The Evolution of Mammalian Molars to and from the Tritubercular Type

Evolutionof Mammalian Molar8. 1067

THE EVOLUTION OF MAMMALIAN MOLARS TO AND FROM THE TRITUBERICULAR TYPE.'

BY HENRY FA IRFIELD OSBORN.

THE dentitionin the recentMammalia is so diversethat the most sanguineevolutionist of fifteenyears ago could not have anti- cipatedthe discoveryof a commontype of molar,in both jaws, as universal among the Mmamalia of an earlyperiod as the pentadactylefoot, and as centralin its capacityfor development into the widelyspecialized recent types. The tritubercularmolar, discovered by ProfessorCope in the Puerco,is exactlysuch a type,and may be consideredwith the pentadactylefoot as playinga somewhatanalogous r6le in mammalian history,with this importantdifference-the unmodified pentadactylefoot was probablyinherited direct from the reptiles,and its subsequentevolution, with a few exceptions,has been in the direc- tion of the greateror less reductionof primitiveelements towards special adaptation,as, to borrow an extreme illustration,in the transition.from Phenacodus with 26 elementsin the manus to Eqmucswith only 12 such elements. On the other hand, the tritu- berculartooth was not inherited,but in all probabilitydeveloped withinthe mammalianstock, from a hypotheticalform with almost, if not quite simple conical molars,implanted by single fangs,in a nearlyhoinodont series.2 No such primitivetype of mammalian dentitionis actuallyknown, although Dromotheriurnapproximates it; but the apparentreversion to this typeamong the C'etacea,and apparentretention of it in theEdentcta,3 support all theindependent evidence upon this point derived from the Mesozoic Mamimals. The principleof growthwas the regular addition of new parts to the simple cone,not at random,but accordingto a certaindefinite 1 Read in the geological section of the British Association at Bath, September,1888. Read in abstractby Prof. Cope, National Academy is Sciences, at New Haven, Nov., 1888. 2 See Autbor " Structureand Classification of the Mesozoic Mamma- Ila." Jour.Phila. Academy, 1888,p. 240. 3 See OldfieldThomas, " The Homologies and Succession of the Teeth in the Dasyuridm." Phil. Trans., 1887,p. 458.

orderwhich apparentlyprogressed independently in differentphyla, througha seriesof sub-tritubercularstages until trituberculy1 was attained. The tritubercularmolar consistsessentially of three cusps,form- ing what may be called the primitivetriangles, so disposed that the upper and lowermolars alternate. This,when attained, formed a centralstage fromwhich the greatmajority of recentmolar types have divergedby the addition,modification and reductionof cusps; we mustexcept the Monotremes,the Edentates,and possibly the Cetaceans,although there is considerableevidence that the cetacean molarswere once of the triconodonttype.2 Among extinctorders, the Multituberculata (Plagiaulax, Tritylodon,etc.) must also be exceptedfrom this seriesand discussion. The almostuniversal predominance of trituberculyin the early geological periods,is verysignificant of the uniformityof molar origin. Of twentyknown Mesozoic genera,3all exceptthree 4 show trituberculyin someof its stages. As to the Lower Eocene, eighty- two Puerco species,representing twenty-six genera and five orders (Creodonta,Tillodontia, Lemuroidea, Condylarthra,Amblypoda), only fourspecies have quadritubercularteeth, all theremainder are tritubercular.5Prof. Riitimeyerhas recentlypointed out thepre- dominance of this type in the nearlyparallel Egerkingen beds The contemporaryCernaysien fauna in the collectionof Dr. Le- moine at Rheims, recentlyexamined by the writer,shows exclu- sively tritubercularmolars or their derivatives. By the Middle Eocene the linesof divergencetowards the existingtypes of molars were well advanced,but trituberculypersisted in the dentitionof several orders,in whichit is foundto-day (Lemuroidea, Insectivora, Carnivora,and manyMarsupialia). 1 First employed by Rfitimeyer, "Ueber Einige Bezi ehungen zwischen den SaugethierstdmmenAlter und Neuer Welt." Abh. d. .schweiz. pal. gesellsch., Vol. XV., 1888,'p. 54. 2 See Brandt, "Die Fossilen u. Subfoss. Cetacean Europas." Taf. XXXII., figs. 4-9. -3The list given by the writer (op. cit., p. 247) is foundto contain several synonyms. See " Additional observationsupon the Structureand Classificationof the Mesozoic Mammalia." Proc. Phila. Acad., Nov., 1888,p. 292 4Dicrocynodon (Diplocynodon), Docodon, Enneodon, Marsh. 5 Cope, " Synopsis of the Vertebrate Fauna of the Puerco Series," Am. Phil. Soc., 1888.p. 298.

It followsthat it is quite as essentialfor the comparativeanato- mist to thoroughlygrasp the meaning and historyof each of the componentcusps of the tritubercularmolar and of theirderivatives, as it is to perfectlyunderstand the elements of the manus and pes. For, the homologiesof the cusps can now be determinedalmost as certainlyas thoseof the digits. Take a human molar,for example, ever componenttubercle has its pedigree,and it can be demon- strated,almost beyond a doubt,which of thesetubercles is homologous with the single reptiliancone. The writer recently(op. cit., p. 242) proposed the adoption of a distinctnomenclature for the differentcusps of the tritubercularmolar, and offereda series of termsfor the primarycusps based as faras possibleupon the primitive positionand order of development,and in most instancesin accordwith their secondaryposition. This nomenclaturecan be extendedto the secondarycusps in the sextubercularsuperior, and quinquetubercularinferior molars. The termsnow in generaluse are based,for the most part,upon the secondaryor acquiredposi- tion,and in no instanceupon the homologiesof the cusps in the upper and lowermolars, or even in correspondingmolars of different genera, thus involving much confusion. For example, the Antero-internalcusp of the lowermolar of M1iiolcenusis not homologous with the antero-internalcusp of Hyopsodus,nor with the antero-internalcusps of the upper molar of eithergenus. The presentcontribution is based principallyupon the writer's studies amongthe Mesozoic Mammalia, and, with some additions, upon Prof. Cope's numerousessays upon the trituberculartype in the TertiaryMammalia.l Four propositionsmay be laid down fordiscussion: (1.) That trituberculywas acquired duringthe Mesozoic period, in a seriesof stages beginningwith the single cone and attaining to the primitivesectorial type in the Jurassicperiod. 1 ProfessorCope's essays abound with discussions and notes upon the originand successionof the trituberculartype. (See collection,in " Ori- gin of the Fittest"). He has outlined the transitionfrom the single cone to the tritubercularcrown (p. 347); the tubercular sectorial (p. 246); the quadrituberculartype (p. 245 and p. 359); the Spalacotherium molars as a transitionto the trituberoular(p. 259). The acquisition of the superiorand inferiorquadritubercular molar (p. 361). The predic- tion of the discoveryof Carnivora with triconodontmolars (p. :365),and of the simple trituberculartype in both jaws (p. 362).

(2.) The majorityof Mesozoic mammals showedtrituberculy in some of its stages. Presentevidence goes to show thatthe remain- ing,or aberranttypes, if such existed,did not persist. The majority of the persistingforms of laterperiods were derived from the forms, withsimple tritubercular molars, of earlierperiods. It followsthat trituberculywas an importantfactor in survival. (3.) The definitehomologies of the primaryand to some degree of the secondarycusps in the upperand lower molarscan be estab- lished. (4.) The mode of successionof toothforms favors the kinetogen- esis theoryadvanced by Ryder and Coope. There are threegeneral observations to be made: First.-In attemptingto completethe historyof each of the cusps,we naturallyfind that the paleontologicalrecord is not suffi- cientlyperfect to admitof ourfollowing a certaintype along a single phylumback to the primitivetype. We must at the outsetpro- ceed upon the principleof similar effects,similar causes. For example,since the historyof the developmentof the intermediate tuberclesin the superiormolars of the Lemuroidea (Pseudolemu- roidea,Schlosser) is perfectlyclear during the Wasatch and Bridger epochs-it is safe to infer that the intermediatetubercles of the Ungulate molars, which are fully developed in the underlying Puerco,had the same history. Second.-There are in each period Aberranttypes which embrace either incomplete or degeneratetri- tubercularstages, i.e., a high specializationin which the past rec- ord is obliterated,or, finally,stages in non-tritubercular lines of development. Third.-In the parallel evolution of trituberculy in differentphyla we findthat the progressionis by no meansuni- form. In everygeological period in 'whichthe fauna is well known we observeprogressive genera which outstrip the othersin reaching a certainstage of molardevelopment, contrasted with persistent types which representarrested lower stages of development,while between themare the centraltypes which representthe degreeof evolution attained by the majorityof genera. The latter may be said to constitutethe stagewhich is characteristicof the period. The Stages of trituberculymay now be definedas seen in different typesin theirorder of succession

I. Haplodont Type (Cope).' A simple conical crown. The fang usually single and not distinguishedfrom the crown. This typehas not as yet been discoveredamong the primitive Mammalia. A ProtodontTub. Type.2 The crownwith one main cone, and lateral accessoryCuspules; the fanggrooved. There is someques- tion as to the advantageof distinguishingthis as a type,for it stands intermediatebetween typesI. and III. Example, Dromotherium of the AmericanTriassic. II. TriconodontType (Osborn, op. cit., p. 242). The crown elongate,trifid, with one centralcone and two distinctlateral cones The fangdouble. Example, Triconodon. III. Tritubercular(Cope)3. The crowntriangular, surmounted by threemain cusps,the centralcone placed internallyin the upper dollars and externallyin the lower molars. Example, the lower molars of Spalacotheriumand Asthenodon. This typeis rare in its primitivecondition as above defined. The upper and lowermolars are alike in typesI. and II.; in type III. theyhave a similarpattern but with the arrange- mentof the homologouscusps reversed. These typesare all primitive. In the followingsub types,the primitivetriangle formsthe main portion of the crown, to which other "secondary"cusps are added, the homologiesof which in the upper and lower molars are somewhatdoubtful. Parallel and withan intimaterelation to the additionof the secondary cusps,is the divisionof the tritubercularinto a secodontand bunodontseries, according to the assumntionof a purelycut- ting or crushingfunction. In departingfrom the primitive type,the upperand lowermolars diverge in structure,and the homologies of the secondary cusps in each are somewhat doubtful, LOWER MOLARS. A. TubercularSectorial, sub type (Cope). a. The primitive 1" The Homologies and Origin of the Types of Molar Teeth in the Mammalia Educabilia." Journ. Phila. Acad., 1874. The term Homo- dont was previously applied to this type by Rfitimeyer,"Odonto- graphie der Hufthiere,etc." Verh. d. Naturforsch.Gesellsch. in Basel, Band. III., 1863,p. 563. In the writer's opinion this termhas acquired a special significanceas applied to a whole series of teeth,viz., the re- verse of" heterodont,"and may well be retainedin this sense. 2 Osborn,op. cit., p. 222.

This content downloaded from 128.135.012.127 on November 15, 2016 14:32:18 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 1072 Evolutionoj Mammalian Molars. triangle elevatedand its eusps connectedby cuttingcrests; a low posteriorheel. b. This typeembraces a quinquetubercular formin whichthe heel consistsof two cusps,an internaland external.'c. In theBunodont series it developsinto the quadri- tubercularform, by the loss of one of the primitivecusps. UPPER MOLARS. B. Tritubercular.a. The primitivetriangle in thesecodont series purelytricupsid. b. This embracesa quinquetubercularform in which " intermediate" tuberclesare developed,both in the Secodont and Bunodont series. c. In the Bunodont series a postero-internalcusp is added, formingthe sextubercular molar. NOMENCLATUREOF THE Cusps-As above stated,there is no doubt about the homologiesof the three"primary" cusps (protocone, paracone, metacone) in the upper and lower molars They may be giveslthe same terms,with the arbitrarysuffix id, to distingiish tile l cwsr cisa. The first" secondary" cusps (hypucoue-hypoconid),added to the upper and lower molarsof the primitivetriangle, modify the crownfrom a tri- angularto a quadrangularshape, and hencemay be considered homologous. The threeadditional secondary cusps (protoconule, metaconule,entoconid) evidently have no homologywith each other. TERMS NOW IN USE. PROPOSED TERMS.' ABBREV- Upper Molars. Antero-internalcusp ...... Protocone. pr. Postero- " " or 6th cusp...... Hypocone. h. Antero-external"...... Paracone. p. Postero-...... Metacone. m. Anterior Intermediate cusp...... Protoconule. PI. Posterior " ...... " Metaconule. ml. Lower Molars.." Antero-exterinalcusp ...... Protoconid. pr. Hostero- " " ...... Hypoconid. h. Antero-internalcusp, or 5th cusp...... Parrconid. p. Intermediateor antero-internalcusp (in quad- ritubercularmolars) ...... Metaconid. M. Postero-internalcusp ...... Eutoconid. e. Evolution of the Cusps. The cusp evolution in the Mesozoic period has been fully discussedby the writer(op. cit., pp.240-4) I American Naturalist,April, 1883, p. 407. 2 I am much indebted to my colleagues ProfessorsMacloskie and Winans fortie assistance in the selectionof these terms.

/ 2 3 4 J

j~~~~~1 me line'4

pa tja /uasn i /7~~Z A""'~~~Pi /(i' , Av Pat~~~~~~c

8'a a 3 /

F J4. 1.

l7a

2 L)XJeI~Iani) J) t P IC T.lo

FBIG. 2.

FIG. 1.-Molar teeth of Mesozoic Mammalia. FIG. 2.-Molars of oppositejaws in normal mutual relation.

and in the Tertiaryperiod, by ProfessorCope, so that only a brief r~sumiis necessaryhere. In Dromotherium(fig. 1), fromthe upper Triassic,the oldest mammaliantype known, with the exception

FIG. 1.-D~iagram of quadritubercular molars of both jaws in normal mutual relation; the superior cusps double lines; the inferior black. Of Miicrolestes,the molarshave a mainprotoconid with several minutelateral cuspules,differing in size in the differentteeth, but in generalgiving a trifidappearance to the crown. The molars of the contemporaryM~icroconodon (fig. 2) also have unpairedfangs, but distinctlytrifid crowns, with the anteriorand the posterior cusps,or para and metaconids,upon the slopes of the protoconid. This Triconlodonttype reappears, with the additionof a cingulum and pairedfangs, in Amphilestes(fig. 3) and Phascolotherium(fig. 4) of the lowerJurassic and persistsinl Tricondon (fig. 6) of the upper Jurassic. In this successionwe observe especially the relative subsidence of the protoconidand upgrowth of the para- and metaconids. Contemporarywith Amphilestesis the classical genus Arnphitheriumr(fig. 6). A recentexamination of the type specimen by the writer revealed the very interestingfact that themolars of this genus are probablyof the primitivetubercul- ar-sectorial types,-the oldest known example. 'Only the paracone and metaconidsand hypoconidshave been observedheretofore, but one can see the tip of the main externalcusps betweenthe formerpair. This patternis repeated,with a considerableelevation of the heel, in Peramuqs of the upper Jurassic.1 Neither of the two foregoingare ofthe primitiveheelless tritubercular type which is apparently found in Spalacotheri'tnn also upper Jurassic,and in the nearlyrelated if not synonymousPeralestes, Plate XXV. Contemporarywith the above, are numerous genera of the Stylodonorder; among these, Asthenodon is of 1 This genus includes also Septocladus dubius Owen, and Spataco- theriu-mminus Owen. See Proc. Phila. Acad., Nov. 1888,p. 292.

the primitivetritubercular type without the hypoconid,all the remainderpresent various modifications of the tubercularsectorial. This covers our knowledgeof trituberculvin the Mesozoic period. No bnunodont formsare known-they were probably developed during the Cretaceous, for a few are found well developed in the Puerco. In the Sectorial series many of the types do not widelydepart from those seen in the Jurassic,but the Bunodont series are universallycharacterized by the initial or advanced developmentof the proto-and metaconulesin the upper molars and the apperance of theEntoconid upon the inner side of the hypoconidbelow. Theprinciples governing cusp development.-Itis remarkableto notein how manyparticulars the actual successionof molardevelopment in the Mesozoic period coincides with the theoretical schemeof originof trituberculyproposed by Cope' and supported by Wortman2 severalyears ago. At thattime Spalacotherium and the generanow embracedunder the Triconodontida3were the only Mesozoic mammals whose molar structurewas fully known,and the views of these authors were partly speculative and partly deductivefrom recent dental anatomy. Two hypothesesmay be advancedto explain the evolutionof the rituberculartype. The firstis thatthe typehas been acquired by the selectionof accidentalvariations in the productionof new cusps and modellingof old ones. The second is, that the interactionof the upper and lower molars in the movementsof the jaws has resultedin local increaseof growthat certainpoints, resulting first in new cusps,then in a change of position and of formin the cusps. Both hypothesesare open to numerousobjections and are by no meansmutually exclusive, but the whole subject is so com- plicated as to requirea separatetreatment. The balance of evidence in tritubercularevolution seems to favor the second or kinetogenesistheory-as apparentlywitnessed in two laws of cusp development. I. The primarycusps firstappear as cuspules,or minutecones, 1 " The Evolution of the Vertebrata Progressive and Retrogressive," American Naturalist,April, 1885,p. 850. 2 "The ComparativeAnatomy of the Teeth of the Vertebrata," 1886 p. 418.

at the firstpoints of contact between the upper and lowermolars in the verticalmotions of thejaws. II. The modellingof the cusps into new forms,and the acquisitionof secondaryposition, is a concomitantof interferencein the horizontalmotions of the jaws. The secondlaw applies especiallyto the evolutionof the molars after the acquisitionof the tritubercularstage, and has been ably proposedand supportedby Ryder,' principallyin its applicationto recenttypes of teeth. The first,although not heretoforedistinctly formulated,is partlyfounded upon facts and principlesadvanced by Cope,and applies chieflyto the stageswhich have been discussed in this essay. During the Homodont mammalian or sub-mammalianmolar stage,the jaws were probably isognathousand the simple cones alternatedas in the Delphinidce(fig. 1). The firstadditions to the protoconeappeared upon its anteriorand posteriorsurfaces. The growthof the para- and metaconidsinvolved anisognathism,2for we findin the latertriconodonts that the lower molarsclosed inside of the upper (Triconodon,fig. 2). There are several transition formssuch as Tinodon and Mernacodonbetween the primitivetri- conodonttype and Spalacotherium,and it has been assumedby Cope and the writer(op. cit., p. 243) thatthe para- and metaconidswere firstformed upon the anteriorand posteriorslopes of the protoconid and then rotatedinwards, but it is also possible that theywere originallyformed upon the inner slopes. In the complemental formationof the upper and lower trianglesthe jaws remained nearlyisognathous (fig. 4). There is no-evidence as to the originof the hypoconid,which as a rule preceded the hypocone,as it was developedvery early. In the Stylaeodontidce,Phascolestes, Amblo- *1 "iOn the Mechanical Genesis of Tooth Forms." Proc. Phila. Acad., 1878,p. 45. 2As employedby Ryder (op. cit., p. 45). " So as not only to indicate respectivelyparity and disparityin transverse diameterof the crowns of the upper and lower molars, but also the parityor disparityin width transversely,from outside to outside," etc. It is clear that ia the homodont condition,with the teeth simply piercing the food,the greatest comminution(of the food) is effectedby isognathism; in the triconodontstage, the jaws must be anisognathous to close upon each other,but the tritubercularstage admits a returnto isognathismby the alternationof the triangles.

therium,etc., the crowns rapidlyincreased in transversediameter (fig.7) and, in some genera,they (Kurtodon) so far lost the tritubercular aspect that,but for the connectingform A sthenodon(fig. 6), we mighthesitate to place themin this series. The key to the furtherevolution of the crownis seen in the bunodontseries during the lowerEocene period. The superpositionof the lowerand uppermolar patterns brings out manyinteresting facts. First,even in the complexcrowns of the bunodontmolars the primitivetriangles retain their primitive alternatingarrangement. Second, the jaws are somewhatanisog- nathous. Third, in supportof the firstlaw of cusp development, we observethat the protoconuleand metaconuleare developedat the pointsof contactwith the ridgeswhich extend from the hypoconid,and, secondly,that the hypoconeappears at the pointwhere the paraconidabuts againstthe protocone. It followsfrom a com- parisonof numerousspecies of Pelycodusand Iioclanusthat as the hypoconedevelops, the paraconidrecedes, as firstobserved by Cope; a fact difficultto reconcilewith the kinetogenesistheory. In this manner the inferiorprimitive triangle is broken, as the upper molarsdevelop into the sextubercularand the lowerinto thequad- rituberculartype. The complementaldevelopment of the upper and lower molars in the known generaof successivehorizons- is approximatelydisplayed in the subjoinedtable. The Eocene list of generawill be greatlyreduced, especially in the Tritub.-tuberc-.sectorialtype, whenthe upperand lower jaws are found associated,and it must be clearlyunderstood that the sub-typesa, b, c, in the above table, are very closely related by transitionforms. In fact, in the carnivorousforms, the extreme secodontand bunodonttypes are frequentlyseen side by side, as in the firstand second inferiormolars of T)idymictis.The chief dis- tinctionbetween these two seriesis the greaterdevelopment of the secondarycusps and the almost invariableloss of the paraconidin the latter; this is effectedby the broader surfacesof contactin the bunodont crowns. In the secodontseries, on the otherhand, the developmentof the secondarycusps is subordinated,and the metaconid is almostinvariably suppressed.' 1 See Cope: " Origin of the specialized Teeth of the Carnivora." Am' Naturalist, March, 1879.

a'~~~~~~~~~~~~~~~li

FIGE. l.-llery7coe~lurrus nmero78stegus Cope. FIGS. 2.-Jle2-ycoehwrus .super bus LeidyQ R

Adapis and Anapto- o.0 ,t0 ~ 4 nmorphuus. v are examples 0 cn . . . 1. U ofSub-types a, c, associated; forit frequently happens thatthe para- ? . conid. U atrophieswithout ;4?3S = Y . . Q a completeenlargement l c)= Q ~ = z ~ . Eq ..... of the hypocone. A H | R eQ .. . X ..... study of the diagram 9 ] demonstrates,a O Q . . .a however, V.0 the associationof E_( ; ^ ?; > Cq . . -that 0 ! a = . . . o Sub-typesb and c is ir- ___-_._, - - - . - - possible. The recent * - , . n Ymonkeys ; Q t > Tarsiws and c*.*Ld Eoris afforda good il- 0 ?; lustration. of the asso- +3@ 3n g tritubercular

4 -( U2 ?3 ;t) ? C; C ciationof ______*.____ quinquetubercular ;-4 sextubercular A, and

44,F,~Sb-4 . ., quadritubercular

4 - _ o molars, ml _ - o, The subsequentevo- C) C) . . 4$e; solution of the molars,in -___ . -._-. ------t differentorders was ..4 characterized,first, by ,4 ..= o . . . UXC theloss ofthe primary ... cusps,e.g., the metaco- . .. .* * . * ,, . > nid in the Carnivora, > ;> Q Q . Y ,3 . o f S the paraconid in the 0~4 0 ~b v 0 PO Ungulata. Second, > 00o 0 * by B the loss of some of the 0 >; ank p; . *~nX *secondary cusps, e.g., theproto- and metaconulesin theArtiodactyla.' Third,by the met- amorphosisin the formof the cusps. This subjecthas been fully l Schlosser: " Beitrage zur Kenntniss der Stammgeschichteder Huf- thiere," Morph. Jahrb.,1886, p. 123,has especially drawn attention to the probabilitythat the Artiodactylawere derived from sexitubercular forms.

treated by RfUtimeyer,Kowalevsky, Cope, Schlosser and others. The Relation of Trituberculyto the Persistenceof Miammalian Phyla.-The above table shows somewhatindefinitely, but nonethe less positively,the general progressionof the Mammalia,to and fromthe primitivetritubercular type. As alreadystated, even with our presentvery limited knowledge,certain stagesappear to have been characteristicof certain periods,as follows: the triconodont in the lower Jurassic; the primitivetritubercular and tubercular sectorial in the upper Jurassic; the secodont and bunodontsub- typesof trituberculy,predominated in the Puerco; in the Bridger, the Perissodactylungulates had mostlypassed beyond into the lophodont and symborodonttypes, and the Artiodactylswere approximatelyin the stage of sub-typesc; but the Lemuroidea, Creodonta,Ilnsectivora, etc., were, almost withoutexception, tritubercular. There can be little doubt that,parallel with the tritubercular forms,in each period,there were aberrant or degeneratetypes, but it is difficultto determinewhich theseare. Many Mesozoic types, which the writerformerly considered aberrant, have now provento be tritubercular.2The upper Jurassic.genera included under the Diceoeymoodontidce(see Marsh, Amer. Journ.Se., April, 1887, p. 338) are apparentlyaberrant. There are several degeneratetypes among the Puerco and Wasatch Creodonts,such as Dissctcas and Mes6oenyx.But there is a strikingproof of the superiorityof the tritubercularmolar in the factthat, according to our presentknow- ledge at least,the Jurassicmammals possessingaberrant or degen- erate molar types dlid not persist into the Puerco,nor did such typesin the Puerco persistinto the Bridger. There is some doubt as to the persistenceof the sub-tritubercularstage; thiewriter previously consideredthe Thylaciuts molars as triionodont; but Mir. Lydekkerhas called attentionto the probabilitythat the metalonid has disappearedand been replacedby a heel as in the sectorialteeth of the Carnivora. The disappearanceof the degeneratetypes may be attributedto the generalprinciple that rapid specializationand loss of partsleads ultimatelyto extinction,by deprivingthe animal of the means of adaptation to new conditions,or surroundings. 3See " Additional Observationsupon the Structureand Classification of the Mesozoic Mamnmalia." Proc. Phila. Acad., Nov., 1888.

The mechanicalsuperiority of the trituberculartype, over every other has been repeatedlydemonstrated in its plasticcapacity of adaptationto the mostextreme trenchant and crushingfunctions.

THE ARTIODACTYLA.'

BY E. D. COPE.

THE Artiodactylais the suborderof the DiplarthrousUngulata in which the astragalus articulateswith the secondrow of tarsal bones by a ginglymusor hinge,aud in which the thirdand fourthtoes are equally or subequally developed.' It includes the most highly modifiedof the Mamimalia,whether we regard the organsof locomotionor of digestion. The antelopeand deer illus- tratethe greatestspeed to whichthe mammalhas attained. Their extraordinaryapparatus for the digestionof vegetable substances which containbut a small percentageof nutritiousproteids, has given them an extraordinaryadvantage, so that theyare afterthe rodents,the mostabundant of theirclass, iu spiteof the persistent persecutionof the carnivorousspecies. They attain in the genera Giraffaand Bos the largestdimensions in the class, exceptingonly the Proboscidia. The Artiodactylamake their firstappearance in the early or WXasatcliEocene in the genus PantolestesCope. A genus existsat a correspondinghorizon in Europe. No other genus of the sub- order appears with it. Its representativessteadily increase in numbersin the succeedingBridger and Uinta epochs in America, and in the Calcaire grossierand Gypse of Europe. Some of these, esg.,the Anoplotheriideof Europe, diverge fromthe line of succession,while others,e.g., : iphoclontidT, are clearly ancestorsof laterforms. In America,the Pantolestidcleappear as ancestorsof the Camels especially. I now give a synopsisof the familiesof the suborderand theirphylogenetic relations. I. Superiormolars tritubercular (Pantolestoidea). Molars bunodont; fourdigits...... Pantolestidce. I Modified for the Naturalist fronma paper by the author in the Pro- ceeds. of Amer. Philos. Society, 1887,p. 377. 2 See Naturalist,November, 1877.

This content downloaded from 128.135.012.127 on November 15, 2016 14:32:18 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).