SOCIETAS PRO FAUNA ET FLORA FENNICA

lACTA ZOOLOGICA FENNICA

144

Miguel Crusafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs from the Vallesian of the Valles-Penedes Basin, Spain

Helstngl Yllopl ton et Aki.das o

HELSINKI-HELSINGFORS 1976 ACTA ZOOLOGICA FENNICA

1-45 vide Acta Zoologica Fennica 45-50. 46-59 vide Acta Zoologica Fennica 60-93. 60-99 vide Acta Zoologica Fennica 100-125.

100. MARrA REuTER: Untersuchungen iiber Rassenbildung bei Gyratrix hermaphroditus (Turbellaria Neorhabdocoela). 32 S. (1961). 101. MARrA REuTER: Index Generalis Seriei Acta Zoologica Fennica 51-100 (194&-- 1961). 63 s. (1964). 102. WALTER HAcKMAN: Studies on the dipterous fauna in burrows of voles (Microtus, Oethrionomys) in Finland. 64 pp. (1963). 103. A. M. ]. EvERS: Dber die Entstehung der Excitatoren und deren Bedeutung fiir die Evolution der Malachiidae (Col.). 24 S. (1963). 104. ]oHAN REUTER: The international concentration in some hypotrichous Ciliates and its dependence on the external concentration. 94 pp. (1963). 105. GoRAN BERGMAN and KAr Ono DoNNER: An analysis of the spring migration of the Common Scoter and the Long-tailed Duck in southern Finland. 59 pp. (1964). 106. HENRrK OsTERHOLM: The significance of distance receptors in the feeding behaviour of the Fox, Vulpes vulpes L. 31 pp. (1964). 107. BJORN KuRTEN: The Carnivora of the Palestine caves. 74 pp. (1965). 108. BJORN KuRTEN: The evolution of the Polar Bear, Ursus maritimus Phipps. 30 pp. (1964). 109. FRANK S. ToMPA: Factors determining the numbers of song sparrows, Melospiza melodia (Wilson), on Mandarte Island, B. C., Canada. 73 pp. (1964). 110. PoNTUS PALMGREN: Die Spinnenfauna der Gegend von Kilpisjiirvi in Lappland. 70 s. (1965). 111. BJ6RN KURTEN: On the evolution of the European Wild Cat, Felis silvestris Schreber. 29 pp. (1965). 112. PAuL KRi.iGER: Dber die Einwirkung der Temperatur auf das Brutgeschaft und das Eierlegen des Rebhuhnes (Perdix perdix 1.) 64 S. (1965). 113. SAMUEL PANELIUS: A revision of the European gall midges of the subfamily Porricondylinae (Diptera: Itonididae). 157 pp. (1965). 114. Bo-]uNGAR P. WrKGREN: The effect of temperature on the cell division cycle in Diphyllobothrid plerocercoids. 27 pp. (1966). 115. BJORN KuRrtN: Pleistocene bears of North America. 1. Genus Tremarctos, spec­ tacled bears. 120 pp. (1966). 116. ToR G. KARLING: On nematocysts and similar structures in Turbellarians. 28 pp. (1966). 117. BJORN KuRTEN: Pleistocene bears of North America. 2. Genus Arctodus, short­ faced bears. 60 pp. (1967). 118. FRANK S. ToMPA: Reproductive success in relation to breeding density in pied flycatchers, Ficedula hypoleuca (Pallas). 28 pp. (1967). 119. ANN-MARIE FoRSTEN: Revision of the Palearctic Hipparion. 134 pp. (1968). 120. CHARLES HrGHAM: Size trends in prehistoric European domestic fauna, and the problem of local domestication. 21 pp. (1968). 121. ANrTA RosENGREN: Experiments in colour discrimination in dogs. 19 pp. (1969). 122. ToM REuTER: Visual pigments and ganglion cell activity in the retinae of tad­ poles and adult frogs (Rana temporaria L.). 64 pp. (1969). 123. GoRAN GYLLENBERG: The energy flow through a Chorthippus parallelus (Zett.) (Orthopera) population on a meadow in Tviirminne, Finland. 74 pp. (1969). 124. PAuL KRiiGER: Zur Rassenfrage der nordeuropiiischen Igel (Erinaceus europaeus L.). 15 S. (1969). 125. BRIT GoosKE ERIKSEN: Rotifers from two tarns in southern Finland, with a description of a new species, and a list of rotifers previously found in Finland. 36 pp. (1969). ACfA ZOOLOGICA FENNICA 144 EDIDIT SOCIETAS PRO FAUNA ET FLORA FENNICA

BEARS AND BEAR-DOGS FROM THE I I V ALLESIAN OF THE V ALLES-PENEDES BASIN, SPAIN

Miguel Crusafont Pair6 and Bjorn Kurten

Helslngln YUoplston Me Ak rjasto

H E LSI KI - H E L I GF OR March 1976 ISB 95·1-661-013-7 ISS 0001-7299 © Societas pro Fauna et Flora Fennica 1976

Abstract

CRUSAFONT PAIR6, MiGUEL & KuRTEN, BJOR : Bears and bear-dogs from the Vallesian of the Val!es-Penedes basin, Spain. - Acta Zool. Fennica 144:1- 29. 1976. A revised list of ma=alian fossils from the Vallesian localities Can Uoba­ teres and Can Ponsich (Valles-Penedes basin, Spain) is given and the following species of Carnivora described. Indarctos vireti, validated as the most primitive Indarctos known. Ursavus brevirhinus and U. primaevus as associated but distinct species. Protursus simpsoni n .g., n.sp. as an early member of the Ursinae. Amphicyon cf. maior and Canidae indet. Authors' addresses: M. Crusafont Pair6, Institute Provincial de Paleontologia, Sabadell, Spain; B. Kurten, Department of Geology and Paleontology, University of Helsingfors, SF-00170 Helsingfors 17. Finland.

Contents

Introduction ...... 3 Prorursus Crusafont & Kurten, n.g. . . . . 22 List of species ...... 4 Protursu simp oni Cru afoot & Kurten, Systematic descriptions ...... 5 n. sp...... 22 Indarctos vireti Villalta & Crusaf.ont . . 5 Amphicyon d. major De Blainville . . . . 2 Ursavus brevirhinus (Hofmann) ...... 15 Canidae indet...... 27 Ursavus primaevus (Gaillard) ...... 16 References ...... _. . . . 28 otes on the history of Ursavus ...... 18 Appendix: Abbreviation . _. _...... 29 INTRODUCTION

The fossils to be described here come ring about 12.5 m.y. BP (see VA Cou­ from the early Vallesian beds at Can VERrNG, 1972). Although Hipparion and Llobaoeres and C an Ponsich in the Val­ other forms of Oriental origin enoer the ~~~s-Penedes basin (Spain). Both localities Valles-Penedes at this stage, the conti­ are situated in the immediate neighbour­ nuation of ecological conditions essen­ hood of the city of Sabadell; still an­ tially similar to those of the Vindobo­ other fossiliferous exposure of the same nian resulted in a phase of endemism, age is found at nearby Rio Ripoll. Can the Vallesian phase of CRUSAFONT Llobateres is now richer in mammalian (1 958a). Vindobonian relicts are numer­ species than any other locality from the ous in the lower beds but gr adually be­ " Pontian", sensu lato, in all of Eurasia. come more scarce as we ascend through Can Ponsich is the type site of the Val­ the Vallesian beds, which have a maxi­ lesian Land-Mammal Stage. mum thickness of about 400 m. The name Vallesian was introduced The two localities discussed here are by CRUSAFO T (1950) for the lower in the lower Vallesian. The site at Can levels of the "Pontian", characterized Llobateres was discovered in 1926, but by the pre ence of a primitive Hippa­ its Vallesian age was recognized in 1951 rion, giraffids of the genus Palaeotragus, (CRUSA FO T & TRUYOLS, 1951 ). Can and various rel icts surviving from the Ponsich, first discovered in 1946 and preceding Vindobonian stage. Its strati­ described by CRuSAFONT & TRUYOLS graphic posi tion, then, is between the (1947) has yielded the type section of terminal Vindobonian as exemplified by the Vallesian. It has a direct stratigraph­ La Grive aint-Alban, and the upper or ic relationship with the underlying upper "typical" Pontian of Pikermi, the Piker­ Vindobonian beds of Sant Q uirze. mian (CRUSAFO T 1950) or Turolian The Vallesian beds of the Valles­ (CRUSAFO T 1965a). In the Valles-Pene­ Penedes rank with the Siwaliks as the des basin, the late Vindobonian is repres­ main pongid-bearing strata in Eurasia ented by the fossiliferous exposure of the (CRUSAFO T & H tiRZELER 1961; SrMo s railway cut at Sant Quirze, while the & PILB EAM 1965 · CRUSAFO T 1972a). A Turolian is found at Piera; both sites grant from the Wenner-Gren Founda­ have yielded large mammalian faunas. tion for Anthropological Research, as THALER (1966), in his suggested zonation well as the acquisition of the site by a of the continental T e rtiary in Europe, large chemical firm, have greatly furth­ makes an essentially similar arrange­ ered the work at this locality. We grate­ ment his " abadell Zone" corre pond­ fully acknowedge the su pport thus re­ ing to the Vallesian, and his "Teruel ceived. Zone" to the Turolian. Studies of the faunas of Can Llobate­ The Vallesian is an important phase res and Can Ponsich may be found in the in the evolution of the mammalian fau­ papers already referred to; in addition nas: a process of modernization with see CRUSAFO T {1965b, 1972), CRUSA­ the appearance of various new species FO T & CASA OVAS {1973), CRUSAFO T and genera with a datum line provided & HtiRZELER {1969), CRUSAFONT & by the immigration of Hipparion occur- PETTER {1969) GOLPE (1971 ), HARTE - 4 Miguel Crusafont Pair6 and: Bjorn Kurten: Bears and Bear-Dogs

BERGER (1965, 1966), P ETTER (1963, natively classified as Canidae, Ursidae, 1967). and Amphicyonidae) from Can Lloba­ The present paper is a S'tudy of the teres and Can Ponsich. ursids and othe r bear-like forms (alter-

LIST OF SPECIES

An interim list of the species recognized to Can Can date at Can Llobateres and Can Ponsich is pre­ Lloba- Ponsich sented herewith . A comparison with the list teres given for the former locality m CRuSA FONT H. sp. + (1964) wi'il indicate the rate at which the studies Cf. Spermophilus bredai have been progressing. (Meyer 1848) + Citellus sp. + Can Can Miopetaurista aff. grimmi Black + Lloba- Ponsich Cryptopterus crusafonti teres Mein 1970 + + Insecti vora Monosaulax rninutus Postpalerinaceus vireti Crus. & (Meyer 1838) + + ~-1~ + + Steneofiber jaegeri Kaup 1832 + Lantanotherium sanmigueli S. depereti Mayet 1908 + Vill. & Crus. 1944 + + P rogonomys aff. cathalai Galerix socialis Meyer 1865 + + Schaub 1938 + Heterosorex sansaniensis Cricetodon sansaniensis Gaill. 1915 + + Lartet 1851 + Miosorex grivensis Viret & C. (Hispanomys) decedens Zapfe 1951 + Schaub 1925 + Crusafontina endernica Megacricetodon aff. minor Gibert in litt. + (Lartet 1851 ) + + Talpa rninuta Blainv. 1839- 64 + + M. debruijni Plesiodimylus chantrei Gaill. Freudenthal 1968 + 1899 + + M . aff. gregariu (Schaub 1925) + Primates Ruscinomys thaleri Hart. 1965 + + D ryopitbecus piveteaui Crus. & Rotundomys sabadellensis Hlirz. 1962 + + (Hart. 1965) + Hispanopithecus laietanus Crus. R. hartenbergeri Freud. 1%7 + & Vill. 1946 + + Anomalomys cf. gaillarcli Rahonapithecus sabade1lensis Viret & Schaub 1946 + Crus. & HUrz . 1962 + Corimus leemanni Hart. 1965 + Leptodontomys catalunicus Lagomorpha (Hart. 1966) + Prolagus oeningensis Meyer + + Kerarnidomys pertesunatoi H art . 1966 + Rodentia T itanomys fontanesi Dep. + Carnivora Pentaglis sp. + Amphicyon cf. major Muscard:inus (Muscardinus) Blainv. 1841 + + crusafonti Hart. 1966 + I ndarctos vireti M. (Eomuscardinus) vallesensis Vill. & Crus. 1943 + + Hart. 1966 + Ursavus primaevus Gaillard 1899 + Glirulus aff. lissiensis U. brevirhinus Hoffmann 1877 + Hug. & Mein 1965 + Protursus simpsoni Heteroxerus cf. rubricati Crus. & Kurten n.g. n.sp. + Crus. & Vill. 1955 + Ischyrictis (Ischyrictis) petteri H . grivensis (Major 1893) + Crus. 1972 + H . hurzeleri Stehl. & Schaub 1951 + Plesiogu]o sp. + ACfA ZOOLOGICA FENNICA 144 5

Can Can Can Can Lloba- Lloba- Ponsich Ponsich teres teres Circamustela dechasauxi Jourdan 1861 + + Petter 1967 + D. giganteum Kaup 1829 + Marceria santigae Petter 1967 + Manes aff. andersoni Perissodacryla chlosser 1924 + Hipparion catalaunicum M. mellibulla Petter 1963 + Pirlot 1956 + + M. munki Roger 1900 + Macrotherium grande Lartet 1837 + + Taxodon cf. sansaniensis Tapirus priscus Kaup 1832 + + Lartet 1851 + Aceratherium incisivum abadellicti~ crusafonti Kaup 1834 + + Petter 1963 + ?Chilotherium sp. + + Trocharion albanense Major 1903 + Dicerorhinus sansaniensis Promephitis pristinidens (Lanet 1851) + + Petter 1963 + D. sp. + Mesomephitis medius Petter 1967 + Paralutra sp. + Artiodactyla Limnonyx sinerizi Crus. 1950 + Hyotherium soemmeringi Semigenetta mutata Meyer 1834 + (Filhol 1883) + H. palaeochoerus (Kaup 1833) + Progenetta crassa (Dep. 1892) + H . sp. + P. gaillardi Major 1903 + Parachleuastochoerus crusafonti P. montadai (Vill. & Crus 1943) + Golpe 1971 + + Metarctos batalleri Viret 1933 + Listriodon splendens Meyer 1846 + Machairodus aphanistus Conohyus simorrensis Kaup 1833 + + (Lartet 1851 ) + Grivasmilus jourdani Dorcatherium jourdani (Filhol 1883 ) + + Filhol 1883 + D. sp. + Proboscidea Euprox dicranocerus (Kaup 1835) + "Mastodon" sp. + E. furcatus (Hensel 1859) + Gomphotherium angustidens Micromeryx flourensianus (Cuvier 1806) + Lartet 1851 + + Tetralophodon longirostris Capreolus sp. + + (Kaup 1835) + Miotragocerus chantrei Dep. 1887 + Deinotherium laevium M. sp. +

SYSTEMATIC DESCRIPTIONS

Indarctos v ireti Villalta & Crusafont ciform bones. o number, left and right Ms, two left MC 2, left Ci, two caudal vertebrae. from Can Llobateres. lndarctos vireti VILLALTA & CRuSAFO T 1943 :54. Agriotherium insignis pontiensis VILLALTA & CRUSAFO T 1945:94. lndarctos arctoides vireti, THE IUS 1959:284. Diagnosis M~terial. VP 646--647, cru hed skull and right and left rami of mandible. VP 633, pans of A lightly built lndarctos with com­ second skull with left mandibular ramus. P 640, left M!. VP 641-642, left and right M1. paratively long jaws and uncrowded VP 651-652, two left C•. VP 655-656, right premolars; second and third premolars and left Ci. VP 668, left MC 2. VP 683-684, two-rooted relatively large, elongate left and right MC 3 ( probably same individual with high, keeled as VP 668). VP 669, left MC 4 fragment. VP and premolariform 670, left MT 2 fragment. VP 665, r ight humerus. crowns; molars on average somewhat VP 666, right astragalus. VP 667, two left un- smaller than in / . arctoides. 6 Miguel Crusafont Pair6 and Bji:irn Kurten: Bears and Bear-Dogs

FIG. 1. Indarctos vireti, VP 646, skull, Can Llobatere .

Taxonomic history ally dropped from the fauna! lists of the Vallesian by CRUSAFO T & TRUYOLS The species lndarctos vireti was based (1960). by VrLLALTA & CRUSAFO T (1943) on a The species /. vireti was originally right mandible fragment (Col!. Villalta) characterized as follows (VrLLAL TA & with the crowns of M1 and M2, from the CRUSAFO T 1943:54): !ndarctos vecino Vallesian of Can Purull at Viladecaballs del /. arctoides D eperet, del Ponriense in the Valles-Penedes basin; its strati­ de Montredon, aunque de talla clara­ graphic position is considerably higher mente inferior con tubercula a (M2) mas in the stage than that of Can Llobateres. alargada en relaci6n con el M 1 com­ Somewhat later another pan of the parativamente a la especie francesa. Me­ same jaw, with P4, was discovered tac6nido de M 1 mas soldado con la (VrLLALTA & CRUSAFO T 1945). At the punta principaL same time another mandible from the ToBIE (1955) remarked on the simi­ same locality was described under the larity of what was then known of I. vi­ name Agriotherium insignis var. pon­ reti and the species /. arctoides which is tiensis. The species A. insigne is, of characteristic of the Pontian in western course, a well-known Ruscinian guide Europe. THE ru (1959) concluded that fossil; in actual fact the specimen does there wa no valid distinction b~tween belong to /. vireti. The error is pre­ the two and regarded the Vallesiah 'form sumably due to an ancient break in the as a subspecie of /. arctoides. fossil, which passes through the crown The larger material now at hand from of M1 and distorts its shape, giving it a Can Llobateres dispro es this conten­ certain resemblance to the agriothere tion. Indarctos vireti turns out to be a carnassial. The specimen is a right man­ more primitive pecies than any other dible (in the Museum at Sabadell) with in the genus as befit its early date. M1, M2 and the roots of P4. Brief descriptions of both species, on the basis of the same material, were re­ D escription iterated in the survey by CRUSAFONT & VrLLALTA (1951). Agriotherium was fin- The lower canmes are ursiform but ACfA ZOOLOGICA FE ICA 144 7

somewhat more recurved than usual in in the anterior half of the tooth in this bears. The internal keel is strongly de­ as in the other premolars. The ridge on veloped and situated more internally the posterior slope of the tooth is well than in Ursus; near the enamel basis the developed and widens backward; on the keel curves backward in a characteristic anterior slope it is faint and slightly way. There is an incurvation of the curved, convexity to the lingual side. enamel base in front of the keel, instead P2 and P3 hav·e two roots and thus of right at its base. The p osterior edge is differ from I. arctoides and /. atticus. rounded. Their outline in occlusal view is elongate The upper canines have a well deve­ and tending to the rhomboid. The main loped anterointernal ridge. but it is not cusps are h igh and pointed, as for in­ quite as sharp as in C. At the enamel stance in P4 of Ursus americanus. There base it terminates in a convexity of the is no acces ory cusp. A cingulum en­ enamel border, as in Felis, for instance, circles the tooth but it is weakly deve­ rather than an incurvation. There is an lcped except along the posterointemal incurvation in front, however. pa rt. The front and hind slopes are Both the upper and lower canines keeled, but the edges are not particularly show a d is tinct sexual dimorphism (see sharp. below). These teeth represent a more primitive P1 is preserved only in the jaw VP condition than that seen in other species 633. It has one root and looks rather like of Indarctos. In I. arctoides, P3 is two­ its homologue in Tremarctos. The outline rooted and may have resembled its ho­ is an elongate oval. The crown is flatten­ mologue in /. vireti (unfortunately only ed, with a posterior and posterointernal the alveolus is known in the Pfaffstetten cingulum. The point of the m ain cusp is jaw described by THENrus (1959)), but

TABLE 1. Measurements of lower canines and premolars of Indarctos.

C; L B

I ndarctos vireti Can Llobateres VP 647 rde~t. 18.5 13.0 10.8 6.7 13.5 8.0 16 .9 9.4 9.5 \. sm. 19.2 13.2 11.1 6.7 13.5 8.0 16.7 9.5 9.4 VP 633 sm. 16 11! 10.4 9.9 6.5 12.2 7.7 16.3 8.9 VP 656 sin. 22.7 15.3 VP 655 dext. 21.8 16.6 VP 654 dext. 24.3 16.5 o number Slll. 20.1 12.8 Can PuruU, type dexr. 14.7 8.5 8.1 I ndarctos arctoides Montredon, type Mtn 14 sin. - 161/! 12.0 10.2 Pfaffstetten (After THENrus) 18.4 13.8 ( 8.5) 9.7 6.1 (11.0) - 15.8 9.6 I ndarctos atticus amo (After HELsr G) 11.8 8.9 21.0 12.8 13.7 Maragha (After DE MEcQUENEM) 10 9 7 13 91/! 23 14 Indarctos anthracitis

Monte Bamboli, cast dext. 9.2 6.1 19 10.5 8 Miguel c,:usafont Pair6 and Biorn Kurten: Bears and Bear-Dogs

FIG. 2. Indarctos vireti, VP 647, mandible, Can Llobateres.

P2, separated from P3 by a long dia­ parastyle, and indeed this element is un­ stema, is one-rooted and much reduced. common in Indarctos, though present in In I. atticus, P3 is also one-rooted. /. atticus from Concud. The paracone P1 resembles P1 rather closely, and has a blunt edge fore and aft, and there similarly P2 and P3 resemble P2 and PJ; is a ridge running from its apex to tihe but their anterior keel is more deflected front end of the protocone. The meta­ inwards, as usual in upper premolars. cone forms a broad, blunt carnassial Again the large size and homodonty of blade. In outline the tooth shows a deep the second and third premolars is a pri­ notch in front of the protocone, but none mitive condition. P 1 has one root, P 2 and behind it, where there is a well develop­ P 3 have two. ed internal cingulum. The external cingu­ P.o~ is somewhat larger than P2-P3 but lum is weak and only present in the otherwise rather similar, even to the ab­ metacone region, in contrast with / . arc­ sence of acces·sory cusps. The homo­ toides, where it is well developed. donty of the series P2-P.o1 is more primi­ M1 has a simply built trigonid, in tive than the condition in any other which all the three cusps are low-crown­ species of the genus. ed and single. The entoconid on uhe P4 is quite large. Th'e protocone is well talonid is duplicated with a smaller developed and set off from the blade. posterior part. There is no hypoconulid There is a tendency to protocone dupli­ bordering the talonid basin at the back. cation, the anterior pan being the small­ The hypoconid is single but there is a er. Such a tendency is not uncommon in small cusp between the protoconid and I. arctoides and I . atticus. There is no the hypoconid similar to that seen in ACTA ZOOLOGICA FENNICA 144 9 tremarctines (see KuRTE 1966). At the without a distinct paraconid. The talo­ same point, that is to say at the external nid forms a larger basin, from the centre junction between the talonid and the of which there radiate a number of small trigonid, there is a weak cingulum. The grooves to the notches between the prin­ enamel surface of the tooth is smooth cipal cusps of the surrounding wall, the without wrinkling. hypoconid and the double entoconid (of M 1 has a five-cornered outline, the which the posterior part is the smaller, inner wall being produced into an obtuse as in M1). There is no posterior talonid angle by the large cingulum. There are cusp. also external and posterior cingula. The M 2 has a large paracone. The meta­ paracone is the highest external cusp. cone is small, and the protocone is ridge­ The inner cusps are confluent and form like, continuing into the small hypocone. a ridge with a somewhat individualized There are no talon cusps apart from the but obtuse and small hy pocone. Between last-mentioned, but there is a weak ridge the inner ridge and the cingulum, the encircling uhe talon field, which is al­ enamel is transversely wrinkled in a most smooth except for a few small characteristic ursid manner. The middle wrinkles. There are more wrinkles be­ field, on the other hand, is almost tween the inner ridge and the cingulum, 1 smooth, except for a few grooves. as in M • The outer cingulum is large in M2 has the front part somewhat VP 646 smaller in VP 633. The talon broader than the hin d. The main cusps appears fairly long; perha ps this im­ are the protoconid and metaconid, which pression is determined in pan by the are connected by a transversal ridge relatively small transverse dimensions with a notch in the middle, just as in of the cheek teeth. Arctodus. The front part is basined and MJ is nearly circular or ovoid. This is bounded in front by a semicircular wall the only tooth with a system of enamel

TABLE 2. Measurements of lower molars of Indarctos.

M2 M3 L Lt L Ba Bp L B

I ndarctos vireti Can Llobateres VP 647 { de~t. 30.0 11.3 14.7 21.2 23 .7 16.2 14 .8 14.2 12.7 " t Stn. 30.2 11.4 14.9 21.0 23 .6 16.2 14.8 VP 633 sm . 32.7 12.8 16.8 25 .0 17.2 16.0 14.8 12.4 VP 640----641 SlD. 12.9 26 .1 18.0 16.1 a17 .2 VP 642 dext. a32 1/t 17.1 14.9 Can PuruU, type dext. 30.8 12.2 15.1 21.4 22 15.6 13.4 u u pontiensis dext. a34 1/t 13 .8 17.3 25 .6 17.7 16.2 Indarctos arctoides Montredon, type Mtn 14 dext. 35.7 14 .8 18.3 25 .2 19.0 17 .2 Pfaffstetten (After THE ms) 30.8 15.8 24 .1 16.5 Westhofen (After ToBJE 32 .6 17.3 25 .2 17.4 I ndarctos atlicus amos (After HELBJ G) 41.6 22 .2 32.4 23.3 21.5 17.4 Maragha (After DE MEcQuE EM ) 44 17 22 32 23 22 Con cud SlD. 42.4 17.8 22 .2 I ndarctos anthracitis Monte Bamboli, cast in. 33.2 13.5 16.7 24 .0 23 .7 16.5 15.5 13.5 12.0 10 Miguel Crusafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs

FIG. 3. Indarctos vireti, same specimen as fig. 2. wrinkles on the occlusal surface similar Dimen ions of the kull and teeth are to that in more advanced ursids. The given in table 5. It may be noted that middle field is surrounded by a slightly the presumably female specimen VP raised waH without any distinct cusps, 646-647 is on! slightly shorter than only the protoconid forms a weak swel­ the adult Samos II individual described ling, bounded posteriorly by a notch. A by THENTUS (1959) though its teeth are faint ridge runs from the protoconid to much smaller. The presumed male VP the centre of the field. 265, probably exceeded the amos form The only preserved skull, VP 646, is in skull and mandible length. Despite the so crushed and distorted that little in­ small size of the teeth it would seem formation on its original configuration that the head of I. vireti was about as can be obtained. The basicranial region long as that of I. atticus. In con equence, resembles that in I. atticus described by there i no rowding of the tooth row, THE IUS (1949a) but i much less short­ so that the premolars rand in a traight ened. The facial part of the skull is re­ line. latively much longer than in I. atticus. On the other hand the head was cer­ The mandibles are better preserved. tainly narrower in /. vireti than in I. As in other Indarctos, the premasseteric atticus, and the animal was more slen­ fossa is absent. The lower border i der-jawed doubtle s a primitive charac­ strongly convex, and the ramus tapers ter. This tenden y to lightness in build forward. These are characters generally is a! o found in the post ranial keleton. found in the genu , and likewi e the Mea uremenrs of the limb bones of I. position of the mental foramen below vireti from Can Llobareres are com­ P1 is typical. The mandibles are however pared in Table 6 with data on I. atticus noticeably slimmer and more elongate from Con ud · on the large short-faced than in most Indarctos, especially I. atti­ bear of the orth Am rican Plei rocene, cus, and the teeth are small in relation to Arctodus simz~s, from Rancho La Brea the length of the ramus. (after MER RI AM & TOCK 1925) · and on

, ACfA ZOOLOGICA FE ICA 144 11

TABLE 3. Measurements of upper canines and premolars of Indarctos.

Cs p' p2 p3 p4 L B L L B L B B L H

[ndarctos vireti Can Llobateres VP 633 dext. 22.7 16.0 10.0 10.2 6.2 11.9 6.8 23.7 17 .7 VP 646 19.5 13.5 11.8 6.2 12.4 6.7 21.8 17.4 11.7 VP 651 Sin. 25.6 17.1 VP 652 sin. 16.1

[ ndarctos arctoides Montredon (After HELBING) 27.4 21.0 Orignac (After HELBING) 24.6

I ndarctos attic us Samos (After HELBI G) 7.1 7.8 - (9.5) 29.1 24.3 18.5 Samos II (After THENIUS) 24.4 17.8 (7.2) 9.5 7.0 11.3 7.9 29.5 22.5 Con cud - 12.0 a specimen of the modern American The specimen is of about the same black bear, Ursus americanus, from Flo­ length as that of a large black bear. It is rida (see KuRTE 1966). much smaller than the humerus of Arc­ The humerus from Can Llobateres be­ todus but resembles the latter more than longs to a young individual and lacks it does Ursus. As in the short-faced bear, the proximal epiphyse. From its small the supinator ridge is weaker than in size and its age it could well belong to Ursus but the deltoid ridge is more de­ the same individual as VP 646-647. veloped than in Arctodus. The entepi­ This may also be true for some other condylar foramen is larger than in Arc­ bones. todus, and the medial ep icondyle is more

FIG. 4. I ndarctos vireti, VP 633, mandible. Can Llobateres. Part restored in plaster. 12 Miguel Crttsafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs

FIG 5. Indarctos vireti, same specimen as fig. 4. prominent. In the distal articulation the vily built and presumably comes from a capitulum is less clearly set off from the male. This is probably also true for the trochlea than in Arctodus. MC 4. If the latter belonged to the same The only carpal bone represented is individual as the large MC 2, the diffe­ the unciform, which is somewhat closer rence between the lengths of the two in proportions to Ursus than to Arcto­ bones would amount to 6.4 per cent of dus, as the homologous bone in the latter the MC 2 length. Corresponding figures is much extended proximo-distally. are 9.7 for Arctodus and 13.3 for U. Four metacarpal bones are preserved. americanus. It may perhaps be suggested The two MC 2 may represent male and that the outer fingers in /. vireti were female individuals resp ec tively; an ad­ less enlarged, relatively to the inner, ditional specimen is broken. The shorter than in the Pleistocene and recent forms; MC 2 measures some 24.1 per cent of the as a corollary it would apparently fol­ length of the humerus, which is more low that the typical toe-in position of than in Ursus (20.9) or Arctodus (19.5) the bears had not yet evolved in the Val­ and indicates that the hand was more lesian form. The length of the meta­ elongate in lndarctos than in the later carpals may suggest that this animal was ursids. not as fully plantigrade as modern bears. The fragmentary MC 3 is rather hea- The astragalus is represented by one

TABLE 4. Measurements of upper molars of I ndarc tos.

M1 M 2 Ba L Bp Hp Hm Ba L Bm

I ndarctos vireti Can Llobateres VP 633 dext. 24.5 19.5 20.6 27 .9 19.0 17.2 VP 646 22.4 19.6 21.0 91/! 9.8 27.3 19.7 17.8 I ndarctos arctoides Montredon Mtn 17 Stn. 24 22 .2 22.2 29.3 20.5 18.8 sin.1 27.0 22.5 24 .0 31.0 24.0 Orignac dext.1 24 .6 22.0 28.2 22.0 Indarctos atticus Samos1 29 .0 26 .0 28.5 15.2 34 26 Samos IP 28.3 262 30.0 26.5

1 After HELBING. 2 After THE rus. ACTA ZOOLOGICA FE ICA 144 13

TABLE 5. Measurements of skull and mandible of Indarctos.

I. I. I . I. vireti arctoides atticus anthracitis VP 646-7 VP 633 Type Pfaff.l Samos 2 Samos3 Bam~ll

Skull Basal length 300 e310? e325 Palatal length 150 e151 e162 Width over condyles 65.0 e74

Mandible Length, C to condyle 218 a265 e246 e243 Height, coronoid process e115 e98 e111 Depth under P4 43 a53 a41 44.0 54.0 46 Depth under M2 48 60 46 49.5 61.0 46 Length Mt-Ms 67 72 a65 68

1 Pfa£fstetten, after THENIUS. 2 Samos I, juvenile, after HELBING. 3 Samos II, adult, after THENIUS. specimen, which may be compared with but they are plumper and heavier. They the corresponding bone in I. atticus from are much shortened relative to the Con­ Concud, in Arctodus, and in the black cud MC 2, so that the relationship in bear. Of all these the Can Llobateres I. atticus resembled that of modern bears bone is by far the least bear-like, with a more closely than I. vireti. longer and more constricted neck than the others. The Indarctos bone from Concud shows a more advanced bear Sexual dimorphism type, although its neck is still somewhat longer than in Ursus. Both forms of There is an apparent sex dimorphism Indarctos have large sustentacular facets in the size of the jaws and teeth (Tables of identical shape, but the ectal facet is 1, 3, 5) and also in the postcranial more reduced in I. atticus. bones. VP 633 is probably a male, while The graceful build of the Can Lloba­ the skull and mandible VP 646-647 is teres astragalus again emphasizes the probably a female. The isolated canines slenderness of the I. vireti skeleton and VP 651-656 are quite large and pro­ its heritage from more nearly digitigrade bably male, while the 'Specimen without ancestors. a number is probably female. At Can MT 2 is the only metatarsal bone pre­ Purull, too, both sexes are probably re­ served. It is rather long and powerful, presented, the type of I. vireti being fe­ exceeding the largest MC 2 by 9.0 per male, and the second jaw, with marked­ cent of the length of the latter bone. In ly larger teeth, male. Such a degree of Arctodus the corresponding figure is on­ sex dimorphism would be equivalent to ly 4.7 per cent, while in the black bear that found in larger members of the the metatarsal is shorter than the meta­ genus Ursus (KuRTE 1955). carpal. Again, this is obviously a trait tetained from more fast-running an­ cestors. Phyletic position The specimens of MT 2 from Concud are about as long as the I. vireti bone, I ndarctos vireti stands out as the most 14 Miguel Crusafont Pair6 and Biorn Kurten: Bears and Bear-Dogs

primtttve known species of the gen us. predaceous carnivore than the modern In the dentition, the premolars are less bears in general. reduced than in other !ndarctos, and In an evolutionary perspective, / . vi­ the head and jaws are slimmer a nd more reti retains many characters from the lightly built. The limb bones indicate ancestors of the bears. Its closest relative a lightness and fleetness of foot very is evidently / . arctoides of the later different from the orthodox b ear image. Pontian, which most probably was a It may well have been a more active, direct descendant of the Vallesian spe-

TABLE 6. Measuremenns of limb bones in I ndarctos and other Ursidae.

Indarctos Arctodus _Ursus 'lmerrcanus Rancho Can Llobateres Con cud UF 5639 La Brea Florida

Humerus Length e290- 497 304 Shaft width 24.3 43.5 29.6 Distal width 75.4 126.8 84 Dist. artic. width 50 .6 97 55 11• Unciform Dist. anteropost. 27.9 25.2 38.5 21.3 Dist. transverse 21.2 22.3 39.7 19.0 Depth 24.5 24.6 50.8 24 .2 MC II Length 70 78 115 118 96.8 63.7 Prox. anteropost. 27.8 25.4 24 .0 29.3 28.8 30 19.2 Shaft width 12.0 12.8 13.8 14.0 14 .0 13.4 10.9 Distal width 20.0 19.5 22.9 17.3 MC Ill Length 104 68.8 Prox. anteropost. 25.5 31.6 21.5 Shaft width 14.4 14 11.1 MC IV Length e83 106.2 72 .2 Prox. anteropost. 24.5 31 20.0 Shaft width 15.7 15.8 11.8

A:s tragal us Width 36 43 .1 86.2 42 .0 Length 45.2 48.8 79 39.4 Width, trochlea 29.0 32.0 31.4 Width, head 25 32.4 25.3 Width, sust. facet 17.7 20.8 16.8 Width, ectal facet 14.0 7.8 14.7 MT II Length e85 87 86 101.3 58.8 Prox. anteropost. 28 31.0 30.5 36.8 18.3 Shaft width 14.5 15.0 17.8 16.3 10.7 ACTA ZOOLOGICA FE ICA 144 15 cies. The posltlon of the Italian form form it lies just anterior to the notch I. anthracitis (= laurillardi) is at present between paracone and metastyle. The uncertain; it may be a precociously spe­ protocone lobe is rather small, and the cialized form. Indarctos atticus of the actual cusp lies in the posterior pan of Turolian, on the other hand, is more ad­ the lobe; it is low and sharply pointed. vanced with larger cheek teeth, more re­ M1 has a trigonid which retains much duced premolars, heavy and powerful of the canid carnassial facies with a jaws, and ursine limb proportions. This high protoconid, a rather large para­ species, too, may well have arisen from conid and a small, simple metaconid I. vireti. lying somewhat posterior to the proto­ conid. The talonid is a simple, basined structure, much broader than the trigo­ Ursavus brevirhinus (Hofmann) nid, with a large hypoconid and a low, ridge-like, bipartite entoconid. The ena­ For synonymy see THENIUS (1949b). mel is smooth; there is a small external Material. Teeth of two individuals (Can Llo­ 4 cingulum. bateres): Specimen A, left and right P1 and P , left M1 and M2, right M1 fragment, left and M2 is smaller than M1. The anterior right C;, Ps, P•, M1-M2, left Ms; specimen B, lobe is broader than the posterior. The left P<, Mt, M2, Mt; left and right 0. protoconid and metaconid, which form the summit of the crown, lie beside each other and are connected by a transverse Description ridge, w hich divides the middle field into a small anterior and a large poste­ For dimensions see tables 7-9. rior basin. Notches in the outer and in­ The lower canines are less flattened in ner walls separate the trigonid and talo­ cross section than in Ursus, and more nid. Of the talonid cusps, only the ento­ pointed and higher crowned. The poste­ conid is individualized. The middle field t ior edge is sharp, and there is a promi­ is nearly smooth. nent crest along the lingual side from The single MJ is somewhat damaged the tip to the cingulum at the enamel but clearly had an ovoid outline and an base. The uppe::r canines are also sharply almost flat, slightly wrinkled surface pointed and have a comparatively broad surrounded by a rim without any indi­ base. Their posterior crest is very sharp, vidual cusps. while the anterior ridge, which arises M 1 is broadly quadrangular with a from the basal cingulum, is less develop­ large paracone, a smaller metacone and ed than in the lower canines. two inner cusps showing up along the The anterior premolars are dorso­ inner ridge. There is no distinct para­ ventrally flattened, one-rooted unicus­ style, but the anteroexternal corner of pids with a longitudinal ridge. P4, which the crown protrudes distinctly. The in­ has two roots, has a distinct posterior ner cingulum is large, the outer small; cusp in addition to the main cone; both the enamel is mostly smooth but has a lie on the same longitudinal ridge. The slight tendency to w rinkling in the val­ enamel of this tooth tends to wrinkle leys and along the cingular border. on the inner cingulum. M 2 is not unlike a miniature replica of 4 P is comparatively large, and the size its homologue in Indarctos. It has a large difference between specimens A and B paracone with a longitudinal ridge and is marked. Presumably A, which tends smooth outer wall; it is separated from to larger dimensions in all the teeth, is a the metacone by a deep notch. The meta­ male, while B is a female. The proto­ cone is smaller but of the same type. An 4 cone of P shifts backward in the evo­ oblique ridge extends from the metacone lution of the Ursidae; in the present to the inner (protocone) ridge, separating 16 Miguel Crusafont Pair6 and Bji:irn Kurten: Bears and Bear-Dogs

FIG. 6. Ursavus brevirhi­ nus, left lower cheek teeth, Can Uobateres specimen A. 1 1/2 nat. size.

FIG . 7. Ursavus brevirhi­ nus, same specimen as fig. 6. 11/2 nat. size.

FIG. 8. Ursavus brevirhi­ nus, same specimen ~s figs. 6-7. 1 1/2 nat. size.

the talon from the trigon. The protocone but is more elongated, clearly lower itself is only developed as a lingual crowned, and with a relatively longer flexure in the inner ridge. The middle paraconid. As in V. brevirhinus, the field is only faintly wrinkled, and the talonid carries one outer and two inner flat talon, which slopes gently outward, cusps. The talonid basin forms a straight is quite smooth except for the low bord­ fore-and-aft furrow which is continued ering rim and the small protuberance of up to the hind edge of the protoconid, the internal cusp. A small cingulum which thus has a vertical excavation of borders the paracone and qJ.etacone; the this face of the cusp. inner cingulum is somewhat larger. M2 is much larger than in V. brevir­ hinus and its talonid shows a greater de­ velopment of the cusps. There are two Ursavus primaevus (Gaillard) well separated inner cusps and one outer cusp in this part of the tooth. The trans­ For synonymy see THEmus (1949b). verse ridge between protoconid and me­ Material. Right and left M2 (the former de­ scribed and figured by CRuSAFONT 1959), Can taconid is again present and is notched Uobateres. Right lower jaw with Mt-Ms, Can in the middle. As in V. brevirhinus it Ponsich. separates the anterior and posterior ba­ sins, both of which are well developed. M3 is also larger than in V. brevir­ Description hinus and forms a basin rimmed by a slightly raised border on which only the M1 resembles that of V. brevirhinus protoconid has a certain individuality. ACTA ZOOLOGICA FENNICA 144 17

Near the centre of the basin there is a low eminence which is confluent with the protoconid. The two M 2 from Can Llobateres are almost perfect mirror images of each other. They differ greatly from their ho­ mologues in V. brevirhinus in size as well as morphology. In V. brevirhinus this molar is of the bunodont type nor­ mally seen in ursids. The M 2 of V. pri­ maevus, in contrast, has a distinct ten­ dency to lophodonty. The paracone is the highest cusp but is formed by a sinuous longitudinal rid­ ge. A faint crest runs from its summit to the basal external cingulum, while an­ other, still weaker crest leads inward to the middle field, where it blends with FIGS . 9-12: FIG the numerous . 9. Ursavus brevirhinus, :left wrin'kles in this area. The P4, Can Llobateres. 11/2 nat. size; FIG. 10. metacone, which is also ridge-shaped, Ditto, left M1, Can Llobateres. 11/2 nat. size; 2 has a similar crosscrest system. The pro­ FIG. 11 . Ursavus primaevus, right M , Can tocone is a weak protuberance on a low Llobateres. 11/2 nat. size; FIG. 12. Ursa­ vus brevirhinus, left M2 , Can Lkibateres. 1 1/2 ridge, curving from the paracone around nat. size. the front end of the tooth and enclosing the middle field. At the junction be­ break in both the external and internal tween the talon and the trigon there is a ridges; the internal talon cusp is pro-

TABLE 7. Measurements of upper P4 and molars of Ursavus. p4 M1 M 2 L B H L Ba Bp Hp Hm L Ba Bm H Ursavus brevirhinus Can Llobateres 1 , A sin. 12 h 8.6 8.4 13.0 11.5 11.9 13 .7 10.4 9.5 4.6 " " dext. 12.8 8.1 8.2 -a12.0 B 10.3 7.3 6.7 11.6 10.1 10.1 5.8 5.1 12.2 9.9 8.7 4.4 Kieferstadt 12.4 10.6 12.7 10.1 Sreinheim a. Alb. 11.4 8.0 6.7 13.2 10.8 11.1 6.0 5.4 13 .9 10.0 4.8 Ursavus primaevus Can Llobateres 17.2 12.4 11.8 5.6 La Grive a 16.0 11.2 14.4 12.7 16.8 12.1 5.8 15.8 13.3 Oppeln 14.2 9.4 13.8 12.1 6.8 6.0 15.5 11.5 4.9 " 13 .2 9.0 14.2 11.8 6.4 U rsavus de pereti

Sobl~y 19.7 13 .1 18.9 15.7 16.7 8.6 7.7 23.4 16.2 14.6 7.3 21.3 14.5 10.6 19.9 16.3 17.2 8.1 7 .8 21.3 15.8 9.3 Ursavus sp . Soblay 12.9 8.6 8.8 18 Miguel Crusafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs

TABLE 8. Measurements of canines and precarnassial premolars in Ursavus.

pt Ci L B L B H

Ursavus brevirhimts Can Llobateres, A 6.9 4.8 7.6 6.2 4.3 9.2 5.4 5.4 Steyregg, type (After THENrus) 6.6 3.6 8.0 3.9

Ursavus primaevus Oppeln 9.6 5.1 " 9.4 4.9

minent, but there is no outer cusp, and The masseteric fossa extends forward the talon slopes outward as in U. brevir­ to a point below the front end of MJ. hinus. The middle field is intensely The depth of the ramus below M2 is 24.8 wrinkled, mostly in the transverse di­ and below MJ 24.0 mm., measured on rection, while the talon has a system of the internal side. The width of the ramus radiating wrinkles in its posteroexternal beneath M2 is 10.3 mm. part. There is a distinct cingulum at the The size of the teeth compares closely base of the paracone, and a large lingual with those from La Grive (see Tables 7, cingulum extends from the protocone to 9). the talon. The mandibular ramus is markedly curved in shape, probably indicating Notes on the history of Ursavus that the condyle was raised well above the occlusal plane of the cheek teeth. Vallesian members of the genus Ursavus

TABLE 9. Measurements of lower molars of Ursavus.

Mt M~ M3 L B Lt L Ba Bp L B

Ursavus brevirhinus Can Llobateres, A 17.6 8.6 11.6 13 .1 9.1 8.4 8.6 " B 15.0 7.5 9.9 Steyregg, type1 16.2 7.2 11.8 7.4 Goriach1 11 .7 8.4 11.8 8.7 6.8 6.5 Ursavus primaevus Can Ponsich 21.2 9.5 14.7 15.9 9.9 9.2 9.7 8.5 La Grive 20.6 10.4 14 .0 14.8 10.8 8.3 15.1 9.2 8.8 Deinothere Sands 20.4 10.1 Op~ 19.2 10.1 14.4 9.6 9.4 19.4 10.0 14.2 9.6 95 Ursavus depereti Luzinay 11.4 17.6 10. 11.2 Melchingen2 23 .5 12.5 17.3 11.8 10.6

1 After THENIUS. 2 After FRICK. ACTA ZOOLOGICA FE ICA 144 19

FIG . 13 . Ursavus primaevus, right mandible, Can Pon­ sich. 1/1 nat. size. have hitherto been little known. It now by one of us (Kurten) in the Museum of appears that both the typical Vindo­ the University, Munich. bonian species U. brevirhinus and U. The next larger species, U. brevir­ primaevus survived into Vallesian times, hinus, has been found at several loca­ in common with many other Vindo­ lities of Vindobonian age. Original mea­ bonian holdovers. surements were taken by Kurten on ma­ To facilitate comparison, metric data terial from Kieferstadt and Steinheim on various ursavi have been collected am Albuch. The Kieferstadt specimen in tables 7-9. (KOKEN 1888; ScHLOSSER 1899) is a cast By far the largest local sample of a of a right maxilla in the Base! Museum single Ursavus species available at pre­ (NHMB A.S. 88); the Steinheim speci­ sent is that of U. elmensis Stehlin from men, also a cast of a right maxilla, is in the early Burdigalian fissure filling at the same museum (NHMB Sth. 751). Wintershof-West (DEHM 1950). This is The latter was regarded as a subspecies also the most primitive species of the of U. primaevus by DEPERET & GoMEZ genus, and so is a useful standard of (1928) but we regard it as U. brevir­ comparison. The material was studied hinus. In addition, measurements of the

FIG. 14. Ursavus primaevus, ame specimen as fig. 13 . 111 nat. ize. 20 Miguel Crusafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs

type (from Steyregg) annd of specimens from Goriach have been provided by THE IUS (1949b ). Of the late Vindobonian V . primae­ vus from La Grive (DEPERET & GoMEZ 1928), an M 2 was measured in the British Museum ( atural History), BM 5318, while casts of various other teeth were seen in the Munich collection. Here were also measured casts of the specimens from Oppeln ( eudorf) described by WEG ER (1913 ). A lower carnassial from the Deinothere Sands in Gau Weinheim, in the Museum of Mainz, is evidently also V. primaevus. WErTZEL & ToBrE (1952) referred the Deinothere Sands Vrsavus to V. depereti, but they are di­ stinctly inferior in size to that species and agree well with V. primaevus. Since the Can Llobateres and Can Ponsich material proves that this species surviv­ ed into post-Vindobonian (Vallesian) times, there is no reason to refer the W einheim material di fferently. FIG. 15. Ursavus primaevus, same specimen as Vrsavus depereti Schlosser, the Turo­ figs. 13-14. 1/1 nat. size. lian species, is represented by the lower

TABLE 10. Measurements of the dentition of Ursavus elmensis, Wintershof-West. N O.R. M S.D. V P4, length 5 9.5-11.5 10.52 ± 0.33 0.74 7.0 width 5 6.3- 7.4 6.94±0.20 0.45 6.5 height (external} 5 5.7- 6.4 6.14±0.13 0.29 4.7 M1, length 16 9.9-12.3 10.80±0.14 0.56 5.2 width 15 8.8-11.5 10.39±0.16 0.61 5.9 height of paracone 13 4.5- 5.2 4.82±0.06 0.22 4.3 heigth of metacone 13 4.1- 4.8 4.49 ± 0.07 0.24 5.3 M~, length 4 9.2-10.0 9.50±0.18 width 4 8.3- 8.8 8.60±0.12 height of paracone 4 3.4- 3.8 3.55±0.09 P4, length 4 8.0- 8.6 8.32±0.14 width 4 3.5- 4.0 3.85 ±0.12 height (externaJ) 4 4.5- 5.1 4.80±0.17 M1, length 13 12.6-14.3 13.63±0.19 0.69 5.1 width 13 5.3- 6.7 6.22±0.11 0.41 6.6 length of trigonid 10 8.2- 9.7 9.02±0.20 o.64 n height of protoconid 9 5.7- 6.6 6.30±0.11 0.32 5.1 height of hypoconid 11 3.4- 4.4 3.86±0.08 0.28 7.3 M!, length 10 9.7-11.2 10.38±0.16 0.50 4.8 anterior width 9 5.9- 6.7 6.40 ± 0.12 0.36 5.6 posterior width 10 5.6- 6.3 6.04±0.0 0.26 4.4 height of protoconid 6 3.6- 3.9 3.72±0.06 0.15 4.0 height of hypoconid 5 3.1- 3.3 3.20±0.03 0.07 2.2 height of metaconid 6 3.5- 4.0 3.78±0.08 0.19 5.1 Ms, length 3 5.3- 6.9 6.20±0.47 width 3 4.2- 5.4 4.87 ± 0.35 ACTA ZOOLOGICA FE ICA 144 21

TABLE 11 . Average lengths of cheek teeth in samples of Ursavus expressed as percentages of means for U. elmensis.

P• Mt M! p. Mt M2 M3

Burdigalian U. elmensis 100 100 100 lOO 100 100 100 Vindobonian U. brevirhinus 108 119 140 96 119 113 110 U. primaevus 137 135 170 114 145 141 174 Vailesian• U. brevirhinus 114 113 136 111 120 126 139 U. primaevus 181 153 153 156 Turolian U. ehrenbergi 138 153 205 U. depereti 195 180 235 172 167 U. sp. 123 teeth from Luzinay (DEPERET & GoME Z Spanish material. It is now evident that 1928), while the Melchingen data are the two species existed side by side in from PRICK (1926). The upper dentition Europe duri ng a long period of time, is represe nted by the Soblay material in which constitutes definitive proof of Lyon (VIRET & MAZENOT 1949); here is specific d is tinction. also found the small'er Ursavus sp. figur­ A measure of the evolution and dif­ ed by VIRET (1949). feren tiation in Ursavus can be obtained The Turolian species U. ehrenbergi by expressin g the lengths of the cheek (Brunner) is only known from a skull tee th in local and temporal samples as a fragment with the upper dentition, de­ percentage of the corresp onding mean scribed anew by THENIUS (1947). values for the Burdigalian U. elmensis W e do not fo ll ow STROME R (1928, from Wintershof-W es t (see T able 10). 1940) and THENIUS (1949b) in uniting The data are summarized in T able 11 . the tw o species U. brevirhinus and U. In the Vindobonian, the two species primaev us. The combined variation in U. brevirhinus and U. primaevus are these would greatly exceed that in o ther d is tinct but not very strongly separated. species of Ursavus, and the two evident­ Both are larger than U. elmensis and ly represent quite d is tinct foci, in size as have changed so mewhat in relative pro­ w ell as morphology. The sep aration of portions; for ins tance, P4 is relatively the two speci es is evident in the lengths small in both, and M 2 is elongated. So and widths of M 1 and M 2 (see T able 7). is MJ, but only in U. primaevus. Important morphological differences In the Vallesian, the si tuation is were noted in the description of the basically similar, but both species are

TABLE 12 . Average size index for cheek rooth lengths in samples of Ursavus, compared with U. elmen- sis.

Species Locality Age Size

U. depereti Various Turolian 188.5 U. ehrenbergi Euboea Turolian 162.8 U. sp. ob lay Turolian 123 U. primaevus Can Llobateres Vallesian 161.5 U. primaevus La Grive Upper Vindobonian 152.6 U. primaevus Oppeln Lower Vindobonian 135.3 U. brevirhinus Can Llobateres Vaile ian 118.3 U. brevirhinus Various Vindobo ·an 115.7 U. elmensis Wintershof-West Burd.igalian 100 22 Miguel Crusafont Pair6 and Bjorn Kurten: Bears a nd Bear-Dogs

TABLE 13 . Measurements of M2 in early Ursinae.

Ursus (Ruscinian) Perpignan Protursus Baroth- simpsoni Weze Kopecz Lyon Paris

Length 11.8 ea. 181/2 18.4 20.5 20.7 Width, anterior 6.9 11.2 11.2 12.2 12.8 Width, posterior 6.0 11 .0 12.0 13 .2 Height, pr otoconid 3.7 7.6 7.3 7.5 Height, hypoconid 2.6 6.0 7.1

slightly larger. Ursavus brevirhinus now bears from some form more closely re­ shows elongation of MJ and the two lated to the U. brev irhinus branch. species differ a little more in size. If so, however, the derivation pro­ In the Turolian of Soblay, Luzinay, bably dates weH back into pre-Vallesian Melchingen and Euboea a new ·situation times; for, as will be shown below, a is found. An Ursavus of brevirhinus more Ursus-like form was already in type still survives (the U. sp. of Soblay), existence in the Vallesian. but the primaevus type seems to have vanished, replaced by two still larger species, U. ehrenbergi and U. depereti. Protursus Crusafont & Kurten, new The former shows progressive charac­ genus ters in the reduction of P 4 and elonga­ 2 1 tion of M relative to M , while U. de­ Type pecies. Protursus simpsoni, new ·species. pereti is more conservative in relative Diagnosis. Small Ursidae in the Ur­ proportions, though well distinguished savus size range but with talonid of M2 by its large size. more elongate, clearly narrower than Information on phyletic growth may trigonid, and tapering posterad; talonid be summarized by averaging out the with five distinct cusps, of which two percentages for individual teeth (Table external, one posterior, and two inter­ 12). It is suggested that U. primaevus nal, and a fl at middle field; trigonid and U. brevirhinus arose by branching with the three original cusps plus a from U. elmensis. Both species flourish­ fourth, anteroexternal cusp, together en­ ed in the Vindobonian and Vallesian, closing a tran sverse basin . increasing slightly in size. Ursavus bre­ Age. Vallesian. virhinus may have given ri·se to the U. sp. of Soblay. The two large species, U. ehrenbergi and U. depereti, both of Protursus simpsoni which show a tendency to lophodonty Crusafont & Kurten, new spec1es in the upper molars, may have been de­ rived from U. primaevus. This charac­ Type. Left M! (Fig . 16-17). ter, as well as the size and complication Derivatio nornini . We take plea ure in nam­ of P 4 in U. depereti, makes it unlikely ing this pecies for George Gaylord impson, that that species was ancestral to later leader in contemporary paleomammalogy. ursids (VrRET 1949). THE rus (1947) Diagnosis. Sole known spec ies of ge­ considered that U. ehrenbergi may have nus; for dimensions of ty pe see Table been close to the Ursus line. Comparison 13. of dental morphology would however Age and locality. V allesian, Can Llo­ seem to favour a derivation of ursine bateres, Spain. ACTA ZOOLOGICA FE ICA 144 23

Description

The crown is completely prese rved and unworn, except for a wear facet on the summit of the metaconid. The an­ terior root is broken; it was fairly large and expanded transversely. The posterior root is expanded an teropost­ eriorly but c?mpressed laterally, and tapers to a pom t. The outline of the tooth is elongate and somewhat irregular. Its maximum width is at the middle of the trigonid, from where it tapers pestered, but its 7 8 long axis is distinctly curved, so that the outer wall is concave and the inner FIG. 16. Protursus simpsoni, type, left M~, Can convex. The front end is rounded. There Llobareres. 2/1 nat. size. is no cingulum. The cusp pattern is generally bear­ distinct, faintly defined cusps. The like. The protoconid and metaconid, middle field or talonid basin is practi­ which are at the same level, are the cally featureless. largest cusps. The protoconid carries a sharp anteroposterior keel, while the metaconid is more faintly keeled; its Comparison summit is obliterated by wear in this specimen. A transverse ridge from the In comparison with Ursavus this protoconid to the metaconid is inter­ tooth appears progressive because of the rupted by the deep valley between the relatively weak development of the two cusps. The valley connects with the main cusps (protoconid and metaconid), transverse valley in front of the proto­ the great elongation of the talonid, and conid, which in turn is bounded in front the large size of the posterior talonid by the low, transverse paraconid. The cusps. The well-developed posterior paraconid is separated from the meta­ "hypoconid" and "entoconid" are only conid by a notch and from the antero­ represented in Ursavus by small acces­ external cusp by another. The last­ sory elements, while in progressive Ur­ mentioned cusp lies at the anteroexter­ sus they are the major talonid cusps. nal corner of the tooth and so completes The condition in Protursus, in which the the enclosure of the vaHey, while in anterior and posterior cusps are sub­ Ursus there is usually an outlet in this equal is thus intermediate. region. The trigonid thus carries four When compared with a Ruscinian distinct cusps. specimen of Ursus (a jaw from Perpig­ The talonid is elongate, with low re­ nan in the atural History Museum, lief. Externally there are two well-de­ Paris), Protursus appears more primi­ fined cusps, of which the anterior cor­ tive. The talonid is still quite narrow, responds to the main hypoconid of Ur­ while in the Ruscinian specimen it is savus, while the posterior may corre­ actually broader than the trigonid (but spond to the element usually termed a in other Ruscinian Ursus the talonid hypoconid in Ursus. The hind edge of may be narrower than the trigonid: the tooth carries a blunt hypoconulid. see Table 13). The posterior "hypo­ On the inner rim there are again two conid" is quite dominant in the Rusci- 24 Miguel Crusafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs

latter. Its tem poral position is correct for a link of th is kind. It is much young­ er than the earliest known Ursav us, which dat es from the Burdigalian, and may thus be de ri ved from an early form of that g-enus. This would allow all of the Vindobonian, and p erhaps p art of the Burdigalian, for its evolution. A gain it is distinctly older than any true Ursus or any of the tremarctine bears. Both of these groups make t heir appearance in the Ruscinian-H emphillian, so that the­ 13., 14 15 1 re is all of the intervening Turolian to J . account for the evolution of higher ur­ FtG. 17 . Protursus simpsoni, same specimen as sids. fig. 16. 2/1 nat. size. Such inferences, however tempting, must yet remain provisional, for the evidence obtained from a si ngle tooth nian form, and the anterior or original can hardly justify very far-reaching hy poconid may be represented by a conclusions. The peculiar twist of the small section of the outer rim just be­ tooth axis may, for instance, indicate hind the protoconid, bounded by two that the C an Llobateres species was off small notches. On the other hand, the the main line of ursine evolution. Should two entoconid elements are s till of this be so, perhaps another sp ecies of about the same size in the Ruscinian Protursus was the true ancestor of Ursus. bear, just as in Protursus. The trigonid It is therefore necessary to await the of the Ruscinian form is more reduced discovery of additional material. Mean­ in front, and the anteroexternal cusp, while, the new genus Protursus may be which is presented in Protursus, has provisionally referred to the subfamily been lost. (It is however retained in Ursinae as its earliest and most primi­ H elarctos.) tive representative. The twist of the tooth axis, which is so pronounced in the Protursus speci­ men, cannot be matched in Ursavus or Amphicyon cf. maJor D e Blainville Ursus. The length of the specimen is the same Material. Left and right M2 , left C , right as in Ursavus brevirhinus, but it is con­ humerus, Can Llobateres. Left C;, Can Ponsich. siderably narrower and probably came from a somewhat smaHer ahimal, per­ haps about the size of U. etinensis (in D escription which the width is about the same). It The second molars from Can Lloba­ is less than two- thirds the size of the teres, which are only slightly worn, re­ Ruscinian Ursus M2 (Table 13). semble entirely the specimens from Sansan described and figured by G1 s­ Phyletic posi tion BURG (1961 : pi. II fig. 1 pi. IV fig. 1 ). The dimensions of the two specimens Protursus, being more advanced than are 23.0 X 17.2 (right) and 23.6 X 17.1 Ursavus and more primitive than Ursus, (left). Sansan lengths are 24.5 and 25.0 may represent a s tep in the sequence mm. leading from the former genus to the The lower canines differ from those ACTA ZOOLOGICA FENNICA 144 25 in Indarctos in several characters. The root is fl atter a nd shorter, there is no internal cingulum and the inner ridge is stronger, more nearly s traight and close r to the midline of the tooth; the crown curves sharply outward. The same characters are seen in the S ansan material. The canine from Can Llobateres, wh ich is much worn, measures 21.5 X 15.0 at the base of the enamel. The cor­ responding data for the Can Ponsich tooth a re 19.5 X 13.2. The latter tooth is unworn, but the tip is broken . The morphological resemblance is so com­ plete that we tentatively r ecord the sp e­ cies from Can Ponsich on the b asis of

Ftc. 19. Cf. Amphicyon, same specimen as fig . 18. Ca . 3/8 nat. size.

this canine, but it is also possible that the canine belongs to th e other canid to be desc ribed below. D imensions of C; from San san cited by G1 SBURG are again slightly larger (lengths 23.4 and 26 mm.). Can Llobateres has also yielded a humerus which may belong to the same species as the teeth. The proximal end is damaged so that only about three­ fourths of the bone is left; the preserved pan is about 250 mm. long. The least transverse width of the shaft i'S 28 .0, while the greatest distal w idth is app­ FtG . 18. Cf. Amphicyon, right humerus, Can roximately 90 mm. 3/8 nat. size. The size of this bone, wh ich exceeds 26 Miguel Crusafont Pair6 and Bjorn Kurten: Bears and Bear-Dogs that in I. vireti, shows that it belonged (2) The humerus from Can Lloba­ to a large carnivore, while its morpho­ teres is Hemicyon, while the teeth are logical charact·ers preclude reference to Amphicyon; in this case the determina­ the other great predator of this locality, tions in HDRZELER's study should be Machairodus aphanistus Kaup. Since exchanged. this humerus is of the paraursoid typ e (3) The humerus from Sansan ascrib­ (as defined in VrLLALTA & CRUSAFONT ed by Ginsburg to Amphicyon is in 194 3) it would seem na rural to refer it reality Hemicyon, in which case, again, to the Amphicyon. However, its cha­ the humerus from Can Llobateres would racters differ considerably from those be Amphicyon like the teeth. found in the Sansan A. major as de­ We favour alternatives (2) and (3) scribed by GrNSBURG (1961: fig. 9, pl. but are unable to choose between them. XVII fig. 1, pl. XVIII fig. 1 ). The deltoid ridge is powerfully de­ According to the figures and desc rip­ veloped in the Can Llobateres humerus, tion, the olecranon fossa in A. major and runs down to the lower third of the is small, elongate, deep and triangular bone; in this respect the resemblance to in shape. In our specimen it is larger, A. major (as figured by Gr SBURG) is rather shallow, short and broad: its great. On the other hand, marked diffe­ shape is that of an ·equilateral triangle rences are seen in the supinator ridge. rather than the isosceles triangle on a Its alar portion is relatively shorter narrow base in the Sansan form. It is of than in the Sansan form, with a notable course possible that the intra- or in ter­ development of the part corresponding specific variation in the genus may to the extensor carpi; while the area of suffice to accommodate both variants, origination of the extensor communis for this character in our specimen is quite digitorum on the external side forms a similar to that in a humerus of A. cf. marked projection not visible in the lemanensis Pomel described and figured Sansan humeri. by HDRZELER (1945). Oddly enough, The pan corresponding to the entepi­ the humerus of H emicyon sansaniensis condylar ridge is poorly preserved, and figured in the same paper by Hiirzeler the foramen cannot be studied. The dis­ shows a fossa of the same type as that tal articulation is horizontal and cylind­ in A. major described by Gr SBURG. rical in shape, as normally in ursids and GrNSBURG (1961) does not ascribe any paraursoid dogs. The surface for the postcranial material from San an to insertion of the m. coracobrachialis, al­ Hemicyon. though not completely preserved, seems The following alternative explana­ relatively large. tions of this situation are suggested: The size of this specimen is only (1) The variation within Amphi­ slightly inferior to that in the humeri cyon accommodates both types of hu­ from Sansan, for which GINSBURG gives meri, in which case the humerus from the length of 333 mm.; it agrees fairly Can Llobateres may belong to the same well with the size of A. crassidens de­ species as the teeth. scribed and figured by H ELBr G (1929).

TABLE 14. Measurements of canid calcaneus from Can Pon ich.

Greatest length from tuber to cuboid facet ...... 93.2 mm . Greatest depth ...... 39.8 Greatest width at level of posterior a tragalar facet ...... 39.2 Distance from upper border of anterior facet to cuboid facet ...... 49.8 Least width of tuber calci ...... 16.3 Depth of tuber calci ...... 28.3 ACTA ZOOLOGICA FE ICA 144

TABLE 15. Measurements of caudal vertebrate.

Canid'

Length 52.5 39.4 45 .6 Least width at middle 15.3 10.0 8.0 Least depth at middle 14.2 8.3 7.2

Unfortunately Helbing's figure does not ductor of the fifth digit is weakly de­ show the olecranon fossa. veloped, in contrast with Amphicyon. Since the characters of this bone do The cuboid facet is triangular in not agree well with the material from shape but much narrower and deeper Sansan, there remains some uncertainty than in A. major. The posterior articular as to its taxonomic status and indeed as facet for the as tragalus has a shorter to that of the dental material. For this radius of curvature than in A. major. reason, our reference to the genus and The groove for the insertion of the ac­ species Amphicyon major must remain cessorius of the m. perforans on the in­ ten tative. ner face is well marked, but does not extend back to the tuber calci as in the Sansan form. The anterior astragalar Canidae indet. facet is much longer than in Amphicyon, and its extension reaches the border of Material. Right calcaneus, caudal vertebra, the cuboid facet, as in the hemicyonines. d.istal end of metapodium, all from Can Ponsich. The bone thus differs greatly from the type seen in A. major and A crassi­ dens as well as the ursids; the great Description length from the anterior astragalar facet to the cuboid facet is especially note­ In addition to the canid C tentative­ worthy. The combination of characters ly attributed to Amphicyon major, a in the specimen indicates that its pos­ few postcranial bones of a large canid sessor was digitigrade rather than plan­ have been unearthed at Can Ponsich. tigrade as the amphicyonids. There are The most significant specimen is the various incompletely known large ea- right calcaneus, which is certainly not A . major and probably not amphicyo­ nine at all; it is entirely different from the Sansan material (G1 SBURG 1961 ) and from the A. major calcaneus found at Burro! (CRuSAFONT & TRuYoLs 1957). The bone is long and slender in pro­ portions, with a narrow and elongate sustentacular part quite different from that in Amphicyon. The groove for the tendon of the flexor digitorum is deep and well developed, as in primitive ca­ nids generally, while Amphicyon is bea r-like in this respect. The posterior end of the tuber calci is compressed not expanded as in A. major, and the FIG. 20. Canidae indet., right calcaneu , Can insertion on the plantar si de for the ab- Pon ich. Ca . 9/10 nat. ize. 28 Miguel Crusafont Pair6 and Biorn Kurten: Bears and Bear-Dogs

nids from the Vi ndobonian and Pontian CRUSAFO T, M. & PETTER, G. 1969: Contri­ bution a !'etude de Hyaenidae. La sous­ that m ight fill the b i ll , for insta nce famille des Ictitherii nae . - Ann. Paleont. Agnotherium or some other me mber of 55:89-128. . the Thau mastocyoninae (Hu RZELER CRuSAFO T, M. & TRUYOLS, ]. 1947: Sobre el 1940). descubrimiento de un nuevo yacimiento del Me6tico en el Valles. - Bol. Inst. The dimensions of the c alcaneus are Geol. Min. E paila 60:71-107. listed in tab le 14. - »- 1951 : Hallazgo del Plesiodimylus chantrei The caudal vertebra evidently c omes Gaill. en el Me6tico del Valles. - otas from an animal of a bout the same size Corn. In t. Geol. Min. E pai'ia 22 :3-30. as the calcaneum. It is much lar ger than -»- 1957 : Descubrimiento del primer yaci­ miento de mamiferos miocenicos de la the specimens of I. vireti from Can Llo­ cuenca valenciana. - otas Corn. In t. bater es (Tab le 15 ). The d is tal fragment Geol. Min. Espai'ia 48:1-20. of metapod ial also b elonged to a carni­ -•- 1960. obre la caracterizaci6n del Valle­ vore of this size; its gr eatest transverse siense. - otas Corn. Inst. Geol. Min. diameter is 22.2 mm. Espaiia 60:109-126. CRuSAFO T, M. & VILLALTA , ]. F. 1951: Los nuevo mamiferos del neogeno de Es­ pai'ia . - otas Corn . Inst. Geol. Min. R eferen ces Espaiia 22:1-25. D EH~1, R. 1950: Die Raubtiere aus dem Mirtel­ CRuSAFO T, M. 1950: La cuesti6n del llamado Miozan (Bu rdigalium) von Winter hof­ Me6tico espaiiol. - Arrahona 1950 ( 1): We t bei Eich tart in Bayern. - Abh. 1-10. Sabadell. Bayeri chen Akad. Wiss., Math.- at. Kl., -•- 1958a: Endemism and Paneuropeism in ( .F .)58: 1-141. Spanish fossil mamma·lian fauna , with DEPERET, Ch. & GoMEZ, F. 1928: Sur l'Indarctos special regard to the Miocene. - Com­ arctoides et la phylogenie des ur ides. - ment. Bioi. Soc. Sci. Fennica 18:1-31. Bull . oc. Geol. France, (4) 28:149-160. - »- 1958b: uevo hallazgo del pongido valle­ PRICK , C. 1926: The Hemicyoninae and an Ame­ siense Hispanopithecus. - Bioi. Inform. rican Tertiary bear. - Bull. Amer. Mus. AEPV 13-14:37-44. at. Hist. 56(1): 1-119. -•- 1959: Primer hallazgo en Espaiia del ge­ GINSBURG , L. 1961: La faune des carnivores nera Ursavus (Carnivora, Ursidae).- No­ miocenes de Sansan (Ger ). - Mem. Mus. tas y Corn. Inst. Geol. Min. Espaiia 55: at. H i t. at., n.s. C9: 1- 190. 137- 144. GoLPE, J. M. 1971 : uiformes del Terciario es­ - •- 1964: La biota de Can Llobateres (Saba­ paiiol y sus yacimientos. - Paleont. y del1) y su signilicaci6n biol6gica. - CUis. Evol. 2:1-197. y Con£. Inst. Lucas Mallada 9:177-179. lliRTE BERGER, J. L. 1965: Les Cricetidae de - •- 1965a: Observations a un travail de M. Can Llobatere ( eogene d'Espagne). - Freudenthal et P. Y. Sondaar sur des Bull. oc. Geol. France (7)7:487-498. nouveaux gisements a Hipparion d'Espag­ - •- 1%6: Le rongeur du Vallesien (Mio­ ne.- Proc. Kon. Akad. Wet. Amsterdam, cene uperieur) de Can Llobateres ( aba­ B 68 (3):121- 126. dell, Espagne): Gliridae et Eomyidae. - -»- 1965b: El desarrollo de lo caninos en al­ Bull. oc. Geol. France (7)8:5%--604. gunos driopitecidos del vallesiense de Ca­ HELBING , H . 1929: Zur Osteologie van Amphi­ taluna. - Notas Corn. Inst. Geol. Min. cyon crassidens Pomel. - Verh. aturf. Espaiia 80: 179- 192. Ges. Base! 40:280--294. -»- 1972a: Los P rimates f6sHes de Espaiia. ­ Hi.iRZELER, J. 1940: Ober felinoide Caniden des Enciclopedia Mundo. Barcelona. europai chen Miocaen . -Verb. chweiz. - »- 1972b: Les Ischyrictis de la transition aturf. Ge . 1940:150-151. Vindobonien-Vallesien. - Palaeovertebra­ - •- 1945: Zur Kenntni des Extremitaten- ta 5: 253-260. skelettes einiger oligocaener und miocae­ CRUSAFO T, M. & CAsA ovAs, L. 1973: Mamma­ ner Carnivoren Europa . - Eel. Geol. lia Tertiaria Hispaniae. - Fo s. Catal. I. Helv. 38:635-655. Animalia 121:1-198. Kou , E. 18 : Ueber die miocanen augethier CRuSAFONT, M. & Hi.iRZELER , J. 1961: Les pon­ -Re te van Kieferstiidtl in Oberschle ien gides fossiles d'Espagne. - C. R. Acad . und iiber Hyaenarctos minutus hlo er Sci. Paris 252:919-920. M . - irzber. Ges. aturf. Fr. Berlin -»- 1969: Catalogo comentado de Ios P6ngi­ 1888:44--49. dos f6siles de Espaiia. - Acta Geol. His­ KuRTE , B. 1955: Sex dimorphism and size panica 4:44-48. trends in the cave bear, Ursus spelaeus ACTA ZOOLOGICA FE ICA 144 29

Rosenmu1ler and Heimoth. - Acta Zool. _ ,._ 1949b: Die Camivoren von Gi:iriach (Stei­ Fennica 90:1-48. ermark). Beitrage zur Kennrnis der Sauge­ 1966 : Pleistocene bears of North America. tierreste des steirischen Tertiars IV. - 1. Genus Tremarctos, spectacled bears.­ Sitzber. Akad . Wiss. Wien 158:695- Acta Zool. Fennica 115:1-120. 762 . MERRIAM, J. C. & STOCK, C. 1925: Relationships _ ,. _ 1959: I ndarctos arctoides (Carnivora, and structure of the short-faced bear, Mammalia) aus dem Pliozan Osterreichs Arctotherium, from the Pleistocene of nebst einer Revision der Gattung. - N. California. - Carnegie Inst. Washington Jahrb. Geol. Pal., Abh. 108:270-295. Pub b. 347:1-35. TosiEN, H. 1955: Neue und wenig bekannte Car­ P ETTER, G. 1963: Contribution a l'etude des nivoren aus den unterpliozanen Dinothe­ mustelides des bassins neogenes du Val­ riensanden Rhei nhessens. Notizbl. les-Penedes et de Calatayud-Teruel. - Hess. Landesamt. Bodenforsch. Wiesba­ Mem. Soc. Geol. France n.s. 42(97 ): 1- den 83:7-31. 44. VA CouVERING, ]. A. 1972: Radiometric calib­ -»- 1967: Mustelides nouveaux du Vallesien ration of the European Neogene. - In: de Catalogne. - Ann. Paleont. 53:93- BISHOP, w. w. & MILLER, J . A. (ed.): 114. Calibration of hominoid evolution: 247- ScHLOSSER, M. 1899: Ober die Baren und baren­ 271. Edinburgh. ahnlichen Formen des europaischen Ter­ VILLALTA,]. F. & CRUSAFONT, M. 1943 : Indarc­ tiars. -Palaeontographica 46:95-147. tos vireti, nova sp. un nuevo ursido del SIMO s, E. L. & PILBEAM, D. R. 1965: Prelimi­ grupo de !os hemici6nidos del Mioceno nary revision of the Dryopithecinae (Pon­ superior del Valles-Penedes. - Consejo gidae, Anthropoidea). -Folia primat. 3: Sup. Invest. Cien. Madrid 194}:45-62. 81-152. 1945: uevas aportaciones al conocimien­ STROMER , E. 1928: Wirbeltiere im obermiozanen to de !os carnivoros potienses del Valles­ Flinz Munchens. - Abh. Bayer. Akad. Penedes. - Miscellanea Almera, Pub!. Wiss. 32(1): 1-71. Inst. Geol. Barcelona 7:81-124. -»- 1940: Die jungtertiare Fauna des Flinzes VIRET, ]. 1949: Observations complementaires und des Schweiss-Sandes von Munchen. sur quelques mammiferes fossiles de Sob­ Nachtrage und Berichtigungen. - Abh. lay. - Eel. Geol. Helv. 42:469-476. Bayer. Akad. Wiss. n.s. 48:1-102. VIRET, J. & MAZENOT, G. 1949: Nouveaux res­ THALER, L. 1966: Les rongeurs fossiles du Bas­ res de mammiferes clans le gisement de Languedoc clans leur rapports avec l'his­ lignite pontien de Soblay (Ain). - Ann. toire des faunes et la stratigraphie du ter­ Paleont. 34:1-42. tiaire d'Europe. - Mem . Mus. Nat. Hist. WEGNER, R. T_ 1913: Tertiar und umgelagerte at. n.s. C 17:1-295. Kreide bei Oppeln (Oberschlesien). - Pa­ THEN IUS, E. 1947: Ursavus ehrenbergi a us dem laeontographica 40:175-274. Pont von Euooa (Griechenland) . - Sitz­ \YhrrZEL, K. & Tos!EN, H. 1952: Indarctos und bet. Akad. Wiss. Wien 156:225-249. Ursavus (Carnivora, Mamm .) aus den un­ -»- 1949a: Ober die Gehi:irregion von Indarc­ terpliozanen Dinotheriensanden Rhein­ tos (Ursidae, Mamm.). - Sitzber. Akad. hessens. - otizbl. Hess. Landesamt. Bo­ Wiss. Wien 158:647-653. denforsch. Wiesbaden 6(3):7-14.

Appendix : Abbreviations

The following abbreviations appear in the posterior, Lt trigonid length, Hp paracone tables of measurements: height, Hm metacone height. All cusp heights L length, B breadth, H height, a anterior, p measured from enamel base to summit.

126. ANN-MARm FoRSTEN: Variation in and between three populations of Mesa­ hippus bairdii Leidy from the Big Badlands, South Dakota. 16 pp. (1970). 127. BENGT WESTERLING: Rumen ciliate fauna of semi-domestic reindeer (Rangifer tarandus L.) in Finland: composition, volume and some seasonal variations. 76 pp. (1970). 128. WoLFRAM NoooT: Zur Okologie der Copepoda Harpacticoidea des Kiistengebietes von Tvarminne (Finnland). 35 S. (1970). 129. R. B. ANGus: A revision of the beetles of the genus Helophorus F. (Coleoptera: Hydrophilidae). Subgenera Orphelophorus d'Orchymont, Gephelophorus Sharp and Meghelophorus Kuwert. 62 pp. (1970). 130. ELAINE ANoERSON: Quaternary evolution of the Genus Martes (Carnivora, Musteli­ dae). 132 pp. (1970). 131. BJORN KuRTEN: The Neogene wolverine Plesiogulo and the origin of Gulo (Carnivora, Ma=alia). 22 pp. (1970). 132. C.-H. von BoNSDORFF, T. FoRssTEN, M. K. S. GusTAFSSON and B.-J. WrKGREN: Cellular composition of plerocercoids of Diphyllobothrium dendriticum (Cestoda). 25 pp. (1971). 133. RAINER RosENGREN: Route fidelity, visual memory and recruitment behaviour in foraging wood ants of the genus Formica (Hymenoptera, Formicidae). 106 pp. (1971). 134. GoRAN GnLENBERG and A. J. P. GoRE: Some preliminary models for the energy flow of a Chorthippus parallelus (Zett.) (Orthoptera) population. 25 pp. (1972). 135. BRIT GoosKE BJORKLUND: The rotifer fauna of rock-pools in the Tvarminne archipelago, Southern Finland. 30 pp. (1972). 136. MARTIN MEINANDER: A revision of the family Coniopterygidae (Planipennia). 357 pp. (1972). 137. ANN-MARrE FoRsTEN: Size and shape evolution in the cheek teeth of fossil horses. 31 pp. (1973). 138. M. K. S. GusTAFSSON: The histology of the neck region of plerocercoids of Triaeno­ phorus nodulosus (Cestoda, Pseudophyllidea). 16 pp. (1973). 139. H. SrLFVERBERG: A revision of the genus Prosmidia Weise (Coleoptera, Chrysomeli­ dae). 54 pp. (1973). 140. BJORN KuRrtN: A history of Coyote-Like Dogs (Canidae, Mammalia). 38 pp. (1974). 141. KARt VEPSALAINEN: The life cycles and wing lengths of Finnish Gerris Fabr. species Heteroptera, Gerridae). 73 pp. (1974). 142. GoRAN BYLUND: The taxonomic significance of embryonic hooks in four European Diphyllobothrium species (Cestoda, Diphyllobothriidae). 22 pp. (1975). l 143. RA uNo LrNNA vu oR I: Revision of the Cicadellidae (Homoptera) of the Ethiopian Region Ill. Deltocephalinae, Hecalini. 37 pp. (1975). 144. MIGUEL CRuSAFONT PAIR6 and BJORN KuRTEN: Bears and Bear-Dogs from the Vallesian of the Valles-Penedes Basin, Spain. 29 pp. (1976). tngln Y I opi ston R··· kh:· to

Exchange- Austauscb - Echange SOCIETAS PRO FAUNA ET FLORA FENNICA Snellmanin.katu 9-11 Snellmansgatan 00170 Helsinki 17 Helsingfon

For sale- Verkauf- En vent Akateeminen Kirjakauppa - Akademiska Bokhandeln 00100 Helsinki 10 Helsingfon

ISBN 951-661-013-7 ISSN 0001- 7299 PRINTACO