PaleoBios 19(3): 15-36, December 15, 1999 © 1999 University of California Museum of Paleontology Evolution of enamel microstructure of archaic ungulates ("Condylarthra") and comments on some other early Tertiary mammals CLARA STEFEN Department of Integrative Biology and Museum of Paleontology, University of California, Berkeley, CA 94720-4780 cmstefen@aol.com. The enamel structure of early Tertiary mammals, and in particular the Hunter-Schreger bands (IISB), is surveyed and discussed in light of the correlations between IISB occurrence and body size, chewing mechanisms and Other factors. The previously proposed positive correlation between body size and MSB occurrence is generally supported, as well as the previously stated exceptions; a causal relationship is not assumed. A positive correlation between the occurrence of IISB and divergence from the structure ol the primitive tribosphenic molar and insectivorous feeding is also observed, but exceptions do occur. Poorly-developed HSB, well developed IISB, and a specialized bending of prisms in advanced periptychids probably evolved independently and directly from radial enamel. Observations indicate convergent evolution of HSB within different families of archaic ungulates as well as in other mammalian groups, such as Pantolesta, Pantodonta, Primates, Rodentia, and Tillodontia. Within Ungulata underlying synapomorphies or developmental constraints may facilitate parallel acquisition of HSB in several families. It is hypothesized that co occurring changes in dental morphology, body size, overall ecology, and chewing pattern in eutherian mammals of the early Tertiary resulted in the evolution of HSB. However, combinations of these factors differed in different lineages. Through biomechanical constraints, either one factor or a combination of factors could have caused the alteration of the enamel structure to HSB or specialized bending. INTRODUCTION The evolution of mammalian enamel structure involved two phases: the development of prisms and the develop- ment of schmelzmuster (Kocnigswald 1997a). Schmelzmuster refers to the distribution of different enamel types within individual teeth. An enamel type is character- ized by the repetitive spatial arrangement of prisms and interprismatic enamel in a volume of enamel (Kocnigswald and Clemens 1992, Kocnigswald and .Sander 1997). The occurrence of prisms in early mammals has been studied in detail (Carlson 1990, Wood and vStcrn 1997, Wood et al. 1999). All prisms in primitive prismatic enamel are parallel and rise straight from the enamel-dentine junc- tion (EDJ) towards the outer surface of the tooth forming radial enamel (Kocnigswald and Clemens 1992, Fig. 1A). Subsequent evolution has led to enamel types in which prisms follow different courses from the EDJ to the out- side. Prisms are no longer parallel, but show different angles to each other and cross or decussate from the EDJ to the outer surface. The most common enamel type with decussating prisms in mammals is Hunter-Schreger bands (HSB) in which prisms arc parallel in one layer and oriented at angles to the prisms of the adjacent layers (e.g., Figs. IB, 3C, D). The bands usually run horizontally around the tooth crown. In contrast to the relatively detailed knowl- edge of prism occurrence in early mammals, the evolution of I ISB in early Tertiary mammals is not well understood. A preliminary study of early Tertiary mammals indicates a positive correlation between an increase in body size and the occurrence of HSB at the beginning of their dietary diversification (Kocnigswald et al. 1987). An evolutionary sequence from radial enamel to poorly-developed decussa- tion to well-developed decussation was also postulated by these authors. Poorly-developed decussation is character ized by prisms showing low angles between their courses in the middle of the enamel but parallel prisms at the EDJ and towards the outside of the tooth (Kocnigswald et al. 1987, Fig. 1C). Well-developed IISB run through the whole thickness of the enamel and the angle between the prisms of adjacent bands is greater. Although authors have ques- tioned the co-evolution of body size and HSB and the proposed evolutionary sequence (Maas and Thcwissen 1995), Maas and O'Leary (1995) demonstrated the occur- rence of HSB with increasing body size in some primate lineages. In this study enamel from a variety of early Tertiary mammals was examined in order to clarify how body size, HSB evolution, and other factors like food and tooth use are related. Archaic ungulates ("Condylarthra'') were cho- sen as the main study group for several reasons: (1) they are the most taxonomically diverse group of early Tertiary North American therian mammals; (2) the enamel of some "Condylarthra" is already known (Kocnigswald et al. 1987, Maas and Thcwissen 1995, Stefen 1997b); (3) material of several other species was available for study; and (4) some "condylarths" are thought to be the basal members of modern groups (e.g., Van Valen 1978). Data on represen- tatives of other groups such as Artiodactyla, Perissodactyla, Tillodontia, Pantodonta, Pantolcstidida, Leptictida, Pri- mates, and Rodentia are presented for comparison to the "condylarths." 16 PALEOBIOS, VOL. 19, NUMBER 3, 1999 ~" STEFEN-ENAMEL MICROSTRUCTURH IN ARCHAIC; UNGULATES 17 The "Condylarthra" is not a monophyletic group but rather a waste basket for taxa with similar dentition (Prothero 1993). Several authors have suggested total aban- donment of the name (Prothero et al. 1988, Archibald 1998). "Condylarthra" is used here to refer to a vast group of archaic ungulates because the taxa included in this group continue to be well known under this name even though the taxonomy and phylogeny of several "condylarth" groups remains poorly understood. Hollowing McKcnna and Bell (1998) Condylarthra sensu stricto as used here includes Hyopsodontidae, Mioclaenidae, Phenacodontidae and Pcriptychidae, while Arctocyonidae and Oxyclaenidae be- long to the Procreodi. However, see Archibald (1998) for alternative classification. The "Condylarthra" vary in body size with ml lengths ranging from about 3 mm (Protungulatum Sloan & Van Valen 1965) to about 10 mm (PcriptyclmsCope 1881) but reach 33 mm or more in Harpaijolcstcs Woltman, 1901. The dentitions of "Condylarthra" are as diverse as their range in size. Most have dentitions best described as unspeciali/.cd and adapted for an omnivorous diet. The primitive Protungulatum retains a dentition well adapted for piercing and crushing, which is similar to the basic tribosphenic molar patterns. Specializations towards a more shearing and crushing dentition occur in mesonychids. Phcnacodontids develop additional cusps and lophs, which in modern mammals are associated with hcrbivory. Periptychids developed blunt premolars that allowed for increased compressional chewing forces (Rensberger 1986). MATERIALS AND METHODS The material used in the study is listed in Appendix 1; the material studied with scanning electron microscopy is listed in Table 1. Although it would be ideal and most consistent to compare homologous teeth or, at least, teeth of the same dental category for all species, the fossil material is not always available for sectioning. 'Therefore, it is assumed that the enamel microstructure seen in one or several teeth of one or a few specimens of a taxon is representative for all teeth of a given taxon. All specimens were examined for USB using light mi croscopy. In most cases IISB can be observed in the enamel of unprepared teeth with a binocular microscope using angled light (Boyde 1976, Koenigswald and Pfretzschner 1987). Hither light or dark bands appear depending on the angle of light to the prisms. Natural fracture planes and sectioned teeth were analyzed using an Environmental Scan- ning Microscope (ESEM; manufactured by Electro Scan Corporation), which does not require coating of the speci- mens with a conductive material. Surfaces of the original fracture planes or slightly ground specimens (see below) were scanned. When possible, specimens were embedded in polyester resin and cut vertically or horizontally in relation to the long axis of each tooth. Tangential sections were cut tangential to the curvature of the outer surface of the teeth at different heights along the tooth crown. All sections were made on an Isomer low-speed saw (Buehler, with a diamond blade 11 4244). Sections and some fracture planes were ground with a 5 micron polishing powder (levigates Aluminia Polishing Compound, Buehler) to obtain a level surface for easier detection of prism decussation. All stir faces viewed with the ESEM were etched with HO. (2%) for 2-4 seconds (rime varied with the preservation of the teeth) and cleaned in an ultrasonic bath before analysis in the HSHM. Images were taken at 5 kV and recorded on Polaroid film. Description of enamel—The terms and acronyms used here for describing enamel follow Koenigswald and Sander (1997). Prism shapes refer to the shapes of prisms as viewed in sections perpendicular to their long axes and are de- scribed from tangential sections cut at the outer surface of the teeth. I he angular relationship between prisms and interprismatic enamel was judged at high magnifications from individual crystallites or general crystallite grain orien- tation (Fig. ID). Different types ol' HSB were distinguished depending on their amplitude of waviness in the horizontal plane as viewed in unprepared teeth by light optical means (Stefen 1997a). Undulating I ISB show angles of > 140 , acute angled bands show angles of 140 -70 and zigzag HSB show angles of < 70 at the wave crests and wave troughs. In zigzag HSB additional vertical connections between wave troughs and wave crests respectively are visible. Abbreviations—ESEM, Environmental Scannning Elec- tron Microscope; HSB, Hunter-Schreger band(s); EDJ, Enamel-dentine junction; SDSMN, San Diego State Mu- seum of Natural I listory, San Diego; UCMP, University of -^ Fig. 1. Scanning election micrographs of radial enamel and HSB. A. Protungulatum, UCMP 171 SI 1/12, vertical section of lower molar showing radial enamel. Dentine d Scale bar - 50 urn. B. Eoconodon, UCMP 171513, vertical section (mesio distal) of upper molar showing I ISB on paracolic. Scale bar 50 urn. C. Mioclaenus, U( "M P 36533, vertical fracture plane of upper molar showing poorly developed I ISB. Decussation of prisms occurs only in the middle zone of the enamel. At the EDJ and towards the outer surface radial enamel (rad) is present. Scale bar = 100 urn. D. Protungulatum, UCMP 171511/12, vertical section of lower molar showing, the angle between the crystallites of the interprismatic enamel (Ipe) and the prism (pr) course. As judged from the overall crystallite grain, the angle between them is about 40 45 . Scale bar - 5 Jim. E. Protungulatum, UCMP 171511/12, vertical section in lingual buccal direction of lower molar. Slight deviations of prisms can be observed at metaconid (arrows indicate directions of prisms). Scale bar = 50 urn. P. Protungulatum, UCMP 171511/12, langetial view of a lower molar. Prisms have closed (e) or open (o) prism sheaths. Scale bar 20 um. oe Table 1. Material studied by scanning electron microscopy and the results listed in alphabetical order, irrespective of systematic affiliations. Abbreviations: rad - radial enamel; apr, apiismatic enamel; cl, closed; ent, entoconid; hex, hexagonal; polyg, polygonal; hors, horseshoe-shaped; hyp, hypoconid; Ipe:pr, angle between crystallites of interprismatic enamel and rpsism course; lb, lingual-buccal; met, metacone/conid; par, paracone/conid; pr-sheath, form of prism sheath; prot, protocone/conid; rad, ra- dial enamel; tan, tangential; vert, vertical. Taxon Specimen No. Tooth Section HSB rad apr Ipc:pr seam prismshapc pr-sheath A nisonclms gillianus UCMP94532 1M vert, lb, tan - + 40-45 hors open Baioconodon (= Rajjnarok) UCMP134681 m vert, lb, met & prot, tan - + 40-45 hors open Ccifsioptyclms UCMP89701 P3 vert, md, prot - + 40-45 ChHacus UCMP111955 ml tan, buccal + - 40-45 round-hex cl-opcn Conacodon cophater UCMP89720 ml vert, lb, hyp 8c ent, tan - + 40-45 hors open Conacodon entoconus UCMP68667 in vert, oblique, lingual-mesial - + 40-45 Coiypbodoii UCMP44215 m vert, hor, tan, various fractures ... + polygonal (cl-)opcn Desmatoclaenus UCMP69279 p-frag. vert, lb, hyp & ent - + 40-45 round-hors cl-open Diacodexis UCMP43518 ml vert, lb, par & met, tan ¦•¦ (+) " 40-45 round-hors cl-open Ectocion UCMP39848 M vert, md, prot ~ + 40-45 Ectoconus cf. majuscnlns UCMP89964 Mfiag. vert, various directions - + 40-45 Ectoconus ditrigomts UCMP31313 M vert, various, tan - + 40-45 hors open Eoconodon UCMP171513 Mfiag. md, par & met, tan mesial at prot + - 40-45 round-hors cl-opcn Eoconodon UCMP156100 M2 vert, par + - 40-45 Esthonyx UCMP44420 M vert, md i - 40-45 round closed Hetnithlaeus UCMP89676 in vert, lb, tan - + (+) 40-45 hors open Heptadon UCMP118409 M vert, lb, prot 6k parastyl, + - 40-45 hors (round) open -(cl) Hyopsodus UCMP132801 m lan vert, lb, met, tan _ + 40-45 round-polyg cl-open Hyracotherinm UCMP43596 in vert, lb, par & prot, tan - + 40-45 hors open Hymcotberium UCMP43613 P4 vert, lb oblique + - t- 1 5 I to c Table 1. (cont.). Material studied by scanning electron microscopy and the results listed in alphabetical order, irrespective of systematic affiliations. Taxon Lamdotberimn Lamdotbcrium Litaletes laennatus Litaletes stemberjjeri Loxolopbns byattianus Meniscotberium tapiacetum Mimatuta Mioclaenus turgidus Mioclaenns turgidus Oxyacodon Oxyprimns erickseni Pacbyaena Pantolambda batbmodon Periptycbus carinidens Periptycbus coarctatus Periptycbus Phenacodns intermedins Platychoerops Plesiadapis Plesiadapis cf. rex Pvocerberus Protu ngida tn m Tetraclaeno'don pitercensis Tetraclaenodon pnercensis Specimen No. Tooth Section HSB rad apr Ipe:pr seam prismshape pr-sheath UCMP43101 M?frag. vert, md, met, tan + + - 40-45 - hors irregular open UCMP43601 m3 vert, lb, mesial of par + ~ - UCMP114364 ml tan, buccal 40-45 round-hors cl-open UQMP69283 m vert, lb, distal ofent - + 40-45 - UCMP36632 Ml vert, lb, prot & met, tan - + (+) 40-45 (+) round-hors open UCMP171507 m vert, lb, tan + + - 40-45 (+) hors (cl-)opcn UCMP134682 m3 vert, lb, hyp & ent, tan - + - 40-45 - hors open-(cl) UCMP36533 Ml vert, lb oblique, prot-par, tan, distal at prot + - - 40-45 - round-hex cl-(open) UCMP36624 m2 vert, lb prot & met + + 40-45 UCMP31307 P vert, lb oblique - - 40-45 - UCMP134635 M1 vert, oblique, lb, prot, met, tan - + - 40-45 round cl-open UC.MP43446 in vert, lb oblique - - - 40-45 - round-hex closed UCMP89927 Mfrag. vert - + - 40-45 UCMP171506 P vert - + - 40-45 - UCMP74974 m vert, tan - - - 40-45 hors open UCMP89701 P3 vert, md, prot - + - 40-45 UCMP39870 m & M vert, lb, md, tan + ¦i- (+) 40-45 round-hors cl-open UCMP71438 Mfrag. vert - - (+) 40-45 UCMP61595 I hor + - 40-45 - UCMP114345 m3 vert, oblique - + - 40-45 UCMP171509 mlor2 vert - - + 0-20 UCMP171511/12 m & M vert, hor, tan - + (+) 40-45 - round-hors (cl-)open UCMP36536 ml vert, lb oblique met & + - 40-45 UCMP36538 in prot tan, lingual . round-hors cl-open 20 PALEOHIOS, VOL. 19, NUMBER 3, 1999 California Museum of Paleontology, Berkeley; USNM, United States Museum of Natural History, Washington. RESULTS The observations for all taxa studied arc given in Appen- dix 1 and Table 1. HSB occur in some early Tertiary representatives of the families Procreodi, Condylarthra, Artiodactyla, Perissodactyla, Pantolesta, Pantodonta, Pri- mates, Rodentia, and Tillodontia; no USB were observed in the early Tertian' representatives of the Cimolcsta, Hrinaceomorpha, Ix'ptictida, I.ipotyphla, or Didelphidae examined in this study. These results support previous work summarized by Koenigswald (1997b). Ungulata Protunjjulatum has radial enamel and an additional thin layer of aprismatic enamel at the outer surface, both visible in the basal parts of its molars. The prisms generally run straight from the EDJ following a radial course towards the apex and the outside of the tooth (Fig. 1A). In some areas the prisms are slightly bent (m2, metaconid, and M2, paracolic and mctacone), or are more strongly bent (M2, mesial side of paracolic) close to the base of the enamel. In few areas a very slight prism deviation seems to occur at the metaconid of a lower molar (Pig. IK). The crystallites of the interprismatic enamel are oriented at about 45 to the prism axes. In cross section the prisms are circular to horseshoe-shaped with open or complete prism sheaths (Pig. IP). Procreodi Oxyclaenidae—The oxyclaenid Chriacus Cope 1883 has HSB, whereas 'Ibryptacodon Matthew 1915, Oxyclaenus Cope 1884 and Oxyprimus Van Valen 1978 have radial enamel. In Chriacus prisms show rounded to horseshoe- shaped cross sections with closed and, less often, open prism sheaths. In Oxyprimus the prisms are dominantly rounded with closed prism sheaths, but open ones do occur. Arctocyonidae —Among the arctocyonids only Anacodon Cope 1882 has HSB, whereas Loxolophus Cope 1885 and Baioconodon Gazin 1941 (= Ragnarok\rAn Valen 1978) have radial enamel only. The prism shapes observed in Dcsmatoclacnus Gazin 1941 and Anacodon are rounded to horseshoe-shaped with closed or open prism sheaths respectively. Baioconodon (= Rajjnarok) has dominantly horseshoe shaped prisms with open prism sheaths. HSB may occur in Dcsmatoclacnus, but only a few fragments of Dcsmatoclacnus were available for analysis and the light optical and scanning electron microscopic obser- vations gave contradictory results. Using compound light microscopy HSB were visible on the outside of a premolar fragment, whereas in a vertical section of the same fragment all the prisms appear to be parallel when viewed with the F.SKM. Etched surfaces did not show HSB with light optical means. Thus, Dcsmatoclacnus is considered to have radial enamel, but study of more material may show other- wise. Condylarthra Hyopsodontidae—No HSB were observed in represen- tatives of this family (Hyopsodus Leidy 1870, Hapalctcs Simpson 1935, Haplomylus Matthew 1915, Apbcliscus, Cope 1875, and Litomylus Simpson 1935). As far as the prism shape could be analyzed in Hyopsodus, in cross sec- tion the prisms arc rounded with closed to open prism sheaths. Mioclaenidae—Most of the mioclacnids examined have radial enamel (Litalctes .Simpson 1935, Promioclacnus Trouessart 1904, Protoselenc Matthew 1897, Ellipsodon Scott 1892 and the European species Ortbaspidotbcrium Lemoine 1885 and Plcuraspidothcrinm Lemoine 1878). Only Mioclacnus Cope 1881 has poorly-developed HSB (Pig. 1C). The thickness of the inner and outer zones of radial enamel varies within the oblique Ungual-buccal frac- ture of an upper molar, and thus can vary around the tooth. The HSB in the middle zone of the enamel consist of approximately 5-7 prisms. The interprismatic enamel is oriented at about 45 to the prism axes. The prisms in Mioclacnus are rounded to slightly po- lygonal with complete prism sheaths. Litalctes, however, shows rounded prisms with closed prism sheaths and horse- shoe-shaped prisms with open prism sheaths. Periptychidae—The periptychids studied (Hcmitblacus Cope 1882, Ectoconus Cope 1884, Pcriptycbus (= Carsioptychus Simpson 1936), Conacodon Matthew 1897, Mimatuta Van Valen 1978, Anisoncbus Cope 1881, Haploconus Cope 1882, and Oxyacodon Osborne ik Earle 1895) have radial enamel. Some specimens of Pcriptycbus have a pronounced bending of the prisms in the form of a sigmoidal wave close to the EDJ. In these specimens three- layers arc distinguished in the enamel (Pigs. 2A-D) and clearly visible in vertical -sections (Pigs. 2A, C): (1) at the EDJ the prisms start towards the outside of the tooth ai angles of about 50° to the EDJ; (2) the prisms then bend sharply either perpendicular to the EDJ or slightly inclined towards the base of the tooth; (3) the prisms bend vertically again, not as sharply as before, to continue their course at angles of about 70° to the EDJ towards the outside of the tooth (Pig. 2, Table 2). In vertical sections a prism can often be followed throughout the innermost layers 1 and 2, but in layer 3 a stronger lateral bend occurs in the horizon tal plane. Thus, in layers 1 and 2, prisms bend predomi- nantly in the vertical plane but in layer 3 they change their angle in the horizontal and vertical plane. Horizontal sec- tions also show the subdivision of the enamel into three parts (Figs. 2B, D). Phis pronounced sigmoidal bending of the prisms oc- curs only in some premolar fragments of Pcriptycbus carinidens Cope 1881 (Table 3). The associated molars S'EEFEN-ENAMEL MICROSTRUCTURE IN ARCHAIC UNC.ULATHS 21 Tabic 2. Angles of' bending prisms in a Feriptychns carinidens premolar fragment. The prism measurements are taken ai irregular intervals from the worn tip, which is at the top of the table. MS = number of measurement; a = angle between the EDJ and the rising prism (l'r); c = angle between a projection of the initial course of ihe prism in the innermost enamel and the new course in this layer ol the enamel; a' ~ angle between a line projrected parallel to the EDJ and the prism course in the third layer of the enamel. For detail sec schematic diagram. MS a c a' 1 53 43 67 2 53 45 68 3 50 45 63 4 48 45 68 5 46 57 68 6 52 57 68 _ 51 55 68 8 53 56 i 1 y/?X EDJ show radial enamel with a single (meaning just one change of direction), slight bending of the prisms in some areas. The prisms on the lingual and buccal sides of the lower molars bend slightly inward towards the middle of the talonid basin. Periptychus (= Carsioptychus Simpson 1936) lias few po- lygonal prisms with complete prism sheaths, but have prima- rily horseshoe to irregularly horseshoe-shaped prisms with open prism sheaths (big. 2E). Even in a single tangential section at an enamel ridge the prisms show slightly different patterns of spatial distribution. They are arranged either in nearly horizontal bands, where prisms of the vertically adjacent lines alternate, or in weakly pronounced horizontal lines, with the prisms stacked vertically. Other periptychids, Mimatnta, Hemithlaeus., and Ectoconus, have rounded prisms with closed prism sheaths and horseshoe-shaped prisms with open prism sheaths. In Anisonchus horseshoe-shaped prisms with open prism sheaths dominate. Phenacodontidae—The phenacodoiitids Ictraclaenodon Scott 1892, Phenacodus Cope 1873, Ectocion Cope 1882, and Meniscotherium Cope 1874 have HSB. Results from 'ictraclaenodon are equivocal (see Appendix 1). No HSB were seen with light microscopy, but several specimens viewed with the ESEM showed HSB. Therefore, at least some individuals of 'Ictraclaenodon have HSB. However, future research may reveal a complex picture with older species lacking HSB and younger species displaying HSB. Ectocion has HSB and, towards the apex of the crown, a thin layer of radial enamel apposed at the outer surface. Meniscotherium has HSB (Fig. 3A) in which the angle between prisms of adjacent bands is not as pronounced as in Phenacodus, A variable layer of radial enamel may be apposed towards the outside of the tooth. Phenacodus-has HSB and radial enamel towards the outside, which is also of varying thickness (Fig. 3B). In the upper molar the enamel thickness increases towards the tips of the cusps, which may- be a feature of all teeth. In the talonid basin, especially between hypoconid and entoconid, very little decussation is detectable. Cete Triisodontidae - The triisodontid Eoconodon Matthew & danger 1921 has HSB (Fig. IB). These run from the EDJ to the outside, some are straight and parallel to each other, while others bend slightly along their course and arc not parallel to the neighboring band. Analysis of Eoconodon teeth by light microscopy suggests an increase in waviness in the horizontal course of the HSB towards the apex of the tooth. However, due to the dark enamel of the specimens available the amount of variation is difficult to judge. Mesonychidae—The mesonychids Pachyaena Cope 1874, Harpagalcstcs Wortman 1901, and Mesonyx Cope 1872, have HSB. The waviness of the HSB in the horizon- tal plane in mesonychids varies. There is a transition from undulating HSB at the base of the tooth crown to zigzag HSB in the apex of the tooth crown (see also Stefen 1997a, b). This transition can be seen in Pachyaena and Mesonyx. Only zigzag HSB were observed in the worn teeth of Harpajjokstes. DISCUSSION The goal of this study was to analyze the pattern of occurrence of HSB in "Condylarthra" and to relate these occurrences to body size, chewing mechanisms and other factors. The occurrences of HSB within the "Condylarthra" in a phylogenetic, stratigraphic framework and in relation to body size found here and that of other workers is summarized in bigs. 4 and 5. The observations on the enamel structure of other early Tertiary mammals (Appen- dix 1 and Table 1) are presented for comparison to outgroups. The presence of HSB was difficult to determine for the two condylarths, Dcsmatoclacnnsand 'Ictraclaenodon. HSB 22 PALEOHIOS, VOL. 19, NUMBER .?, 1999 r in Desmatoclaenus were visible using light microscopy, but were not seen with the ESEM. Since the observation of IISB could not be repeated after etching the surface of the Desmatoclaenus material, it is assumed that Desmatoclaenus has radial enamel. However, future studies of more suitable material may change this determination. HSB in Tetraclaenodon were not always visible with light micros- copy but were apparent using scanning electron micros- copy. In other groups ambiguous results also were obtained with light microscopy (see Appendix 1), e.g., Pareumys, Pelycodus, Plesiadapis, Paramys, and Sciuravus. 1 ISB usually appear as light and dark bands with trans- mitted light, but can be obscured by the enamel color and difficult to see. Scanning microscopy allows one to see the course of the prisms directly, allowing a more reliable way to look for prism decussation and 1 ISB. The observations made here of the occurrence of HSB within "Condylarthra" generally support those made by Koenigswald et al. (1987) and Maas and Thewissen (1995). However, some contradictory observations were obtained STEFEN-ENAMEL MICRO-STRUCTURE IN ARCHAIC, UNGULATES 23 Fig. 3. A. Meniscotheriutn, UC'MP 171507, lower molar in oblique vertical section showing 1ISB. Scale bar _ 100 pm. B. Pbenacodus intermedins, UC'MP 39870, lower molar in vertical section showing I ISB and radial enamel towards the outer surface of the tooth. Scale bar = 100 pm. C. Esthonyx, UCMP 44420, vertical fracture plane of an upper molar showing prism decussation. Scale bar = 100 pm. D. Coryphodon, UC'MP 44215, fragment of upper molar in oblique vertical section. The MSB show a very irregular course from the KDJ to the outer surface of the tooth. Scale bar "~ 200 (tin. during the present study and include the following: (1) Loxolopkus Cope 1885 and Oxyclaenus Cope 1884 were found to lack I ISB, whereas Kocnigswald et al. (1987) stated that HSB were present in Loxolophus byattians and Oxyclaenus and (2) HSB could not be found in Thryptacodon Matthew 1915 and Eitafctes, whereas Maas and Thewissen (1995) observed poorly developed decussa- tion in 'Ebryptacodon and well developed decussation in Litaletes (Fig. 4). These ambiguous observations and contradiction,' results point to some difficulties and limitations in enamel research, particularly studies based on fossils. Difficulites include: (1) problems identifying fragmentary material correctly, when frequently this is the only material available for sectioning, (2) resolving differences between light and electron microscopic observations, especially if there is limited material available to confirm a particular observation (e.g., Desmatoclaenus), (3) studies involving groups of mammals in which HSB evolution has taken place, it can be expected that early representatives of a genus might lack HSB while younger representatives have them, and this has been demonstrated for some primates (Maas and O'Leary 1995), and (4) the assumption that enamel microstructure is uniform within a genus is open to question. For example, in Tctaclacnodon a more detailed study of a stratigraphically controlled sequence of samples might demonstrate the evolution of HSB within the genus. 24 PALEOBIOS, VOL 19, NUMBER 3, 1999 m m m prisms with closed prism sheaths prisms with closed and open prism sheaths prisms with open prism sheaths radial enamel poorly developed HSB in the middle zone of the enamel with radial enamel at the EDJ and at the outside well developed HSB well developed HSB with radial enamel towards the outside of the enamel different in previous studies Protungulatum Desmatoclaenus Chriacus Deuterogonodon Loxolophus Anacodon Arctocyon Thryptacodon Haplaletes Haplomylus Litomylus Hyopsodus Oxyprimus Ellipsodon Litaletes Mioclaenus Promioclaenus Protoselene Apheliscus Mimatuta Hemithlaeus Ectoconus Carsioptychus Periptychus Anisonchus Haploconus Oxyacodon Ragnarok Oxyclaenus Eoconodon Dissacus Pachyaena Harpagolestes Mesonyx Cetaceae Tetraclaenodon Phenacodus Ectocion Meniscotherium Paenungulata am am nm m m am m m w m am am mi mi am mi mi mi am m m wrii' SJ? m N m STEVEN-ENAMEL MICROSTRUCTURE IN ARCHAIC UNGULATES 25 Table 3. UCMP material of Periptychusanalyzed by light microscopy, and, where possible by scanning electron microscopy. lb=lingual- buccal, md~mcsio-distal. Spec. no. Loc. no. Taxc Tooth Enamel 31268 V2813 P. rbabdodon: tn3 30016 V3103 P. rhabdodorii ml 30016 V3103 P. rbabdodon: m2 30016 V3103 P. rhabdodorii p3 30016 V3103 P. rhabdodorii C 36572 V6526 P. car hi id em P4 31268 V2813 P. car mid ens m2 36564 V6526 P. carinidens m3 36567 V6526 P. car mid ens ml frag. 171506 V87142 P. carinidens p/P-frag. 171506 V87142 P. carinidens m-frags. 51786 V5707 Periptychus sp. Ml 89701 V2811 P. coarctatns P2 74974 V65396 P. coarctatns frag. 74974 V65396 P. coarctatns m 74974 V65396 P. coarctatns P4 vertical fracture mesial and basal lb; no bending various fractures; no bending mesial fracture; no bending basal, no bending horizontal fracture, lb; no bending vertical fracture, lb md, in mesio-lingual part of tooth; no bending fracture at metaconid and protoconid; no bending vertical fracture, lb mesial: no bending vertical fracture, lb; single bending in talonid basin oblique vertical sections; bending vertical sections, lb: no bending vertical fracture, md; no bending oblique vertical fracture, difficult to judge; slight bending close to the ED J no bending vertical fracture, lb; single bending in talonid oblique vertical fractures; single bending observable at few points Evolution ofHSB Primitive radial enamel is characterized by parallel prisms that run radially out from the ED J and are straight (Koenigswald and Sander 1997), or slightly convex. Maas and Thcwissen (1995) suggest that a combination of radial and decussating enamel could be primitive for ungulates. The cladogram on which they mapped the occurrence of enamel types is based on Prothero ct al. (1988) in which the Arctocyonidae are the basal group and the clade includ- ing Prot/iiijjit/iitum, Hyopsodontidae and Periptychidae is considered more derived. This mapping ot radial and de- cussating enamel requires an evolutionary reversal from decussation to radial enamel in Protunjjulatnm and Hyopsodontidae (Maas and Thcwissen 1995, p. 1161). In a more recent analysis of archaic ungulates Archibald (1998) considers Protunjjulatnm to be at the base of the "Condylarthra" (Fig. 4). On the basis of this proposed phylogeny it can be argued that radial enamel, as seen in Protnnnii/atitm, is primitive for "Condylarthra" and decus- sation evolved from that condition, as had been proposed by Koenigswald ct al. (1987). Another argument for the evolution of decussating enamel from radial enamel is strati- graphic occurrence (Fig. 5): radial enamel is seen in all early Puercan (Pul) forms including Protunjjulatnm, Mimatttta, Baiaconodon (= liajjnarok), and Oxyprimus. These groups were assumed to be ancestral to several "condylarth" fami- lies by Van Valen (1978). HSB first occur in species of the later Puercan (Pu2-3). An evolutionary sequence from radial to poorly devel- oped and then to well-developed decussation has been proposed by Koenigswald et al. (1987). However, Maas and Thcwissen (1995) question this sequence because most species with HSB that they studied show well-developed decussation. Their interpretation is supported by the present study. Poorly-developed HSB were observed only in Mioclaenus. These HSB probably evolved from radial enamel and thus would support the proposed evolutionary sequence. As no geologically younger representatives of this clade are known, the possible further evolutionary sequence to well-developed HSB cannot be tested here. The HSB of the late Puercan species of Eoconodon, how- ever, are well-developed and these data do not support the evolutionary sequence proposed by Koenigswald et al. (1987) unless one assumes a yet unknown ancestor with poorly-developed HSB. Thus, the current observations do not support a clear evolutionary sequence from radial enamel to poorly-developed decussation and then to well- developed decussation. It seems more likely that both, poorly-developed HSB and well-developed HSB originated -^ Fig. 4. The occurrences of HSB and prism shape are mapped on the cladogram of "Condylarthra" and selected other rlicrians as proposed by Archibald (1998). The data for Anacodon, Arctocyon, and Dissacusarc from Stefen, 1997b. The different observations in previous studies marked with * are based on Koenigswald et al. (1987) and Maas and Thcwissen (1995) and are discussed in the text. 26 PALEOBIOS, VOL. 19, NUMBER 3, 1999 09 tx: Z3 R^gnarok X LX rx c 7\/\/\/xx X IX. r 5! Protungulatum \ \ \ \_X \ \ \ \~x ¦S S S \ S ES_7S SZX SI Oxyclaerus i\ /' Eoconodon \ \ \ \ S. \ \ \ \ \ Cxyprimus di \ m Desmatoclaenus \ m Prothryptacodon /\/\/\/\/\/\/ \Z\/\/\/\/\/\/\/\ n Loxolophus K~y\ /\ 7\ /\ /\ /\ /W\ /\ /\ A A A A^2 A vtccyon \ \ \ \ \ \ \ 1A /\ /N /\ /\ AAA/N /\ 7\ \ \ \ \ X~^1 Hap/a fetes \ \ \ \ \ VI Litomylus Litaletes E,/\/\/\zx Mioclaenus \\\\\\\\\l Promioclaenus \ \ \ \ Zl Mimatuta ? a E P Hemithlaeus Ectoconus Carsioptychus -x—"s—\ v \ \ \ \~\~1 Periptychus \ s \\ s —:.----^----:.----:.---- Conacodon Oxyacodon /\ /\ /\ /X s zx \ \ \ XH Protoselbne _J Anisonchus Haploconus XI Tetraclaenodon \ A A A A A A \ /\ /\ /\ /\ /V7\ o Q) \ \ \ \ \ \ \~^XX /\ /\ /\ /X7\l Dissacu$ X z\ A A A A A A A A \\\\\\\\ \\\\\\\\ Chriacus Anacodon Thryptacodon 2 Pachyaena 3 Mesonyx Haplomylus ] Hyopsodus K \ I radial enamel r^^i hsb x\ x\ /\ x\ /\ /\ /\ A . /\ /\. /\. AAAAAAA/n ] Phenacodus ] Ectocion ] Meniscotherium Fig. 5. Occurrence or IISB mapped on the stratigraphic ranges of the genera analysed. Range data from Archibald (1998). SIEFEN-V.NAMFA. MICROS TRUCTURK IN ARCHAIC UNGULAThS 27 directly, and independently from radial enamel as Mass and Thevvissen (1995, p. 1161) argued. HSB are thought to have originated independently in several different lineages of mammals (Koenigswald and Clemens 1992, Maas and O'Lcary 1995; Stefen, 1997a). Analyzing the occurrence of HSB within "Condylarthra" (Procrcodi, Condylarthra, and CctC specifically; Fig. 5) there are IISB in Arctocyonidae and Oxyclaenidac, no HSB in Hyopsodontidae, HSB in Mioclaenidae, no I ISB in Peripivchidae, and HSB in Triisodontidae, Mesonychidae, and Phcnacodontidac. | If"the contradictory observations of Oxyacodon Osborne & Earle 1895 by Koenigswald et al. (1987) are accepted this would imply the presence of HSB even in the Periptychidae. Accepting presence of HSB in Loxolophus (Koenigswald et al. 1987), Litaletes, and Tbryptacodon (Maas and Thewissen 1995), which were not duplicated in this study, would not change the overall picture but add more species with HSB to the respective families. ] This pattern of occurrence indicates that HSB evolved several times within "Condylarthra" (Procreodi, Conylarthra, and Cete sensn McKcnna and Bell [ 1998]). Parallel evolution of enamel structures has been observed at different levels. Prisms evolved in several mammalian lineages and one reptilian genus, Urmnnstyx. The direction of the crystallites of the interprismatic enamel relative to the axes of prisms probably changed several times from parallel to the prisms to angled (Koenigswald 1997a). Also HSB types evolved in similar ways in different lineages (e.g., zigzag I ISB evolved in members of the Carnivora, Creodonta, Artiodacyla (Entelodontidae) and Cetaceans in the broadest sense | Stefen 1997a, b]). Koenigswald (1997a, p. 229) pointed out that some steps in the evolution of enamel structures may best be described by underlying or incomplete synapomorphies (Sxther 1979), rather than parallel evolution. Underlying synapomorphies are defined as "close parallelism as a result of common inherited genetic factors" (Ssether 1983, p. 344) where, "The apomorphic character alternative is not present in the whole group, not necessarily even in the most plesiomorphic taxa of the group" (Sxther 1983, p. 374). Following this argument, the developmental mechanism producing I ISB can be assumed to be a synapomorphy of the Ungulata that is either not expressed in all taxa or shows early reversals. However, the assumption of an underlying synapomorphy makes phylogenctie analysis difficult, because it is untestable whether taxa share an underlying synapomorphy that is not expressed or reversed or whether the trait is not there. The data presented here support parallel evolution of I ISB in several lineages. This may or may not be the result of an underlying synapomorphy. The position of Artiodactyla in relation to "Condylarthra" was not clearly resolved by Archibald (1998); artiodactyls are either at the base of the cladogram and primitive relative to ProPuvgulatum, or branch off the lineage leading to mesonychids and ultimately to cetaceans. The relationship between mesonychids, cetaceans, and ar- tiodactyls is currently debated (Milinkovitch and Thevvissen 1997, Thewissen 1998), with no clear picture emerging. The earliest representatives of the Artiodactyla examined here by light microscopy, Dincodcxis and Hcxacodus, have HSB which supports observations by Koenigswald and Pfretzschner (1987) and Maas and Thevvissen (1995). No extant artiodactyl is known to lack HSB (Koenigswald 1997b). Considering the various suggested phylogenctie positions of Artiodactyla (Archibald 1998), the evolution of HSB within Artiodactyla seems to be convergent to the evolution of HSB in the other groups of "condylarthra." Convergent evolution of HSB is also supported by their occurrence in other, non closely related mammalian groups such as Rodentia, Primates, Tillodontia, Pantolesta, and Pantodonta. HSB and Cetacean origins The mesonychids Dissncits, Pachyctena, Harjmjjolcstcs, and Mesonyxshow a transition from undulating I ISB in the lower part of the tooth to zigzag HSB in the upper part of the crown (Stefen 1997b). Eoconodon, the stratigraphically oldest species with HSB, seems to show an increase of waviness in the HSB towards the apex of the crown, but as the enamel of all available specimens is so dark these observations cannot really be confirmed. It is possible that Y.oconodon has some zigzag HSB in the tips of the teeth. These observations are interesting with respect to the origin of whales, whose phylogenetic relationships are still debated. Based on morphological evidence whales are con- sidered to be closely related to mesonychids and, in particu- lar, forms close to Pachyacna (Gingerich et al. 1994). Molecular data support the proposal of a sister group relationship with Artiodactyla (Milinkovitch and Thewissen 1997). Maas and Thewissen (1995) find vvell-develeoped I ISB in Pakicetus, the earliest archaeocete whale. Sahni and Koenigswald (1997, p. 189) recently documented well developed HSB which are "...transversely oriented and show a slight undulation..." in other archaeocete whales. Neither reported the presence of zigzag I ISB. As Paciryacna and other mesonychids show the more derived zigzag I ISB it seems unlikely on the basis of enamel structure that these mesonychids arc directly ancestral to archaeocete whales lacking zigzag HSB. Assuming a close relationship between mesonychids and whales these data suggest that both groups diverged from a common ancestor with undulating HSB. However, another possibility is loss of zigzag HSB very early in whale phylogeny, with further evolutionary reduc- tion of other enamel characters leading to younger tooth forms having aprismatic enamel (Ishiyama, 1987, Sahni and Koenigswald 1997). HSB and body size In order to compare the results of this study with those of Koenigswald et al. (1987) and Maas and O'Leary (1995), 28 PALEOBIOS, VOL 19, NUMBER 3, 1999 HSB presonl <\ i ? ? ? ? • • 12 3 4 J r. 7 8 9 IP 11 12 13 • :s . i. a • • •• • ? ? HSB absent A ConQylarthra It'll! ml length others HSB present 1-2 3 4 5 Condylanhra 9 10 11 12 13 mrn •"••V» »i« • ml width Condylanhra B Fig. 6. Occurrence of HSB in some "condylarths" and other animals in relation to the length (A) and width (B) of ml as an indicator of body size. Taxa and measurements are given in Appendix 2. the length and width measurements of ml and m.2 are used as indicators of body size. The proposed positive correlation between increasing body size and HSB occurrence (Koenigswald ct al. 1987) is supported for "Condylanhra" in general (Fig. 6, Appendix 2). IISB seem to occur with m 1 lengths of more than about 6 mm and widths of more than 4.5 mm. However, as Koenigswald et al. (1987) noted and reiterated by Maas and Thewisscn (1995), exceptions occur. For example, the periptychids Ectoamus and Periptyclms with mis longer than 6 mm have no HSB. Including observations on repre- sentatives of other lineages points to exceptions to the size- increasc/HSB-occurrencc "rule" at both ends of the spectrum of body sizes. The relatively small artiodactyl Diacodexis has HSB, while the relatively large taeniodont Onychodectes does not. In younger and larger members of the Taeniodonta the enamel is completely lost. Similarly, whales reduce the thickness of their enamel and lose HSB as they become larger. In some fossil archaeocetes well- developed HSB were described (Sahni 1981, Maas and Thewisscn 1995), whereas some extant odontocetcs have a combination of radial and aprismatic enamel, still others have completely aprismatic enamel (Ishiyama 1987, Sahni and Koenigswald 1997). The small rodents Paramys, Sciuravus, and Tapomys at the other end of the size spec- trum have MSB. In functional terms, IISB have been interpreted as an adaptation to prevent stress induced crack formation and crack propagation when the enamel is loaded during mas- tication (e.g., Pfretzschncr 1987, Koenigswald and Pfretzschner 1991). Taxa using high chewing forces, such as bone crushing hyaenas, show zigzag HSB. This arrange ment is considered to be best suited for withstanding the high tensions (Rcnsberger 1995, Stefen 1997, Stefen and Rensberger in press). Thus, it can be expected that a change in chewing mechanism would lead to a change in tonsil strength of the enamel and would induce an evolutionary change on the enamel structure. HSB and chewing mechanisms The appearance of HSB in early Tertian' euthcrian mam mals coincides with a radiation of basic molar types in mammals. Primitive tribosphenic molars are adapted for shearing and puncture-crushing (Crompton and Kielan Jaworowska 1978) and their chewing motion was basically characterized by two phases. First, an uplift of the lower jaw leading to a tooth-food-tooth contact resulting in puncture-crushing. Second, a translational movement of the jaws with the teeth occluded leading to tooth-tooth contact and shearing (Crompton and Hiiemae 1970, Kay and Hiiemae 1974). During the Cretaceous the basic tribosphenic pattern of occlusion changed verv little but slight changes did occur in some taxa (Crompton and Kielan-Jaworowska 1978). In Cimolestcs Marsh 1889 verti- cal shear was accentuated and puncture-crushing was im- portant, but the teeth were not well adapted for crushing. The occlusal pattern of Procerberns Sloan & Van Valen 1965 is assumed to be similar (Crompton and Kielan Jaworowska 1978), however, Rensberger (1986) points to compression, translative action and shearing in Procerberns. New wear facets were added in Gypsonictops Simpson 1927 (Kay and Hiimae 1974, Crompton and Kielan Jaworowska 1978), and a power stroke which moved the lower jaw downwards and slightly antero-medially resulting in the floor of the talonid being drawn across the trigon leading to grinding. The newly acquired wear facets are poorly devel- oped in Gypsonictops but in the Puercan plesiadapid Purcfatorius Sloan & Van Valen 1965 and later in the primate Pclycodus (= Cantins) they are more pronounced. In Protttnjjitlatiim the shear angle was reduced, but other- wise its dentition differs little from the insectivore pattern (Rensberger 1986). 'Zhelestids,' a basal group of ungulates known from western Asia and North America, retain several more primi- tive characters as compared to archaic ungulates from North America. Most importantly in 'zhelestids,' the mandibular condyle is lower than the occlusal level of the teeth, whereas in archaic ungulates the condyle has moved to a position above the occlusal level of the teeth (Nessov et al. 1998). A more dorsal location of the mandibular condyle is associ- ated with herbivory and the increased importance of the masseter and pterygoid muscles. This indicates a change in the overall chewing pattern within archaic ungulates com pared to the more primitive 'zhelestids.' Nessov et al. STEFEN-V.KAMV.L MICROSTRUCTURE IN ARCHAIC; UNGULATES 29 (1998) also point to some apomorphies of ungulates, more specifically the Ungulatomorpha, comprising 'zhelestids' and ungulates, which indicate divergence from the basic tribosphenic molar pattern. These are: (1) expansion of the talonid, (2) lowering of trigonid height in relation to the talonid, and (3) appression of the trigonid cusps. These features may all be interpreted as the first morphological shifts toward a different tooth use, towards herbivory. Thus, even the most primitive archaic ungulates were de- rived relative to the tribsophenic chewing pattern. At first glance it appears that evolution of HSB occurred in concert with changes in the basic pattern of tribosphenic occlusion and chewing. For example, insectivores do not change their pattern of chewing much and retain radial enamel, while rodents change their chewing mechanism to one dominated by translational movement and evolve HSB. However, analysis of the results of broader studies reveals that the picture is not that simple. Periptychids and phenacodontids show adaptations for compression-domi- nated chewing motions (Rensberger 1986). Although both groups show a tendency to develop a similar loading re- gime, the enamel structure is different: periptychids retain radial enamel, phenacodontids evolve HSB. Microstructure Shape of prisms—In tangential sections at the outer surfaces of their teeth most "Condylarthra" show circular to horseshoe-shaped prisms with closed to open prism sheaths. For Protnnjjulatum this supports the observations by Sahni (1979) and the observations of Maas and Thewissen (1995) for other "Condylarths." Prisms running radially outwards from the KDJ with a strong inclination tow aids the tip of the tooth appear to be more likely to give an image of open prism sheaths in tangential sections, because of the angle at which they are viewed (Dumont 1995). For HaplomylusMaas and Thewissen (1995) report a unique enamel that consists entirely of closed prisms. Microstructure of periptychid enamel—The pro- nounced bending in the form of a sigmoidal curve of prisms in the inner part of the enamel in periptychids (Fig. 2) could only be observed clearly in fractured enamel of Periptychus carinidens from the Torrejonian. A premolar from the Puercan Periptychus (= Carsioptychus) showed a slight bending of the prisms, while molars of different specimens (Table 3) showed no or only one bending on their course from the EDJ to the outside as compared to the two seen in Fig. 2. There could be two possible explanations for these observations. First, the pronounced bending could be a newly emerging evolutionary trend in the Torrejonian species. As such, the bending might be related to the strong grooving of the surface of the teeth, which first occurred in the premolars (Clemens 1998, personal communication). This hypothesis is supported by the occurrence of just one single bending of the prisms in the molars that develop the strongly grooved tooth surface later and also by the slight bending in early specimens of Periptychus. A relation between the enamel microstructure and the grooving of the tooth surface, and the functional significance of the grooving itself, remains to be clarified. A biomechanical advantage of the grooved surface is not known. The grooving might form a greater depositions! surface during tooth formation and thus make it possible to deposit the enamel faster. Interpreted biomechanically, the bending of prisms might deflect cracks and thus reduce their energy and could make the enamel more resistant to tensions than radial enamel with straight prisms. The other possibility is that the pronounced bending is a peculiarity of this particular population of Periptychus from this one site. The sharp bending of all prisms in parallel is not very common. A simultaneous change in the direction of parallel prisms (i.e., simultaneous prism deviation) has been de- scribed for several marsupials, eutherians in the Cricetinae and Arvicolinae (Rodentia) (Koenigswald 1994), and for ptilodontid multituberculates (Sahni 1979). In simulta- neous prism deviations the prisms usually change their direction from "...a radial and steeply rising direction into a horizontal and lateral direction..." (Koenigswald 1997a, p. 213). In the periptychid sample the prisms bend primarily in the vertical plane, changing from a steeply rising direc- tion into the horizontal, with only little change in the lateral direction from layer 2 to layer 3. A characteristic bending of prisms has also been described in wombats, where they bend from tangential enamel to FISB (Ferrcira et al. 1990). CONCLUDING REMARKS Mammals coexisted with dinosaurs during the latter part of the Cretaceous (Kumar and Heges 1998). Only after the extinction of the nonavian dinosaurs were different mam malian groups able to undergo a marked radiation, use different food sources, occupy different ecological niches, and 'experiment' with body size. Complex changes of different body parameters probably occurred at the same time and their causal links are not obvious from the fossil remains. With an increase in body size, food items that were inaccessible before because of their large size could have become available later. At the same time, a change in diet probably induced an evolutionary change in tooth use, biting, and loading of the teeth as well as accompanying changes in tooth morphology. A change in overall ecology and habitat probably occurred with a change in diet. There- fore, correlating HSB with one factor is very difficult. A positive correlation of the occurrence of HSB with increas- ing body size and with changing chewing mechanisms can be noted, but no single factor can be identified as the only cause for changes in enamel structure. This study and others indicate HSB evolved several times within different lineages and that the evolutionary sequence leading from radial enamel to well developed HSB is not yet clearly documented. 30 PA LEO BIOS, VOL. 19, NUMBER 3, 1999 Similar developmental constraints or underlying synapomorphies within eutherian mammals may have facili- tated the evolution of HSB in ungulates and other orders. However, the Periptychidae with specialized bending show that even with an increase in body size and deviation in chewing mechanisms from the tribosphenic pattern, radial enamel may be retained and acquire differentiated features. This and the near lack of IISB in marsupials indicate that other responses to changing loading conditions of the teeth and in the enamel were possible. We still lack detailed information on the enamel micro- structure of many early Tertiary mammals. It would be necessary to assess the variability within the dentition and within a species or genus to accurately address the evolu- tion of I ISB in different lineages. Especially, stratigraphically-controlled sequences of specimens of early taxa that evolved HSB, such as Eoconodon and 'l'ctraclaenodon, could shed more light on the rate and mode of HSB evolution. Also results show caution has to be used when characters of tooth enamel are included in phylogenetic analyses because they are related to function and convergent evolution does occur. ACKNOWLEDGEMENTS I wish to thank W. A. Clemens for the opportunity to work in his lab and his help with the English; the Depart- ment of Integrative Biology and the Museum of Paleontol- ogy, University of California, Berkeley for their hospitality and use of material and resources. I also want to thank Dr. R. J. Emry (USNM) and Dr. H. M. Wagner (SDSMN) for the opportunity to study specimens in their museums. This work was made possible through a scholarship by the Deutsche Forschungsgemeinschaft (grant ST 798/1-1), whose support is gratefully acknowledged. Observations in the USNM were made during a previous visit to the USA which was supported by the Deutsche Akademische Austauschdienst. This is UCMP Contribution No. 1687. LITERATURE CITED Archibald, J.D. 1982. A study of Mammalia and geology across the Cretaceous-Tertiary boundary in Garfield County, Mon tana. University of California Publications in Geological Sci- ences 122:1-288. Archibald, J.D. 1998. Archaic ungulates ("Condylarthra"). pp. 292-329 in CM. Janis, K.M. Scott, and 1..L. Jacobs (eds.). Evolution of Tertiary Mammals ot North America Vol. 1: Terrestrial Carnivores, Ungulates and Ungulatelike Mammals. Cambridge Univ. Press, Cambridge. Boyde, A. 1976. Enamel structure and cavity margins. Operative Dentistry 1:13-28. Carlson, S.J. 1990. Vertebrate dental structures, pp. S31-56 in S.G. Carter (ed.). Skeletal Biomineralization: Patterns, Pro- cesses and Evolutionary Trends I. Van Nostrand Reinhold, New York. Crompton, A.W., and K. Hiiemae. 1970. Molar occlusion and mandibular movements during occlusion in the American opossum, Didclphis marsupialis I,. Zoological Journal of the Linnean Society 49( 1 ):21-47. Crompton, A.W., and Z. Kiclan Jaworowska. 1978. Molar struc turc and occlusion in Cretaceous therian mammals, pp. 249- 87 in P.M. Butler and K.A. Joysey (eds.). Development, Function and Evolution of Teeth. Academic Press, London. Dumont, K.R. 1995. The effects of sectioning angle on measure ments of enamel prisms size and spacing. Archives of Oral Biology 40:956- 66. Ferreira, J.M., P.P. Phakey, J. Palamara, and VV.A. Rachinger. 1990. Electron microscopic investigation relating the occlusal morphology to the underlying enamel structrure of molar teeth of the wombat ( Vombatns ursinus). Journal of Morphol- ogy 200:141-49. Fortelius, M. 1985. Ungulate cheek teeth: Developmental, func tional , and evolutionary interrelations. Acta Zoologica I'ennica 180:1-76. Gingerich, P.D., S.M. Raza, M. Arif, M. Anwar, and X. Zhou. 1994. New whale from the Eocene of Pakistan and the origin of cetacean swimming. Nature 368(6474):844-47. Ishiyaina, M. 1987. Enamel structure in odontocete whales. Scanning Microscopy 1:1071-79. Kay, R.F., and K.M. Hiiemae. 1974. Jaw movements and tooth use in recent and fossil primates. American Journal of Physical Anthropology 40(2):227-56. Koenigswald, W. von, and W.A. Clemens. 1992. Levels of com- plexity in the microstrucrure of mammalian enamel and their application in studies of systematics. Scanning Microscopy 6(1): 195-218. Koenigswald, W. von, and P.M. Sander, 1997. Glossary' of terms used for enamel microstructure. pp. 267 80 in VV. von Koenigswald and P.M. Sander (eds.). Tooth Enamel Micro- structure. A. A. Balkema, Rotterdam. Koenigswald, W. von, J.M. Rensberger, and H.U. Pfrctzschner. 1987. Changes in the tooth enamel of early Paleocene mammals allowing increased diet diversity. Nature 328(6126): 150-52. Koenigswald, VV. von, and H.U. Pfrctzschner. 1987. Hunter- Schreger-Bander im Zahnschmelz von Saugetieren: Anordnung und Prismenverlauf. '/.oomorphology 106:326 38. Koenigswald, W. von, and H.U. Pfretzschner. 1991. Biomechan ics in the enamel of mammalian teeth, pp. 113-25 in N. Schmidt-Kittler and K. Vogel (eds.). Constructional Mor- phology and Biomechanics. Springer, Berlin/I leidelberg. Koenigswald, W. von. 1994, Differenziertingen im Zahnschmelz der Marsupialia im Vergleich zu den Verlialrnissen bei den Placentalia (Mammalia). Berliner gcowisscnschaftlichc Abhandlnngcn E13:45-81. Koenigswald, VV. von. 1997a. Evolutionary trends in the differen- tiation of mammalian enamel ultrastructurc. pp. 203 35 »'»W. von Koenigswald and P.M. Sander (eds.). Tooth Enamel Microstructure. A. A. Balkema, Rotterdam. Koenigswald, VV. von. 1997b. Briel survey of enamel diversiy at the schmclzmuster level, pp. 137-61 inW. von Koenigswald and P.M. Sander (eds.). Tooth Enamel Microstructure. A. A. STEFEN-ENAMEL MICROSTRUCTURK IN ARCHAIC UNGULATES 51 B.ilkcma, Rotterdam. Kumar, S. and S.B. Hegcs, 1998. A molecular timescale fro vertebrate evolution. Nature 392:917-920. Maas, M.C., and M. O'Leary. 1995. Evolution of molar enamel microstructure in North American Notharctidac (primates). Journal of Human Evolution 31:293-310. Maas, M.C., and J.G.M. Thewissen. 1995. Enamel microstruc- ture of 1'akicctus (Mammalia: Archaeocetae). Journal qfPttle- ontology 69(6):114l-63. McKenna, M., and S.K. Bell. 1998. Classification of mammals above the species level. Columbia University Press, New York. Milinkovitch, MX',., and J.G.M. Thewissen. 1997. Even-toed fingerprints on whale ancestry. Nature 388(6643):622-23. Nessov, L.A., J.D. Archibald, and X. Kielan-Jaworowska. 1998. Ungulate-like mammals from the Late Cretaceous of Uzbekistan and a phylogenetic analysis of Ungulatomorpha. Bulletin of the Carnegie Museum ofNatural History 34:40 88. Pfretzschner, H.-U. 1986. Structural reinforcement and crack propagation in enamel, pp. 133-43 in D.E. Russel, J.-P. Santoro and D. Sigogncau-Russell (eds.). Proceedings of the Vllth International Symposium on dental morphology. Paris 1986. Memoires Museum of Natural History Paris Series C 53. l'rothero, D.R. 1993. Ungulate phylogeny: Molecular vs. mor- phological evidence, pp. 173-81 in F.S. Szalay, M.J. Novacek and M.C. McKenna (eds.). Mammal Phylogeny—Placentals. Springer Vcrlag, New- York. Prothero, D.R., P.M. Manning, and M. Fischer, 1988. The phylogeny of the ungulates, pp. 201-34 in M.J. Benton (ed.). The Phylogeny and Classification of Tcrrapods, Vol.11. Syt. Assoc. Special Vol. 35B. Rensberger, J.M. 1986. Early mechanisms in mammalian herbi- vores. Paleobiology 12(4):474-94. Rensberger, J.M. 1995. Determination of stresses in mammalian dental enamel and their relevance to the interpretation of feeding behavior in extinct taxa. pp. 151-72 in J. Thomason (ed.). Functional Morphology in Vertebrate Paleontology. Cambridge University Press, Cambridge. Sanhcr, O.A. 1979. Underlying synapomorphies and anagenic analysis. Zoologica Scripta 8:305-12. S.vthcr, O.A. 1983. The canalized evolutionary potential: Incon- sistencies in phylogenetic reasoning. Systematic Zoology 32:343-59. Sahni, A. 1979. Enamel structure in certain North American Cretaceous mammals. Paleontographica A 166:37-49. Sahni, A. 1981. Enamel ultrastructure of fossil Mammalia: Eocene Archaeoceti from Kutch. Journal of the Pa/eon tological Society India 25:33-37. Sahni, A., und W. von Koenigswald. 1997. The enamel structure of some fossil and recent whales from the Indian subconti- nent, pp. 177-91 in W. von Koenigswald and P.M. Sander (eds.). Tooth Enamel Microstructure. A. A. Balkema. Stefen, C 1997a. Differentiations in Hunter-Schreger bands of carnivores, pp. 123-36 in W. von Koenigswald and P.M. Sander (eds.). Tooth Enamel Microstructure. A. A. Balkema. Stefen, C. 1997b. The enamel of Creodonta, Arctocyonidae and Mesonychidae (Mammalia) with special reference to the ap- pearance of Hunter-Schreger bands. Pnlaontologiscbe Zcitscbrift 71(3/4):291-303. Stefen, C, and J.M. Rensberger. In press. The specialized struc- ture of hyaenid enamel: Description and development within the lineage- including percrocutids. Scanning Microscopy. Thewissen, J.G.M. 1990. Evolution of Paleocene and Eocene Phenacodontidae (Mammalia, Condylarthra). University of Michigan Papers on Paleontology 29:1-107. Thewissen, J.G.M. 1998. Whale ankles and evolutionary relation ships. Nature 395:452. Uhen, M.D., and P.D. Gingerich, 1995. Evolution of Coryphodon (Mammalia, Pantodonta) in the late Paleocene and early Eocene of Northwestern Wyoming. Contributions from the Museum of Paleontology, Univ. Michigan 29:259-89. Van Valcn, L. 1978. The beginning of the age of mammals. Evolutionary Theory 4:45-80. Wood, C.B., E.R. Dumont, and A.W. Crompton, (1999). A review of enamel microstructure in non-therian mammals: Implications for enamel prisms as a synapomorphy of mammalia. Journal of Mammalian Evolution 6:177-214. Wood, G.B., and D.N. Stern, 1997. The earliest prisms in mammalian and reptilian enamel, pp. 63-83 in W. von Koenigswald and P.M. Sander (eds.). Tooth Enamel Micro- structure. A. A. Balkema. 32 PALEOBIOS, VOL. 19, NUMBER 3, 1999 Appendix 1. Material used Tor light optical analysis and presence or absence of HSB. Taxa are listed alphabetically within the different orders. Classification follows McKcnna and Bell (1998). Taxon PROCREODI Aiincodmi Cope, 1XS2 Haioconodon Ga/in, 1941(~ Rnjjnarob) Chriacus Cope, 1883 Chriaaisgallinaci Matthew, 1915 Clacnodon cf. procyonoides Matthew, 1937 Dcsmatoclaoius hermaeus Gazin, 1941 Desmatoclaenusi Gazin, 1941 Loxolophus Cope, 1885 Loxolophiis {--- Protqgonodon) hoxolophus hyattianus Cope, 1885 Loxolophus prisms (. 'ope, 1888 Oxyclaenus cuspidatus Cope, 1884 Oxyclaenus simplex (Cope, 1884) Oxyprimns erikseni Van Valen, 1978 Oxyprimus Van Valen, 1978 Thryptacodon sp. Matthew, 1915 Thryptacodon, cf. CONDYLARTH RA Apheliscus insidiosus Cope, 1874 Anisonchus Cope, 1881 A nisoncbus jiillianns ((lope, 1882) Conacodon Matthew, 1897 Conacodon cophater (Cope, 1884) Conacodon entoconus Cope, 1882 Ectocion Cope, 1882 Ectocion osborneanus Cope, 1882 Ectocion ralstoncnsi Gazin, 1956 Ectoconus Cope, 1884 Ectoconus cf. majusculus Matthew, 1937 Ectoconus ditrigonus Cope, 1882 EUipsodoni sternbergeri Eoconodon Matthew & Granger, 1921 Haplaletes Simpson, 1935 Haplocoiins ailfllisrns Cope, 1880 Haploconus speirianus Hapbmylus Matthew, 1915 Haphmylus speirianus Cope, 1880 Hemithlaeus kowalevskianus Cope, 1882 Hemithlaeus Cope, 1882 Hemithlaeus apienlatus Hyopsodits I .cidy, 1870 Hvopsodns loomisi Mekenna, 1960 Hyopsodus miticulus Cope, 1874 Litaletes Simpson, 1935 Litnleies sternbergeri (Gazin, 1935) Litomylus scapbiscus Gazin, 1956 Meniscotherium Cope, 1874 Family Collection no. Arctocyonidac UCMP 137364 Arctocyonidae UCMP 116537, 116538, 116540, 116542, 116543, 134681 Oxyclacnidae UCMP 111954, 111955,36516,44269, 11955,30005, 1185 09 Oxyclaenidac UCMP 46639 Arctocyonidae UCMP 114412 Arctocyonidac UCMP 69279 Arctocyonidae UCMP 69293 Arctocyonidac UCMP 36632, 31285, 69252 Arctocyonidae UCMP 113130, 69285, 69286, 69287 Arctocyonidae UCMP 36632 Arctocyonidae UCMP 124411 Oxyclacnidae UCMP 63755 Oxyclacnidae UCMP 36479 Oxyclacnidae UCMP 134635,933576 Oxyclacnidae UCMP 133575, 133576, 133577, 133578, 133851. 134636, 134638 Oxyclaenidac UCMP 118508, 114435, 111913, 111923 Oxyclacnidae UCMP 68868 1 Ivopsodonlidae UCMP 46493, 46494 Periptychidae UCMP 89700 Periptychidae UCMP 89699, 94532, 124408 Periptychidae UCMP 68667 Periptychidae UCMP 30025 36502, 36618, 36619, 89689, 89968, 89720 Periptychidae UCMP 124417, 31308, 36497, 36580, 68667, 89934 Phcnacodontidae UCMP 39848 Phcnacodontidae UCMP 39783 Phenacodontidac UCMP 53943 , 53944, 53991 Periptychidae UCMP 36543 Periptychidae UCMP 89964 Periptychidae UCMP 30021, 31313, 36543, 89964 Mioclaenidac UCMP 47243, 47244 Triisodontidac UCMP 156100 1 [yopsodontidae UCMP 72095, 72096, 74158 Periptychidae UCMP 36614 Periptychidae UCMP 44134 [{yopsodontidae UCMP 43543 1 [yopsodontidae UCMP 44132 Periptychidae UCMP 89676 Periptychidae UCMP 89676 Periptychidae UCMP 124415 1 [yopsodontidae UCMP 132801,43543,43676 1 [yopsodontidae UCMP 44110, 44142 I [yopsodontidae UCMP 44137, 44138, 60346 Mioclaenidac UCMP 69283, 124418 Mioclaenidac UCMP 69283 H yopsod< mtidae UCMP 111918, 111925 Phcnacodontidae UCMP 171507 HSB i'i STEFEN-ENAMEL MICROSTRUCTURE IN ARCHAIC UNGULATES 33 Appendix 1. (coin.) Material used for light optical analysis and presence or absence of HSB. Taxa are listed alphabetically within the different orders. Classification follows McKcnna and Bell (1998). Mimatuta Van Valcn, 1978 Mioelaenus turgidus (.'ope, 1 cS81 Orthaspidolherium Lemoinc, 1885 Oxyacodon Osborne & Earle, 1895 Oxyacodon agapetiilus (Cope, 1884) Oxyacodon apienlains Osborne & Earlc, 1895 Periptychus carinidtns (lope, 1881 Periptychus eoarctatus Cope, 1883 Periptychus Phcnacodiis Cope, 1873 Phenacodas cf. primaevus Cope, 1873 Phenacodas intermedins Granger, 1915 Ptcuraspidotberitim aiistralis Promieclaenus acolytus (Cope, 1882) Protoseltne Matthew, 1S97 (= Dracoclaenus Gazin, 1939) Protostleni opisthaens (Cope, 1882) Protiinjjiilatiim Sloan & Van Valcn, 1965 Tttrttclaenodon Scott, 1892 Tttraclaenodon Tetraciaenodon puercensis (Cope, 1881) 1 'etraclaenodon pnercensis CETE Dissacus nnvnjorins C lope, 1881 Harpajjolestes, cf. Wormian, 1901 Metonyx obi widens Cope, 1872 Pachyaentt Cope, 1874 ARTIODACTYLA Diacodexis Cope, 1882 Diacodexis robustus Sinclair, 1914 Ilexneodns prlodrs Gazin, 1952 CIMOLESTA Apatcwys cl. //i7/m.< Marsh, 1872 Didelphodits absarokae Cope, 1881 ER1NACEOMORPHA Aiinpisorexjinndrvi Lemoinc, 1883 Adapisorex gaudryi Crypholtstts vannbni(Novacek, 1973) Litolestes lacunatus Gazin, 1956 Seenopaflnscf. prisctts(Marsh, 1872) SORICOMORPHA Centetodon cf. aztecus Lillegravcn ct al., 1981 I.KPTICTIDA Gypsouictops bypoconus Palaeictops Matthew, 1899 Family Collection no. IVriptvchidac UCMP 132089, 134682 Mioclacnidae UCMP 30003, 36533, 36537, 36624 Mioclacnidac UCMP 62109, 62117, 62137, 62187, 62204, 62225, 62260 IVriptvchidac UCMP 31307 lYriptycliidac UCMP 36585, 68683 IVriptvchidac UCMP 31271, 31815 IVriptvchidac UCMP 30006 IVriptvchidac UCMP 31268, 31282, 74775, 89701 IVriptvchidac UCMP 74974 Phcnacodontidac UCMP 114422, 39870, 43531 Phcnacodontidac UCMP 39870 Phcnacodontidac UCMP 39870 Mioclacnidac UCMP 114435,61519 Mioclacnidac UCMP 118506 Mioclacnidac UCMP 47241,69289 Mioclacnidac UCMP 30009, 30010 HSli Phcnacodontidac Phcnacodontidae Phcnacodontidac Phcnacodontidae Mesonychidac Mesonychidac Mesonychidac Mesonychidac Dichobunidae Dichobunidae Dichobunidae Apatcnividac Cimolcstidae Ad.ipisoricidae Adapisoricidae Scspcdcctinae Hrinaccidac Scspcdcctinac Geolabididac Leptictidae Lcptictidae UCMP 135009, no \'r. UCMP 36534 UCMP 36536, 68817 UCMP 36536, 36538. 68819 UCMP 36536 UCMP SDSMN 50575 USNM 299484 UCMP 43446 UCMP 43553, 43518, 43538 UCMP 43513, 46707 UCMP 40799 SDSMN 50574 UCMP 44027, 44916, 45965 UCMP 62057 UCMP 62056, 62058 62063 SDSMN 54984, 54985, 55160 UCMP 114364, 114434 SDSMN 54986 54994, 55161, 55162, 55164, 55287, 55288 SDSMN 54997 55000, 55102, 55165, 55166 UCMP, uncataloged, loc. V73080 UCMP 46672, 59159-59161 34 PALEOBIOS, VOL. 19, NUMBER .?, 1999 Appendix 1. (cont.) Material used for light optical analysis and presence or absence of HSB. Taxa are listed alphabetically within the different orders. Classification follows McKenna and Bell (1998). Coryphodontidac IK ',M P 4421S Pantolambdidae UCMP 89927 Taxon Family Prodiacodon crustulum Leptictidac PANTODONTA Coryphodon Owen, 1845 Pantolambda bathmodon Cope, 1882 PANTOLESTA Pantalettes Cope, 1872 Pantolcstidae Pavtolestes Pantolcstidae Paleosinopn, cf. Matthew, 1901 Pantolcstidae PERISSODACTYI.A Dilophodm Scott, 1883 Hyracodontidae Heptodon Cope, 1882 Tapiroidcac Homojialax Hay, 1899 [sectolophidac Hymcotherium Owen, 1840 Equidac Lambdotherium Cope, 1880 Brontothcriidac PRIMATES Hemiacodon Marsh, 1872 Omomyidac Microsyops Ixidy, 1872 (¦ Cynodimtomys) Microsyopidae Notbarctus crassus (Marsh, 1872) Adapidac Omomys cf. carteri Lcidy, 1869 Omomyidac Pclycodus Cope, 1875 Adapidac Pclycodus rabtoni Matthew, 1915 Adapidac Pcfacodus; Adapidac Phenacoltmur Matthew, 1915 Paromyidae Pbcnncolem nrt Pare >m< miyidae Pbenacolemuv jepscni Simpson, 1955 Paromyidae Plagiomtne muhicuspidiis Matthew, 1918 Plagiomenidac Platycboerops Charleswoth, 1855 Plesiadapidac PUsiadapis Gervais, 1877 Plesiadapidac Plesiada/iis- Plesiadapidac Pksiadapiscf. ir.v(Cidlcy, 1923) Plesiadapidac Plcsiadapis tricuspideus Gervais, 1877 Plesiadapidac Uintasorcx montcziimicus Liilcgravcii, 1976 Purgatoriidac RODENTIA Metanoinmys ajjorus Chiment & Korth, 1996 Eomyidac Microparamyscf. miliums (R. Wilson, 1937) Sciuravidac Paramys Lcidy, 1871 Ischyromyidae Paramys [schyromyidac Pareumys Peterson, 1919 Cylindrodontidac Pareumys C ylindrodontidac Sciuravuspowayensis R. Wilson, 1940 Sciuravidac Sciuravus powayensis Sciuravidae TapemysA.E. Wood, 1962 1 Leptotomys) Ischyromyidae TAENIODONTA Onychodectes Cope, 1888 Woirwanin otariidens (Cope, 1885) Collection number UCMP uncataloged, loc. V5620 HSB UCMP 155999 UCMP 55587 UCMP 55596 SDSMN 38617 UCMP 118409,43703 UCMP 72459 72461 UCMP 43566, 43567, 43596, 43613, 43644, 55655 UCMP 43590, 43101, 43599, 43601, 43614, 43616, 43627, 55657 UCMP 55582 UCMP 59366, 59581, 59582 UCMP 55579 SDSMN 55108, 55109, 55110, 55111 UCMP 46704 UCMP 59495 UCMP 39850 UCMP 44843 UCMP 39839 UCMP 59357, 59359 UCMP 73366 UCMP 71438 UCMP 61595 UCMP 114346 UCMP 114345 UCMP 62300, 62304 SDSMN 47251-47256, 47258 47262 SDSMN 62172-62179, 62182-62187 SDSMN 37625, 47263-47267, 45839-45847 UCMP 59713, 59716, 59983, 59991, 59992, 59715, 40241 UCMP 59984, 59985, 59714 SDSMN 31774 31777, 40994 SDSMN 31778 SDSMN 39350, 39353, 47268, 47269, 49351 SDSMN 55181 SDSMN 3482, 37617, 37618, 37622 UCMP 36514, 89695, 92156, 68668, 31817 UCMP 89289 (') i STEFEN-Y.NAMV.L MICROSTRUCTURK IN ARCHAIC UNGULATES 35 Appendix 1. (cont.) Material used for light optical analysis and presence or absence of HSB. Taxa are listed alphabetically within the different orders. Classification follows McKenna and Bell (1998). Taxon Family Collection number H.SH TTLLODONTA Esthonyx Cope, 1874 MARSUPIALIA Peradectes Matthew & Granger, 1921 Peratherium Aymard, 1850 Perutherium Thaler, 1967 Tillotheriidae UC:M1' 26600, 44420 Didclphidae SIXSMN 54979, 54980, 55286 Didelphidae SDSMN 54969, 55152, 55278 Didclphidae UCMP 106384 Appendix 2. Table of length (I.) and width (VV) measurements of m 1 and m2 of some of the analysed genera. Where values have been taken from other studies a literature citation is included. Uhen and Gingerich has been abbreviated to Uhen & Gins> Taxon A intention A n ison chits jjillisoti in ('.hrinnis ('Jttniiidnn ('onacodon cophater ('.onacodon nitncoitns < ''.onacodon entoconus Ectocion osborneanus Ectoconus Eetoconus ditrigonus Haploconus nnjiiiltus llnploiiiy/iis 11'nil itbinfits a pint lit tits I {yopsodus Litaletes I.itoiiivlits sntp/jisctts hoxohphus bvaitiijiitis Loxolophus hyattiiintis hoxohphus prints Mniiscotlicriiitii Mimatttta? Mioclnniiis tttrflititts Mioiiiinuts tuvnidits Orthaspidotheriutn Orthaspidotheriuttt ( X\'\rln nuts cuspidal its Osypviiniis eriksnii Ptriptychus cariniiens Periptychus coarctatus Phenacodus cf. primaevus Promioclacnus Protogonodon Protoselen e opistha nts [SB Collection no. Lml Wml I, m2 Wm + UCMP 137364 12.9 10.5 - UCMP 124408 4 3 3.5 3.2 + UCMP 111955 7.2 4.5 - UCMP 114412 ~ 5.2 8.2 6 - UCMP 36502 3 2.7 3 2.9 UCMP 36580 4.8 4.1 4.2 4.! - UCMP 124417 5 4.2 4.7 4.7 i UCMP 39783 7.2 6.7 7 7 - UCMP 36543 10 9.6 11.5 - UCMP 36543 10 9 i: 10.9 - UCMP 36614 4.4 3.2 3.8 3 - UCMP 43556 2.8 2.2 3 2.2 - UCMP 124415 3.5 3 3.5 3.3 - UCMP 43543 4 3.2 3.9 3.2 - UCMP 124418 5 4 5.9 4.5 - UCMP 111918 3 2.5 - UCMP 60467 5.2 4 - UCMP 31824 5.9 5 - UCMP 124411 6 5.1 6.5 6 i UCMP 171507 (i 4 5 4 UCMP 145065 3.2 2.9 3.8 3 • UCMP 36624 6 5.9 • UCMP 36534 6.2 6 UCMP 62187 4.2 3 ¦ UCMP 62260 4.4 3 - UCMP 63755 3.1 : I 6.3 4.9 . UCMP 30006 9 1 1 9.2 UCMP 31268 9 7.8 7.8 + UCMP 39870 11.9 10 12.5 10.5 - UCMP 114413 2.8 2.2 3 2.8 - UCMP 6925 8.5 7.4 - UCMP 30009 5.2 4.4 m inf. Literature Thcwisscn, 1990 36 PALEOBIOS, VOL. 19, NUMBER 3, 1999 Appendix 2. (cont.) Table of length (L) and width (W) measurements of m 1 and m2 of some of the analysed genera. Where values have been taken from other studies a literature citation is included. Uhen and Gingerich has been abbreviated to Uhen tk ding. Taxon HSB Collection no. Lml W ml Lm2 Wm2 in inf. Protosckne opisthaeus - UCMP 30010 5 3.9 Protungulatum - UC'.MP 135009 3.1 3 3.9 3 Rnpuarot - UCMP 134681 5.1 Rctflnavok - UCMP 134681 4.9 3.8 Tetracktenodon + UCMP 36S36 8.5 7 9.5 7.7 Thryptacodon attslralis UCMP 114435 6.1 5.] Adapisorex - UCMP 62055 2.4 1.5 Adapisorex - UCMP 62056 2.2 1.6 Diacodcxis + UCMP 43553 4 3 4.2 3.8 Entomolestes - UCMP 59098 1.9 x 1.2 Esthonyx + UCMP 26600 8 6 V.N 5.9 llepladon caciculus + UCMP 43703 7.2 4.8 8.3 5.9 Homojjalax - UCMP 39828 10 7 Hyracolhcrium - UCMP 55655 6.7 4.9 7 5 Lamdotherium - UCMP 113957 III 7.4 11 8.2 lAtoltstes lacunatus - UCMP 114434 2.9 2.1 2.2 2.1 MUrosyops (= Cynodontomys) - UCMP 59591 2.9x2 No than ms - UCMP 55579 6.8 5.2 Oncliydectes - UCMP 170848 6.9 4.9 6 5 Palaeictops - UCMP 59169 2.2x1.9 Paltosinopa - UCMP 55596 3.6 2.6 Paramys + UCMP 59823 2.2x2 Pclyaidus - UCMP 39850 4.1 3.5 Pelycnditi UCMP 59462 3.8x3.2 Perutherium - UCMP 106384 2 1 PhenucoUmur UCMP 44843 2 2.2 ('.oiypbodoji proterus + 29 22 36.7 26.1 Literature Archibald, 1982 Uhen & Gins',., 1995