817) FAMILY PEDIOMYIDAE (SIMPSON, 1927b) FAMILY CF. NYCTITHERIIDAE SIMPSON, 1928 Pediomys sp. GENUS PARANYCTOIDES FOX, 1979c INCLUDED SPECIMEN: UW 15533, isolated right trigonid of m2 or m3. Paranyctoides megakeros, new species LOCALITY: Meteors locality, UW V-81066, Figure 13G; table 8 "Mesaverde" Formation, Bighorn Basin, HOLOTYPE: AMNH 109420 (fig. 13G), iso- Wyoming. lated left ml. COMMENTS: The single specimen (probably TYPE LOCALITY: Case Site 8, "Mesaverde" m2 or m3) is identified as Pediomys sp. on Formation, Bighorn Basin, Wyoming. the basis of placement of the anterior ter- REFERRED SPECIMEN: AMNH 109469, iso- mination ofthe cristid obliqua, well labial to lated trigonid of mx (Case Site 2, Bighorn the notch between the crests of the protoco- Basin). nid and metaconid. Identification to the level KNoWN DISTRIBUTION: "Mesaverde" For- of species is impossible, though it is probable mation (Judithian), Bighorn Basin, Wyo- that the specimen (LTRI, 1.24; ANW, 1.24) ming. would fall within the range of variation ofP. ETYMOLOGY: Gr. megas, large, plus Gr. clemensi (see Fox, 1979b, table 1, p. 106 for keros, horn; after discovery in the Bighorn mean values). Basin. The rarity ofspecimens ofPediomys in the DIAGNOSIS: Lower molar morphology as in "Mesaverde" Formation comes as somewhat Paranyctoides sternbergi, but size of new of a surprise; it is a fairly common member species about 56 percent greater. of more northerly faunas, and is known from DESCRIPTION, COMPARISONS, AND DISCUS- Late Cretaceous faunas as far south as New SION: AMNH 109420 (fig. 13G) and 109469 Mexico and Baja California, Mexico (see follow without exception the description of Clemens, 1979, table 11-2, p. 194). Whether lower molars of Paranyctoides sternbergi by this represents an artifact of collecting bias Fox (1979c, pp. 119-121); molar construc- or a real ecological differentiation is un- tion seems to be more delicate than in P. known. Clemens (1973b, p. 163), however, also noted a comparative rarity of pedi- omyids within the Hunter Wash local fauna TABLE 8 ("Edmontonian") of the Kirtland and Fruit- Measurements of Teeth Referable to land formations of New Mexico. Paranyctoides megakeros, New Species

INFRACLASS EUTHERIA GILL, 1872 AP LTRI ANW POW ORDER INSECTIVORA CUVIER, 1817; ml AMNH 109420 2.71 1.05 1.73 1.75 mx AMNH 109469 - 0.99 1.63 - BOWDICH, 1821 AMERICAN MUSEUM NOVITATES NO. 2840 maleficus (see Fox, 1984, pp. 12-13). The River Basin are identified as G. lewisi on the teeth from Wyoming (table 8), however, are basis of stratigraphic occurrence (Judithian), more than half again as large as those from comparable size of teeth, and lack of mor- Alberta (Fox, 1979c, table 1, p. 116; Fox, phological differences from other specimens 1984, table 1), and it seems highly unlikely referred to the species. that the same species are represented. p4. UW 17047 (fig. 13H-J) agrees in size The only descriptive details that can be (AP, 2.24; W, 1.25) and major features of added beyond those by Fox (1979c, 1984) construction with AMNH 77428, a tooth have to do with tooth wear. AMNH 109420 from the Judith River Formation identified shows that the talonid (at least of ml) wears by Sahni (1972, fig. 15D-F, p. 392) as the much more heavily than the trigonid and that penultimate premolar of Gypsonictops lewisi. wear is principally horizontal on the crests Fox (1979c, p. 117) suggested that because that rim the talonid basin, not on the more known species ofGypsonictops have five low- vertical shearing surfaces. Postvallid wear on er premolars, the designation of p1-2-3-4-5 the trigonid greatly exceeds prevallid wear, should replace Clemens' (1973a, p. 6) rec- but the cristids receive the heaviest wear of ommended use of pa-b-c-3-4; Fox's termi- the trigonid. nology is applied here. Clemens (1973a, pp. The new specimens from Wyoming pro- 7-8 and fig. 5, p. 11) documented the high vide no additional information on the phy- degree ofindividual variation in p4 structure logenetic affinities of Paranyctoides. Al- in the Lancian species of Gypsonictops hy- though we use the higher-category systematic poconus; marked variation thus may also be hierarchy suggested by Fox (1979c, p. 119, expected in p4s of G. lewisi. Species identi- 1984, p. 9), we neither defend nor challenge fication at this point, however, is tentative. the existence of true nyctitheres in Late Cre- UW 17047, of robust construction, has a taceous time. The new specimens do, how- small anterior basal cusp set well above the ever, serve to document further the extent of enamel-dentine boundary immediately lin- diversification ofeutherian mammals during gual to the anterior ridge of the tooth's main Late Cretaceous time in North America (see cusp. A short but strongly developed cingu- discussions by Fox, 1979c, p. 124, 1984, p. lum descends ventrad and curves posterola- 16). biad from the basal cusp to terminate on the anterolabial base ofthe main cusp. The main cusp has a sharp vertical ridge that is slightly SUPERFAMILY INCERTAE SEDIS convex labially and runs from the apex to the GENUS GYPSONICTOPS SIMPSON, 1927a labial base of the anterior basal cusp. There is no hint of a metaconid. A strong postero- Gypsonictops lewisi Sahni, 1972 lingual ridge runs ventrad from the apex of Figure 13H-J the main cusp to terminate at the talonid. HOLOTYPE: AMNH 77429, isolated p4. Immediately lingual to that point of termi- TYPE LOCALITY: Clambank Hollow local- nation, an equally strong anteroposterior tal- ity, Judith River Formation, Montana. onid ridge forms the lingual edge of a small LOCALITIES REPRESENTED IN PRESENT and narrow but distinctly basined talonid. A STUDY AND REFERRED SPECIMENS: Wind Riv- second vertical ridge, less pronounced and er Basin, right p4 (p3 of traditional desig- more undulating in course, runs from the apex nation), UW 17047 (fig. 13H-J), UW locality of the main cusp down its posterolabial cor- V-81006 and lingual fragment ofMx, AMNH ner to terminate at the base of the cusp. 108682, "Barwin Quarry." The posterior edge ofthe talonid has enam- KNowN DISTRIBUTION: Oldman Forma- el broken away and its morphology is thereby tion (Judithian), Alberta; Judith River For- impossible to determine fully. It appears, mation (Judithian), Montana; and "Mesa- however, that the posterior edge of the tal- verde" Formation (Judithian), Wind River onid basin was closed by a transverse ridge Basin, Wyoming. that continued ventrolabially, then anterior- DESCRIPTION AND COMPARISONS: The two ly, to terminate at the ventrolabial base of specimens of Gypsonictops from the Wind the main cusp. The talonid basin is bordered 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 39 on its labial side by a sharp ridge that runs UCMP 125340, 125343 forward to terminate just labial of the pos- UW 15515, 15529, 17048, 17068 terolingual ridge of the main cusp. mx Wear on the tooth is almost entirely re- AMNH 80014, 80018, 86314-86315, stricted to its posterior half, especially on the 86319, 86334, 86339, 86348, 86349, back wall of the main cusp, on the posterior 86358-86359,86365,88487,88489- talonid rim, and on the broad labial slope of 88490, 109413, 109433-109434, the talonid. One small wear facet is present 109448-109449, 109459, 109476- on the labial wall of the main cusp, imme- 109477 diately ventral to its apex. An important wear UCMP 125346, 125348 facet is on the posterior wall ofthe main cusp, UW 15525, 15527, 15581, 17051, 17058, broad near the cusp's apex and narrowing 17060, 17062, 17069, 17076-17077 ventrally, with convergence of the facet onto Px the posterolingual ridge of the main cusp at AMNH 109416 about two-thirds of the way down its length. DP3 Strong posterolabially directed wear is also UCMP 125339 present on the ventrally sloping posterior edge Mx of the talonid and its broad labial slope. AMNH 86361 Mx. AMNH 108682 is a moderately worn UCMP 125341-125342 upper molar fragment with protocone, con- UW 17097 ules, and lingual cingula. The specimen pro- Edentulous jaw fragments, lower and upper vides no new information on morphological AMNH 80006, 80009-80010, 80019- variation in species of Gypsonictops. 80020,86340,86353-86354,88498- DISCUSSION: Unfortunately, neither UW 88499 17047 nor AMNH 108682 provides new in- UW 15518, 17096, 17099 formation on the phylogenetic affinities of Gypsonictops lewisi. If the identification is NEW TAXA AND correct, a range extension southward ofsome RANGE EXTENSIONS 575 km is documented for the species. The present research has allowed advances in knowledge oftaxonomic diversity, greater appreciation ofgeologic and geographic ranges UNIDENTIFIED SPECIMENS OF THERIANS of species, and increased biostratigraphic In addition to the specimens discussed utility of the fossils; results are summarized above, a number of isolated teeth and jaw below. fragments remain unidentified and are listed below. Although most are fragmentary or NEW GENERA heavily worn, some are excellently preserved Three previously unknown genera are rec- and should be identifiable once larger sam- ognized, though none is formally named ples and better comparative series become herein. They are: (1) "Multituberculata, new available. genus and species, unnamed," representing a Isolated incisors and canines, lower and up- presently unidentifiable species of extraor- per dinarily large body size; (2) "Dryolestidae, AMNH 80005, 86322, 86325 (2 speci- genus and species unidentified," a phyloge- mens), 88335, 88492-88493, 109415, netic relict represented by materials inade- 109440, 109455, 109460, 109471, quate for diagnosis; and (3) Falepetrus bar- 109488 wini, an advanced therian being described UW 15517, 17101, 17084, 17095 elsewhere (Clemens and Lillegraven, in press). px AMNH 80015, 86317, 86326, 88494- NEW SPECIES OF PREVIOUSLY 88495, 108681, 108685, 109418, DESCRIBED GENERA 109450, 109453-109454, 109458, Three new species of previously described 109472, 109486-109487 genera are named. They are: (1) Alphadon 40 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 9 3. Recognized for the first time both in Wy- Species-Level Taxa in "Mesaverde" Formation in oming and in northern Montana, Fale- Bighorn and Wind River Basins of Wyoming petrus barwini. Wind 4. Probably recognizable in northern Mon- Bighorn River tana and southern Alberta as well as Wy- Taxon Basin Basin oming, Alphadon sahnii. Mesodma primaeva + + Mesodma sp. + + GEOLOGIC RANGE EXTENSIONS INTO Cimolomys clarki + + + + JUDITHIAN RECORD Meniscoessus intermedius + + In addition to the new taxa summarized Paracimexomys priscus + Multituberculata, n. gen. above, ranges of Paracimexomys priscus and sp., unident. + (species only) and Alphadon lulli (species only) Dryolestidae, gen. and sp. are extended from rocks of Lancian age into unident. + the Judithian record. Falepetrus barwini + + Alphadon russelli + + + + JUDITHIAN OCCURRENCES PRESENTLY Alphadon halleyi + + + + RESTRICTED TO "MESAVERDE" Alphadon sahnii + + + + FORMATION OF WYOMING Alphadon lulli + + Alphadon attaragos + Including the new genera and species, the Pediomys sp. + following taxa are presently recognized in the Paranyctoides megakeros + Judithian only from the "Mesaverde" For- Gypsonictops lewisi + + mation of Wyoming: Paracimexomys pris- Symbols: +, present; + +, present, but also known cus, Multituberculata, new genus and species, from north of Wyoming in northern Montana and/or unnamed, Dryolestidae, genus and species southern Alberta. unidentified, Alphadon lulli, Alphadon attar- agos, and Paranyctoides megakeros. sahnii, most closely comparable to the Lan- cian species A. wilsoni and A. marshi; (2) Al- TAXA COMMON TO BOTH BIGHORN AND phadon attaragos, with the smallest body size WIND RIVER BASINS yet described for the genus; and (3) Para- If we look only at the Wyoming Judithian nyctoides megakeros, with a body size sig- record, counts from the existing faunal list nificantly larger than P. sternbergi. Para- (table 9) show that six species-level taxa are nyctoides megakeros represents the only common to both the Bighorn and Wind Riv- known occurrence of the genus south of Al- er basins, four are known only from the Big- berta. horn Basin, and six are known only from the Wind River Basin. Of the 12 species-level NEW RECORDS FOR WYOMING AND taxa known from the more southerly Wind GEOGRAPHIC RANGE EXTENSIONS River Basin, at least 8 also occur in rocks of All species-level taxa except Alphadon lulli Judithian age from farther north in Montana are reported from Wyoming for the first time. and/or Alberta (table 10); in all probability, These include both new taxa (summarized the Judithian mammalian fauna ofthe coast- immediately above) and geographic range ex- al western interior was essentially homoge- tensions. Speaking only for the Judithian, the neous geographically, at least as far as from following range extensions are documented: southern Alberta to central Wyoming. 1. Extended into Wyoming from northern Montana, Mesodma primaeva, Cimolo- BIOSTRATIGRAPHY mys clarki, Alphadon halleyi, and Gyp- sonictops lewisi. THE NATURE OF NORTH AMERICAN 2. Extended into Wyoming from southern LAND-MAMMAL "AGES" Alberta, Meniscoessus intermedius and The pioneering effort ofWood et al. (1941) Alphadon russelli. established a continent-wide framework for 1 986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 41

TABLE 10 Mammals in More Important Late Cretaceous Localities of North American Western Interior To conserve space on the printed page, summary information for areas 1-13 is presented immediately below. See text concerning nature of "Edmontonian." See figure 1 for geographic reference. Presumed "age" Area Formation Province or state represented Primary references 1 Upper part of Milk Alberta Aquilan Fox, 1970, 1971a, 1971b, 1972, River Fm. 1976a, 1980, 1981, 1982, 1984; Clemens et al., 1979 2 Lower part of Oldman Alberta Judithian Fox, 1976b; Clemens et al., 1979 Fm. 3 Upper part of Oldman Alberta Judithian L. S. Russell, 1952; Fox, 1974, Fm. 1979a, 1979b, 1979c, 1981; Clemens et al., 1979; this paper 4 Judith River Fm. Montana Judithian Sahni, 1972; Fox, 1979c, 1981; Clemens et al., 1979; Clemens and Lillegraven, in press; this paper 5 "Mesaverde" Fm. Wyoming (Bighorn Judithian this paper Basin) 6 "Mesaverde" Fm. Wyoming (Wind Judithian this paper River Basin) 7 St. Mary River Fm. Alberta "Edmontonian" L. S. Russell, 1975; Sloan and L. S. Russell, 1974; Clemens et al., 1979 8 Kirtland and Fruitland New Mexico "Edmontonian" Clemens, 1973b; Armstrong-Zieg- formations and Lancian ler, 1978; Clemens et al., 1979; Archibald, 1982; Lehman, 1984; Flynn, in press 9 Scollard Fm. Alberta Lancian Fox, 1974; Lillegraven, 1969; Clem- ens et al., 1979 10 Frenchman Fm. Saskatchewan Lanciana Johnston, 1980 11 Hell Creek Fm. Montana Lancian Sloan and Van Valen, 1965; Van Valen and Sloan, 1965; Estes and Berberian, 1970; Novacek and Clemens, 1977; Van Valen, 1978; Clemens et al., 1979; Lup- ton et al., 1980; Archibald, 1982; Archibald and Clemens, 1984 12 Type Lance Fm. Wyoming Lancian Clemens, 1963, 1966, 1973a; Fox, 1974; Clemens et al., 1979; Ar- chibald, 1982 13 North Horn Fm. Utah Lancian Clemens, 1961; Clemens et al., 1979 Area Taxon 1 2 3 4 5 6 7 8 9 10 11 12 13 Triconodonta Alticonodon lindoei + ?Triconodontidae, n. gen. and sp. + Multituberculata Mesodma formosa + + Mesodma cf. M. formosa + Mesodma hensleighi + + + 42 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 10-(Continued) Area Taxon 1 2 3 4 5 6 7 8 9 10 11 12 13

Mesodma primaeva + + + Mesodma senecta + Mesodma thompsoni + + + Mesodma cf. M. thompsoni + Mesodma sp. + + + + + Mesodma? sp. + ?Neoplagiaulax burgessi + ?Neoplagiaulacidae, gen. and sp. undet. Cimexomys antiquus + Cimexomys minor + + Paracimexomys magister + Paracimexomys cf. C. judithae + Paracimexomys judithae Paracimexomys magnus + Paracimexomys priscus + + + Kimbetohia campi + Ptilodontidae, gen. and sp. undet. Cimolodon electus Cimolodon similis + Cimolodon nitidus + + + + Cimolodon sp. + + Catopsalis joyneri + Stygimys kuszmauli Stygimys aff. S. kuszmauli ?Cimolomys sp. A + ?Cimolomys sp. B Cimolomys clarki + + + Cimolomys gracilis + + + Cimolomys trochuus + Eucosmodontidae, n. gen. and sp. + Meniscoessusferox + Meniscoessus intermedius + + Meniscoessus major + + Meniscoessus conquistus + Meniscoessus robustus + + Essonodon browni + + + Viridomys orbatus + Suborder and family incertae sedis + Symmetrodonta Symmetrodontoides canadensis 4H Dryolestoidea Dryolestidae, gen. and sp. unident. + Therians of metatherian-eutherian grade cf. Deltatheroides sp. + + + Potamotelses aquilensis 4+ Picopsis pattersoni H 1 986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 43

TABLE 10-(Continued) Area Taxon 1 2 3 4 5 6 7 8 9 10 11 12 13 cf. Picopsis sp. Tribosphenic upper molar + Falepetrus barwini + Bistius bondi + Marsupialia Alphadon creber + Alphadon halleyi + + + + Alphadon russelli + + Alphadon lulli + + + Alphadon sahnii + + + Alphadon marshi ?+ + + + Alphadon cf. A. marshi Alphadon wilsoni + + + Alphadon attaragos + Alphadon praesagus + + Alphadon rhaister + + + Alphadon cf. A. rhaister Alphadon sp. + + Alphadon?, n. sp. + cf. Peradectes sp. Glasbius intricatus + Glasbius twitchelli + Albertatherium primus + Didelphidae, gen. and sp. indet. + Pediomys exiguus + Pediomys clemensi + + Pediomys sp. A Pediomys prokrejcii Pediomys cooki + + Pediomys cf. P. cooki + + Pediomys elegans + ± + Pediomys cf. P. elegans + Pediomysflorencae + + ?Pediomys cf. P. florencae + Pediomys hatcheri + + + + Pediomys cf. P. hatcheri + Pediomys krejcii + + + Pediomys cf. P. krejcii + Pediomys sp. + ?Pediomyidae Aquiladelphis incus Aquiladelphis minor + + ~+ Eodelphis cutleri + Eodelphis browni + + + Eodelphis sp. Eodelphis? sp. + Didelphodon vorax + + + Didelphodon? sp. + Eutheria Gypsonictops lewisi + + + Gypsonictops, n. sp. + 44 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 10-(Continued) Area Taxon 1 2 3 4 5 6 7 8 9 10 11 12 13

Gypsonictops sp. + Gypsonictops hypoconus + + + Gypsonictops illuminatus + Gypsonictops cf. G. illuminatus + Cimolestes sp. + cf. Cimolestes sp. + Cimolestes incisus + + Cimolestes magnus + + + Cimolestes cerberoides + Cimolestes cf. C. cerberoides + + Cimolestes propalaeoryctes + + Cimolestes stirtoni + + Batodon tenuis + + + Procerberusformicarum + Palaeoryctidae, gen. and sp. undet. + Telacodon laevis + Paranyctoides maleficus + Paranyctoides sternbergi + Paranyctoides megakeros + Insectivora, n. gen. and sp. A + Insectivora, n. gen. and sp. B + Protungulatum donnae + Protungulatum cf. P. donnae + Protungulatum gorgun + Ragnarok harbichti + Oxyprimus erikseni + Mimatuta morgoth + Purgatorius ceratops + Eutheriaif? + a Although received too late to be incorporated within the table, Johnston and Fox (1984) reported the following mammals from the base of the Ravenscrag Formation (which overlies the Frenchman Formation), latest Lancian in age, from southwestern Saskatchewan: Mesodma thompsoni, Catopsalis, n. sp., Cimexomys cf. C. hausoi, Alphadon sp., Pediomys elegans, Procerberus cf. P. formicarum, Protungulatum cf. P. donnae, Oxyprimus cf. 0. erikseni, ?Oxyprimus sp., Ragnarok sp., and ?Mimatuta sp. temporal correlation ofNorth American Ter- Despite their long-standing success in prac- tiary nonmarine rocks that was based on as- tice, NALMAs do not fit precisely within semblages of mammalian fossils. Although definitions of any formalized category estab- heavily modified since its publication (see, lished for purposes of correlation or descrip- for example, Woodburne, in press), the North tion by the general geological community (see American Provincial Ages developed by the ACSN, 1970 and references therein; and Wood committee have been used extensively NACSN, 1983). The key elements of NAL- and successfully (see excellent review by Ted- MAs are specified within the following quo- ford, 1970) in geological and paleobiological tation, often overlooked, from the Wood research across the continent. Savage (1962) committee report (1941, p. 6): suggested modification ofthe name to North American Land-Mammal "Ages," leading to . . . they are meant to cover all ofTertiary time, the acronym (NALMA), commonly used to- without reference to whether each year, century, day. or millennium is represented by a known fau- 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 45

nule or stratum. The type ofeach age necessarily DORF's LANCIAN AGE belongs to it, and the sequence and approximate scopes of the ages are thus intended to be def- Dorf(1942, p. 105) recognized the need for initely fixed (barring new discoveries which a temporal term for the time of deposition should lead to radically different interpreta- of the Lance Formation: tions). However, the ages are not necessarily coextensive with their types, and the precise Definition ofLancian age-There does not exist limits between successive ages are intended to at present a clearly defined temporal term for be somewhat flexible and may presumably be the latest Cretaceous of the Rocky Mountain modified in the of later discoveries. region. For the practical purposes ofclarity and light precision it is here proposed to use Lancian age as a convenient provincial time term, based on It is clear that NALMAs were intended to the Lance Formation at its type locality near represent time units, as indicated above and Lance Creek, Niobrara County, Wyoming. This time unit is delimited below by true Fox Hills by the very use ofthe term "ages." According time (i.e., latest Montanan age, characterized by to present formalized usage (NACSN, 1983), marine sandstones comprising the well-defined however, an "age" (Art. 80) is a geochro- Sphenodiscus zone), and delimited above by the nologic unit that ". . . corresponds to the time beginning ofPaleocene time. The terrestrial sed- span of an established chronostratigraphic iments of Lancian age carry the characteristic unit (articles 65 and 66), and its beginning mammals' and dinosaurs2 of the Triceratops and ending corresponds to the base and top zone, as well as the plants here described. of the referent." According to the NACSN I Wood, H. E., et al., Bull. Geol. Soc. Amer., (1983, Art. 66), a chronostratigraphic unit vol. 52, 8, 1941. ". . . is a body of rock established to serve as 2 Russell, L. S., Proc. Amer. Philos. Soc., vol. the material reference for all rocks formed 69, 139-141, 1930. during the span of the same time." Under that definition, the boundaries oftime ofany It is debatable whether Dorf thereby de- geochronologic unit (age, epoch, period, etc.) fined a true age, a NALMA, or merely a geo- are synchronous (i.e., occur at the same time) chron (temporal duration of a rock unit-in with its chronostratigraphic (material) ref- this case, the Lance Formation). We believe erence rock unit. that Dorfwas attempting to create a NALMA As unequivocally stated in the above quo- in the manner ofthe Wood committee (Wood tation from the Wood committee, the tem- et al., 1941), which he cited, but that he failed poral limits of a NALMA are not restricted because his Lancian age is, in reality, either: by the time represented within the type sec- (1) bounded earlier by Fox Hills time and tion (chronostratigraphic referent unit) ofany later by Paleocene time; or (2) is the time given NALMA. As discussed by Tedford during which the type Lance Formation was (1970), that is the reason why Savage (1962) deposited. The fossils of the Lance Forma- suggested use ofquotation marks around the tion were used to identify this time, to be word "age" when used in reference to NAL- sure, but they do not define it. MAs; the term does not match exactly the concept used by the general geological com- THE NATURE OF RUSSELL'S STAGES munity. The formalized terminology ofa geo- L. S. Russell (1964, 1975) attempted to chronologic unit (NACSN, 1983, p. 848) and establish a system of biological correlation a NALMA agree in being purely conceptual, for nonmarine rocks of Early and Late Cre- dealing solely with geologic time, but differ taceous age in the North American western in that the former has limits defined by a interior. Although similar in some respects material (lithostratigraphic) referent while the to the concept of NALMAs as used for the latter has flexible limits. Cenozoic, Russell's system ofunits was more The definition of a NALMA does not fit broadly based, including in the definitions within concepts of biostratigraphic units as nonmammalian vertebrates, freshwater mol- considered by the NACSN (1983, Arts. 48- luscs, and a few plants in addition to mam- 54); no form ofbiozone is implied within the malian assemblages. Russell's (1975) ap- definition of a NALMA. proach to terminology and concept also 46 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE 11 Summary of Stratigraphic Distributions of Mammalian Taxa from the North American Western Interior Late Cretaceous See table 10 for more detailed information. See text concerning nature of "Edmontonian." "Edmon- Taxon Aquilan Judithian tonian" Lancian Alticonodon lindoei x ?Triconodontidae, n. gen. and sp. x Mesodma senecta x Mesodma sp. x x x x Cimexomys antiquus x Paracimexomys magister x Cimolodon electus x Cimolodon similis x ?Cimolomys sp. A x ?Cimolomys sp. B x Meniscoessusferox x Viridomys orbatus x Symmetrodontoides canadensis x Potamotelses aquilensis x Picopsis pattersoni x cf. Picopsis sp. x Tribosphenic upper molar x Alphadon creber x Alphadon sp. x x x Albertatherium primus x Pediomys exiguus x Aquiladelphis incus x Aquiladelphis minor x Eodelphis cutleri x x ? Eodelphis sp. x x Paranyctoides maleficus x Insectivora, n. gen. and sp. A x Insectivora, n. gen. and sp. B x Eutherian? x Mesodma primaeva x Paracimexomys judithae x Paracimexomys magnus x Paracimexomys priscus x x Cimolodon sp. x x Cimolomys clarki x Meniscoessus intermedius x Meniscoessus major x Multituberculate, suborder, family incertae sedis x Dryolestidae, gen. and sp. unident. x cf. Deltatheroides sp. x x Falepetrus barwini x Alphadon halleyi x Alphadon russelli x Alphadon lulli x x Alphadon sahnii x Alphadon attaragos x Alphadon praesagus x Pediomys clemensi x Pediomys sp. A x Pediomys prokrejcii x 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 47

TABLE 1 -(Continued) "Edmon- Taxon Aquilan Judithian tonian" Lancian Pediomys cf. P. elegans x Pediomys cf. P. hatcheri x Pediomys sp. x ?Pediomyidae x Eodelphis browni x Gypsonictops lewisi x Palaeoryctidae, gen. and sp. undet. x Paranyctoides sternbergi x Paranyctoides megakeros x Mesodma cf. M. thompsoni x Mesodma? sp. x Paracimexomys cf. C. judithae x Kimbetohia campi x Cimolodon nitidus x x Cimolomys gracilis x x Eucosmodontidae, n. gen. and sp. x Meniscoessus conquistus x Meniscoessus robustus x x Alphadon marshi ? x Alphadon cf. A. marshi x Alphadon rhaister ? x Alphadon?, n. sp. x cf. Peradectes sp. x Pediomys cf. P. cooki x Peidomys cf. P. krejcii x Eodelphis? sp. x Didelphodon? sp. x Gypsonictops, n. sp. x Gypsonictops sp. x Cimolestes sp. x cf. Cimolestes sp. x Mesodma formosa x Mesodma cf. M. formosa x Mesodma hensleighi x Mesodma thompsoni x ?.Neoplagiaulax burgessi x ?Neoplagiaulacidae, gen. and sp. indet. x Cimexomys minor x Cimexomys cf. C. hausoi x Ptilodontidae, gen. and sp. undet. x Catopsalis joyneri x Catopsalis n. sp. x Stygimys kuszmauli x Stygimys aff. S. kuszmauli x Cimolomys trochuus x Essonodon browni x Bistius bondi x Alphadon wilsoni x Alphadon cf. A. rhaister x Glasbius intricatus x Glasbius twitchelli x Didelphidae, gen. and sp. indet. x Pediomys cooki x 48 AMERICAN MUSEUM NOVITATES NO. 2840

TABLE I1 -(Continued) "Edmon- Taxon Aquilan Judithian tonian" Lancian Pediomys elegans x Pediomys florencae x ?Pediomys cf. P. florencae x Pediomys hatcheri x Pediomys krejcii x Didelphodon vorax x Gypsonictops hypoconus x Gypsonictops illuminatus x Gypsonictops cf. G. illuminatus x Cimolestes incisus x Cimolestes magnus x Cimolestes cerberoides x Cimolestes cf. C. cerberoides x Cimolestes propalaeoryctes x Cimolestes stirtoni x Batodon tenuis x Procerberusformicarum x Procerberus cf. P. formicarum x Telacodon laevis x Protungulatum donnae x Protungulatum cf. P. donnae x Protungulatum gorgun x Ragnarok harbichti x Ragnarok sp. x Oxyprimus erikseni x Oxyprimus cf. 0. erikseni x ?Oxyprimus sp. x Mimatuta morgoth x ?Mimatuta sp. x Purgatorius ceratops x differed from those involved in NALMAs. and utility to what had been done so suc- Russell considered his units to be chrono- cessfully for the Cenozoic by way of NAL- stratigraphic in nature (stages in the sense of MAs. He was quite incorrect, however, in ACSN, 1970), relating times of existence of stating that the NALMAs as originally de- specific organisms to the physical limits of fined ". . . are, more accurately, the equiva- particular rock units. Such usage is exhibited lents ofthe European stages, now considered in his text-figure 6 (1975, p. 158) in which to be time-rock terms, designating the time discrete time gaps are seen between the lim- of deposition of certain formations charac- its, for example, ofthe Aquilan and Judithian terized by distinctive faunas." NALMAs, in or the Judithian and Edmontonian; each stage contrast to stages, are not restricted in their in Russell's scheme relates to faunal assem- definition by time represented within a par- blages within strata representing deposition ticular stratigraphic type section. For histor- during distinct time units. No such gaps be- ical reasons based in part upon a misconcep- tween the various "ages" are seen in plate 1 tion, therefore, Russell's subdivisions for the of Wood et al. (1941). Cretaceous (his stages) differ in fundamental It seems clear from Russell's (1975, p. 138) concept, and thereby in practical utility, from introductory remarks that he intended the Cenozoic North American land-mammal provincial subdivision of the Cretaceous "ages" (of Wood et al., 1941). (Russell, 1964, 1975) to be similar in concept A further problem exists in that Russell's 1 986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 49 stages (of 1964 and 1975) were not properly semblages, knowledge of the faunas (except formalized under the then-existing (ACSN, perhaps for the Lancian) is low in comparison 1970, Art. 26a) or presently existing (NACSN, with those characteristic of most Cenozoic 1983, Art. 76) rules of stratigraphic nomen- "ages." For that reason, the previously used clature; the upper and lower limits of the system of faunal definitions (i.e., index fos- stages were not defined at specified type sec- sils, first appearances, and characteristic fos- tions within type areas in order to provide sils) must be given careful consideration. For standards for the units. Without such stan- example, immediately pre-Aquilan mam- dards, the units are meaningless. Thus, malian faunas are unknown, negating reality speaking practically, Russell's stages were in- within the Aquilan for such categories as in- adequately defined and, technically, they are dex fossils, first appearances, and character- invalid. istic fossils. Present listings and discussion are limited to the categories of first appear- NEW LATE CRETACEOUS ances (if the taxon is also known from youn- LAND-MAMMAL "AGES" ger rocks), last appearances (if the taxon is GENERAL DIscuSSION also known from older rocks), and unique occurrences (ifthe taxon is known only from Fox (1978) updated the vertebrate paleon- a particular age). tological definitions ofRussell's (1964, 1975) The reader is cautioned to recognize the nonmarine subdivisions of the North Amer- primitive nature of the unique occurrences ican Late Cretaceous. As Fox pointed out, as category (based largely upon which species knowledge of Late Cretaceous mammals has are unknown from older or younger rocks) increased, characterization of Russell's units and to realize that monographic inconsis- has shifted nearly exclusively to mammalian tencies in taxonomy make the category less assemblages. We follow that procedure in the reliable than what is ordinarily implied for sections below, in which we revise Dorf's and the concept of an index fossil. For example, Russell's terminology from a chronostrati- Mesodma sp. (see tables 10 and 1 1) occurs graphic basis to the philosophy employed in all of the named Late Cretaceous "ages," within the North American land-mammal but available material is inadequate to per- "ages." As has been the practice with terres- mit: (1) recognition and diagnosis of a new trial faunas described from rocks ofCenozoic species; or (2) recognition of conspecificity age, paleontochronologic ranges ofnonmam- with members of younger or older faunas. malian organisms can readily be considered Nevertheless, knowledge of Late Cretaceous within the framework ofNALMAs. Only Late mammalian faunas from the western interior Cretaceous (i.e., Aquilan-Lancian) ages are is on the threshold of practical utility for de- considered, and the summaries (presented as tailed geologic correlation ofnonmarine rock much as possible in the style employed by units. Wood et al., 1941) are derived from data in It could also be argued, with considerable tables 10 and 1 1. justification, that it is premature even to at- We employ the same nomenclature (i.e., tempt definition of Late Cretaceous NAL- Aquilan-Lancian) established by Dorf( 1942) MAs. Basically, we agree with that thought. and Russell (1964, 1975). Although a case Nevertheless, the terms have existed for more could be made for using new terms to reduce than two decades, and are commonly used in confusion between the concepts employed published evolutionary and geologic re- here versus those originally defined by Dorf search. Our purposes are merely to: (1) sum- and by Russell, in practice most usage to date, marize the present state of knowledge on the rightly or wrongly, has not differed signifi- subject; (2) provide a practical framiework of cantly from the intention implied within constraints upon which to build and refine NALMAs. We feel, therefore, that confusion geochronologic concepts as available infor- would be less by retaining Dorf's and Rus- mation grows; (3) clarify specifically where sell's terminology than by coining new terms. additional research is needed most critically; Despite extensive recent research on North and (4) encourage stability in concept and American Late Cretaceous mammalian as- practice in use of the "ages." so AMERICAN MUSEUM NOVITATES NO. 2840

The Edmontonian (Russell, 1964, 1975) correlations between the two areas are ten- causes a special problem. As can be seen from uous. study of table 11, an Edmontonian NALMA Problems exist both at the beginning and is presently indefensible as defined on the at the end of the Lancian NALMA. Defini- basis of fossil mammals. Most species from tion of the earlier extreme is hampered by rocks referred to L. S. Russell's Edmontonian the paucity of information available to limit stage (table 10) are either: (1) conspecific with an Edmontonian "age," as discussed above. representatives ofthe Lancian "age" (ofpres- The end of the Lancian (see Clemens et al., ent context); and/or (2) so poorly known as 1979; Archibald, 1982), as observed in the to have low degrees of certainty of identifi- Hell Creek Formation in northeastern Mon- cation. Nevertheless, comparisons of dino- tana, is presently under active research by saurian assemblages (see, for example, Lang- William A. Clemens, J. David Archibald, and ston, 1959; D. A. Russell, 1967, 1970, 1972; others. The problem of the age of the "Bug D. A. Russell and Chamney, 1967; Dodson, Creek faunal facies" (see Clemens et al., 1979, 1971; L. S. Russell, 1983) derived from strata pp. 47-50; Smit and van der Kaars, 1984) referred to L. S. Russell's Edmontonian stage has yet to be resolved. The following is a suggest the reality of a discrete interval of quotation from a letter from Clemens to Lil- geologic time intermediate in age between the legraven (November 1983) that summarizes Judithian and Lancian "ages." the situation as it presently stands: "Argu- Thus we are suggesting that, with further ments on stage of evolution of Bug Creek fieldwork, an Edmontonian NALMA may faunal facies organisms (multituberculates eventually be documentable on the basis of and condylarths) indicate they are older than fossil mammals; as presently represented in the oldest, certainly Puercan (post-dinosaur) collections, however, it cannot be justified. faunas. However, are they ofCretaceous age? In short, we are placed in the unsettling po- I still suspect they are but cannot demon- sition of recognizing the probable existence strate it. Therefore I suggest that you indicate of a discrete interval of geologic time that is the Lancian taxa that are limited to the Bug not yet documentable using the criterion of Creek faunal facies." Lancian taxa restricted choice (i.e., mammalian species) employed to or beginning in the "Bug Creek faunal fa- in the present paper. We suspect that the sit- cies" are marked with asterisks in the defi- uation is analogous to how the Clarkforkian nition of the Lancian "age," below. Mam- NALMA of the North American late Paleo- malian taxa (see footnote to table 10) closely cene-early Eocene was considered prior to related to those otherwise known only from rejuvenated fieldwork, development ofa sol- the "Bug Creek faunal facies" were reported id biostratigraphic framework, and subse- recently by Johnston and Fox (1984) from quent thorough review ofthe concept by Rose the base of the Ravenscrag Formation of (1981). southwestern Saskatchewan in association Similar, but less severe, problems exist with with several varieties ofdinosaurs. Johnston the Aquilan, Judithian, and Lancian NAL- and Fox (1984, p. 215) interpreted the faunas MAs as defined below. The Aquilan is de- as representing very latest Lancian time. fined by a single fauna with an extremely lim- ited geographic distribution, presently known DEFINITIONS OF "AGE" TERMS only from a small area of southern Alberta. The Judithian is defined according to faunas As was the procedure ofWood et al. (1941, from northerly realms (central Wyoming to p. 8) some doubtfully referred or inadequate- southern Alberta). Because of uncertainties ly documented specimens are intentionally of identifications of mammalian fossils, we ignored; the lists, therefore, are less complete are unsure of how faunas from the upper than those provided in tables 10 and 11. The Fruitland and lower Kirtland formations of system used here differs from that of Wood northwestern New Mexico (see Clemens et et al. (1941) in our emphasis on species-level al., 1979, p. 42; Flynn, in press) relate in time (rather than generic) taxonomy. The pro- to the more northerly assemblages. Present posed time units are presented in order of 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 51 decreasing relative geologic age. "Character- UNIQUE OCCURRENCES: Mesodma primae- istic fossils" are not defined. Because ofprob- va, Paracimexomysjudithae, Paracimexomys lems discussed above, an Edmontonian "age" magnus, Cimolomys clarki, Meniscoessus in- is not defined on the basis of mammals as a termedius, Meniscoessus major, Multituber- new provincial time term. culata (suborder and family incertae sedis), Dryolestidae (genus and species unidenti- AQUILAN "AGE" fied), Falepetrus barwini, Alphadon halleyi, Alphadon russelli, Alphadon sahnii, Alpha- The Aquilan "age" is defined as a new pro- don attaragos, Alphadon praesagus, Pedi- vincial time term, based on mammalian fau- omys clemensi, Pediomys sp. A, Pediomys na from upper part ofMilk River Formation prokrejcii, Eodelphis browni, Gypsonictops of southern Alberta, faunal type area in Ver- lewisi, Palaeoryctidae (genus and species un- digris Coulee, 29 km east of village of Milk determined), Paranyctoides sternbergi, Para- River. The name Aquilan is in reference to nyctoides megakeros. the Eagle Sandstone with which the Milk COMMENT: The Judithian "age" is highly River Formation intergrades (Sahni, 1972). diagnostic in its mammalian fauna; only three PRINCIPAL MAMMALIAN FAUNAL CORRELA- into TIVES: None known. species are recognized that continue FIRST APPEARANCE: Eodelphis cutleri. younger strata. UNIQUE OCCURRENCES: Alticonodon lin- doei, ?Triconodontidae (new genus and LANCIAN "AGE" species), Mesodma senecta, Cimexomys an- tiquus, Paracimexomys magister, Cimolodon The Lancian "age" is a new (or modified) electus, Cimolodon similis, ?Cimolomys sp. provincial time term, based upon mamma- A, ?Cimolomys sp. B, Meniscoessus ferox, lian fauna from type Lance Formation ofeast- Viridomys orbatus, Symmetrodontoides can- central Wyoming in Niobrara County, faunal adensis, Potamotelses aquilensis, Picopsis type area in valleys of Lance Creek and its pattersoni, cf. Picopsis sp., Alphadon creber, tributaries north of town of Lance Creek. Albertatherium primus, Pediomys exiguus, Principal mammalian faunal correlatives: Aquiladelphis incus, Aquiladelphis minor, In- Faunas from Scollard Formation (south-cen- sectivora (new genus and species A), Insec- tral Alberta), Frenchman Formation (south- tivora (new genus and species B), eutherian? western Saskatchewan), basal Ravenscrag fauna Formation (southwestern Saskatchewan), COMMENT: The Aquilan mammalian Hell Creek Formation (northeastern Mon- is unique, with only one species presently tana), Lance Formation (Bighorn Basin and known to occur in a younger "age." Rock Springs Uplift, Wyoming), Laramie Formation (northeastern Colorado), and JUDITHIAN "AGE" lower part ofNorth Horn Formation (central The Judithian "age" is a new provincial Utah). time term, based upon mammalian fauna LAST APPEARANCES (including taxa record- from Judith River Formation of north-cen- ed from strata referred to Russell's, 1964 and tral Montana, faunal type area adjacent to 1975, Edmontonian stage; see tables 10 and Chouteau and Blaine county line in breaks a 11): Paracimexomys priscus, cf. Deltathe- few kilometers north ofMissouri River. Prin- roides sp., Alphadon lulli, Cimolodon nitidus, cipal mammalian faunal correlatives: Faunas Cimolomys gracilis, Meniscoessus robustus, from Oldman Formation (southern Alberta), Alphadon marshi, Alphadon rhaister. Judith River Formation (northernmost UNIQUE OCCURRENCES (taxa restricted to Montana; unpublished fauna under study by or beginning with "Bug Creek faunal facies" W. A. Clemens), "Mesaverde" Formation marked with asterisks): Mesodma hensleighi, (Bighorn and Wind River basins, Wyoming). ?Neoplagiaulax burgessi, Cimexomys minor, FIRST APPEARANCES: Paracimexomys pris- Catopsalis joyneri, *Stygimys kuszmauli, cus, cf. Deltatheroides sp., Alphadon lulli. *Stygimys aff. S. kuszmauli, Cimolomys tro- LAST APPEARANCES: None recognized. chuus, Essonodon browni, Bistius bondi, Al- 52 AMERICAN MUSEUM NOVITATES NO. 2840 phadon wilsoni, Glasbius intricatus, Glasbius out clearly stated methodology (e.g., Guo, twitchelli, Pediomys cooki, Pediomys ele- 1984). gans, Pediomys florencae, Pediomys hatch- Generally, as studies have progressed in eri, Pediomys krejcii, Didelphodon vorax, recent years, the postulated ages ofthese Late Gypsonictops hypoconus, Gypsonictops illu- Cretaceous sites have become younger and minatus, Cimolestes incisus, Cimolestes younger within a Late Cretaceous context, magnus, Cimolestes cerberoides, Cimolestes with the Djadokhta Formation coming to be propalaeoryctes, Cimolestes stirtoni, Batodon regarded as approximately equal in age to the tenuis, *Procerberus formicarum, Telacodon Judith River Formation of North America; laevis, *Protungulatum donnae, *Protungu- the other Asian Late Cretaceous sites, be- latum cf. P. donnae, *Protungulatum gor- lieved to be younger still, are therefore gen- gun, *Ragnarok harbichti, *Oxyprimus erally thought to be compressed into the late erikseni, Mimatuta morgoth, *Purgatorius Campanian and Maastrichtian. Thus the Ba- ceratops (questionable record). run Goyot Formation, including the redbeds COMMENT: As discussed above, the youn- of Khermeen Tsav (Khermeen Tsav For- ger limit of the Lancian "age" (see Johnston mation of Kielan-Jaworowska, 1974a: see and Fox, 1984) is under active research by Kielan-Jaworowska, 1979, p. 6; Rozhdest- W. A. Clemens, J. David Archibald, and oth- venskii, 1977, p. 109), is generally regarded ers. Tentatively, Clemens and Archibald are to be slightly younger than the Djadokhta recognizing "Premantuan" and "Mantuan" Formation, although the physical relation- faunal assemblages that are post-Lancian in ships are unknown and the type Barun Goyot age and appear to straddle the Mesozoic-Ce- is some distance from the type Djadokhta. nozoic boundary. Some species from the "Bug Rozhdestvenskii (1977, p. 109) assigned the Creek faunal facies" (marked with asterisks Khermeen Tsav beds a Campanian-Maas- in list ofUnique Occurrences, above) appear trichtian age. In the same year, Barsbold to continue upward into the "Premantuan" (1977) assigned an early Campanian age to and younger faunas (see Archibald, 1982, and the lower part of the Barungoiotskaya Svita Archibald and Clemens, 1984, for prelimi- near Khermeen Tsav and referred to the up- nary evaluations). Thus, upon further docu- per Barungoiostkaya Svita as late Campani- mentation of post-Lancian faunas, some an. species listed above as unique occurrences In any case, the Nemegt Formation phys- may be considered first occurrences. ically overlies the Barun Goyot Formation (Gradzinski et al., 1977, fig. 12) and is, there- CORRELATION OF fore, aside from any possible time-transgres- MAMMAL-BEARING sive boundary problems, younger than the LATE CRETACEOUS FORMATIONS Barun Goyot Formation. Thus far, the Ne- OF MONGOLIA megt Formation has not yielded mammals, but its age has a direct bearing on the age of In the Mongolian People's Republic (MPR), the underlying Barun Goyot Formation and, fossil mammals have been recovered from therefore, indirectly on that ofthe Djadokhta rocks ofLate Cretaceous age at a few localities Formation. The Bugeen Tsav locality has not (see Clemens et al., 1979, pp. 34-37 for re- been fitted with certainty into the stratigraph- view) in the nonmarine Djadokhta and Barun ic scheme. The multituberculate Buginbaa- Goyot formations and possibly at a Late Cre- tar, the sole fossil mammal reported from taceous site of undetermined stratigraphic Bugeen Tsav, has been assigned the age of position near Bugeen Tsav. For geographic late Maastrichtian or early Paleocene (Kie- details and discussion of the faunas, see lan-Jaworowska and Sochava, 1969; Kielan- Gradzinski et al. (1977, fig. 1). Attempts to Jaworowska, 1974b). Trofimov (1975) and determine ages offaunas recovered from these Badamgarav and Reshetov (1976) regarded Asian Late Cretaceous continental sediments Buginbaatar as questionably early Paleocene have not had the benefit of radioisotopic or in age. In a more recent paper, Gradzinski et magnetostratigraphic data and have involved al. (1977, p. 302) regarded the beds at Bugeen biostratigraphic methods alone, usually with- Tsav (at Khaichin-Ula I. part of the Svita 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 53

Dzunmod of Russian authors) as biostrati- the previously published comment on Mon- graphic equivalents of the Nemegt Forma- golian Late Cretaceous units. tion on the basis of (mainly unpublished) Novozhilov (1954), Rozhdestvenskii Russian work. Rozhdestvenskii (1977, p. 109) (1957), and Van Valen (1966) concluded that has stated that the age is the same (Maas- the mammals of the Djadokhta Formation trichtian in his view) as that of the might not have been derived from that unit "Nemegetu." Much work remains to be ac- at all, but rather were let down onto the pres- complished before the stratigraphic and bio- ent Djadokhta surface from some now-miss- stratigraphic relationships of these vaguely ing, overlying, younger unit of Tertiary age, defined units can be stated more clearly. The possibly as young as the North American following historical summary represents an Clarkforkian near the Paleocene/Eocene attempt to keep track of what has been said, boundary (Van Valen, 1966, p. 51). How- often in obscure reports, and is admittedly ever, this divergent train ofthought was soon more exhausting than exhaustive. We begin put to rest by Kielan-Jaworowska (1969), who with the Djadokhta Formation. reported the same species ofmammals in situ The Djadokhta Formation (Berkey in within the Djadokhta Formation. Granger and Gregory, 1923, p. 8; Berkey and Sahni (1968, p. 343) remarked in passing Morris, 1927) yielded the first Cretaceous that the Djadokhta beds are Cenomanian- mammals known from Asia. The type local- Turonian in age. His estimate postulated the ity of the Djadokhta Formation is at Bayn oldest age ofwhich we are aware. It may have Dzak (lat. 4401 2'N, long. 103°44'E) northeast reflected unpublished opinion of the time, ofBulgan, MPR. Estimates ofage ofthe Dja- but no rationale was given. Later, Sahni (1972, dokhta Formation have ranged widely. Early p. 325) repeated the estimate ofCenomanian papers (e.g., Simpson, 1925, 1928; Gregory or Turonian age, attributing it to Kielan-Ja- and Simpson, 1926; but also as late as Ma- worowska ([1968], actually 1969). rinov, 1957) wisely referred the Djadokhta Gradzinski et al. (1969, p. 37), evidently to Late (or Upper) Cretaceous, without at- on the basis of a draft of the manuscript of tempting further refinement. However, the Kielan-Jaworowska's (1969) paper rather primitive nature ofthe "pre-ceratopsian" di- than the corrected published version, re- nosaur Protoceratops andrewsi from the Dja- marked in passing that the Djadokhta For- dokhta Formation led many workers to sus- mation is of Turonian or Santonian age. pect that the Djadokhta was not of latest However, at the time of publication of her Cretaceous age (Granger and Gregory, 1923). papers, Kielan-Jaworowska (1969, 1970a, No attempts were made in these early papers 1970b; see also Clemens, 1970 and Lefeld, to relate the Djadokhta Formation to Cre- 1971) estimated the Djadokhta Formation to taceous marine stages/ages based on Euro- be Coniacian or Santonian in age. This de- pean stratotypes, nor could this have been termination was based on fossil mammals done. In the absence of fossiliferous marine which, at that time, were thought interme- interbeds, radioisotopic dating, or magneto- diate in stage of evolution between those of stratigraphy, attempts since the 1920s to place the Paluxian (Slaughter, 1965; L. S. Russell, the Djadokhta Formation into a temporal 1975) Forestburg site of Texas and the Ju- framework have remained crude. Age esti- dithian of Montana and Alberta. In this con- mates have been based on "stage of evolu- clusion, Kielan-Jaworowska was influenced tion" ofAsian forms and on biostratigraphic by the opinions of Dr. R. E. Sloan (Kielan- comparisons of various sorts (Gradzinski et Jaworowska, 1969, p. 173). Thus, she was al., 1977; Fox, 1978), generally with North considering North American biostratigraphic American forms rather than with European units believed to be equivalent in age to the ones. However, these efforts have been dif- European-based marine Albian and mid- ficult because of endemism of many organ- Campanian, respectively, as calibration points isms involved. Moreover, correlation with on which to base an estimate of an inter- marine stages/ages unfortunately cannot be mediate age. accomplished directly, although such corre- At the same time, McKenna (1969) esti- lations are incorrectly implied by much of mated, on the basis of his existing view of 54 AMERICAN MUSEUM NOVITATES NO. 2840 stage ofevolution ofits vertebrate fauna, that ska-Tworzydlo (1983), Kielan-Jaworowska the Djadokhta Formation might be equiva- (1982, 1984) returned to earlier views, ten- lent in age to the marine Cenomanian stage. tatively correlating the Djadokhta Formation However, as Fox (1978) noted, this estimate with the boundary (or slightly older or youn- is much too old. In his discussion ofthe fossil ger) between Coniacian and Santonian. Thus, vertebrates ofthe Judith River Formation of even now, there is no universal agreement on Montana, Sahni (1972, p. 325) misquoted the age of the Djadokhta Formation. Kielan-Jaworowska's " 1968" (actually 1969) The Barun Goyot Formation (originally paper, saying that the Cretaceous mammal known as the Lower Nemegt Beds: Gradzin- localities of Djadokhta Formation were re- ski et al., 1969, p. 37) is based on the type garded as Cenomanian or Turonian in age. locality at Khulsan (lat. 43°30'N, long. 1017- In actuality, Kielan-Jaworowska had esti- 8'E). This formation was variously estimated mated by the time the paper was published to be: (1) possibly contemporaneous with, or that they are Coniacian-Santonian. Rozh- somewhat younger than, the Djadokhta For- destvenskii (1971, 1972, 1977, table 1) placed mation (ibid.); (2) possibly Campanian (Kie- the Djadokhta Formation at about the San- lan-Jaworowska and Barsbold, 1972); (3) tonian-Campanian boundary, but regarded Campanian, and possibly lower Campanian the "Belly River" of Canada as contempo- (Kielan-Jaworowska, 1974a, 1974b); (4) near rary. In the standard text on the geology of or questionably mid-Campanian (Kielan-Ja- Mongolia, Marinov et al. (1973, p. 479) cited worowska, 1975a, 1975b; Gradzinski et al., unspecified work by Barsbold that the Dja- 1977; Osmolska, 1980; Kielan-Jaworowska dokhta Formation is "Santonian-Campani- and Trofimov, 1980); (5) Santonian in the an." lower part and Campanian in the upper part The next estimates ofage ofthe Djadokhta (Shuvalov and Chkhikvadze, 1975); (6) ap- Formation by Kielan-Jaworowska were proximately equivalent in age to the North Coniacian (1974a, 1974b) and then Santo- American St. Mary River and Two Medicine nian (1975a, 1975b). Sulimski (1975) called formations (Maryanska and Osmolska, 1975); the Djadokhta Formation ?Santonian. How- and (7) in the younger half ofthe Campanian ever, Sloan (1976) returned the Djadokhta (Judithian or younger; Fox, 1978). As with Formation to the Coniacian, whereas Bars- the Djadokhta Formation, Fox's and Os- bold (1977) returned to the more general term molska's correlations were based on a large "lower Senonian." Somewhat younger esti- variety of fossil vertebrates correlated to mates have been made by: (1) Rozhdestven- North America and then indirectly to Eu- skii (1974, 1978) and Belyaeva et al. (1978), rope, not directly from Asia westward to Eu- Campanian; (2) Fox (1977, 1978), medial rope. Campanian (in terms of nonmarine "ages,"" However, Stankevitch and Khand (1976, closer to Judithian than to Aquilan); (3) p. 361, abstract in English) dated the Barun Gradzinski et al. (1977) and Kielan-Jawo- Goyot Formation as Santonian-Maastrich- rowska and Sloan (1979), ?upper Santonian tian on the basis of ostracodes, and Kar- and/or ?lower Campanian; and (4) Osmolska czewska and Ziembinska-Tworzydlo (1983) (1980), ?lower Campanian. Rozhdestvenskii dated it as late Santonian (tentatively ac- based his estimate on biostratigraphy of di- cepted by Kielan-Jaworowska, 1984) on the nosaurs. Fox's and Osmolska's correlations basis ofcharophytes. Ifthe assumption is ac- were based on a variety of fossil vertebrates cepted that the Barun Goyot Formation is believed closely related to, and in some cases equal in age or slightly younger than the Dja- probably identical to, taxa from nonmarine dokhta Formation, then the Djadokhta would Cretaceous rocks in North America. These also be late Santonian or even older if the vertebrates, however, cannot be correlated charophyte data are to be relied upon. directly with marine cephalopods and micro- The Nemegt Formation originally was fossils ofthe European type Santonian, Cam- known informally as the Nemegetu section panian, or Maastrichtian. ofObruchev (Marinov, 1957) and later more Recently, following evidence from studies formally as the Upper Nemegt Beds (Grad- ofcharophytes by Karczewska and Ziembin- zinski et al., 1969, p. 37). These were rede- 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 55 fined as the Nemegt Formation by Gradzin- ger), as Fox (1978) would have it, the Nemegt ski and Jerzykiewicz (1974). Discovery of Formation would have to be equivalent to spectacular dinosaur remains in these beds the Judithian (or younger, as Rozhdestven- following World War II by Efremov and oth- skii, 1971, would have it). We prefer Rozh- ers (Efremov, 1954; Chudinov, 1966) made destvenskii's interpretation, but the matter is these deposits justly famous. not settled. The Nemegt Formation overlies the Barun At present, interpretations of stratigraphic Goyot Formation and was dated loosely as ranges of fossil vertebrates in Cretaceous Campanian or Maastrichtian by its original Mongolian formations yield biostratigraphic formal describers. Later, it was dated as late correlations that conflict with the correlation Campanian or early Maastrichtian (Nowin- based on charophytes from the same for- ski, 1971; Kielan-Jaworowska and Barsbold, mations. However, we tentatively accept 1972; Osmolska et al., 1972). Rozhdestven- Fox's correlation on the basis of the whole skii (1971, 1972, 1974, 1977, 1978) placed vertebrate fauna of the mammal-bearing the Nemegt Formation in the Maastrichtian, Djadokhta and Barun Goyot formations. We but nevertheless believed the Nemegt to be believe that the Djadokhta and Barun Goyot older than the Lance Formation of North formations are approximately equivalent in America. Shuvalov and Chkhikvadze (1975) age, respectively, to the older and younger also correlated the Nemegt Formation with parts of the North American Judithian land- the Maastrichtian. Gradzinski et al. (1977) mammal "age." However, we stress that the considered the Nemegt Formation to be ?up- evidence, especially that from fossil mam- per Campanian and ?lower Maastrichtian. mals, is slim. Moreover, in terms of the Eu- Fox (1978) made a similar estimate ofits age, ropean-based Campanian and Maastrichtian possibly equivalent to the Edmontonian of marine stages, which can be correlated with North America (late Campanian to early southern Mongolia only by indirect methods, Maastrichtian in his view). Osmolska (1980) we believe that the Djadokhta and Barun placed the Nemegt Formation in the ?late Goyot formations probably correlate with the Campanian to ?Maastrichtian. Maryanska upper Campanian and/or lower Maastrich- and Osmolska (1981) regarded Saurolophus tian of the type areas in Europe, where the angustirostris, a Nemegt dinosaur, as "prob- stage/age framework had its origin. In terms ably not older than the late Campanian." ofradioisotopic dating, that would have been Barsbold (1971, 1972) regarded the gastro- about 75 million years ago (see Berggren et pods and pelecypods (Martinson, 1961) of al., in press). the Nemegt Formation as characteristically Maastrichtian (see also Martinson, 1973, 1975; Martinson et al., 1969). Karczewska SUMMARY and Ziembinska (1970) as well as Kyansep- This is the first systematic account ofmam- Romashkina (1975) thought that the Nemegt malian remains from the Late Cretaceous charophytes suggest undifferentiated Cam- (Judithian) "Mesaverde" Formation of the panian-Maastrichtian age. Recently, how- North American western interior. The fossils ever, Kielan-Jaworowska (1982) placed the are ofbiogeographic importance because they Nemegt Formation earlier, correlating it with represent the most southerly, well-docu- the North American Aquilan and the Euro- mented record ofJudithian mammals known pean-based lower Campanian. from the continent. Fossils were recovered Although no mammals are known from the from nonmarine rocks in Wyoming in mul- Nemegt Formation, it superposes the Barun tiple localities in the southern Bighorn Basin Goyot Formation; thus, if the Nemegt were and in a single stratigraphic level (two local- as old as early Campanian, as Kielan-Jawo- ities) in the southeastern Wind River Basin. rowska (1982) would have it, the Barun Go- Faunal comparisons were emphasized yot Formation would also be of that age or among mammalian assemblages from the older, as would the Djadokhta Formation. "Mesaverde" (Wyoming), Judith River On the other hand, if the Barun Goyot were (Montana), and Oldman (Alberta) forma- equivalent in age to the Judithian (or youn- tions. A tacit assumption was that the ani- 56 AMERICAN MUSEUM NOVITATES NO. 2840 mals lived essentially contemporaneously; States, and the Gulf Coast sections. The thus most differences observed within species boundary at each place is at the top ofthe G. collected from the various areas were con- calcarata (total-range) Zone. sidered to be a result of individual or pop- Correlations based on range zones of geo- ulational variation rather than evolutionary graphically wide-ranging species of forami- change expressed through a significant inter- niferans show that the mammal-bearing strata val ofgeologic time. These assumptions seem of the Oldman, Judith River, and "Mesa- justified because all three nonmarine mam- verde" formations, which almost universally malian assemblages are temporal equivalents have been considered solidly Campanian in ofapproximately the same cephalopod-based age, lie precariously close to the revised Cam- more easterly marine biostratigraphic unit, panian-Maastrichtian boundary, and even the Baculites gregoryensis Zone (or as many possibly overlapping into the early Maas- as four zones higher, about 2 million years trichtian. It is clear that the generally recog- younger), part of the standard zonation of nized Campanian-Maastrichtian boundary, Upper Cretaceous rocks of the North Amer- which is based on molluscan faunas within ican western interior. Little evidence exists the western interior, must be relocated earlier to suggest that the limits of the marine in- in Cretaceous time in light of new evidence vertebrate zonation are significantly time- from the discipline of micropaleontology. transgressive from one geographic area to The mammalian fossils from the Oldman, another. Judith River, and "Mesaverde" formations The Red Bird Silty Member of the Pierre were deposited essentially contemporane- Shale as seen at Redbird, eastern Wyoming, ously in nonmarine sediments near the west- holds the largely endemic, temperate-water, ern shoreline of the Western Interior Seaway invertebrate assemblage characteristic of the during the regressive phase of the Claggett Baculites gregoryensis (cephalopod) Zone. cyclothem. The probable proximity ofthe Ju- The rock unit also yields a newly described dithian faunas to the Campanian-Maastrich- planktonic foraminiferal assemblage that we tian boundary suggests that the mammals correlate to the warm-water upper Taylorian lived within the late phases of geomagnetic and/or lower Navarroan foraminiferal stages Polarity Chron 33 or the early part ofPolarity of the Gulf Coast. Comparison of micro- Chron 32. This part ofthe Upper Cretaceous faunal species ranges between the two areas section ofthe western interior, however, lacks suggests correlation of the Red Bird Silty a reliable magnetostratigraphic zonation. The Member with the Globotruncana elevata Judithian mammals correlate stratigraph- Subzone and/or the next higher Rugotrun- ically with the lower part of the Aquilapol- cana subcircumnodifer Subzone of the Gulf lenites quadrilobus palynomorph Interval Coast. The upper part of the G. elevata Sub- Zone of the northern Rockies. zone ofthe GulfCoast is characterized by the Sixteen species-level taxa are recognized total-range zone for G. calcarata (considered from the "Mesaverde" Formation of Wyo- locally as a "Zonule" ofthe G. elevata "Sub- ming. Three previously unknown genera are zone"). reported, but none is formally named. These The top of the Globotruncana calcarata include a presently unidentifiable multitu- Zone is recognized as the boundary between berculate of extraordinarily large body size, the Campanian and Maastrichtian European a phylogenetically relictual dryolestoid (pre- stages at Gubbio, Italy. The rocks at Gubbio viously unknown in North America from represent an important reference section for post-Jurassic rocks), and a new variety ofad- the European Late Cretaceous in terms of vanced therian referable neither to the mar- correlation to the presently recognized stra- supials nor to the eutherians. Three new totypes for the Campanian (France) and species of previously described genera are Maastrichtian (Netherlands) stages. Identical named. They include a variety of Alphadon species-level criteria are used by micropa- (a marsupial) close to Lancian species, a tiny leontologists in defining the Campanian- form ofAlphadon, and a large version of an Maastrichtian boundary at Gubbio, the insectivore previously unknown south ofAl- northern Atlantic seaboard of the United berta. 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 57

All species-level taxa except Alphadon lulli resented at the type sections) as time-rock are reported from Wyoming for the first time, (chronostratigraphic) units. Thus, Russell's and new records involve both geographic and stages were inadequately defined, technically geologic range extensions. Four species (two invalid, and based in part on a misconcep- multituberculates, one marsupial, and one tion. eutherian) are extended southward from Using mammalian assemblages exclusive- northern Montana. Two species (multituber- ly, we revise Dorf's (1942) and Russell's culate and marsupial) are extended into Wy- (1964, 1975) terminology from geochrono- oming from southern Alberta. A "therian of logical and chronostratigraphic bases, re- metatherian-eutherian grade" and a marsu- spectively, to the philosophy employed with- pial are recognized for the first time both in in North American land-mammal "ages" (as northern Montana and central Wyoming. defined by Wood et al., 1941); only Late Cre- Two species (multituberculate and marsu- taceous ages are considered. pial) typical of Lancian rocks are extended In a balance ofjudgment based on recent into the Judithian. Six species-level taxa of usage and with the intention of minimizing Judithian age are presently recognized only confusion, we employ the same nomencla- from the "Mesaverde" Formation of Wyo- ture (i.e., Aquilan-Lancian) established by ming. Dorf (1942) and Russell (1964, 1975). Aqui- Comparison ofJudithian species-level taxa lan, Judithian, and Lancian North American known from up and down the western coast- land-mammal "ages" are defined, using the line of the Western Interior Seaway suggest criteria offirst appearances, last appearances, that the mammalian fauna was essentially and unique occurrences of mammalian homogeneous, at least as far as from southern species-level taxa. An Edmontonian "age" Alberta to central Wyoming. (chronologically intermediate between the The equivocal position of the "Mesa- Judithian and Lancian) is not named because verde," Judith River, and Oldman mam- such a term is presently indefensible on the malian faunas with regard to local boundaries basis of mammalian systematics. Neverthe- of the European-based Campanian and less, comparisons of dinosaurian assem- Maastrichtian stages underlines the need for blages derived from strata referable to Rus- provincial geochronological terminology ap- sell's Edmontonian stage suggest to us the plicable to Late Cretaceous nonmarine fau- reality ofa discrete unit ofgeologic time that nas of the North American western interior. should, with increased information, be defin- Russell (1964, 1975) defined the terms (in able on the basis ofmammalian assemblages. decreasing relative geologic age) Aquilan, Ju- Definition of the younger boundary of the dithian, Edmontonian, and Lancian specifi- Lancian "age" is under active research by cally for that purpose, on the basis of fossil others, and the present faunal criteria soon vertebrates, invertebrates, and plants. These will be modified. terms were intended to be strict chronostrati- Although admittedly premature in at- graphic units (i.e., stages, restricted to time tempting to define Late Cretaceous North represented within the deposition of the American land-mammal "ages," Russell's stratotype sections). stage terms are longstanding, and the names Unfortunately, Russell's stages were not commonly are used incorrectly in application properly formalized under the rules of strati- to published evolutionary and geological re- graphic nomenclature; the upper and lower search. Our purposes, therefore, were to: (1) limits of the stages were not defined at type summarize the status of knowledge on the sections. Furthermore, Russell's intention of subject; (2) provide a practical framework of usage resembled that for North American constraints upon which to build and refine land-mammal "ages" (in the senses ofWood geochronologic concepts as available infor- et al., 1941; Savage, 1962; Tedford, 1970) as mation grows; (3) clarify specifically where applied successfully to the Cenozoic. How- additional research is needed; and (4) en- ever, he incorrectly interpreted the Cenozoic courage stability in concept and practice in "ages" (intended to be pure time units with the use of "ages." flexible boundaries not restricted to time rep- Correlation of the Late Cretaceous rocks 58 AMERICAN MUSEUM NOVITATES NO. 2840

of Mongolia with those of North America R. Barwin provided useful information and and Europe is a highly sought-after objective advice. Steven S. Barrell of the Worland Of- that has been attempted by almost all authors fice of the Bureau of Land Management who have dealt with the Mongolian Creta- helped with logistics. Otto Simonis cleaned ceous. Unfortunately, few authors have stat- and made casts of the specimens. Chester ed the premises upon which their efforts rest. Tarka provided many ofthe illustrations, in- The history of these attempts is a depressing cluding all photographs of casts (preferred one, full of contradictions and conclusions over photos of original specimens because that go far beyond available evidence. More- confusion caused by internal reflections or over, stratigraphic practices and degrees of mottling could be reduced). We thank Fulton adherence to international rules vary from Jameson of Riverton, Wyoming, for permis- author to author. sion to work on land under his control, for We concur with Fox (1978) that the Dja- the use of his ranch buildings, and for the dokhta of eastern Asia and the Judith River loan of his ancient but capable bulldozer. and Oldman faunas of North America are Funds for Lillegraven's part of the effort essentially of the same age and that, more were provided by NSF grants DEB-8 105454 indirectly, the Djadokhta Formation is there- and EAR-8205211. Funds for McKenna's fore approximately late Campanian or even work were provided by the Frick Laboratory early Maastrichtian in age. This compresses Endowment Fund ofThe American Museum the faunas of the Barun Goyot and Nemegt of Natural History. formations, plus that of the poorly known Finally, we thank Zofia Kielan-Jaworow- Bugeen Tsav locality in southwestern Mon- ska, Karl Waage, William A. Clemens, David golia, into an interval of time equivalent to S. Kite, Brent H. Breithaupt, Linda E. Lil- the Judithian and/or Lancian North Ameri- legraven, Dale A. Russell, John J. Flynn, and can land-mammal "ages" and to the latest Lawrence J. Flynn for suggesting improve- Campanian or, more probably, Maastrich- ments in the manuscript. We also thank each tian stages as typified in western Europe. We other for amazing displays of tolerance. believe that the vertebrate evidence of this correlation, weak as it is, is nevertheless LITERATURE CITED stronger than that provided by charophytes and ostracodes for more ancient age assign- ACSN (American Commission on Stratigraphic ments. Nomenclature) 1970. Code of stratigraphic nomenclature. Tulsa, Amer. Assoc. Petrol. Geol., 22 ACKNOWLEDGMENTS PP. We thank the following people for aid in Alvarez, Walter, Michael A. Arthur, Alfred G. under our direction: Brent H. Fischer, William Lowrie, Giovanni Napoleone, the fieldwork Isabella Premoli Silva, and William M. Rog- Breithaupt, Linda E. Lillegraven, Jane E. genthen Hartman, Thomas H. Rich, Patricia V. Rich, 1977. Upper Cretaceous-Paleocene magnetic Julian Kadish, George 0. Whitaker, Ronald stratigraphy at Gubbio, Italy. V. Type P. Ratkevich, Richard P. Ratkevich, Eli Min- section for the Late Cretaceous-Paleo- koff, Marilyn Galusha, Philip Aranow, and cene geomagnetic reversal time scale. Priscilla C. McKenna. Gerard R. Case pro- Geol. Soc. Amer., Bull., vol. 88, pp. 383- vided many specimens resulting from his 389. fieldwork in the Bighorn Basin. J. Howard Archibald, J. David Hutchison and Michael T. Greenwald ex- 1982. A study ofMammalia and geology across tended our of the mammals at the Cretaceous-Tertiary boundary in knowledge Garfield County, Montana. Univ. Cal- Fales Rocks in the Wind River Basin. Donald ifornia Publs. Geol. Sci., vol. 122, xvi + Baird was helpful in establishing provenience 286 pp. of some of the specimens from the Bighorn Archibald, J. David, and William A. Clemens Basin and in other ways. William E. Frerichs, 1984. Mammal evolution near the Creta- Zofia Kielan-Jaworowska, William R. Kee- ceous-Tertiary boundary. In W. A. fer, J. David Love, Dale A. Russell, and John Berggren and J. A. Van Couvering (eds.), 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 59

Catastrophies and earth history. The new tion in the southern part of the Wind uniformitarianism. Princeton, Prince- River Basin, Wyoming. Wyoming Geol. ton Univ. Press, pp. 339-371. Soc. Guidebook, 16th Field Conf., p. Armstrong-Ziegler, Judy G. 173. 1978. An aniliid snake and associated verte- Belyaeva, E. I., V. Yu. Reshetov, and B. A. Tro- brates from the Campanian of New fimov Mexico. Jour. Paleont., vol. 52, pp. 480- 1978. Klass Mammalia. Mlekopitayushchie. 483. In V. A. Shimanskii and A. N. Solov'ev Badamgarav, D., and V. Yu. Reshetov (eds.), Razvitie i smena organicheskogo 1976. 0 novom Mestonakhozhdenii rannetre- mira na rubezhe Mezozoya i Kaino- tichnykh mlekopitayushchikh v zaal- zoya. Moscow, Akad. Nauk SSSR, Iz- taiskoi Gobi (Mongoliya). In N. N. Kra- datel'stvo "Nauka," pp. 103-123. marenko et al. (eds.), Paleontologiya i Berggren, William A., Dennis V. Kent, and John biostratigrafiya Mongolii. Sovmestnaya J. Flynn Sovetsko-Mongol'skaya Paleontologi- In press. Paleogene geochronology and chrono- cheskaya Ekspeditsiya, Trans., vol. 3, stratigraphy. In N. J. Snelling (ed.), pp. 265-268. Geochronology and the geological rec- Barrier, J. ord. Geol. Soc. London, Spec. Paper. 1980. A revision of the stratigraphic distri- Bergstresser, Thomas J. bution ofsome Cretaceous coccoliths in 1981. Foraminiferal biostratigraphy and pa- Texas. Jour. Paleont., vol. 54, pp. 289- leobathymetry of the Pierre Shale, Col- 308. orado, Kansas and Wyoming. Unpubl. Barsbold, R. Ph.D. diss., Univ. Wyoming, Laramie, 1971. Nekotorye krupnye bryukhonogie mol- Wyoming, ix + 337 pp. lyuski iz verkhnemelovykh otlozhenii Berkey, Charles, and Frederick K. Morris yugo-zapadnoi Mongolii. In N. S. Zait- 1927. Geology of Mongolia. Natural history sev et al. (eds.), Fauna Mezozoya i ofCentral Asia. New York, Amer. Mus. Kainozoya Zapadnoi Mongolii. Nat. Hist., vol. 2, pp. 1-475. Sovmestnaya Sovetsko-Mongol'skaya Birkelund, T., J. M. Hancock, M. B. Hart, P. F. Nauchno-Issledovatel'skaya Geologi- Rawson, J. Remane, F. Robaszynski, F. Schmid, cheskaya Ekspeditsiya, Trudy, vol.3, pp. and F. Surlyk 14-20. 1984. Cretaceous stage boundaries-propos- 1972. Biostratigrafia i presnevodnye molluski als. Geol. Soc. Denmark Bull., vol. 33, verkhnego mela gobiiskoi chasti MNR. pp. 3-20. Moscow, Akad. Nauk SSSR, Izda- Butler, Percy M. tel'stvo "Nauka," pp. 1-88. 1939. The teeth of Jurassic mammals. Zool. 1977. Kinetizm i osobennosti stroeniya Soc. London, Proc. (ser. B), vol. 109, chlyustnogo apparata u oviraptorov pp. 329-356. (Theropoda, Saurischia). In R. Bars- bold, E. I. Vorob'eva, B. Luvsandanzan, Caldwell, W. G. E., and B. R. North L. P. Tatarinov, B. A. Trofimov, V. Yu. 1984. Cretaceous stage boundaries in the Reshetov, B. B. Rodendorf, and M. A. southern interior plains ofCanada. Geol. Shishkin (eds.), Fauna, Flora i Bio- Soc. Denmark, Bull., vol. 33, pp. 57- stratigrafiya Mezozoya i Kainozoya 69. Mongolii. Sovmestnaya Sovetsko- Caldwell, W. G. E., B. R. North, C. R. Stelck, and Mongol'skaya Paleontologicheskaya J. H. Wall Ekspeditsiya, Trudy, vol. 4, pp. 34-47. 1978. A foraminiferal zonal scheme for the Barwin, John R. Cretaceous System in the interior plains 1959. Facies of the Mesaverde Formation. of Canada. In C. R. Stelck and B. D. E. Amer. Assoc. Petrol. Geol., Rocky Mtn. Chatterton (eds.), Western and Arctic Sec., Geol. Rec., pp. 139-142. Canadian biostratigraphy. Geol. Assoc. 1961a. Stratigraphy of the Mesaverde Forma- Can., Spec. Paper no. 18, pp. 495-575. tion in the southeastern part ofthe Wind Chudinov, P. K. River Basin, Fremont and Natrona 1966. Unikal'noe mestonakhozhdenie pozd- counties, Wyoming. Unpubl. M.A. the- nemelovykh presmykayushchikhsya v sis, Univ. Wyoming, Laramie, Wyo- Bayan-Khongorskom Aimake. In N. A. ming, vi + 78 pp. Marinov (ed.), Materialy po Geologii 1961b. Stratigraphy of the Mesaverde Forma- Mongol'skoi Narodnoi Respubliki. 60 AMERICAN MUSEUM NOVITATES NO. 2840

Moscow, Izdatel'stvo "Nedra," pp. 74- Cobban, William A. 78. 1958. Late Cretaceous fossil zones ofthe Pow- Clemens, William A. der River Basin, Wyoming and Mon- 1961. A Late Cretaceous mammal from Drag- tana. Wyoming Geol. Assoc. Guide- on Canyon, Utah. Jour. Paleont., vol. book, 13th Field Conf., pp. 114-119. 35, pp. 578-579. Cobban, William A., and John B. Reeside, Jr. 1963. Fossil mammals ofthe type Lance For- 1952. Correlation of the Cretaceous forma- mation, Wyoming. Part I. Introduction tions ofthe western interior ofthe United and Multituberculata. Univ. California States. Geol. Soc. America, Bull., vol. Publs. Geol. Sci., vol. 48, vi + 105 pp. 63,pp. 1011-1044. 1966. Fossil mammals ofthe type Lance For- Cope, Edward Drinker mation, Wyoming. Part II. Marsupialia. 1882. Mammalia in the Laramie Formation. Ibid., vol. 62, vi + 122 pp. Amer. Nat., vol. 16, pp. 830-831. 1970. Mesozoic mammalian evolution. Ann. Cross, C. W., A. C. Spencer, and C. W. Purington Rev. Ecol. Syst., vol. 1, pp. 357-390. 1899. Description ofthe La Plata Quadrangle 1973a. Fossil mammals ofthe type Lance For- (Colorado). U.S. Geol. Surv., Geol. At- mation, Wyoming. Part III. Eutheria and las, Folio no. 60, pp. 1-14. summary. Univ. California Publs. Geol. Cushman, Joseph A. Sci., vol. 94, vi + 102 pp. 1946. Upper Cretaceous Foraminifera of the 1973b. The roles of fossil vertebrates in inter- GulfCoastal region ofthe United States pretation of Late Cretaceous stratigra- and adjacent areas. U.S. Geol. Surv., phy ofthe San Juan Basin, New Mexico. Prof. Paper 206, iii + 241 pp. Four Corners Geol. Soc. Memoir, 1973, Dalrymple, G. Brent pp. 154-167. 1979. Critical tables for the conversion ofK-Ar 1979. Marsupialia. In J. A. Lillegraven, Z. ages from old to new constants. Geol- Kielan-Jaworowska, and W. A. Clem- ogy, vol. 7, pp. 558-560. ens (eds.), Mesozoic mammals: the first Dodson, Peter two-thirds of mammalian history. 1971. Sedimentology and taphonomy of the Berkeley, Univ. California Press, pp. Oldman Formation (Campanian), Di- 192-220. nosaur Provincial Park, Alberta (Can- 1980. Gallolestes pachymandibularis (Theria, ada). Palaeogeog., Palaeoclimatol., Pa- incertae sedis; Mammalia) from Late laeoecol., vol. 10, pp. 21-74. Cretaceous deposits in Baja California Dorf, Erling del Norte, Mexico. PaleoBios, no. 33, 1942. Upper Cretaceous floras of the Rocky 10 pp. Mountain region. II: Flora ofthe Lance Clemens, William A., and Zofia Kielan-Jawo- Formation at its type locality, Niobrara rowska County, Wyoming. Carnegie Inst. 1979. Multituberculata. In J. A. Lillegraven, Washington, Publ. no. 508, pp. 79-159. Z. Kielan-Jaworowska, and W. A. Dowling, D. B. Clemens (eds.), Mesozoic mammals: the 1915. Southern Alberta. Can. Geol. Surv., first two-thirds of mammalian history. Summary Rpt. for 1914, no. 8, pp. 43- Berkeley, Univ. California Press, pp. 99- 51. 149. Efremov, I. A. Clemens, William A., and Jason A. Lillegraven 1954. Paleontological investigations in the In press. New Late Cretaceous, advanced the- Mongolian People's Republic (prelimi- rian mammals from North America that nary results of the expeditions in 1946, fit neither the marsupial nor eutherian 1947, and 1949). Trudy Mongol. Co- molds. Contrib. Geology, Special Pa- missii., vol. 59, pp. 3-32 (in Russian). per 3. Emry, Robert J., J. David Archibald, and Charles Clemens, William A., Jason A. Lillegraven, Ev- C. Smith erett H. Lindsay, and George Gaylord Simpson 1981. A mammalian molar from the Late Cre- 1979. Where, when, and what-a survey of taceous of northern Mississippi. Jour. known Mesozoic mammal distribution. Paleont., vol. 55, pp. 953-956. In J. A. Lillegraven, Z. Kielan-Jawo- Estes, Richard D., and Paul Berberian rowska, and W. A. Clemens (eds.), Me- 1970. Paleoecology of a Late Cretaceous ver- sozoic mammals: the first two-thirds of tebrate community from Montana. Bre- mammalian history. Berkeley, Univ. viora (Harvard Univ.), no. 343, pp. 1- California Press, pp. 7-58. 35. 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 61

Fisher, D. J., C. E. Erdmann, and John B. Reeside, Assoc. Can., Spec. Paper 18, pp. 577- Jr. 594. 1960. Cretaceous and Tertiary formations of 1979a. Mammals from the Upper Cretaceous the Book Cliffs, Carbon, Emery, and Oldman Formation, Alberta. I. Alpha- Grand counties, Utah, and Garfield and don Simpson (Marsupialia). Can. Jour. Mesa counties, Colorado. U.S. Geol. Earth Sci., vol. 16, pp. 91-102. Surv., Prof. Paper 332, iv + 80 pp. 1979b. Mammals from the Upper Cretaceous Flynn, Lawrence J. Oldman Formation, Alberta. II. Pedi- In press. Late Cretaceous mammal horizons omys Marsh (Marsupialia). Ibid., vol. from the San Juan Basin, New Mexico. 16, pp. 103-113. Amer. Mus. Novitates. 1979c. Mammals from the Upper Cretaceous Forester, R. W., W. G. E. Caldwell, and F. H. Oro Oldman Formation, Alberta. III. Eu- 1977. Oxygen and carbon isotopic study of theria. Ibid., vol. 16, pp. 114-125. ammonites from the Late Cretaceous 1980. Picopsis pattersoni, n. gen. and sp., an Bearpaw Formation in southwestern unusual therian from the Upper Cre- Saskatchewan. Can. Jour. Earth Sci., vol. taceous ofAlberta, and the classification 14, pp. 2086-2100. of primitive tribosphenic mammals. Fox, Richard C. Ibid., vol. 17, pp. 1489-1498. 1970. Eutherian mammal from the early 1981. Mammals from the Upper Cretaceous Campanian (Late Cretaceous) ofAlber- Oldman Formation, Alberta. V. Eo- ta, Canada. Nature, vol. 227, pp. 630- delphis Matthew, and the evolution of 631. the Stagodontidae (Marsupialia). Ibid., 1971a. Early Campanian multituberculates vol. 18, pp. 350-365. (Mammalia: Allotheria) from the Up- 1982. Evidence ofnew lineage oftribosphenic per Milk River Formation, Alberta. Can. therians (Mammalia) from the Upper Jour. Earth Sci., vol. 8, pp. 916-938. Cretaceous of Alberta, Canada. Geo- 1971b. Marsupial mammals from the early bios, mem. spec. 6, pp. 169-175. Campanian Milk River Formation, Al- 1984a. Paranyctoides maleficus (new species), berta, Canada. In D. M. Kermack and an early eutherian mammal from the K. A. Kermack (eds.), Early mammals. Cretaceous ofAlberta. In M. R. Dawson Zool. Jour. Linn. Soc., vol. 50, suppl. 1, (ed.), Papers in vertebrate paleontology pp. 145-164. honoring Robert Warren Wilson. Car- 1972. A primitive therian mammal from the negie Mus. Nat. Hist., Spec. Publ., no. Upper Cretaceous ofAlberta. Can. Jour. 9, pp. 9-20. Earth Sci., vol. 9, pp. 1479-1494. 1984b. A primitive, "obtuse-angled" symme- 1974. Deltatheroides-like mammals from the trodont (Mammalia) from the Upper Upper Cretaceous of North America. Cretaceous of Alberta, Canada. Can. Nature, vol. 249, p. 392. Jour. Earth Sci., vol.21, pp. 1204-1207. 1976a. Additions to the mammalian local fau- Frerichs, William E. na from the Upper Milk River For- 1980. Age ofthe western interior Clioscaphites mation (Upper Cretaceous), Alberta. chouteauensis Zone. Jour. Paleont., vol. Can. Jour. Earth Sci., vol. 13, pp. 1 105- 54, pp. 366-370. 1118. Frerichs, William E., and Nancy B. Dring 1976b. Cretaceous mammals (Meniscoessus in- 1981. Planktonic foraminifera from the Smoky termedius, new species, and Alphadon Hill Shale of west central Kansas. Jour. sp.) from the lowermost Oldman For- Foram. Res., vol. 11, pp. 47-69. mation, Alberta. Ibid., vol. 13, pp. 1216- Gill, James R., and William A. Cobban 1222. 1965. Stratigraphy ofthe Pierre Shale, Valley 1977. Notes on the dentition and relationships City and Pembina Mountain areas, ofthe Late Cretaceous insectivore Gyp- North Dakota. U.S. Geol. Surv., Prof. sonictops Simpson. Ibid., vol. 14, pp. Paper 392A, iii + 20 pp. 1823-1831. 1966a. The Red Bird section ofthe Upper Cre- 1978. Upper Cretaceous terrestrial vertebrate taceous Pierre Shale in Wyoming. Ibid., stratigraphy of the Gobi Desert (Mon- no. 393A, iv + 73 pp. golian People's Republic) and western 1966b. Regional unconformity in Late Creta- North America. In C. R. Stelck, and B. ceous, Wyoming. Ibid., no. 550B, pp. D. E. Chatterton (eds.), Western and 20-27. Arctic Canadian biostratigraphy. Geol. 1973. Stratigraphy and geologic history of the 62 AMERICAN MUSEUM NOVITATES NO. 2840

Montana Group and equivalent rocks, Jeletzky, J. A. Montana, Wyoming, and North and 1968. Macrofossil zones of the marine Cre- South Dakota. Ibid., no. 776, iii + 37 taceous of the western interior of Can- PP. ada and their correlation with the zones Gradzinski, Ryszard, and T. Jerzykiewicz and stages of Europe and the western 1974. Sedimentation ofthe Barun Goyot For- interior ofthe United States. Geol. Surv. mation. Palaeont. Polonica, no. 30, pp. Canada, Paper 67-72, v + 66 pp. 111-146. Jepsen, Glenn L. Gradzinski, Ryszard, J. Kazmierczak, and Jerzy 1940. Paleocene faunas of the Polecat Bench Lefeld Formation, Park County, Wyoming. 1969. Geographical and geological data from Amer. Phil. Soc., Proc., vol. 83, pp. 217- the Polish-Mongolian Palaeontological 340. Expeditions. Ibid., no. 19, pp. 33-82. Johnston, Paul A. Gradzinski, Ryszard, Zofia Kielan-Jaworowska, 1980. First record ofMesozoic mammals from and Teresa Marayanska Saskatchewan. Can. Jour. Earth Sci., vol. 1977. Upper Cretaceous Djadokhta, Barun 17, pp. 512-519. Goyot and Nemegt formations ofMon- Johnston, Paul A., and Richard C. Fox golia, including remarks on previous 1984. Paleocene and Late Cretaceous mam- subdivisions. Acta Geol. Polonica. vol. mals from Saskatchewan, Canada. Pa- 27, pp. 281-318. laeontographica, Abt. A, Bd. 186, pp. Granger, Walter, and William K. Gregory 163-222. 1923. Protoceratops andrewsi, a pre-ceratop- Karczewska, Jadwiga, and Maria Ziembinska- sian dinosaur from Mongolia. Amer. Tworzydlo Mus. Novitates, no. 72, 9 pp. 1970. Upper Cretaceous Charophyta from the Gregory, William K., and George Gaylord Simp- Nemegt Basin, Gobi Desert. Paleont. son Polonica, no. 21, pp. 12 1-144. 1926. Cretaceous mammal skulls from Mon- 1983. Age of the Upper Cretaceous Nemegt golia. Ibid., no. 225, 20 pp. Formation (Mongolia) on charophytan Guo, Fuxiang evidence. In Z. Kielan-Jaworowska and 1984. Classification of the Asian non-marine H. Osmolska (eds.), Second Symposium Cretaceous System. Geol. Soc. Den- on Mesozoic Terrestrial Ecosystems. mark, Bull., vol. 33, pp. 115-122. Acta Palaeont. Polonica, vol. 28, pp. Hancock, John M., and Erle G. Kauffman 137-146. 1979. The great transgressions ofthe Late Cre- Kauffman, Erle G. taceous. Geol. Soc. London, Jour., vol. 1968. The Upper Cretaceous Inoceramus of 136, pp. 175-186. Puerto Rico. In J. B. Saunders (ed.), Harland, W. B., A. V. Cox, P. G. Llewellyn, C. A. Fourth Caribbean Geological Confer- G. Pickton, A. G. Smith, and R. Walters ence, Transactions. Arima, Trinidad, 1982. A geologic time scale. Cambridge, Cam- and Tobago, Caribbean Printers, pp. bridge Univ. Press, xi + 131 pp. 203-218. Hatcher, John B. 1970. Population systematics, radiometrics 1904. An attempt to correlate the marine with and zonation-a new biostratigraphy. the non-marine formations of the Mid- N. Amer. Paleont. Conv., Proc., pt. F, dle West. Amer. Phil. Soc., Proc., vol. pp. 612-666. 43, pp. 341-365. 1973. Cretaceous Bivalvia. In A. Hallam (ed.), Hatcher, John B., and T. W. Stanton Atlas ofpalaeobiogeography. New York, 1903. The stratigraphic position ofthe Judith Elsevier, pp. 353-383. River Beds and their correlation with 1975. Dispersal and biostratigraphic potential the Belly River Beds. Science, n.s., vol. ofCretaceous benthonic Bivalvia in the 18, pp. 211-212. western interior. In W. G. E. Caldwell Hayden, F. B. (ed.), The Cretaceous System in the 1871. Geology of the Missouri Valley. U.S. western interior ofNorth America. Geol. Geol. Surv. Terr., Fourth Ann. Prelim. Assoc. Canada, Spec. Paper No. 13, pp. Rpt., pt. 2, chap. 7, pp. 85-98. 163-194. Holmes, W. H. 1977. Geological and biological overview: 1877. Geological report on the San Juan dis- western interior Cretaceous basin. trict. Ibid., Ninth Ann. Rpt., for 1875, Mountain Geol., vol. 14, pp. 75-99. pp. 241-276. 1979. Cretaceous. In R. A. Robison and C. 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 63

Teichert (eds.), Treatise on invertebrate Postcranial skeleton in Zalambdalesti- paleontology, Part A, introduction: fos- dae. Ibid., no. 38, pp. 3-4 1. silization (taphonomy), biogeography, 1982. Marsupial-placental dichotomy and pa- and biostratigraphy. Geol. Soc. Amer., leogeography of Cretaceous Theria. In pp. 418-487. E. M. Gallitelli (ed.), Proc. First Int. Keefer, William R., and Ernest I. Rich Meeting on "Palaeontology, essential of 1957. Stratigraphy of the Cody Shale and historical geology"; held in Venice, younger Cretaceous and Paleocene rocks Fondazione Giorgio Cini, 2-4 June, in the western and southern parts ofthe 1981, Inst. Paleont., Univ. Modena, It- Wind River Basin, Wyoming. Wyo- aly, pp. 367-383. ming Geol. Soc. Guidebook, 12th Field 1984. Evolution of the therian mammals in Conf., pp. 71-78. the Late Cretaceous of Asia. Part VII. Kennedy, William J. Synopsis. Palaeont. Polonica, no. 46, pp. 1984. Ammonite faunas and the 'standard' 173-183. zones ofthe Cenomanian to Maastrich- Kielan-Jaworowska, Zofia, and Rinchen Barsbold tian Stages in their type areas, with some 1972. Narrative of the Polish-Mongolian pa- proposals for the definition of the stage laeontological expeditions 1967-1971. boundaries by ammonites. Geol. Soc. Ibid., no. 27, pp. 5-13. Denmark, Bull., vol. 33, pp. 147-161. Kielan-Jaworowska, Zofia, Jeffrey G. Eaton, and Kennedy, William J., and William A. Cobban Thomas M. Bown 1976. Aspects of ammonite biology, biogeog- 1979. Theria of metatherian-eutherian grade. raphy, and biostratigraphy. London, In J. A. Lillegraven, Z. Kielan-Jawo- Palaeont. Assoc., Spec. Papers Pa- rowska, and W. A. Clemens (eds.), Me- laeont., no. 17, v + 94 pp. sozoic mammals: the first two-thirds of Kennedy, William J., and Gilles S. Odin mammalian history. Berkeley, Univ. 1982. The Jurassic and Cretaceous time scale California Press, pp. 182-191. in 1981. In G. S. Odin (ed.), Numerical Kielan-Jaworowska, Zofia, and Robert E. Sloan dating in stratigraphy. Part I. New York, 1979. Catopsalis (Multituberculata) from Asia Wiley, pp. 557-592. and North America and the problem of Kielan-Jaworowska, Zofia taeniolabidid dispersal in the Late Cre- 1969. Preliminary data on the Upper Creta- taceous. Acta Palaeont. Polonica, vol. ceous eutherian mammals from Bayn 24, pp. 187-197. Dzak, Gobi Desert. Palaeont. Polonica, Kielan-Jaworowska, Zofia, and A. V. Sochava no. 19, pp. 171-191. 1969. The first multituberculate from the up- 1970a. New Upper Cretaceous multitubercu- permost Cretaceous ofthe Gobi Desert, late genera from Byan Dzak, Gobi Des- Mongolia. Ibid., vol. 14, pp. 355-371. ert. Ibid., no. 21, pp. 35-49. 1970b. Unknown structures in multitubercu- Kielan-Jaworowska, Zofia, and Boris A. Trofimov late skull. Nature, vol. 226, pp. 974- 1980. Cranial morphology of the Cretaceous 976. eutherian mammal Barunlestes. Ibid., 1974a. Multituberculate succession in the Late vol. 25, pp. 167-185. Cretaceous of the Gobi Desert (Mon- Kraus, Mary J. golia). Palaeont. Polonica, no. 30, pp. 1979. Eupantotheria. In J. A. Lillegraven, Z. 23-44. Kielan-Jaworowska, and W. A. Clem- 1974b. Migrations of the Multituberculata and ens (eds.), Mesozoic mammals: the first the Late Cretaceous connections be- two-thirds of mammalian history. tween Asia and North America. Ann. Berkeley, Univ. California Press, pp. South African Mus., vol. 64, pp. 231- 162-171. 243. Krause, David W., and Donald Baird 1975a. Preliminary description of two new eu- 1979. Late Cretaceous mammals east of the therian genera from the Late Cretaceous North American Western Interior Sea- ofMongolia. Palaeont. Polonica, no. 33, way. Jour. Paleont., vol. 53, pp. 562- pp. 5-16. 565. 1975b. Evolution of the therian mammals in Kyansep-Romashkina, N. P. the Late Cretaceous ofAsia. Part I. Del- 1975. Nekotorye pozdneyurskie i melovye tatheridiidae. Ibid., no. 33, pp. 103-13 1. kharofity Mongolii. In N. N. Krama- 1979. Evolution of the therian mammals in renko et al. (eds.), Iskopaemaya Fauna the Late Cretaceous of Asia. Part III. i Flora Mongolii. Sovmestnaya Sovet- 64 AMERICAN MUSEUM NOVITATES NO. 2840

sko-Mongol'skaya Paleont. Eksped., McKenna, Malcolm C. Trans., vol. 2, pp. 181-204. 1969. The origin and early differentiation of Lambe, Lawrence M. therian mammals. New York Acad. Sci., 1902. New genera and species from the Belly Ann., vol. 167, pp. 217-240. River series (mid-Cretaceous). Geol. McLean, J. R. Surv. Can., Contrib. Canad. Palaeont., 1971. Stratigraphy of the Upper Cretaceous vol. 3, pt. 2, pp. 25-81. Judith River Formation in the Cana- Langston, Wann, Jr. dian Great Plains. Saskatchewan Res. 1959. Alberta and fossil vertebrates. Alberta Con., Geol. Div., Rpts., no. 11, xi + 96 Soc. Petrol. Geol., 9th Ann. Field Conf. pp- Guide Book, pp. 8-19. Marinov, N. A. Lanphere, Marvin A., and David L. Jones 1957. Stratigrafiya Mongol'skoi Narodnoi 1978. Cretaceous time scale from North Respubliki. Moscow, Akad. Nauk SSSR, America. In G. V. Cohee, M. F. Glaess- 268 pp. ner, and H. D. Hedberg (eds.), Contri- Marinov, N. A., L. P. Zonenshain, and V. A. Bla- butions to the geologic time scale. Tulsa, gonravov (eds.) Amer. Assoc. Petrol. Geol., Studies in 1973. Geologiya Mongol'skoi Narodnoi Res- Geol., no. 6, pp. 259-268. publiki. Vol. I. Stratigrafia. Izdatel'stvo Lefeld, Jerzy "Nedra," 583 pp. 1971. Geology ofthe Djadokhta Formation at Marks, Peter Bayn Dzak (Mongolia). Palaeont. Po- 1984. Proposal for the recognition of bound- lonica, no. 25, pp. 10 1-127. aries between Cretaceous stages by Lehman, Thomas M. means of planktonic foraminiferal bio- 1984. The multituberculate Essonodon browni stratigraphy. Geol. Soc. Denmark, Bull., from the Upper Cretaceous Naashoi- vol. 33, pp. 163-169. bito Member ofthe Kirtland Shale, San Marsh, Othniel C. Juan Basin, New Mexico. Jour. Vert. 1879. Notice ofnew Jurassic mammals. Amer. Paleont., vol. 4, pp. 602-603. Jour. Sci., ser. 3, vol. 15, pp. 396-398. Lillegraven, Jason A. 1889. Discovery of Cretaceous Mammalia. 1969. Latest Cretaceous mammals of upper Ibid., ser. 3, vol. 38, pp. 81-92. part of Edmonton Formation of Alber- Martinson, G. G. ta, Canada, and review of marsupial- 1961. Mezozoiskie i Kainozoiskie mollyuski placental, dichotomy in mammalian kontinental'nykh otlozhenii Sibirskoi evolution. Paleont. Contrib. Univ. platformy, Zabaikal'ya i Mongolii. Tru- Kansas, art. 50 (Vertebrata 12), 122 pp. dy Baikal'skoi. Limnol. Stantsii Akad. Love, J. David, and Ann Coe Christiansen Nauk SSSR, Sibir. Otdel, Vostochno- 1983. Preliminary geologic map ofWyoming. Sibir. Filial, vol. 19, 318 pp. U.S. Geol. Surv., Open-File Rpt. 83- 1973. 0 stratigrafii Yurskikh i Melovykh 802, scale 1:500,000. otlozhenii Mongolii. Izv. Akad. Nauk Lowrie, William, and Walter Alvarez SSSR, Geol. Ser., no. 12, pp. 89-95. 1977. Upper Cretaceous-Paleocene magnetic 1975. To the question about principles ofstra- stratigraphy at Gubbio, Italy. III. Upper tigraphy and relation of Mesozoic con- Cretaceous magnetic stratigraphy. Geol. tinental deposits. In Stratigraphy ofMe- Soc. Amer., Bull., vol. 88, pp. 374-377. sozoic deposits of Mongolia. Joint Lupton, Carter, Diane Gabriel, and Robert M. Soviet-Mongolian Geol. Expeds., West Trans., vol. 13, pp. 7-24. Leningrad. 1980. Paleobiology and depositional setting of Martinson, G. G., A. V. Sochava, and R. Barsbold a Late Cretaceous vertebrate locality, 1969. 0 stratigraficheskom raschlenii verkh- Hell Creek Form-ation, McCone Coun- nemelovykh otlozhenii Mongolii. Dokl. ty, Montana. Contrib. Geol., Univ. Wy- Akad. Nauk SSSR, vol. 189, pp. 1081- oming, vol. 18,pp. 117-126. 1084. McGookey, Donald P., John D. Haun, Lyle A. Maryanska, Teresa, and Halszka Osmolska Hale, H. G. Goodell, Donald G. McCubbin, 1975. Protoceratopsidae (Dinosauria) ofAsia. Robert J. Weimer, and George R. Wulf Palaeont. Polonica, no. 33, pp. 133-18 1. 1972. Cretaceous system. In W. W. Mallory 1981. Cranial anatomy of Saurolophus an- et al. (eds.), Geologic atlas ofthe Rocky gustirostris with comments on the Asian Mountain region. Denver, Rocky Hadrosauridae (Dinosauria). Ibid., no. Mountain Assoc. Geol., pp. 190-228. 42, pp. 5-24. 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 65

Matsumoto, T. Olsson, Richard K. 1980. Inter-regional correlation of transgres- 1964. Late Cretaceous planktonic foraminif- sions and regressions in the Cretaceous era from New Jersey and Delaware. Mi- period. Cretaceous Res., vol. 1, pp. 359- cropaleontology, vol. 10, pp. 157-188. 373. Osmolska, Halszka Molenaar, C. M. 1980. The Late Cretaceous vertebrate assem- 1983. Major depositional cycles and regional blages of the Gobi Desert, Mongolia. correlations ofUpper Cretaceous rocks, Mem. Soc. geol. France, n.s., vol. 59, southern Colorado Plateau and adjacent no. 139, pp. 145-150. areas. In M. W. Reynolds and E. D. Osmolska, Halszka, E. Roniewicz, and R. Bars- Dolly (eds.), Mesozoic paleogeography bold of the west-central United States. Den- 1972. A new dinosaur, Gallimimus bullatus n. ver, Soc. Econ. Paleont. Min., Rocky sp. (Ornithomimidae) from the Upper Mountain Sect., Rocky Mountain Pa- Cretaceous of Mongolia. Palaeont. Po- leogeography Symp. 2, pp. 201-224. lonica, no. 27, pp. 103-143. NACSN (North American Commission on Strati- Ostrom, John H. graphic Nomenclature) 1965. Cretaceous vertebrate faunas of Wyo- 1983. North American stratigraphic code. ming. Wyoming Geol. Assoc. Guide- Amer. Assoc. Petrol. Geol., Bull., vol. book, 19th Field Conference, pp. 35- 67, pp. 841-875. 41. Nichols, Douglas J., S. R. Jacobson, and R. H. Palmer, Allison R. Tschudy 1983. The Decade of North American Geol- 1982. Cretaceous palynomorph biozones for ogy 1983 geologic time scale. Geology, the central and northern Rocky Moun- vol. 11, pp. 503-504. tain region of the United States. In R. Pessagno, Emile A., Jr. B. Powers (ed.), Geologic studies of the 1967. Upper Cretaceous planktonic Forami- Cordilleran Thrust Belt: Denver, Rocky nifera from the western Gulf Coastal Mtn. Assoc. Geol., vol. II, pp. 721-733. Plain. Palaeontographica Americana, North, B. R., and G. E. Caldwell vol. 5, pp. 245-445. 1975. Foraminiferal faunas in the Cretaceous 1969. Upper Cretaceous stratigraphy of the system of Saskatchewan. In W. G. E. Western Gulf Coast area of Mexico, Caldwell (ed.), The Cretaceous system Texas, and Arkansas. Geol. Soc. Amer., in the Western Interior ofNorth Amer- Mem. 1l1, xiii + 139 pp. ica. Geol. Assoc. Can., Spec. Paper no. Premoli Silva, Isabella 13, pp. 303-331. 1977. Upper Cretaceous-Paleocene magnetic Novacek, Michael J., and William A. Clemens stratigraphy at Gubbio, Italy. II. Bio- 1977. Aspects of intrageneric variation and stratigraphy. Geol. Soc. Amer., Bull., evolution of Mesodma (Multitubercu- vol. 88, pp. 371-374. lata, Mammalia). Jour. Paleont., vol. 51, Prothero, Donald R. pp. 701-717. 1981. New Jurassic mammals from Como Novozhilov, N. I. Bluff, Wyoming, and the interrelation- 1954. Mestonakhozhdeniya mlekopitayu- ships ofnon-tribosphenic Theria. Amer. shchikh nizhnego eotsena i verkhnego Mus. Nat. Hist., Bull., vol.,167, pp. 277- paleotsena Mongolii. Trudy Mongol- 325. skoi Komissii, Akad. Nauk SSSR, no. Rawson, Peter F., Dennis Curry, Frank C. Dilley, 59, pp. 33-46. John M. Hancock, William J. Kennedy, John Nowinski, Aleksander W. Neale, Christopher J. Wood, and Bernard C. 1971. Nemegtosaurus mongoliensis n. gen., n. Worssam sp. (Sauropoda) from the uppermost 1978. A correlation ofCretaceous rocks in the Cretaceous of Mongolia. Palaeont. Po- British Isles. Geol. Soc. London, Spec. lonica, no. 25, pp. 57-81. Rpt. no. 9, 70 pp. Obradovich, John D., and William A. Cobban Reeside, John B., Jr. 1975. A time-scale for the Late Cretaceous of 1924. Upper Cretaceous and Tertiary forma- the western interior of North America. tions ofthe western part ofthe San Juan In W. G. E. Caldwell (ed.), The Creta- Basin, Colorado and New Mexico. U.S. ceous System in the western interior of Geol. Surv., Prof. Paper 134, pp. 1-70. North America. Geol. Assoc. Can., Spec. Riccardi, A. C. Paper no. 13, pp. 31-54. 1983. Scaphitids from the upper Campanian- 66 AMERICAN MUSEUM NOVITATES NO. 2840

lower Maastrichtian Bearpaw Forma- Russell, Dale A. tion of the western interior of Canada. 1967. A census of dinosaur specimens col- Geol. Surv. Can., Bull., vol. 354, pp. 1- lected in western Canada. Can. Natl. 103. Mus. Nat. Hist. Papers, no. 36, 13 pp. Rich, Ernest I. 1970. Tryannosaurs from the Late Cretaceous 1958. Stratigraphic relation of latest Creta- ofwestern Canada. Can. Natl. Mus. Nat. ceous rocks in parts of Powder River, Sci., Publs. Palaeont., no. 1, viii + 34 Wind River, and Big Horn basins, Wy- pp- oming. Amer. Assoc. Petrol. Geol., Bull., 1972. Ostrich dinosaurs from the Late Cre- vol. 42, pp. 2424-2443. taceous of western Canada. Can. Jour. Rose, Kenneth D. Earth Sci., vol. 9, pp. 375-402. 1981. The Clarkforkian land-mammal age and Russell, Dale A., and T. Potter Chamney mammalian faunal composition across 1967. Notes on the biostratigraphy of dino- the Paleocene-Eocene boundary. Univ. saurian and microfossil faunas in the Michigan, Papers Paleont., vol. 26, x + Edmonton Formation (Cretaceous), Al- 197 pp. berta. Natl. Mus. Canada, Nat. Hist. Pa- Rozhdestvenskii, A. K. pers, no. 35, 22 pp. 1957. A short conclusion on the study ofMon- Russell, Loris S. golian fossil vertebrates. Vertebrata 1937. New and interesting mammalian fossils PalAsiatica, vol. 1, pp. 169-185 (in from western Canada. Roy. Soc. Can., Russian, with Chinese and English sum- Trans., 3rd ser., Sec. 4, vol. 30, pp. 75- maries). 80. 1971. Izuchenie dinozavrov Mongolii i ikh rol' 1952. Cretaceous mammals of Alberta. Natl. v raschlenenii kontinental'nogo Mezo- Mus. Canada, Bull., vol. 126, pp. 110- zoya. In N. S. Zaitsev, B. Luvsandan- 119. zan, V. V. Menner, T. G. Pavlova, A. 1964. Cretaceous non-marine faunas ofnorth- V. Peive, P. P. Timofeev, 0. Tomur- western North America. Roy. Ontario togoo, B. A. Trofimov, and A. L. Yan- Mus., Life Sci., Contrib. 61, 24 pp. shin (eds.), Fauna Mezozoya i Kaino- 1975. Mammalian faunal succession in the zoya zapadnoi Mongolii. Sovmestnaya Cretaceous system of western North Sovetsko-Mongol'skaya Nauchno-Iss- America. In W. G. E. Caldwell (ed.), ledovatel'skaya Geologicheskaya Ek- The Cretaceous System in the western speditsiya, Trudy, vol. 3, pp. 21-32. interior ofNorth America. Geol. Assoc. 1972. Razvitie Dinozavrovykh faun b Azii i Can., Spec. Paper no. 13, pp. 137-16 1. na drugikh materikakh i paleogeografi- 1983. Evidence for an unconformity at the ya Mezozoya. Izvestiya Akad. Nauk Scollard-Battle contact, Upper Creta- SSSR, Ser. Geol., no. 12, pp. 115-133. ceous strata, Alberta. Can. Jour. Earth 1974. A history of the dinosaur fauna from Sci., vol. 20, pp. 1219-123 1. Asia and other continents and some problems of paleogeography. In R. Russell, Loris S., and R. W. Landes Barsbold, E. I. Vorob'eva, B. Luvsan- 1940. Geology ofthe southern Alberta plains. danzan, L. P. Tamarinov, B. A. Trofi- Can. Geol. Surv., Mem., vol. 221, pp. mov, V. Yu. Reshetov, B. B. Roden- 1-223. dorf, and M. A. Shishkin (eds.), Ryer, Thomas A. Mesozoic and Cenozoic faunas and bio- 1983. Transgressive-regressive cycles and the stratigraphy of Mongolia. Trans. Joint occurrence of coal in some Upper Cre- Soviet-Mongolian Paleont. Exped., vol. taceous strata ofUtah. Geology, vol. 1, 1, pp. 107-131. Acad. Sci. USSR, Pa- pp. 207-210. leontological Institute, Moscow (in Sahni, Ashok Russian). 1968. Community structure of Campanian 1977. A study of dinosaurs in Asia. Jour. Pa- (Upper Cretaceous) vertebrates from the laeont. Soc. India, vol. 20, pp. 102-119. Judith River Formation in north central 1978. Nadotryad Dinosauria. Mlekopitayu- Montana. Proc. Internat. Paleont. shchie. In V. N. Shimanskii and A. N. Union, 23rd Internat. Geol. Congr., pp. Solov'ev (eds.), Razvitie i smena orga- 339-344. nicheskogo mira na rubezhe Mezozoya 1972. The vertebrate fauna ofthe Judith River i Kainozoya. Moscow, Akad. Nauk Formation, Montana. Amer. Mus. Nat. SSSR, Izdatel'stvo "Nauka," pp. 5 7-82. Hist., Bull., vol. 147, pp. 321-412. 1986 LILLEGRAVEN AND McKENNA: "MESAVERDE" MAMMALS 67

Savage, Donald E. (Albian) of Texas. Peabody Mus. Nat. 1962. Cenozoic geochronology of the fossil Hist., Yale Univ., Postilla, no. 93, 18 mammals of the Western Hemisphere. pp- Rev. Mus. Arg. Cien. Nat. "Bernardino Sloan, Robert E. Rivadavia," Cien. Zool., vol. 8, pp. 53- 1976. The ecology of dinosaur extinction. In 67. C. S. Churcher (ed.), Essays on palaeon- Schulz, Max-Gotthard, Gundolf Ernst, Hartmut tology in honour ofLoris Shano Russell. Ernst, and Friedrich Schmidt Roy. Ontario Mus., Life Sci., Misc. 1984. Coniacian to Maastrichtian stage Publs., pp. 134-154. boundaries in the standard section for 1981. Systematics of Paleocene multituber- the Upper Cretaceous white chalk ofNW culates from the San Juan Basin, New Germany(Lagerdorf-Kronsmoor-Hem- Mexico. In S. G. Lucas, J. K. Rigby, Jr., moor): definitions and proposals. Geol. and B. S. Kues (eds.), Advances in San Soc. Denmark, Bull., vol. 33, pp. 203- Juan Basin paleontology. Albuquerque, 215. Univ. New Mexico Press, pp. 127-160. Shapurji, Soli S. Sloan, Robert E., and Loris S. Russell 1978. Depositional environments and corre- 1974. Mammals from the St. Mary River For- lation of the Mesaverde Formation, mation (Cretaceous) of southwestern Wind River Basin, Wyoming. Wyo- Alberta. Roy. Ontario Mus., Life Sci., ming Geol. Assoc. Guidebook, 13th Contrib., no. 95, pp. 1-21. Field Conf., pp. 167-180. Sloan, Robert E., and Leigh Van Valen Shuvalov, V. F., and V. M. Chkhikvadze 1965. Cretaceous mammals from Montana. 1975. Novye dannye o pozdnemelovykh che- Science, vol. 148, pp. 220-227. repakhakh yuzhnoi Mongolii. In N. N. Smit, J., and S. van der Kaars Kramarenko et al. (eds.), Iskopaemaya 1984. Terminal Cretaceous extinctions in the fauna i flora Mongolii. Sovmestnaya Hell Creek area, Montana: compatible Sovetsko-Mongol'skayaPaleont. Eksped., with catastrophic extinction. Ibid., vol. Trans., vol. 2, pp. 214-229. 223, pp. 1177-1179. Simpson, George Gaylord Sohl, N. F. 1925. Mesozoic mammal skull from Mongo- 1967. Upper Cretaceous gastropods from the lia. Amer. Mus. Novitates, no. 201, 11 Pierre Shale at Red Bird, Wyoming. U.S. PP. Geol. Surv., Prof. Paper 393B, pp. 1- 1927a. Mammalian fauna of the Hell Creek 46. Formation of Montana. Ibid., no. 267, Stankevitch, E. S., and E. Khand 7 pp. 1976. Ostrakody Barungoyotskoi Svity verkh- 1927b. Mesozoic Mammalia. VIII. Genera of nego Mela Zaaltaiskoi Gobi (MNR). In Lance mammals other than multitu- N. N. Kramarenko et al. (eds.), Paleon- berculates. Amer. Jour. Sci., ser. 5, vol. tologiya i biostratigraphiya Mongolii. 14, pp. 121-130. Trudy Sovmestnaya Sovetsko-Mon- 1928. Further notes on Mongolian Cretaceous gol'skaya Paleontologicheskaya Ekspe- mammals. Amer. Mus. Novitates, no. ditsiya, vol. 3, pp. 159-161 (English 329, 14 pp. abstr.). 1929. American Mesozoic Mammalia. New Stanton, T. W., and John B. Hatcher Haven, Yale Univ. Press, Peabody Mus. 1903. The stratigraphic position of the Judith (Yale Univ.), Mem., vol. 3, pt. 1, xv + River Beds and their correlation with 171 pp. the Belly River Beds. Science, n.s., vol. 1936. Additions to the Puerco fauna, Lower 18, pp. 211-212. Paleocene. Amer. Mus. Novitates, no. 1905. Geology and paleontology ofthe Judith 849, 11 pp. River beds. U.S. Geol. Surv., Bull., vol. Sissingh, W. 257, pp. 1-128. 1977. Biostratigraphy of Cretaceous calcar- Steiger, R. H., and E. Jager eous nannoplankton. Geologie en 1977. Subcommission on geochronology: Mijnbouw, vol. 56, pp. 37-65. convention on the use of decay con- 1978. Microfossil biostratigraphy and stage- stants in geo- and cosmochronology. stratotypes ofthe Cretaceous. Ibid., vol. Earth Planet. Sci. Lett., vol. 36, pp. 359- 57, pp. 433-440. 362. Slaughter, B. H. Sulimski, Andrzej 1965. A therian from the Lower Cretaceous 1975. Macrocephalosauridae and Polygly- 68 AMERICAN MUSEUM NOVITATES NO. 2840

phanodontidae (Sauria) from the Late Van Valen, Leigh, and Robert E. Sloan Cretaceous of Mongolia. Palaeont. Po- 1965. The earliest primates. Science, vol. 150, lonica, no. 33, pp. 25-102. pp. 743-745. Surlyk, Finn Waage, Karl M. 1984. The Maastrichtian Stage in NW Europe, 1975. Deciphering the basic sedimentary and its brachiopod zonation. Geol. Soc. structure of the Cretaceous System in Denmark, Bull., vol. 33, pp. 217-223. the western interior. In W. G. E. Cald- Tedford, Richard H. well (ed.), The Cretaceous System in the 1970. Principles and practices of mammalian western interior ofNorth America. Geol. geochronology in North America. N. Assoc. Can., Spec. Paper no. 13, pp. 55- Amer. Paleont. Conv., Proc., pt. F, pp. 81. 666-703. Weimer, Robert J. Trofimov, Boris A. 1960. Upper Cretaceous stratigraphy, Rocky 1975. Novye Dannye o Buginbaatar Kielan- Mountain area. Amer. Assoc. Petrol. Jaworowska et Sochava, 1969 (Mam- Geol., Bull., vol. 44, pp. 1-20. malia, Multituberculata) iz Mongolii. Williams, Gordon D., and C. F. Burk, Jr. In N. N. Kramarenko et al. (eds.), 1964. Upper Cretaceous. In R. G. McCrossan Iskopaemaya fauna i flora Mongolii. and R. P. Glaister (eds.), Geological his- Sovmestnaya Sovetsko-Mongol'skaya tory of western Canada. Calgary, Al- Paleont. Eksped., Trans., vol. 2, pp. 7- berta Soc. Petrol. Geol., pp. 169-189. 13. Wood, Horace E., 2nd, Ralph W. Chaney, John Vail, P. R., R. M. Mitchum, Jr., and S. Thompson Clark, Edwin H. Colbert, Glenn L. Jepsen, John III B. Reeside, Jr., and Chester Stock 1977. Seismic stratigraphy and global changes 1941. Nomenclature and correlation of the of sea level, part 4: global cycles of rel- North American continental Tertiary. ative changes of sea level. In C. E. Pay- Geol. Soc. Amer., Bull., vol. 52, pp. 1- ton (ed.), Seismic stratigraphy-appli- 48. cations to hydrocarbon exploration. Woodburne, Michael 0. (ed.) Tulsa, Amer. Assoc. Petrol. Geol., Mem. In press. Vertebrate paleontology as a disci- 26, pp. 83-97. pline in geochronology. Berkeley, Univ. Van Hinte, J. E. California Press. 1976. A Cretaceous time scale. Amer. Assoc. Young, Keith Petrol. Geol., Bull., vol. 60, pp. 498- 1963. Upper Cretaceous ammonites from the 516. Gulf Coast of the United States. Univ. Van Valen, Leigh Texas, Publ. no. 6304, ix + 373 pp. 1966. Deltatheridia, a new order ofmammals. Zapp, A. D., and William A. Cobban Amer. Mus. Nat. Hist., Bull., vol. 132, 1962. Some Late Cretaceous strand lines in pp. 1-126. southern Wyoming. U.S. Geol. Surv., 1978. The beginning of the age of mammals. Prof. Paper 450D, pp. 52-55. Evol. Theory, vol. 4, pp. 45-80.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available l free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, New York 10024.