A newarceoscelid , Zarcasaaras tanyderus, fromthe Gutler Formation (Lower ) of north-centralNew Mexico byDonald B. Brinknan,Curator, Tyrrell Museum of Palaeontology,Drumheller, Alberta, David S Berman,Associate Curator, Section of Vertebrate Fossils, Carnegie Museumoi NaturalHistory, Pittsburgh, PA, and DavidA. Eberth,Graduale Student, Department ol Geology, University of Toronto, Toronto, 0ntario A new and species of araeoscelid that Petrolacosauruswas related to the cap- , and are interpreted as an adaptation reptlle, Zarusaurus tanuderus, is described torhinomorphs.It alsohas been argued thlt to a specializeddiet that probably included on the basisof a partiai jaw and disarticu- Petrolacosaurusbelongs to a group of archaic protectedby a heavy exoskel- postcranial invertebrates lated elements from the Cutler mammal-likerep- importantlv this theory is ex- (Lower edaphosaursof the early eton. Most Formition Permian) of north-cen- a lower tral New Mexico. It is the first reoresenta- tile (Stovall, Price, and Romer, tendedto suggestthat the absenceof tive of this family to be reported from that 1956).Most recently,a redescriptionof Pe- temporal fenestra in Araeoscelismay be the state, and it is considered closely related to trolacosaurusbv Reisz0.977 , 1981\has estab- result of a secondaryclosure. These conclu- .Z. tanyderus possesses a com- lishedfirmly 6oth its positionas a primitive sions prompted a somewhat revised classi- bination of primitive and derived features diaosid and its close relationship with Ar- fication in which Araeosceliswas assignedto with respeci to Araeoscelis, indicating that aeoicelis.Previously undescribed specimens the AraeoscelidaeWilliston, 1910, and Petro- they could not have had a direct ancestor- of Araeoscelispresently under study (Reisz, lacosaurusto the PetrolacosauridaePeabody, descendant relationship. Berman,and Scott,in preparation)have pro- 1952;both families were united under the vided additionalinformation that greatlyex- suborderAraeoscelidia Williston, 1913.Other pands upon our assessmentof the generathat have been included in this sub- Introduction phylogenetic relationships of Araeoscelis,a order, but which are consideredtoo poorly The phylogeneticpositions of Araeoscelisand summary of which can be given here. Cur- known to be assignedto either famiiy, ar-e Petrolacosaurusare important to our under- rent evidenceindicates that Araeoscellsshares I(adaliosaurusCredner, 1889, and Dictybolus standingof the early diversificationof diap- a more recent common ancestorwith-that Olson, 1970. sid .Araeoscelis is a small,lightly built is, it is more closelyrelated to--Petrolacosau- The new araeosceliddescribed herc as Zar- reptile from the Lower Permianof Texasand rus than with the captorhinomorphs,and that casaurustanyderus consists of a partial, dis- Oklahoma (Vaughn, 1955;Simpson, 1979). both genera share a more recent common articulatedskeleton collected by the authors In the most recent detailed descrintion of ancestorwith eosuchiansthan with the cap- during the summer of 1982from the Cutler Araeoscelis,Vaughn (1955)concluded that it torhinomorphs. Formation (Lower Permian) near the village is bestconsidered a -likeexperiment of Together, Araeoscelisand Petrolacosaurusof Arroyo del Agua in Rio Arriba County, the captorhinomorph reptiles. He representthe earliest,most primitive known north-centralNew Mexico (Fig. 1). The dis- argued that, exceptfor such featuresas the stagein theadaptive radiation of the covery was reported in a preliminary paper upper temporal fenestra,specialized denti- reptiles. This relationship is strongly sup- on the sedimentology and paleontology of tion, elongatelimbs, and elongate cervical ported by their postcranialskeletons that ex- the Lower Permianred-bed deposits of north- vertebrae,Araeoscelis resembles in most re- hibit only minor differences,but, most central New Mexico (Eberth and Berman, spectsthe primitive membersof the group. importantly, they share numerous derived 1983).Z. tanyderusrepresents the first record On the basis of better preservedearly cap- feafures.Cranial differencesbetween the two of an araeosceloidfrom New Mexico. It is torhinomorphs, Carroll (1959)agreed wiih genera,on the other hand, aremore marked. more closely related to Araeoscelisthan to Pe- Vaughn's assessment.Romer (1956, 1966, They are recognizedas providing Araeoscelistrolacosaurusand is, therefore, assigned to 1967),on the other hand, placed Araeoscelis with a more massive, sturdily constructed the Araeoscelidae. with a number of poorly known, early (Per- mian and ) forms generally referred to asprotorosaurs, which he regardedas an- cestralto the sauropterygians(plesiosaurs, nothosaurs,and placodonts).The protoro- saurs,as defined bv Romer,are now believed to be a ratherdiverse assemblage of distantly related forms (Kuhn-Schnyder, 1967; Brink- man, 1981;Chatterjee, 1980; Carroll , 7981), and there is little evidenceto suggestthat Araeoscelisis closely related to any of them (Vaughn,1955). The controversyof the phylogeneticaffin- ities of Araeoscelishas seemingly been re- solved with the demonstration of its close relationship with ,a diapsid reptile of lizard-likeproportions from the Late Pennsylvanianof Kansas.Petrolacosaurus was originally describedby Peabody(1952), who believedit to be closelyrelated to the eosu- chians,the most primitive group of diapsid reptiles, with strong ties with the captorhi- nomorphs.Vaughn (1955)recognized the close similarity between Araeoscelisand Petrolaco- L--.-- saurusand even suggestedthat they may be- Iong to the same family, thereby implying FIGURE l-Locality of Zarcasnurustanyderus, Rio Arriba County, north-central New Mexico.

May 1984 Nau MexicoGeology Systematicpaleontology of an ancient river svstem. The limited ex- TABLE 1-Measurements (in mm) of vertebrae of Class Rsprrlre tents of the crevasse channels suggest the Zarcasaurustanyderus gen. et sp. nov., CM 47704. 2 is the height SuborderAnaeoscrlron Williston, 1913 proximity of interchannel pond and./or marsh f. is the greatestlength of centrum, the centrum at the anterior end, 3 is the trans- Family AnaroscEuoanWilliston, 1910 environments where such channels would of width of the centrum at the anterior end, 4 Zanceseunus, new genus verse quickly lose their U-shape morphology. is the height of the centrum at the posterior end, Tvpr spnclss:Zarcasaurus tanyderus nsp. DIacNosrs-Same as for the genus. 5 is the transverse width of the centrum at the Eryuor-ocv-Refers to the Agua Sarca, a posterior end, and 5 is the length of the neural smallcreek about 3 km northwest of the type Description and comparison arch pedicle. locality;pronounced Zarca and spelledas such All of the elements assigned to the holo- on early maps and publications (Langston, type, CM 41704,werc collected from the sur- 1es3). face and had eroded out of a single slope. Axis 2.8 3.5 7.5 DracNosrs-A small araeoscelid closely re- Although only two elements (dorsal verte- Midcervical 77.2 4.0 4.6 3.5 3.8 8.5 lated to Araeoscelisand characterized by the brae) were articulated, all the elements were Posteriorcervical 8.6 3.5 3.7 3.1 3.9 4.3 following derived features: cervical verte- found closely associated. None of the parts Dorsal 5.9 2.8 3.9 3.2 3.9 3.4 brae are longer, with central lengths exceed- is duplicated and all are of an appropriate Dorsai 5.0 3.4 3.5 3.3 4.0 3.4 ing those of the dorsals by as much as two size to have belonged to a single individual. times; absence of a prominent notch on the No bones were found associated with CM distal edge of the astragalus that separates 41704 that would indicate the presence of the articular facets for the fourth distal tarsal another form of similar size. For these rea- tentative, as is identification of the prearti- and centrale. Primitive features that distin- sons, CM 47704is considered to represent a cular-angular contact near the ventral mar- gdsh Zarcasaurusfrom Araeoscelisinclude: the single individual that probably was repre- gin of the jaw fragment. In its possession of humerus lacks ectepicondylar foramen; a sented originally by a greater part of artic- a precoronoid, Zarcasauruswould more closely supracondylar process is present on the pos- ulated skeleton. approach Araeoscelis than Petrolacosaurui. terior condyle of the femur; a concave area Lowrn 1ew-The lower jaw is represented There is no evidence of a postsplenial in Zar- of finished bone separates the lateral and me- by a short section of the left ramus just an- casaurus.The postsplenial is absent in Ar- dial articular surfaces on the proximal end terior to the adductor fossa (Fig. 2). It in- aeoscelis.but is present in Petrolacosaurusas of the tibia; the tibial facet of the astragalus cludes the posterior end of the dentary, which a moderate-sized element wedged between is convex and not developed into a ridge- includes one emptv socket and two incom- the angular and prearticular at the level of and-trough pattern. plete teeth. The more complete tooth is miss- the coronoid. ing only its tip; this is possibly a result of VnnrpnnaE-Three cervical and five dorsal Zarcasaurus tanyderus, new species wear. The tooth is stout, like those of 4r- vertebrae of varying degrees of complete- Ervrr.rorocv-Derived from the Greek tany, aeoscelis,rather than like the slender teeth of ness and fragments of centra comprise the long, and deros, neck, referring to the spec- Petrolacosaurus,though it lacks the lateral, in- vertebral elements. Two of the dorsal ver- imen's very long . cipient cusp seen in the cheek teeth of Ar- tebrae were preserved in articulation. The Holorypr-CM 41704 (collections of the aeoscelis.Little of the bases of the teeth can three cervicati Gig. 3A-C) include an axis, Carnegie Museum of Natural History, Sec- be seen in the lateral view of the jaw due to midcervical, and posterior cervical. Deter- tion of Vertebrate Fossils) is a partial, dis- the low level of implantation of the teeth on mination of their serial positions has been articulated skeleton consisting of the following the medial surface of the dentary. In this fea- based on successive changes in the cervical elements: a short section of the left ramus of ture, Zarcasaurus mote closely approaches series of Arneoscelis and Petrolacosaurus. the lower jaw just anterior to the adductor Petrolacosaurusthan Araeoscelis. As in Araeoscelis and Petrolacosaurus, the fossa; three cervical, including axis, and five Identification of the sutures on the lateral cervical vertebrae of Zarcasaurusare conspic- dorsal vertebrae, and fragments of centra; surface of the jaw fragment presents no se- uously elongate. The centrum of the mid- the distal ends of both humeri; the proximal rious problems. However, with the excep- cervical, the longest and most complete of and distal ends of the left femur and a fras- tion of the coronoid, the interpretation of the the three cervicals, is 1.9 and 2.2 times the ment of the proximal end of the right femui; sutures on the medial surface is uncertain. length of the longest and shortest of the pre- the proximal and distal ends of both tibiae; Though the coronoid of Zarcasaurusis incom- served dorsals (Table 1). The neural spine, the left astragalus; several metapodials and plete, its relative depth, as in Araeoscelis,was nearly preserved in its entirety only in the phalanges; and numerous indeterminate obviously much greater than in Petrolacosau- midcervical. is verv low. In the axis ahd mid- fragments. ras. Identification of a narrow splint of bone cervical vertebrae, the pedicle of the neural Honrzol AND LocALITv-Cutler Forma- contacting the anterior end of ihe coronoid arch extends about three-fourths of the length tion (Lower Permian: Wolfcampian), about 1 as the anterior coronoid, or precoronoid, is km southeast of the small village of Arroyo del Agua in Rio Arriba County, north-central New Mexico (exact locality data on file at the Carnegie Museum of Natural History). The holotype, as is the case with the majority of vertebrate fossils from the of the Arroyo del Agua area, is from the ex- tensive, well-known deposits commonly re- ferred to as lake or pond deposits (Langston, 1953). Preliminary investigations of these deposits (Eberth and Berman, 1983), as well as adjacent lithofacies, have revealed that they are discrete, U-shaped bodies of interbedded mudstones and minor sandstones and have a very limited extent. Their close lateral as- sociation with U-shaped, low-sinuosity channel fills consisting of vertically accreted, rcM medium to coarse sandstones more properly FIGURE 2-Zarcasaurus tanyderus gen. et sp. nov., CM 47704. A is the medial and B is the lateral view suggests that they are actually remnants of of a portion of the left ; angular (A), coronoid (CO), anterior coronoid (COA), dentary (D), crevasse channel fills in an anastomosed reach preaiticular (PRA), and surangular (SA) indicated.

Nm Mexico Geology May 1984 of the centrum, and in the posterior cervical Severalfeatures demonstrate that the cerv- man, and Scott, in preparation) and Petro- the pedicle is restrictedto the anterior half icalvertebrae of Zarcasaurusmore closelyap- Lacosaurus(Reisz, 1981), but in other Paleozoic of the centrum.The zygapophysesof the axis proachthose of Araeoscelis(Fig. 3D) than those tetrapodsas well (Vaughn,1970). The dorsal and midcervicalare narrowly separatedand of Petrolacosaurus(Fig.3E). The most notice- neural arches of Zarcasaurusand Araeoscelis extend very little beyond the lateral margin able feature is that the neural arch pedicles differ from those of Petrolacosaurus:they are of the centrum. The axial prezygapophyses of the axisand midcervicalvertebrae of Zar- slightly wider, have a somewhatswollen ap- aremuch smallerthan those of the midcerv- casaurusand the cervicals of Araeoscelisex- pearance,and they possessmuch lower neural ical, facenearly as much laterallyas dorsally, tend for about three-fourthsof the length of spines.None of the neural spinesshow any and the facetsare convex. The prezygapo- the centrum. In PetrolacosaLffus,on the other signsof the mammillary processesthat have physes of the midcervical extend beyond the hand, the pediclesof all the cervicalsare re- been described in Araeoscells(Vaughn, 1955) end of the centrum, whereas the postzyga- stricted to the anterior half of the centrum, and Petrolacosaurus(Reisz, 1981).The neural pophysesdo not. At the baseof the neural as in the posterior cervical of Zarcasaurus. archesexhibit prominent lateralexcavations. arch, there is a shallow, Iateral excavation. The cerviial zygapophysesdiverge more The zygapophysesare smallerthan those of The anterior ends of the centraare strongly widely and extendmore laterallybeyond the the cervicalsand extend very little beyond beveledto accommodatethe intercentra(not margin of the centruminPetrolacosaurus than the lateral margin of the centrum. The pre- found); in the axis, this featurehas been ex- in Zarcasaurusor Araeoscelis.The neural spine zygapophysesextend only to the anterior end aggeratedby breakage.The transversepro- of the midcervical of Zarcasaurus,as in the of the centrum, whereas the postzygapo- cessesof the axis and midcervicalvertebrae cervicalsof Araeoscelis,is little more than a physesextend beyond the posterior end. The consistof a very small, horizontal ridge near low ridge, whereas in Petrolacosaurus,the zygapophyseal planes diverge only slightly the anterior end of the centrum; a small cos- cervicalspines are moderatelyhigh and blade- from the horizontal. The small transverse tal facet is present near the anterior end of Iike. processesare situatedjust below the baseof the process.The transverseprocess of the The dorsal vertebrae of Zarcasaurus(Fig. the prezygapophysesnear the anterior rim posterior cervical differs from the more an- 3E), as in Petrolacosaurusand Araeoscelis,are of the centrum and project only a short dis- terior vertebraein severalfeatures: its base considerablyshorter than the cervicals.The tance laterally. The elongate axis of the base extendsposterodorsally from the anterior rim lengths of the centra are about equal to, or of the process,as well as its costal surface, of the centrum to the midlength of the base slightly lessthan, the widths. With the marked is inclinedabout 45" from the horizontal,and of the neural arch; it extendsmuch farther exceptionof an isolatedneural arch, the height neither the processnor its costalsurface ex- laterally; and the length of the costal facet is of the dorsalneural spinesis far greaterthan hibit any signs of a capifular-tubercular di- nearlyequal to that of the baseof the process that of the midcervical.The differencein spine vision. There is no beveling of the ventral and expandssomewhat posterodorsally. The height of the dorsalvertebrae may be a result rim of the anterior and posterior ends of the ventral longitudinal keels of the axial and of gradual changealong the column or of a centra to accommodatethe intercentra(not midcervicalcentra are V-shapedin crosssec- regular pattern of alternation in height of found). The ventral halves of the centra are tion, whereas that of the posterior cervical successivespines; the latter featurehas been laterallycompressed, forming a narrow,ven- is U-shaped. reported not only in Araeoscelis(Reisz, Ber- tral, longitudinal ridge that is U-shaped in crosssectron. ApprxorculaR ELEMENTS-Preservedele- ments of the appendicularskeleton include: ffi\ \- the distal ends of both humeri; the proximal \4! and distal ends of the left femur and a frag- \t.'.+{ T\ U ment of the proximal end of the right femur; the proximal and distal ends of both tibiae; the ieft astragalus;and isolatedmetapodials and phalanges.The terminologyused in ori- enting the hindlimb elements is that pro- A-- posedby Rewcastle(1980). The humerus of Zarcasaurus(Fig. 4), as in Petrolacosaurusand most primitive reptiles, €il --<-\ possessesa well-developed,distally-directed \ ,t1 supinator processwhich in dorsal view is \-^ \ \ -i-- -r separatedfrom the ectepicondyleby a deep E ectepicondylargroove. Both the supinator dhr processand ectepicondyleare covered by finished bone, indicating a late adult stage of developmenl.In Araeoscelisthe ectepicon- dylar grooveis bridged over distally to form \# the ectepicondylarforamen. distalto the fust F^ supinator processa prominent, hemispher- l:\ ical radial condyle projects ventrally and slightly distally from the distal margin of the -c\.::x:..:.. ( . |/ numerus. \ :"|o/ Enough of the femur is represented to in- 1., .{ dicate that it was very slender (Fig. 5). For Ii: . \ most of its length the shaft is subcircular f ::rti;;;i' 3 in crosssection. The proximal end of the femur is turned strongly upward, and the entire bone probably had a pronounced sigmoidal flexure. In this feature and in its.slender pro- FIGURE 3-A is .n" and C is the posterior cervical vertebrae of Zarcasaurus portions, the femur of Zarcasaurusis more tanyderusgen. et sp. "o;*3;;;1,nov., CM 41704,in lateral, dorsal, and ventral views; D and E are diagrammatic like that of Araeoscelis than Petrolacosaurus.As lateral views (not to scale)of posterior cervicals of Araeoscelis(after Vaughn, 1955)and Petrolacosaurusin Petrolacosaurus,but in contrast Io Araeos- (after Reisz,1981), respectively; F are lateral views of dorsal vertebrae(relative serial positions unknown) celis,the adductor ridge does not terminate of Z. tanyderus. proximally in a distinct internal trochanter;

May 1984 New Mexico Geology rather, the adductor ridge, which is repre- of the tibia is divided into two areas-a flat, seenin Araeoscelisand Petrolacosaurus.Of the sentedon the shaft of the femur by a faintly lateralarea and a convex,ventrally project- isolated phalanges,one is strongly curved visible line, extends proximally as a sharp ing, mesial area-that are separatedby a and claw-like. crest along the anterior bounding ridge of shallow groove.A similar arrangementis seen the intertrochantericfossa. The length of the in Araeoscelisand Petrolacosaurus. Discussion intertrochantericfossa is greater than the The left astragalus(Fig.D is complete,and The assessmentof Zarcasaurustanyderus as width of the femur head.This compareswell the L-shape is typically reptilian, although a member of the suborder Araeoscelidiais with Araeoscelis,but differs from Petrolacosau- the neck is short. The astragalusof Petrola- basedon its elongatecervical vertebrae, lat- rus where the two measurementsare about cosaurusis alsoL-shaped, but the neck is much eral excavationsof the neural arches, the equal. As in primitive reptiles generally,a longer; in Araeoscelisit is not L-shaped, al- structureof the distal articular surfaceof the distinct ischiotrochanterictuberosity is pres- though the neck is similarly short. A notch tibia, and the long, slender proportions of ent near the posterior margin of the dorsal near the distal end of the calcanealarticular the limb elements.All of these featuresare surfaceof the femur head. surfaceforms the mesialportion of the per- sharedby Araeoscelisand Petrolacosaurus.The The dorsal surfaceof the distal end of the forating foramen.The distalarticular surface astragalusis the only elementthat does not femur is markedby a deepintercondylar fossa. of the astragalusis divided subequallyinto conform to the basic structural pattern of the The dorsal surfaceof the posterior condvle a lateral (fibular), convex, condylarJike area araeosceloids.Apart from this, the holotype extendsdistally a short distancebeyond the and a mesial(tibial), concave,socket-like area. does not possessany features that would level of the tibial facetbelow it as a distinct, Presumablythe convexarea articulated with prevent its assignmentto the suborder. liptike process,here termed the supracon- the fourth distal tarsal,and the concavearea It is clear from the previous section that dvlar process.A narrow strip of finished bone, articulatedwith the centrale.This arrange- Zarcasaurusmore closelyapproaches Araeos- continuouswith that of ihe intercondylar ment is unlike the condition in either .4r- celisthan Petrolacosaurusin overall structure. fossa,separates this processfrom the tibial aeoscelisor Petrolacosaurus,where a notch in Among the more important featuresshared facet. Ridges on the dorsal surface of the the distal edge of the astragalusseparates a by Zarcasaurusand Araeoscelis,but absent in supracondylarprocess mark the site of at- Iateralcondyloid process(the articular sur- Petrolacosaurus,are the shapeand position of tachmentof the extensormusculature of the facefor the fourth distaltarsal) from a mesial the cnemial crest of the tibia, the presence lower leg. In contrastto Zarcasaurus,the in- convexarea (the articularsurface for the cen- of stout teeth, and the great length of the tercondylar fossa is shallower, and the trale). The tibial articular facet of the astra- pedicles of the cervical neural arches. The supracondylarprocess is absent in Araeos- galusof Zarusaurusis a simple convexsurface, former two feafuresare interpreted as shared, celis.The intercondylarfossa of Petrolacosau-with no ridge-and-troughpattern like that rus is well developed. Although the femur figuredby Reisz(1981, fig.22) doesnot show a supracondylarprocess, an extensivearea of unfinished bone at this site may indicate that its absenceis due to the immaturity of the specimen.The poplitealarea on the ven- tral surfaceof the distal end of the femur is a smooth,broadly concavedepression. As in / r'1'''': Araeoscelis,the posterior limited ridge of the / poplitealarea is low; in Petrolacosaurusit is a .,! relatively high, sharp crest. The oroximal articular surfaceof the tibia (Fig. 6i is triangular in outline with a large, oval-shaped, lateral articular surface for the posterior (fibular) condyle of the femur and a small, mesial articular surface that faces somewhat laterally and posteriorly for the anterior (tibial) condyle of the femur. The RAD anterior corner of the proximal end of the tibia forms a knob-shapedcnemial crest sim- ilar to, but lessstrongly developedthan, Ar- aeoscelis;the tibia ofPetrolacosaurlslacks such a crest. Between these three areas of the proximal articular surfaceis a deep, concave areaof finished bone. A similar area of fin- ished bone has not been described in Ar- aeoscelis,nor has it been observed in those Araeoscelisspecimens examined during the courseof this study that are presumedto be adults on the basis of the complete ossifi- cationof their skeletons.The proximal artic- ular surfaceof the tibia of Petrolacosauruslnas not been describedin detail. The preserved proximal ends of the tibiae of Zarcasaurus suggestthat the bone is concavelaterally. This is a primitive feature seenin Petrolaco- I saurus and primitive reptiles generally, but not in Araeosceliswhere the great elongation of the limbs has resulted in the tibia being ICM straight or slightly concavemedially. A ridge on the lateral surface of the tibia presumably FIGURE l,-lsyg45su7y5 tanyderus gen. et sp. nov. Dorsal and ventral views of distal ends of the right, marks the site of attachmentof the interos- A, and the left, B, humeri; ectepicondyle (ECT), ectepicondylar groove (ECT GR), radial condyle (RAD), seousligament. The distal articular surface and supinator process (SUP) indicated.

Neu Mexico Geology May 1984 derived characters and, therefore, under- suborder (Reisz, Berman, and sence of the distal notch in Zarcasaurusis score the close relationship between Zarca- Scott, in preparation). therefore viewed as a derived character. Al- saurus and Araeoscelis.Zarc'asaurus possesses Zarcasaurusexhibits two derived features though closely related to Araeoscelis,the ex- several features that are widespread among not seen in either Araeoscelisor Petrolacosau- istence in Zarcasaurusof both primitive and primitive reptiles and can be considered ras. One is the great length of the cervical derived features with respect to Araeoscelis primitive with respect to the condition in vertebrae of Zarcasaurus, the centra of which rules out the possibility ihat they had an Araeoscelis:there is no ectepicondylar fora- are as much as two times longer than the ancestor-descendant relationship. men on the humerus; a distinct supracon- dorsal centra, whereas in Araeoscelisthe max- It is worth noting that araeosceloid re- dylar process on the posterior condyle of the imum is 1.61 times (Reisz,Berman, and Scott, mains may be present among the isolated -1.55 femur is present; a concave area of finished in preparation) and in Petrolacosaurus limb elements described by Reisz (1980)from bone separates the lateral and medial artic- times (Reisz, 1981). The second derived fea- the Early Permian at Fort Sill, Oklahoma, ular surflces on the proximal end of the tibia; ture of Zarcasaurusis the structure of the dis- and referred by him to the Protorothyridi- and the simple tibial facet of the astragalus, tal articular surface of the astragalus. The dae. This assessment was made solely on the which is a convex surface lacking a ridge- astragali of Petrolacosaurusand Araeoscelis,as basis of primitive features, particularly the and-trough configuration, is present. An ad- in (Peabody, 1951), have a deep presence on the humerus of a prominent, ditional feature that may be considered notch in the distal articular border. This notch primitive within the Araeoscelidia is the ab- separates a condyloid-like structure at the CNEM sence of a distinct internal trochanter of the laterodistal corner of the astragalus from a CNEM femur. This character state is not typical of convex surface mesial to this. The distal notch primitive reptiles, but is present in Petrola- corresponds to the suture between the prox- cosaurus,which in other respects is consid- imal fourth centrale and the tibiale of the ered to be the most primitive member of the amphibian tarsus (Peabody, 1951). The ab-

ISTR

INT FCS

FIGURE 6-Left tibia of Zarcasaurus tanyderus gen. et sp. nov. A are the posteromedial and articular surface views of the proximal and distal ends, B are the anterior and alticular surface views of the proximal end, and C are the posterolateral and ICM oroximal articular surface views of the proximal and distal ends; cnemial crest (CNEM) i;dicated. n#eh , rc!

FIGURE7-Left astragalusoI Zarcawurustanyderus gen. et sp. nov. A is the medial, B is the anterior, FIGURE S-Left femur of Zarcasaurus tanyderus gen. et sp. nov. A is the anterior view of the proximal C is the lateral, D is the posterior, and E is the end; B is the dorsal view and C is the ventral view of the proximal and distal ends; adductor crest distal view; articular surface for centrale (CENT), (ADD CR), intertrochanteric fossa (INT FOS), ischiotrochaniericus tuberosity (IS TR), popliteal area articular surface for (FIB), articular surface (POP), and supracondylar process (SC PR) indicated for tibia (TIB) indicated.

May 1984 Nrw Mexico Geology distally-directed supinator processlocated just grant from the Cordelia Scaife May Chari- vanian reptile from Kansas: University of Kansas Pa- Confributions, Vertebrata, Art 1, 44 pp proximal to the radial condvle. Reisz noted table Trust (to D. S B.). leontological Reisz, R R , 7977, Petrolacosarrus,the oldest known diap- that th" Fort Sill humerus and femur are sur- sid reptile: Science, v 196, pp 1,091.-1.,093 prisingly similar to humeri and femora of References -r 1980,A protorothyridid captorhinomorph reptile Petrolacosaurusand Araeoscells,but he re- Brinkman, D B , 1981, The origin of the crocodiloid tarsi from the Lower Permian of Oklahoma: Royal Ontario Contributions, no 121, pp 1 jected their assignment to an araeoscelid be- and the intenelationships of thecodontian : Museum, Life Sciences Breviora, no 466, pp 1-23 16 protorothyridids, as in the Fort -, cause in the Carroll, R L , 1.968,A ?diapsid (Reptilia) parietal from 1981, A diapsid reptile from the Sill elements, the epipodials are considerably the Lower Permian of Oklahoma: Postilla, no 112, pp of Kansas: Universi$ of Kansas, Museum of Natural shorter than the propodials, whereas in Pe- 1.-7 History Special Publications, no 7,74 pp -, Reisz, R R, Berman, D S, and Scott, D, The anatomy trolacosaurusand Araeoscelisthev are sub- 1969,A Middle Pennsylvanian captorhinomorph, and the intenelationships of primitive reptiles: Journal and relationships of the Lower Permian reptile Araeos- equal in length. This comparison, though of Paleontology, v 4, pp 1,51.-770 celis,in preparation valid, cannot be applied here with any con- -, 7987, Plesiosaur ancestors from the Upper Per- Rewcastle, S C, 1980, Form and function in lacertilian fidence because there is no assurance that mian of Madagascar: Royal Society of London, Philo- knee and mesotarsal joint-a contribution to the anal- ysis locomotion: London, of Zo- the Fort Sill limb bones are from individuals sophical Transactions, Ser B, v 293, pp 315-383 of sprawling Journal Chatterjee, S ,1,980,Malerisarlrs, a new eosuchian reptile ology, v 191, pp 147-770 of equal size, or that they represent a single, from the Late Triassic of : Royal Society of London, Romer, A S , 7956, The osteology of the reptiles: Uni- or even closely related, species. On the other PhilosophicalTransactions, Ser B, v 291,,pp 163 200 versity of Chicago Press, 772 pp hand, the very slender proportions of the Credner, H , 1889, Die Stegocephalen und Saurier aus -, 1966, Vertebrate paleontology, 3rd edition: Uni- Press, 458 pp humeri and femora of the Fort Sill form, Pe- dem Rothliesenden des Plauen'schen Crundes bei versity o{ Chicago Dresden, VIIi, Theil, Kndaliosturus priscus Cred: Zert- -, 7967, Early reptilian evolution re-viewed: Evo- trolacosaurus,and Araeoscellscan be consid- schrift der deutschen geologischen Gesellschaft, v 41, Iution, v 21., pp 821-833 ered a shared, derived characterthat setsthem oo 319 342 Simpson, L C,1,979, Upper Cearyan and Lower Leo- apart from the protorothyridids. The possi- Ebe;th, D A, and Berman, D S , 1983, Sedimentology nardian terrestrial vertebrate faunas of Oklahoma: v 39, pp 3-79 bility that some of the Fort Sill limb elements and paleontologv of Lower Permian fluvial redbeds of Oklahoma Geology Notes, north-central New Mexico-nreliminarv report: New Stovall, J W, Price, L I, and Romer, A S, 1966,The described by Reisz may pertain to an ar- MericoGeology, v \, no 2, pp 21-25' postcranial skeleton of the giant Permian aeosceloid is reinforced by Carroll's (1968) Kuhn-Schnyder, E ,1967, Das problem der : Cotylorhynchusronerli; Museum of Comparative Zool- description of an isolated parietal from the Colloque Internationaux du Centre National de la Re- ogy, Bulletin, v 135, pp 1-30 The Permian reptrle Araeoscelis re- same deposit that indicates the presence of cherche Scientifique, no 153, pp 335-348 Vaughn, P P, 1955, Langston, W , Jr , 1,953,Permian amphibians from New studied: Museum of Comparaiive ZooIogy, Bulletin, v a well-developed, upper temporal opening. Mexico: University of California, Publications in Geo- 1."13,pp 305-457 In retrospect, his conclusion that this parietal logical Sciences, v 29, pp,349-476 -,I970, Alternation of neural spine height in certain may belong to a form structurally, if not phy- Olson, E C, 1970, New and little known genera and Early Permian : Southern California Academy v logenetically, intermediate between primi- soecies of vertebrates from the Lower Permian of Okla- of Sciences, Bulletin, 50, pp 80-85 homa: Fieldiana.Ceology, v 18, pp 359-4J4 Williston, S W, 1910, New Permian reptile-rhachito- tive captorhinomorphs and eosuchians now Peabody, F E , 1951,The origin of the astragalus of rep- mous vertebrae: Journal of Ceology, v 18, pp 585-600 appears even more plausible. tiles: Evoluiion, v 5, pp 339-344 -, 7913, An ancestral lizard from the Permian of AcxNowlrocunNrs-We wish to thank Dr. -t 19521 Petrolacoscurus kansensis Lane, a Pennsyl- : Science, v 38, pp 825-825 ! W Langston and Dr. P. Vaughn for reading this paper and making many useful com- ments leading to its improvement. Field work was supported by grants from the New Mex- ico Bureau of Mines and Mineral Resources (to D. S B. and D. A. E.) and by the M. Graham Netting Research Fund through a

f{ewMerico Mineral Symposium Callfor papers The 5th annual New Mexico Mineral Svmpo- sium will be held at the New Mexico Instiiute of Mining and Technologyin Socorro, New Mexico, on November 10 and 11, 1984.Sponsors of the annual symposium are the New Mexico Bureau of Mines and Mineral Resources,the Albuquerque Gem and Mineral Club, the New Mexico TechMin- eralogical Sociery the New Mexico Museum of Natural History, and the University of New Mex- ico, Department of Geology. This year's symposium will again feature 30- minute papers, a si.lentauction, and a dinner on Saturday night. The sponsors encourage profes- sionals and amateurs alike to submit papers on any one of a variety of topics-new mineral oc- currences within the State. the historv of New Mexico'sclassic rnineral-collecting areas, the geol- ogy of ore deposits, and interesting mineral oc- currencesin adjacent states-to name a few. Abstractsare due by August 15, 7984 Complete papers should be submitted within 6 months after for publication in a symposium vol- lfi::",urto" General registration for the 1984symposium is $10.00;registration for students and senior citi- zens (over age 60) is $5.00.To obtain registration information or to submit an abstract,contact Rob- ert M. North, New Mexico Bureau of Mines and Scepter quartz from Kingston, New Mexico; rhodochrosite from Silverton, Colorado; topaz MineralResources, Socorro, NM 87801,(505) 835- from the Thomas Mountains, Utah; and barite from Superior, Arizona, rePresent minerals 5246. from the four-corners states. Artist: Teresa Mueller

New Mexico Geology May 1984 39