5. >S! IMAGE EVALUATION TEST TARGET (MT-3) 1.0 I.I 150 "■'^^ MtaHB ■^ 1^ 11112.2 m m ■UUu Nil 1.25 1.4 1.6 .4 6" - ► VQ ^5 V / Photographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) 872-4503 ^ \5 ,\ iV \\ 4^ -^ ;\ •^^ io t/. i9 CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. Canadian Institute for Historical Microreproductions / Institut Canadian de microreproductions historiques Technical and Bibliographic Notes/Notes techniques et bibllographiques Tht/ institute has attempted to obtain the best o/lginal copy available for filming. Features of this copy which may be bibliographically unique, which may alter any of the Images In the reproduction, or which may significantly change the usual method of filming, are checlced below. L'Instltut a microfilm* le mellieur exemplaire qu'ii lui a *t* possible de se procurer. Les d6taiis de cet exemplaire qui sont peut-Atre uniques du point de vue bibiiographlque, qui peuvent modifier une image reproduite, ou qui peuvent exiger une raodification dans ia methods normaie de filmage sont indiquis ci-dessous. f-TT Coloured covers/ ALJ Couverture de couleur D Covers damaged/ Couverture endommagAe Covers restored and/or laminated/ Couverture restaurie et/ou pelliculAe Cover title missing/ Le titre de couverture manque r— 1 Coloured maps/ Cartes giographiques en couleur Coloured Ink (i.e. other than blue or blacit)/ Encre de couleur (i.e. autre que bieue ou noire) D □ Coloured plates and/or Illustrations/ Planches et/ou illustrations en couleur □ Bound with other material/ Rail* avec d'autres documents D D D Tight binding may cause shadows or distortion along interior margin/ La reiiure serr6e j;«ut causer de i'ombre ou de la distortion ie long de ia marge int«rieure Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajout6es lors d'une restauration apparalssent dans le texte, . 9is. lorsque ceia Atait possible, ces pages n'ont pas *t* fiimAes. Additional comments:/ Commentalres supplAmentaires; □ Coloured pages/ Pages de couleur □ Pages damaged/ Pages endommag6es □ Pages restored and/or laminated/ Pages restauries et/ou peiilcui^es Pages discoloured, stained or foxed/ Pages d*coior6es. tachet6es ou piqudes □ Pages detached/ Pages d6tach6es □ Showthrough/ Transparence □ Quality of print varies/ Qualit* InAgale de I'impresslon □ Includes supplementary material/ Comprend du materiel supplAmentaire D D Only edition available/ Seule Edition disponlble Pages wholly or partially obscured by errata slips, tissues, etc.. have been refiimed to ensure the best possible image/ Les pages totaiement ou partieiiement obscurcles par un feulllet d'errata. une peiure, etc.. ont «t« filmies A nouveau de fagon d obtenir la meiileure image possible. This item is filmed at the reduction ratio checked below/ Ce document est film* au taux de r*duction indiqu* ci-des«»ous. 18X 22X 10X 14X 12X 16X 20X 26X 30X 24X 28X 3 32X I ier 9 le The copy filmed here has been reproduced thanks to the generosity of: Library of Congress Photoduplicatlon Service The images appearing here are the best quality possible considering the condition and legibility of the original sopy and in keeping with the filming contract specifications. Original copies in printed paper covers are filmed beginning with the front cover and ending on the last page with a printed or illustrated impres- sion, or the back cover when appropriate. All other original copies are filmed beginning on the first page with a printed or illustrated impres- sion, and ending on the last page with a printed or illustrated impression. The last recorded frame on each microfiche shall contain the symbol — ^> (meaning "CON- TINUED"), or the symbol V (meaning "END"), whichever applies. Maps, plates, charts, etc., may be filmed at different reduction ratios. Those too large to be entirely included in one exposure are filmed beginning in the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: L'exemplaire filmA fut reproduit grdce d la g6n«rositA de: Library of Congress Photoduplicatlon Service Lea images suivantes ont 4tt6 reproduites avec le plus grand soln, compte tenu de la condition et de la nettat« de l'exemplaire film6, et en conformity avec les conditions du contrat de fiimage. Les exemplaires originaux dont la couverture en papier est imprimAe sont film^s en commenpant par le premier plat et en terminant soit par la dernidre page qui comporte une empreinte d'impression ou d'illustration, soit par le second plat, selon le cas. Tous les autres exemplaires originaux sont filmAs en commenqant par la premiere page qui comporte une empreinte d'impression ou d'illustration et en terminant par la dernidre page qui comporte une telle empreinte. Un des symboies suivants apparaftra sur la dernidre image de cheque microfiche, s>. !on le cas: le symbole — ► signifie "A SUIVRE ', le symbole V signifie "FIN". Les cartes, planches, tableaux, etc., peuvent dtre film6s d des taux de reduction diff6rents. Lorsque le document est trop grand pour dtre reproduit en un seul clichd, 11 est film« i partir de Tangle supArieur gauche, de gauche d droite, et da haut en bas, en prenant le nombre d'images nAcessaire. Les diagrammes suivants illustrent la mdthode. ate elure, H 32X 1 2 3 1 2 3 4 5 6 j.*i».,-^-Mja'r If H '^ m ?*^"i« i I Quartern'. Pliocene Miocene Loup fork John Day tthitrKi\ri Eocene Bridycr Purrco Crelactoui CeTacvans Jurasiic Triciisic Uppe crwiian Promammalici 'Ihcromorpha ^aiiro-Alammalia. Hypotlii'tical Phylot^euy of the Xortli .hiuriuin Mamvialia. The Monotremes disappear in tlie i.asal llocene. The Marsupia's ch.appear in the upper Cretaceous and reappear in the lower Miocene {Dhi.-l^hys). The aber- rant Placentals branch off in the upper Jurassic. The typical I'lacentals divide into the "Ancient types •• dying out in the Miocene, and the "Modern types' still existing. The aberrant Placentals are given off in niid-Mesozoic times. i m de THE RISE OF THE MAMMALIA IN NORTH AMERICA. IIKNRV FAIRFIKI.I) OSIIORN. TWENTV years ago an era opened in the mammalian palae- on ology of Europe and America. Partly inspired by the Odontosraphie of Rutimeyer, Kowalevsky completed and pub- lished in 1873 his four remarkable memoirs upon the hoofed mammals. He wrote these four hundred and fifty quarto pa^es .n three languages not his own. in French upon Anchitl!crUnn and the ancestry of the horses, in English upon the Ilyopo- tamtdac m German xxvon ^ siccus, Anthracothcrinm and Entclo. don, includmg the first attempt at an arrangement of a great group of mammals upon the basis of the descent theory These memoirs swept aside all the dry traditional fossil lore of Europe; they breathed the new spirit of Darwin, to whom the ch.ef one was dedicated, making principles of descent of more importance than new genera and species. Kowalevsky thus summed up the contemporary palaeontology: "After the splendid osteological investigations of Cuvier had revealed to science a glimpse of a new mammalian world of won- derful richness, h.s successors have been bent rather upon multi- pying the diversity of this extinct creation, than on diligently studying the organization of the fossil forms that successively turned up under the zeal of amateurs and collectors. . . . With the excep- tion of England (referring to Owen, Huxley, Falconer, and others) where the study of fossil mammalia was founded on a sound basis and some glorious exceptions on the continent (referring to Ruti! meyer, Gaudry, Fraas, Milne-Edwards), we have very few good palae- ontological memoirs in which the osteology of extinct mammals has been treated with sufficient detail and discrimination; and thin^js have come to such a pass, that we know far better the osteology of Sou h American Australian, and Asiatic genera of fossil mammals than of those found in Europe." Stud. Biol. Lab. Col. Coll., I, 2. 61 - -tn ,i aww s * J THE RISE OF THE MAMAfALIA At the same time, between 1871 and 1873, the pionfcers of American palaeontology, Lcidy, Marsh, and Cope began the exploration of our ancient lake basins rich in life. The first ten years of their work not only revolutionized our ideas of mammalian descent, but brought together the data for the generalizations of the second decade; for Marsh's demonstra- tion of the laws of brain evolution in relation to survival; for Cope's proof of ungulate derivation from types with the simple foot resting upon the sole, and with the conic or bunodont ancestral molar tooth; and finally for Cope's demonstration of the tritubercular molar as the central type in all the mam- malia. These four generalizations furnished a new working basis for morphology and phylogeny. In these twenty years, thanks to energetic field work, we have accumulated vast materials for the history of the rise of th« mammalia, enough for ten students where there is one, and the question arises: how shall we take best advantage of it, what methods shall we adopt ? In this address, besides bringing before you the more recent achievements of explora- tion and research, I will try to illustrate the advances already made in lines of thought, observation and system in palaeon- tology and indicate other advances which seem to me still desirable. In the problem of how to think and work most effectively, and with most permanent results, all the sciences meet on common ground. Advances in Method. It is to the renown of the veteran Rutimeyer and of Kowa- levsky, so soon unfortunately deceased, that, while their main inductions suffer by American discoveries, their methods of thought have not been displaced. It matters little that their theory, that ungulate molars sprang from lophodont or crested forms, has been disproved; that Kowalevsky's tables of descent are full of errors ; that his main generalization as to the per- sistence of adaptive and extinction of inadaptive foot types does not hold good; that the horses and Anchitherium spring not from Palaeotherium as he supposed, but from Pachynolophus tN NORTH AMERICA. 5 ' and iryracotlicriuui, types which he carefully studied and yet om.tted from the horse line! It is the right system of thought which IS most essential to progress; better in the end wrong results such as the above, reached by the right method, than right results reached hap-hazard by a vicious method If a student asks me how to study palaeontology, I can do no better than direct him to the Vcrsuch cincrnatih lichen Classi- ficatwH dn-fossilcn I luf thieve, out of date in its facts, thor- oughly modern in its approach to ar.cient nature. This work IS a rnodel union of the detailed study of form and function with theory and the working hypothesis. It regards the fossil not as a petrified skeleton, but as moving and feeding; every jomt and facet has a meaning, each cusp a certain significance Rising to the philosophy of the matter, it brings the mechanical perfection and adaptiveness of different types into relation with environment, the change of herbage, the introduction of grasses In this competition it speculates upon the causes of the rise spread and extinction of each animal group. In other words' the fossil quadrupeds arc treated biologically ~^o far as possible m the obscurity of the past. From such models and from our own experience we learn to feel free to abandon traditions in the use of the tools of science, such as mere methods of descrfp. t.on and classification, and to regard priority in nomenclature only. New discoveries continually produce new conditions ; there IS nothmg more obstructive than the reverence for old ideas and systems which have outlived their usefulness. In obser- vation, an old principle was do minimis mn curat lex; now we cannot be too exact. Every cusp and facet has its value' not as a sign-post for a new species, but as suggestive of some function or relationship. Bird's-eye methods of comparison, which, for example, find no difference between a rhinoceros and a loph.odon molar, a^ « of no service now that we are called upon to distinguish between so m-ny lines of ancient mammals crowding in among the ancestors of existing mam- mals. Again, palaeontology is not a science apart ; it has always gone hand in hand with recent osteology ; it must now keep abreast with the embryology of the teeth and skele- 63 a! ' iu;iii;ujitw i a i ugi i aj-_i^ ^i iV i ijj j i j^ i ^^ ff 6 THE RISE OF THE MAMMALIA ton- with the animal mechanics of Marey, Allen, and Muy- bridge; with palaeobotany, geology, and historical-physical geography. In these points we cannot be too broad. All structures should be considered as to their homologies, their mechanics, which throw such a brilliauL ^icht upon their evolu- tion- their relations to the food and soil, and to other parts. This' brings us to the animal as a whole — its tendencies, its place in the system of descent, its relations to its contcmpo- raries, the causes of its progression or retrogression ; finally, into pure speculation. Here I am reminded of a critical saving by the late Professor v. Gudden, the distinguished neu- rologist • "Ein Steinchen der Wahrheit hat mehr Werth als ein grosser Schwindclbau " ; it was in allusion to the tempo- rary character of the great nerve-tract systems of Meynert and Flechsig The great 'Schwindclbau,' literally the 'disappear- ing structure ' of palaeontology, is the phylctic tree which adorns the end of many good as well as superficial papers ; and recently, because of its extremely brief life, has fallen somewhat into disfavor. I do not think the present reaction against these 'trees' is a wise one ; we must remember they are the working hypotheses of our branch of science and serve to m.-st clearly express present knowledge. To illustrate some of these principles of modern methods, let us first look at the evolution of the teeth in the rise of the mammalia. The teeth and the feet are the foci of mammalian evolution, the only direct points of contact with food and the earth Their combined use in phylogeny has increased in interest, because their evolution has proved to be wholly^mde- pendent We recall Cuvier's famous law, of which Balzac said at the time : " Rebuilt like Cadmus cities, from a tooth. No generalization has been more thoroughly routed than that of a necessary law of correlation between looth and foot structure Besides the orthodox clawed carnivores and hooted pachyderms of the t eat French anatomist, we have discovered hoofed carnivores such as Mesonyx. and clawed pachyderms such as Chalicotherium. Even the apparently lastmg barriers of correlation, which Owen raised between the even and odd- toed ungulates, have broken down by Ameghino's discovery ol 64 AV NORTH AMERICA. - a Litoptern odd-toed horse with an cvcn-toed type of astra ga.us Not only is there no correlation of type, but none in the rate of evolution. Hipparion. the most projjressivc horse in tooth-structure, probably owed its extinction to its cons-r vative preservation of its ancestral three toes. For thc^se reasons the teeth and feet, owing to the frequent parallels of adaptation, may wholly mislead us if taken alone: while if considered together, they give us a sure key ; for no case' of exact parallelism in both teeth and feet between two unrelated types has yet been found, or is likely to be. This, I believe is the one esson of later work which reverts to older methods • vve should not base either classification or descent upon the teeth or feet alone. Every additional character diminishes the chances of error. The evolution of foot structure has now become a science and advances have been made in the principles of progression from the plantigrade, pentadactyl serial types to ihe un^ul,- grade, monodactyl alternating types which are of the greasiest .mpoitance in classification and phylogeny. It is sufprising how little attention was given to ungulate foot str'ctur^ between the time of Cuvier and Kowalevsky. Owen's gener- alization as to the Artiodactyland Perissodactyl pes formed the one bright exception. Kowalevsky first directed attention to the importance of the more median metacarpals displacing, or spreading to gam a stronger foothold upon the earp!als as the lateral toes disappeared. Ryder also worked out the laws of reduction. The discovery of Phenacodus led Cope to the final generalization that the primitive ungulates were not only planti- grade but had some of their carpals and tarsals in vertical rows like bricks clumsily set with unstruck joints -and that one grea. Lw of evolution towards digitigradism was to produce itTed? 1 r"'"^ ^""''- "-' '^ ^^""^' ^his alternation differed both m degree and kind in different groups, he revived the comprehensive .Ungulate' of Linnaeus and divided all great 0,27 ""' "'°" ''"' '''' ^'^"^^"^^ ^^^ «-« Rutimeyer and myself have shown that however successful and convenient this system appears, Cope's lines of division 65 8 THE RISE OF THE MAMMALIA ignore the fundamentally different modes of evolution of the fore and hind feet; an animal may be a taxeopod in front nnda diplarth behind or vice versa. Numerous exceptions to Cope's definitions are also found. The discovery of the aberrant ungulate foot types of South America further invalidates Cope's system and sustains the principle that to be permanent classifi- cation must be based upon at least two entirely diverse sets of characters. This does not diminish the importance of the primitive taxeopod plantigrade type as one great key to the still unsolved problems of the primary relationships of the Con- dylarthra, Hyracoidea, Amblypoda, Proboscidia, Toxodontia, Litopterna, Artiodactyla and Perissodactyla. All these orders still stand apart in the dim past like so many mile-posts. While Cope overestimates the feet in these larger divisions/ many writers in Europe still depend wholly upon the teeth and ignore the wide degrees of divergence such as are indicated in the Perissodactyla for example in functional tetra-, tri- and mono- dactylism. By ' functional ' we refer to tendencies which a-e not expressed "in the bare digital formulas — and which have the same relation to the feet that the dental curve h:.s to the teeth. The evolution of a monodactyl tendency is not the work of a century but of a geological period, a principle which • we wholly ignore when we place the monodactyl Anehitheres with the tridactyl Palaeotheres, on the ground that the;r dental type and digital formulae are identical. How many toes an animal has is of far less importance than how these toe^^ are being displaced and reduced. Lower IViesozoic Pro-Mammalia. With the exception of the triassic Theriodcsmus of Se^-lye, no mammal is known by its limbs or skeleton unt" we reacn the basal Eocene ; in studying the first steps in the rise of the mammalia, we are thus practically driven to the teeth au-i jaws alone. In these straits of the fossH-hunter, embryology has lately come famously to aid. Assuming their remote reptilian origin, agreeing with Baur and Kukenthal that the theromorph reptiles were parallel with 66 IN NORTH AMERICA. rather than ancestral to the mammals, and therefore placing before both groups the hypothetical Saurcmammals in or below the Permian, we come to the old question which Huxley discussed m his famous anniversary-address : "VVcs there a succession between Monotremes, Marsupials, and Placcntals or a parallel development from a common promammaiian type'?" Then we look to the newer questions, "When were the Eden- tates and Cetaceans given off ? " Modern tooth-science springs first from the recent demon- stratioi^ of Rutimeyer's hypothesis of 1869. that the teeth of a the mammals centre around a single reptile-derived type With a single exception, which I believe can be disposed of' various stages of trituberculism or a three-cusped condition have become the standard for the teeth, as pentadactyly has ong been for the feet, except that this is developed within the mammalian stem, while our five fingers are a reptilian legacy. Second, it springs from the recent thorough explora- tion of the youngest jaws for evidences as to the primitive form and s-iccession of the teeth. This also supports the reptile theory of tooth descent by proving, what has been in considerable doubt, that the Promammalia had a multiple sue cession of teeth like the reptiles, and that even some of the modern mammals retain dim traces of four series of teeth The brilliant discoveries of Kukenthal. Leche. and Rose begin to show how in various wa,s the mammals early modified the regular succession of all the teeth by suppression of parts of the multiple series ; this is the first thing to consider The next is how heterodontism arose, how the conic rows of teeth were specialized in different parts of the jaw for three or four functions ; as a certain number of teeth took up each function the question arises whether this number or dental formula was ever the same in all the mammals, for we know it is very differ ent now. After the teeth were thus divided, some functions became more important than others, and established a mo- nopoly, causing first a marked difference in the relative develop ment of the series, which we may express in a dental curve resulting finally in a loss of certain teeth. In the meantime ' began the special evolution of the form of the back teeth, or U7 ii iffjfc t fi I i! lO THE RISE OF THE MAMMALIA molars. Was this alike in all mammals, was it tritubercular ? It is surprising how many problems of early lelationship are at stake in these simple processes. Primitive Diphyodontism. What does succession really consist in ? It now appears that Baume was right in denying that the first tooth is the mother of the second; for the teeth of the lower as well as the upper series, spring from the common epithelial dental fold (Schmclz- leiste) which dips down from the surface and extends the whole length of the jaw ; at intervals it buds off the dental caps (Schmelzkcim) of the first series ; after these are sepa- rated off, the dental fold sinks and buds off the dental caps of the second series, always below and inside the first ; thus the fold is the mother and thr caps are sisters, twins, or. triplets, according to the number of the series. In all young mammals, including the traditional monophyodont Cetaceans and Eden- tates, and excepting only the still unexplored Monotreme embryos, traces of two series of teeth have been found. Both Leche and Rose have detected evidence that the dental fold sometimes buds off parts of a third series, thus explaining the occasional reversion of supernumerary teeth on the inner side of the second series, and Leche has seen traces of budding preceding the first series — thus giving us vestiges of four successions ! All our perplexities as to the relations of the milk and permanent teeth, and the ingenious but mistaken hypotheses of Baume, Flower, Wortman, and Cope have sprung from our want of evidence of the regular and complete diphyodontism of the stem mammals. The solution in brief is that the ' milk teeth ' and the ' true molars ' are descended from the first series, while the second series is represented by the 'permanent incisors, canines, and pre-molars' and rudiments of dental caps beneath the true molars. The mammals early began to diverge from this primitive diphyodontism in many ways ; apparently adapt- ing the first and second series, respectively, to their infant and mature feeding habits; losing parts oi all of one series or the 68 V-H„ • IN NOSTH /l.UER/CA. olher. and in some cases pushing teeth of the second series in In the Marsupials (Kulcenthal) almost the entire first series became permanent ; thus from the Jurassic period to he present t,me only a solitary fourth premolar ^f the eeond Observed that an outer upper-incisor also pushes up from the second ser,es ; the remainder of the second series sfil pi, ! as r„d,me„.al dental caps beneath the first, even b nea'h h' firs and second molars I There are wide variations amo^ the Placentals ; thus in the lowest existing forms, the Insectivora Leche finds that in the Shrew (&,r,) the second serier Z' pressed entirely, while in the Hedgehog iEn„aZ of . t twelve permanent teeth in the anterior part of the jaws five belong to the first series and seven to the second. We thu meet w,th the parado., that among the .primitive' Ma;supial and Insecnvores the regular reptilian succession was early in e cession of two senes was retained in the anterior part of the boTh ^"'=='*/'"= r'°*^ "igbly-specialized molar eeh„ both Marsupials and Placentals, the second teeth were ear'v suppressed although in the Edentates, which also orTgira ly teethTvf I,"""""' '"'" '■' " '^Pi-^^' =--»-on of°seven teeth behind the canine. These discoveries prove that tte whale teeth, like their paddles, have acquired as coll adaptive resemblance to those of the Ichyosaurs. Ho," dW 'everot aearl '7'^ f"" "' '"" Edentates and Ce.aeea develop.' Cleaily by retrogression. As Leche points out in the aquatic Carnivora, in which the first series are Segene^a „ the mgle-series condition (monophyodonti.=m) advances stc";; by step with retrogressive simplification of he too h form (homodontism); thus in the true seals, the eared seals and th^ walruses, as the permanent teeth become simpl , t,r ml teeth become smaller. The Edentat<^<; ^r. ,„• i i genetically, parallel the seals i^tnt ' t^., ^LTt,: the same time. We might jump to the conclusion that th^s 12 THE RISE OF THE MAMMALIA gives us an explanation of the homodont and apparently mono- phyodont condition of the toothed whales, especially as it has been supposed they sprang from aquatic Carnivora, but in this Order matters were reversed, for the first series persisted and the second series were suppressed and persist as a rudimental row of tooth caps buried in the jaw. Each dental series has an adaptive evolution of its own, in Erinaceus the first scries has an ancient and the second a modern form; in Ericulus both series are alike; in the Bats the first series is homodont the second is heterodont (Leche); in the Edentates the first series is ancient and heterodont the second is modern and homodont (Thomas, Rheinhardt); so among the Cetacea and Ungulata. What deep and ancient clefts the different laws of succes- sion mark between the Marsupials and these three Placental groups. Primitive Heterodontism and Formula. Now that all mammals are led back to a distant diphyodont stem, it is alsp true that the further we go back both in palin- genesis and embryogenesis, the more widespread heterodont- ism is — all modern homodontism proving to be secondary. The simple conic teeth of the porpoise, for example, bear a misleading resemblance to those of a reptile. Flower, Weber, Julin, and KUkenthal agree that the ancestral whales and edentates were heterodont and had a smaller number of teeth than the existing forms. Heterodontism is then the second problem. When did the division of the teeth into incisors, premolars, and nolars occur, before or after the Monotremes, Marsupials, and Pla- cental separated .> It is well settled that the canine was the first maxillary tooth, and developed from the most anterior bi-fanged premolar; also, from the discovery of complete succession, we must now define the first molar as the most anterior specialized or triconid tooth, not as the most anterior permanent tooth. It seems to me we now find strong evidence that the stem mammals had a uniform number of each kind 70 ' Himsr: '^ fe.'Mfe'. IN NORTH AMERICA. , - only .0 prtaiUveTtcLot. 1" r.^ft"'. '°™^, ■=<""' forward when „e learn whether or noMh U.LT ,"'" » .ho M„,.i.„bercu.are: 2 I Tr'^^TT'"' "=" Tritubcrculatc teeth rr,„. i, , "'"^ degenerate ceous reeks a larkabT Trt h"™?' '"""' '" ""= "^■'- TMacodon: theTw „rth.-, T ""' '"^'"'^ *"-' "™=» Mar.pia,; . .rnrthtt rthXL;:^::^!' T^'hT an a eharg „T; hS ri" •' " ,"' ^'"'''"''ercu.ates'a"- they were poZfytt^^^l'^'^ ^™ " "^^ l"-' '•--< funetionally akin il the n ^T ,' °"'',r"'" ^■•""•'"lly nor .he MierohUrril' ';^~" W^a7^'?"^"' "" '" and a eondvJe evacHv Mt„ .T. , ^ ^ '''""''' m«hanism trace of ^:x^^:^^:::,^^^ -'-v- Izn pI l}l\ ; .f'";'^'- ^<^ntal formula, like this^ Monotrc.es had a ty;^,'^::;:'^^"*"' ^^"^"^^^ ^°^^^' '"^^ Our next step is to unify the typic c , , . r Marsupials with the ? i a /nfh.VK i? ^' ^' '^ °^ ^'^'^^"^ shown in his studie. nf 5 7 ^^''''^'''^- Thomas has probablylost onlr hefo^toic ,""''''f ^'^^ '"^^^ ^^^ observation, fortunately is n. ^f^'^"' /''^"^"^^'-s (//«• 2); this an embryo^ie ^::^ ^"^Tl^:,^ t'' ' '""^ the Jurassie Marsupials. Thomas raised th'^ ,' '""'°'' '^ tral incisors to five tho wT^ ^ """'''*''' «^ '-inccs- Marsupials Lse therefo" T""'' '"™ among recent formity' whin he shol^d Lr? : M^^'"- f ^ '''"''''' ""^- ' a member of the second s n '^'/.^"'".^"P'^' ^S is probably be reckoned with he It Nov -T""' "^' ^^°"'^ ^^ Placentals have lo t one in ' , ""' '''''^''' ^'^^^ ' ^he lost one mcsor. and one molar abundant 7t 14 THE RISE OF THE MAMMALIA evidence of which is found in Otocyon, Centctes and Homo, we derive as the ancestral formula of both orders : Incisors, 4 ; Canines and Premolars, 5 ; Molars, 4. The aberrant placental Cetacea point in the same direction as we read in the conclusion of Weber's fine memoir: "All the Cetacea sprang from a stem with a heterodont, but onlv partly specialized dentition (something like that of Zeuglodon, 3. I. p. & m : 7), . . . not direct from Carnivores or Ungulates, but from a generalized mammalian type of the Mesozoic period, with some affinities with the Carnivora. . . . Zeuglodon itself branched off extremely early from the primitive line, and the heterodont Squalodon (mark its formula, 3. 1. 4- 7-) "branched off later from the toothed whale line, after the teeth had begun to increase in number and before homodontism had set in." It would be easier for us while speculating to take Squalodon and the Odontocetes directly from the Jurassic mammalian formula (3. i. 4. 8.). As for the multiplication of this formula, we have found the way, says Kiikenthal, by which numerous homodont teeth have arisen from a few heterodont molars, it is by the splitting up of the numerous triconid molars of Jurassic ancestors into three. He substitutes this hypothesis for the one advocated by Baume, Julin, Weber, and Winge, that the multiple cetacean teeth represent the intercalation or joint appearance of both the first and second series of teeth, owing to the elongation of the jaw — a view which is now dis- proved by Kukenthal's discovery of the second row beneath the first. Since even by Kiikenthal's hypothesis the typical Mesozoic mammals could not furnish as many teeth as are found in some of the dolphins, a likelier explanation than his seems to be that as the jaws were elongated the dental fold was carried back and the dental caps were multiplied. The Edentates, like the Cetaceans, point back to hetcro- dontism, and somewhat less clearly to a typical dental formula. We are here indebted to Flower, Rhcinhardt, Thomas, Kuken- thal, and Rose. It is their rudimental and useless first series which gives the evidence of heterodontism. while the second series has become adaptively rootless and homodont. The especially aberrant feature is that a double succession exists \ 7» I IN NORTH AMERICA. ,- A.madUlo presents only eight maxillary teeth, seven of ^ ^ y u u u u u u u u U--S -mmfmmmPT----' f i^A^AT vf/^^^ w -/ ^'""""" -'Z'^" ^'''-^t ''"J Second Series of Teeth which are preceded by two-rooted milk teeth (Tomes) ■ in he embrjo Leche finds fifteen dental caps, of whTch ' v .h.rteen are calcified ; this number probably include the Jot 74 I fl' mm i6 THE RISE OF THE MAMMALIA rudimentary incisors observed by Rheinhardt. In the aber- rant Orycteropus (Aard-Vark), witli ten adult teeth, Thomas finds seven milk teeth behind the maxillary suture (thus takmg us into the molar region of the typical hcterodonts). The last of these milk teeth is large, and two-rooted ; behind this are three large permanent posterior teeth, apparently belonging to the first series. The large lateral tooth of Bradypus is sug- gestive of a canine. From this rapidly accumulating evi- dence it appears probable that the ancestral Edentates had four incisors, a canine and eight or more teeth behind it, tie double succession extending well back so that the first scries did not become permanent at the fifth tooth, behind the canme as m the Marsupials and higher Placentals. If these are primitive conditions, as seems probable from comparison with fossil Edentates, they carry the divergence of the Edentates, like that of the Cetaceans, back into the Mesozoic period. Com- parative anatomy and embryology thus point back to highly varied branches of a generalized placental heterodont stem in the Mesozoic, and a much earlier divergence than we formerly imagined. Now let us see what the early Mesozoic mammals point forward to. There are three distinct and contemporary Jurassic types, the Multituberculates, the Triconodonts, and the Trituber- culates Are -not these the representatives of the Protothena, Mctatheria, and Euthcria ? In the archaic Multituberculates we have seen a monotreme type of jaw and vestiges of a typical ancestral formula. The Triconodonts are a newer group, perhaps derived from the Dromotheriidac (incipient Triconodonts) of the Trias although these appear to be aber- rant ; the typical forms extend from Amphilcstes to Tricon- odon, and exhibit the first stages of development of the inflected Marsupial jaw. The Trituberculates include the Am- phitheriidae and Amblotheriidae with true tuberculo-sectonal lower molars, like those of modern Insectivores ; they alone exhibit the typical angular placental jaw, -no reason can be assigned for calling them Marsupials, excepting the traditional reverence for the Marsupial stem theory. Now, it is very significant that the average dentition of these old but highly 74 T IN NOKT/l AMERrCA. '7 <"iits, 4. I. 4. 7.. TntubcTculatcs, 4. ,. .. g. j, ^,„ ,,,„ malrit diC" " '^^ """ ''"^ °' """"^ """^ ■"" '"= Primilive Trilubemilism. whole f,,lK.s and reptiles as well as mammals, ,„ f„,m ,vlv,t are ealled ■ ,„c„„„d„n. • erowns by .he a.ldi.ion „, la„.,..d ct .0 s,mple cones. I„ the mammals alone, these ,„ree e sp pass ,„t„ h.gher stages of evolution, through what t cal ed •tntuberculy; in whieh these eusps form a trian; ?he d ^ covery of pnmitive wide-spread tritubereuly by Cope was a great step forward. In looking over the odonto,ra phiTof Cuv,e, Owen, Tomes and Banme, we find there is no ., ,sp ion of h,s eommon type around whieh the highly diverse 3 n^aban molars een.re. The molars of the elawcd and hoofrd mammals ean now be eompared, as „e compare the h,,n,l „ foot of the bor,,e ,vitb that of the ea*, beeause they ,spri„™ „" a eommon type. All the specialized mammalian series uT gulates, primates, carnivores, insectivores, rodents, marsupials' re found playmg similar yet independent adapti;e varl'io t ^.TT- .'V""' ""'' "" '''y '" ""= comparison oal molars w,h each other, and with the reptile coL ; take the human grmders for e.vample : the anterior outer euspri^ T. upper jawaud the anterior inner cusps in th o 3,: Ire homologous with each other and with the reptil n Jone Leavmg as.de for the moment the Multitube en le, and Monotreraes, every known triassie, Jurassic, cretae™,, a^d basal eocene fossil (excepting Dicroeynodon) is in some sta^e of trttubereuly i all the known cretaceous molars are simple .-.angles above ; all later fossil mammals also eonvere o .'f tubereuly „,, > ,,, f eocene, every molar is tfi u ta- " lar, and the early stages of divergence are so similar that it fZZt Z T' ''' 1° '"■^""^'"»" "-^ "°- "f ™- oy Wf . '""■ ^"■'"yology supports the evidence of these fosstl ser.es ; thanks .0 the recent admirable researches ,3 THE RISE or THE AfAAfAML/A of Ruse and Tackcr. wc find in the primates, ungulates ami marsupials, that every molar in the calcif.cat.on of ,ts den al caps is heraUlcd by f/ Prof. J. .A. .Allen and Dr. J. L. Wortman kindly as.sisted me in this comparison. 20 THE RISE OF THE MAMMALIA Of Thylacinns, in which a tritubercular molar turns back into a triconodont. In the aquatic Carnivora, the seals, eared seals nnd walruses, the triconodont is also retrogressing into the haplodont. The inference is a fair one that the aquatic, like the terrestrial Carnivora. were originally tntubercular. With the Cetace. both palaeontolo-y and embryology take ur, back to a more or less typical triconodorc molar, not to the tritubercular. The Edentates also giv. feebler evidence of ancestral triconodont or tritubercular molar forms Thus, the tendency of late research is to show that all stem mammals were related in their" double succession in their dental formula, and in their primitive molar form. These features point, not to a succession, but to a unity of ancestry of the Monotrtmes, Marsupials, and Placentals. Divergence of the three Groups. The discovery of the complete double series seems to have removed the last straw from the theory of the Marsupial ancestry of the placentals. for the peculiar mode of suppres- sion of the second series in the Marsupials has been constan since the Purbeck ; this difficulty is added to the structure o the iaw, the epipubic bones, the profoundly different mode of foetal nutrition. None the less, any conclusion we can dravv now as to the primary relations of the three great groups is more or les. of a • Schwindelbau,' and I put together the resnits of these later discoveries with a full realization of the temporary character of present conclusions. The Permian Sauro-Mammalia (Baur) with a multiple suc- cession of simple conical teeth divided into : A. Theromorpha. which lost the succession and in some linos acquired a hetero- dont dentition and triconid single-fanged molars ; B, Pro- "^ThThypothetical lower Triassic Promammalia retained a double secession of the teeth ; they became heterodont, with , incipient triconid double-fanged molars ; dental formula approx- imating 4. 1.4-5.8. They gave rise to ciu^e groups: I.^ The Pfototheria which passed rapidly through che tritubercular 78 IN NORTH AMERICA. j, into the multitubercular molars in the line of Multituberculates, and more slowly into tritubcrculy and its later stages in the hne of Monotremes. li. The Metatiicria or Marsupials tended to suppress the second series of teeth, except those inter- calated with the first ; by this and by reduction the formula became 5. 1.3. 4-6 ; the molars passed slowly throu-h the tri- conodont into the typical tritubercular type. III. The Eutheria or Placentals divided early into a number of branches, in which there was heterodontism, but no uniform modification of succes- sion, namely: A, forms suppressing the second scries in the molar region only, and acquiring a typical Eulherian dentition, 3.1.4- 3-4- I. The Insectivores tended to partly suppress the anterior teeth of the second series or intercalate them with teeth of the first series ; the molars became tritubercular. 2 The higher Placentals retained the succession of tlic first and second series as far back as the first mola,r ; the molars entered rapidly into tritubcrculy aiul ils higlier stages. 13, forms retain- . ing the double succession in part of the molar region, and retaining more of the primitive dentition, 4. i. 4. 8 3 The Edentates branched off from an early triconodont or "tritu- bercular diphyodont stage, with numerous molars, and second- arily suppressed the first heterodont series, and established a numerous homodont second series. 4, The Cetacea also branched off from a diphyodont, heterodont stage, and second- arily established a numerous homodont first series, and sup- pressed the second series. Origin and Evolution of TRiTUKERcuLisri. 'Concrescence' is the newest theory of cusp solution — an expansion by Kukenthal and Rose of views eadier e.x pressed by Gaudry, Magitot and Dybowski. As Kukenthal derives three conical Cetacean teeth by splitting apart a triconodont molar, he conversely derives a triconodont molar by bringing together of three reptile cones. Smith Wood- ward has called attention to the support the epidermal structures of the fishes give to this hypothesis, yet as applied to mammalian teeth, it comes from a one-sided Morphology 22 THE RISE OF THE MAMMALIA which regards only the wonderful though mutilated chapters of Embryology when the untorn pages of palaeontology are at hand. Between the Trias and the Puerco, we are, so to speak, ill at the birth of every successive cusp, and can observe positively that the law of cusp evolution is direct upcro-vth from the smooth slopes of the crown or from the cmgulun^ that fertile parent of new cusps. Each new cusp is usually preceded by an abraded surface, and prophesied by an excessively minute hillock. It follows from this that cusps ran-e in size and height directly according to their age — a principle beautifully demonstrated in some of the Mesozoic teeth If the Kiikenthal-Rose theory were correct, the oldest triconodonts should be iso-conid, whereas we know that the three equal cones of Triconodon arc all a very late develop- ment ; the earlier forms show the lateral cones receding to the needle-points of Dromotherium. _ The tritubercular molar owes its survival to the onginal advanta-c of its triangular form, and to the possibilities of free cusp addition -as worked out by Cope. Wortman, Schlosscr, Scott and myself. Riitimeyer's term, 'trigonodont,' best expresses the primitive structure of the upper and -lower teeth as of two interlocking triangles with their open bases turned outward in the upper and inward in the lower jaw. These 'trigons,' cutting past each other, made a shear so perfect that many Inscct.vora retained it without further evolution. But in most Trituberculatcs a talon was next added to the lower molar (Jurassic stage) as a pestle crushing into the upper valley ; this talon gradually widened mto a broad heel supporting three cusps, as found in the Cretaceous. Consider the extreme antiquity of the three homologous cusps borne upon the back part of the human molar. Th's addition gave the opposed molars two shears and one crusher, and was so perfectly adapted to tlie needs oi Lemurs and many Insectivores and Carnivores, in short, of most clawed animals, that they stopped at this point. Not so with the Ilerbivora, which required more extensive crushing surfaces The upper molars, which had remained triangular throuo-h the Cretaceous and into the basal Eocene, began to IN NORTH AMERICA. n chapters ilogy are re, so to \ observe ipjjro'Vth ingulun\ is usually 1 by an lat cusps iir age — Mcsozoic :hc oldest that the ; develop- ceding to e original ies of free Schlosscr, )nt,' best ind • lower pen bases ^wer jaw. shear so .it further was next e crushing led into a Cetaceous, omologous an molar, ■s and one of Lemurs , of most . Not so e crushing triangular ?, began to develop a little talon, like that early seen in the lower molars and at the same time both upper and lower molars entirely sacrificed their primitive cutting powers, and were converted from secodont into bunodont types by bringing the primitive *o*o»o' o-O-oo-O-o B D > «>0> K>o VAVAV '^Ml^r^-'^MM I J Phylctk nutory of Ihe CusJ-s of the U,tgutate Mo!urs. Trhi r T ' ' "^P'°J^°"'' f ^^'"'^''- ^' P^°todont Stage KDrc.otlurium), 7 assjc. r rr^onodont Stage i^AmphiUstes). D, Tritubercular Stage (SpaU- cothertum) ^. 1 ntubercular-tuberculo Sectorial, Lower Jurassic. /; The sZe m Upper Jurassic. G, The same. i„ Upper Cretaceous. //. The s^I. Pue^o Louxr Locene. /, .Vxitubercular-sexitubercular, Puerco. /. SeSubercular- quadntubercular, Wahsatch. ^ ^exituoercular- trigons down to the level of the talons. At the same time the upper molars acquired intermediate tubercles, and the triangular or oblique arrangement of the tubercles was shifted into the quadrangular or transverse arrangement. This out Ime IS the result of fifteen years' observation. 8i 24 THE RISE OF THE MAMMALIA With square crowns {vs. triangular) and six conic cusps above and below the molars of the Artiodactyl and Perisso- dactyl Herbivora ended their first constructive period at that period, and started upon their modernization. From this point wc dnect our attention upon the numerous combinations of three or four forms assumed by these single cones. The important thing now is to determine at what period these combinations were established, for there is wide difference of opinion as to when ungulate divergence began. To this I refcT later. Taeker has recently shown how every modern embryonic lophodont or selenodont molar first exhibits the archetypal cones of the primitive bunodont. This law, together with my own parallel stu.lics of the evolution of the A B The Limits of VariaAon. A, Merychipfus. B, Acemtherlum. Showing the secondary enamel fokUnss of the crests arising from the centres of the ancestral cones. horse and rhinoceros molars, led me to the discovery that these cmbrvonic primitive cones are also the main growth centers, for, in the upper Miocene, long after the perissoda^iyla have separated from each other, wc see the influence of the arche- typal form in the generic and specific variations of the molars. Compare the teeth of Mcrychippus and of Accrathc- rium, and imagine that you see underlying the diverse crests and crescents the simple bunodont molar of such a form as Hyraeotherium leporinum of the London Clay. You will then notice that the characteristic secondary folds and spurs of the Miocene teeth spring from the old bunodont cones, that the tv'o ' cement lakes ' of Merychippus are equivalent to the two 'fossettes' of Aceratherium, because the 'crcscentic LV NORTH AMERICA. 25 spurs' of the horse and the 'crochet' and 'antccrochet ' of the rhinoceros spring alike from the primitive 'intermediate tubercles.' In view of these discoveries of the uniformity of mam- malian molar type, a uniform terminology has' become as necessary for the dental cusps as for the carpal and tarsal elements of the feet. Professor Gaudry's once admirable system, elaborated in his ' Enc/mincmaits,' was based upon the sufjposcd division of the ungulate molar into the 'first lobe' and 'second lobe,' and is still followed in France. Yet it has two drawbacks : it precludes the comparison of the ungulate with the unguiculate molar, for neither lobe includes the complete triangle ; still more inconvenient is the fact that we cannot compare the higher ungulates with the older Cory- phodons and Pcriptychidac in which the molars were developed upon the triangular plan ; these teeth have only the first lobe and half the second. The upper molars of Hipparion and Coryphodon illustrate the advantages of this new system of comparison and of terminology. Scott has made a further advance in Odontology by working- out the laws of premolar evolution or cusp addition. In many groups we know that from one to four of the premolars grad- ually acquire the exact form of the molars in order to further increase the grinding surface, and we should a priori expect that the cusps would be added in the same order, and therefore be homologous with the molars. This, as Schlosser and myself had observed, is not the case. Scott shows the order of cusp development in the premolars is very nearly the same in all the mammals, and yet is entirely different from the order followed in the molars. This law again unexpectedly ties the clawed and hoofed mammals together; the sequence of cusps in palingenesis is smiilar to that observed by Taeker in embryo- genesis, and Scott is justified in proposing a new terminology (protoeone, deuterocone, trittocone, etc.) for the premolar cusps, which will in the end prove to be a great convenience. I alluded above to the well-known extreme and very confus- ing similarity of the tritubercular molars in the early stages of Trituberculism is at once the cause of clearness 83 divergence 26 THE RISE OF THE MAMMALIA and of doubt when we get back to the stem mammals of wiicly different phyla. This has led to strange misconceptions of phyletic affinities as exemplified in Filhol's division < Pachy- Ihnuricns^ q. supposed mixture of lemur and ungulate stock. There was never any such mixture, and the question comes up how to distinguish unlike forms with like teeth ? I have pro- posed to make use of a dental cun'c which will express the incipient atrophy of some parts, and hypertrophy of other parts of the series, a metatrophism which will naturally terminate in the reduction of some teeth, and excessive development of others. This has not been by any means fully worked out, but I believe it will prove to be of great service in directing atten- tion to some of the initial tendencies of divergence, which are not expressed either in the dental formula or in the patterns of the teeth. Below are some of these curves. When worked Homodont Jurassic Insectivora L>emurs Monkeys Modern Carnivores I i t ' / ' C ' P Examples of Dental Curves. out by the composite method, we will find certain primary curves characteristic of the ordinal divisions, and minor curves distinguishing the lesser divisions. Of course the laws of parallelism will also be found in force here ; flesh-eating, insect- eating, and grass-eating animals will be apt to have similar curves even when evolved in different groups, but here the dental formula and succession will come to our aid. 84 L IN NORTH AMERICA. a; Breaks and Links in the Mesozoic Fauna, By our hypothesis all three sub-classes flourished together during the American Mesozoic; the Marsupials disappeared, then the Monotremes, and by the end of the basal Eocene the Placentals were in exclusive possession of the northern con- tinent. Although we have great reason to congratulate ourselves upon the rapid progress of discovery, there still remain great gaps in Mesozoic time between certain horizons and in the lineal phyletic series of both the Mesozoic and Cenozoic. For a time standard we may take advantage of the remarkably con- stant evolution of the Plagiaulacidae in the Mesozoic, and of the Equidae in the Cenozoic — as certain invertebrates are made use of in older rocks. The grooves and tubercles of Plagiaulax and the cusps and styles of the horses are added with Uie precision of clockwork, and supposing that the rate of evolution has been about the same, we can approximately estimate both the periods of deposition and the intervals as below. plagiaulacidae. ^T u r „ , Stonesfield. Purbeck. Laramie. Puerco. Cernaysua. Number of Premolars, ? 4,3 2 2-1 i Grooves on Premolars, ? -j^ i,.,^ ,2_,^ j^ Molar Tubercles: outer ; inner ; ? 4:2 6:4 6:4 9:6 Estimating the geological intervals by dental evolution and fauiial succession, there is first the great gap between the Trias of Microlestes and Dromotherium and the Jurassic of the Stonesfield slate ; there is a relatively shorter interval, but still a considerable one between this and the Purbeck or Atlan- tosaurus beds. Then follows another long and very important interval between the Atlantosaurus beds and the Laramie (Upper Cretaceous). The gap between the Laramie and Puerco was relatively short as indicated by the comparatively limiteJ evolution both of the Plagiaulacids and Trituberculates. The Puerco itself was a long period in which the Plagiaulacids underwent considerable changes. Then follows an interval 8s < < < < W < K H D4 O ^ W H O 55 O 1—4 tn u u to w H 28 ThE RISE OF THE MAMAfALIA •8 I •s ,ll|«^^ o ft. a; !^|l (/5 o lb s K So = *!< 1 a a a 86 !^ ■§ • .■3-n It rt •§•3 1-3 |1 g"§E SI E lis K S (A o UJ o Ui 0. a 3 ^ O it) < 3 -» llJ .J O o (A < E a Q. 3 I I 3 a •C H !| r«-3 c u .•o ■3'i- » « ^ 2 a ^1 ''•a (/) — . Is ^1 •8 — 'i3 (A o llJ c: o e bi 0. Q. 3 ^ O (A < 3 -» llJ .J O o (A (A < E LJ a a 3 /A- NORTH AMERICA. 49 which it is most important to fill by future exploration, for between the Puerco and th<' Wahsatch the differentiation of the even an^J the odd-toed ungulates must have occurred. The Wahsatch proper does not mark a very extensive evolution of the forms it contains. It passes after a slight break into the base of the Bridger (Wind River) and then begins that splendid and almost uninterrupted succession of lake basin.s, terminat- ing in the pliocene. I append a Table, to be compared with that published by Marsh in his admirable address of 1877, and exhibit the great progress of the last si.\teen years. The general faunal succession is marked by the sudden appcarancj and disappearance of certain series and rise and fall of great groups. In the Trias appears the remarkable pro- todont or primitive-toothed Dromothcrium; wc cannot deter- mine its Order at present. We still have no American fauna corresponding to the intermediate Stonesfield 01 England. In the Jurassic Atlantosaurus beds the three supposed repre- sentatives of the Monotremes (multituberculatcs), Marsupials (triconodonts) and Placentals (trituberculates), appear in equal numbers; the latter are generally characterized by the primitive dental formula. In the Laramie the Multituberculatcs continue in great profusion, and the Marsupials and Placentals are also numerous. The serial succession of the Trituberculates from the Meso- zoic is still an unknown chapter; we are utterly unable to connect the Dromotheriidae of the Trias, the Triconodontidae, Amphitheriidae and Amblotheriidae of the Jura with each other, or with any Cretaceous or lower tertiary mammals. The serial relations of the Multituberculatcs, on the other hand, have been made much clearer by the discovery of the Laramie fauna. Cope and Marsh in this country, and Smith Woodward in England, have at last broken into the long barren Cretaceous. In studying the accurate figures pub- lished by Marsh and a large collection of teeth recently made for the American Museum by Wortman and Peterson, I find that this Laramie fauna is widely separated from the Jurassic in its general evolution, and as Gaudry, Lemoine and Cope have observed, it approaches more nearly the basal Eocene 87 jffii i - ji»jy«lf ' 'TrI» ""M 30 T/Zli RISE or THE MAMMALIA of the Pucrco and the Ccrnaysian of France. The Multitu. bcrculates of the Laramie include the riagi.udaeidae, repre- scntcd by I'tilodiis, the form with two prcnmlars^ and Mcnt- scoessus, with two premolars and cresccntic tubercles. Meni- scotissus has a smaller fourth premolar, and is found to k.id off to the huge plaj;iaulacid Polymastodon of the Pucrco. The only other Multituberculates found are those related to Bolodon of the Jurassic and Chirox of the Pucrco. The other mammals of the Laramie range from the mouse to the opossum in size; they have superior molars of the simple tritubercular type — the low cusped or bunodont molar predominating^ in the upper jaw, and the tuberculo-scctorial in the lower. The dental formula is mostly the typical p. 4, m. 3. Yet, judging by the angular region of Mie jaws, we have here both Placentals and Marsupials. Some of the teeth remind us strongly of those in the Puerco ; their determination, however, is very difficult, for the jaws and teeth arc almost entirely isolated. From another exposure of the Laramie, Cope has recently found the remarkable type Thlaeodon — remarkable because it is a highly specialized trituberculate of typical dentition with a jaw which bears resemblance to that of the Multitu- berculates and of Ornithorhynchus. There is no placental angle nor strong marsupial inflection. This raises the sup. position that Thlaeodon may be one of the persistent trituber- culate Monotremes which we are now looking for. In the Puerco or basal Eocene, a very marked change occurs, for the American fauna loses some of its cosmopolitan character, the multituberculates or monotremes die out and the marsu- pials are not found at all ; in fact they do not reappear in North America until the Miocene. Ancient and Modern Placental Differentiation. The Puerco is essentially an archaic fauna and is to be regarded as the climax of the first period of placental differ- entiation, a culmination of the first attempts of nature to establish insectivorous, carnivorous and herbivorous groups. These attempts began in the Cretaceous, and some of the 88 //.' NORTH A.)rERlCA. 3' types thus prodiuc-d died out in tiic Pucrco, sonic in llic Wahsatch ami IJi i(!;;iT ; only a few flosh-ratcis survived to the Miocene. It is most important to -;rasp dearly the idea of this functional radiation in all directions of this old I'uerco fauna, resulting in forms like the modern inscctivorcs, rodents, bears, doj^s and cats, monkeys, sloths, bunodont and seleno- dont ungulates, and lophodont ungulates. This was an inde- pendent radiation of placcntals, like the Australian radiation of marsupials. What was the cause of the wide-spread extinc- tion of these types .' So far as the ancient clawed types are concerned, their teeth and feet seem to be as fully adaptive in many ca.ses as those of the later unguiculates ; the hoofed types were certainly inferior in tooth evolution, for all their molars evolved on the triangular basis instead of the sexitubcr- cular ; the most sweeping defect of both the clawed and hoofed types, was the apparent incapacity for brain growth, their bodies went on developing while their brains stood still. Thus the stupid giant fauna, the Dinocerata, which rose out of this period, gave way to the small but large-brained modern types. It is noteworthy that the latest survivors of this wreck of an- cicnt life were the large-brained Hyaenodons. Some of the least specialized spurs of this radiation appear to have survived and become the centres of the second or mid- Tertiary radiation from which our modern fauna has evolved. Yet we have not in a single case succeeded in tracing the direct connection. To sum up, we find on the North American continent evidence of the rise and decline and disappearance of monotremes and marsupials, and two great periods of placental radiation, the ancient radiation beginning in the mesozoic, reaching a climax in the Puerco and unknown post-Puerco, and sending its spurs into the higher tertiary, and the modern radia- tion reaching its climax in the Miocene, and sending down to us our existing types. Another Eocene centre was h .r South America, which has of late dimmed the prestige of North America in yielding strange forms of life. One theory of this Patagonian fauna is 89 32 THE RISE or HIE AfAAfAfAUA that it was an independent centre of functional radiation like the Pucrco and Australian, lull of adaptive parallels, but not yielding; to Europe or America any of their older tyijes. lUit Ametihino, to whose energetic rcsearehes we are chiefly in- debted, believes that he fintls a lower I-'.ocenc life zone — a sort of south polar centre — which supplied both America and ICurope. The Puerco he believes is no oldi-r than the Santa- cruzian which in turn is very much older than the Parana and Pampean formations, which lUirmeistcr has made so well known. This yields the Ilomunculus pata;jjonicus which paral- lels Cop^^'s Anaptomorphus in presenting a dentition as ad- vanced in reduction as that of man. Ameghino finds here the ancestors of the Macrauchenidac ; he believes the Ilomolo- dontotheridae are the ancestor? of the Chalicotheriidae — thus deriving a buno-sclenodont from a lophndont type ; the Proterotheriidac, he believes, replace the Condylarthra and Ilyracothcrium in the ancestry of the horses. Similarly the Microbiotheriidac are the stem of the creodonts and carnivores. I cannot coincide with any of these views. The Multitubercu- lates are far older and widely different from the Abdcritcs to which Ameghino traces their ancestry. I fully concur with the opinion of Cope, Zittel, Scott and others that this fauna is of somewhat later age, th;it it was directly connected with Australia and somewhat later with North America, supply- ing us, as has always been supposed, with our sloths. I quote from a recent address by Scott : " The oldest mammals from South An ,'rica are those from Pata- gonia, which Ameghino has referred tc the Eocene, but which arc more probably Oligocene or Miocene. This fauna is of extreme peculiarity and isolation ; it is made up chiefly of edentates, rodents and ungulates of those very aberrant types known as IJtopterna and Toxodontia, which are so widely different from the hoofed mammals of the northern hemisphere ; together with some primitive forms of primates, creodonts and marsupials. The marsupials are of extra- ordinary interest, for they comprise not only forms allied to the opossums, but also to recent Australian forms such as Thylacinus, Dasyurus and Hypsiprymnus. This is a most unexpected fact and seems to point unmistakably to a great southern circumpolar con- tinent." IN XORT/r AMERICA, II The Pucrco thus remains the most extensively known and productive lower eocene centre yet we have very slender threads of positive evidence io connect its fauna with the later plarciUal radiation. The Creodonts of Cope occupy the same relation to the modern inscctivores and carnivores that the Condylarthra do tr the ungulates. The American group has been recently enriched by the discoveries of VV'ortman, and the literature by the careful revision of Scott. This author has divided them into eight families, placing the forms which most resend)lc the Inscctivora in the new family, Oxyclaeniilae. These families illustrate superbly the same law of functional radiation later repeated in the placental and marsupial carnivores. The iMcsonyx family presents some analogies to the Thylaeines. The modern bears arc parallelled in the Arctocyons, with their low tubercular molars; VVortman and myself, with fresh materi- als, have recently added Anacodon to this family, a genus which was doubtfully regarded by Cope as an ancient imgulate. The Cats and Hyaenas arc imitated in the Oxyaenas and Ilyaeno- dons, some of the Miocene forms of which Scott suggests developed aquatic habits; as above noted, some of this family acquired large brains and persisted well into the Miocene. A still more remarkable likeness to the Cats is exhibited in the Palaeonictis family, which, unlike the Hyaenodons, forms its sectorials out of exactly the same teeth as the true cats. The first American Palaeonictis was found two years ago by Wort- man, and this author and myself have suggested that this may be the long-sought ancestor of the Felidae. The Civets are anticipated in the Provivcrridae; yet both Cope and Scott, the highest authorities on this subject, believe that the dog-like Miacidae alone formed the connecting link between the Creo- donta and the true Carnivora. The foot structure of the ancient Puerco ungulates is still only partly known. Cope has divided thes,- animals into the Amblypoda and Co:.dylarthra. The Amblypoda are repre- sented in the Puerco by a large form called Pantolambda, with selenodont triangular upper molars, and possibly by Pcrip- 34 THE RISE OF THE MAMMALIA tychus, with bunodont triangular molars. The Pantolambda molars were, as Cope has shown, converted into those of Cory- phodon, the great lophodont Amblypod of the Wahsatch, by a process exactly analogous to that in which the anterior half of a Palaeotherium molar was formed, that is, they acquired outer and anterior crests but no posterior crests. This Coryphodon molar type was still later converted into the Uin- tatherium type by swinging around the outer crest into a transverse crest. I have recently made a careful study of the fore and hind feet of Coryphodon, and have found that while the fore foot was subdigiiigrade like that of the -lephant, the hind foot was fully plantigrade, the entire sole resting upon the ground. The relation or connection between the Bridger Dinocerata and these earlier Amblypoda is still unknown. The Puerco Periptychus left no descendants. The other ungulates of the Puerco were the Condylarthra, the primitive Phena- codontidae, the supposed ancestors of the Artiodactyls and Pcrissodactyls. Much remains to be done to clear up this question. Succession of the Perissodactyls. In the Wahsatch and Wind River we find not only the last of the Phenacodonts and Coryphodonts and the first of the Dinocerata, but the first of the true Artiodactyl-. and Perisso- dactyls. Recent studies of Cope, Schlosser, Pavlow, Filhol have been directed to the phylogeny of the Perissodactyls with very different conclusions. I agree most closely with Schlosser, and have endeavored to show that the molar teeth give us a key to their natural arrangement as shown in this column ^ Titanotheres. {Horses. Palaeotheres. {Tapirs. Lophiodonts. (Helaletes). /- Hyracodonts. J Amynodonts. I Rhinoceroses. Upon one side the Titanotheres pre- sent the seleno-bunofiont extreme with most analogies to the Artiodactyla in tooth structure and in their truly Artio- dactyl fore feet and bony horns. (If, as Cope supposes, the Diplarthra form a natural group, some Perissodactyls should certainly be more Artiodactyl than others.) The Horses and Palaeo- theres diverge from the buno-selenodont IN NORTH AMERICA. 35 type towards the Lophodont ; they were early separated in foot structure. The Tapirs, Lophiodonts, and Hclaletes show well-marked transverse crests and incipient external crests. This brings us to the other Lophodont extreme, the Rhinoceros- like forms, with complete transverse and external crests. There are many other minor characters which support this as the natural arrangement of the Perissodactyls. I think it can be /trctacont " Jrtloeane /"> rari,ru "imfacone - Euprotogonia. 4th I'lr., 1st M. 'nttfaconu/e Ilyracoth^riuiD. 4th P.n. ist U. tcCartccont f^ fr«"jl. ma^/j^ mu.jlyi. 'ijj^Diljh Anchitherium. Coryphodon. Homologies in the Horse and Coryphodon Molars and Premolars. shown conclusively that these eight or nine scries diverged from each other before the Wahsatch, and that all attempts to derive them from each otner in later periods will break down. They will be found to converge into the unknown Sub- Wahsatch period, to stem forms as indicated by the brackets. 93 36 THE RISE OF THE MAMMALIA One of the most decided reforms in the matter of classifica- tion is the use of the family division. Pre-darwinian writers considered animals as arranged in circles ; post-darwinian writers all regard them as in vertical lines, giving off side branches. Classification should keep pace with phylogeny in palaeontology. Yet there are two clcarl> defined schools of classification to-day. The one, led by Flower, Cope, and Lydekker, practically adheres to the old circular system ; according to this, comprehensive families are formed out of members of different lines of descent which happen to be in the same stage of evolution; lor example, among the ungulates, a horse in the first stage of its evolution is called a lophiodont {i. e., it is placed in the Lophiodontidae), in the next stage it is called a palaeothere {i.e., in the Palaeotheriidae). The extreme application of this method by Cope has led to a total misunderstanding abroad of his real phylogenetic views. The other school, including Schlosscr, Scott, Zittel, and myself, adopt the vertical system, according to which a horse is called a horse, a tapir a tapir, a rhinoceros a rhinoceros, from the moment when they clearly appear as such. I have attempted elsewhere to show that the circufir system is con- fusing, that it ignores the divergence of structure which resulted from thousands of years of physiological isolation ; that finally it is only possible when we define families upon the false system of single characters. In England and France the adherence to the circular system is largely due to traditional reverence for the Lophiodontidae, which has become an omnium gatherum for early odd-toed ungulates — just as were the Pachyderms of Cuvier until Owen proved that that term had no meaning. To-day, no one can say exactly what the Lophiodon itself was ; it appears to have been an aberrant and early extinguished line. In the vertical system the great stages of evolution may be indicated by sub- family divisions, as in the following table. It is practically the same system as that which Flower applies to the existing mammals. All the tendency of recent discovery has been to show that these lines are separate as far back as we can now trace them, 94 IN NORTH AMERICA. 11 that is, to the Wahsatch or Suessonian. Such being the case, we are no more justified in placing the ancient tapirs and horses in one family (Lophiodontidae) than the modern. In the matter of genera, opinion divides on different lines. Flower and Lydekker place all the extinct and modern Rhino- ceroses (except Elasmotherium) in one genus, while Cope Pliocene. Pm = M. I I o e n "3 cr W •c ■O (4 "0 rs c o •a .2 Is c o ■a o ■a o 1 < l4 u o c Miocene. Premolars transforming into Molars ; or Pm = M. White River to Loup Fork. Middle and Upper Eocem:. Premolars Simplor than Molars. Wahsatch to Uinta. Basal Eocene. Puerco to Wahsatch. ■i c a. o o rt Ph *3 W u c < u c O O rt c a, _n c o T3 u c o T3 O O rt O K o ■o o c >^ E 1) c c 1> 11 s O S^ c J= .£ a •a o o o % u t 2 Oh (/I ■5, 'X. a^ C/) X' J u v _c c o •a o e >^ != c u o .s Ic Pi 1) c u 01 (4 o .J3 o S M £3 •s 11 Family Series converging into Condylarthra or Stem Perissodactyla. divides the Rhinoceroses into a large number of genera. Here we are dealing not with great separate lines of descent but with stages of evolution in the same or a few closely parallel lines. If we unite a large number of stages into one genus of Rh'io- ceroses, to be consistent we should do the same with the horses 95 38 THE RISE OF THE MAMMALIA and tapirs. Nevertheless, it is very difficult, if not impossible, to agree as to what shall constitute a generic stage. The Titanotheres have been traced by Cope back to Lanib- dotherium in the Wahsatch ; in the Wind River the true Palaeosyops is found, and in the Bridger this becomes the pre- dominant perissodactyl family, and spreads out into a great variety of forms, which have recently been carefully described by Earle. In the Washakie there are some still larger forms, and Marsh has traced t'.ie line through the teeth of Diplacodon of the Uinta to the true Titanotheres. Still the origin of the flattened skull and remarkable anterior pair of horns has never been known ; Hatcher reports species with very small liorns in the base of the Titanothcrium beds (Lower Miocene). Wort- man has just reported to me the brilliant discovery of upper Eocene (Washakie) Palneosyops with a flattened skull and rudimentary horns just iippcaring upon the nasal- ! This forms the desired connecting link. The early history of the horses, probably starting with the Puerco Condylarth luiprotogonia, and passing through Hyraco- therium, Pach)n()lophus, Epihippus, Mesohippus, is now familiar enough. It is the later history which require^elucidation, and is producing the most unexpected number of parallel lines of horses, out of one of which only our modern horse sprang. Here we are especially indebted to Cope, Pavlow, and Scott. By general consent Hipparion comes out of its old position in tlie true line as displaying the most extreme variations in the crowns of the molar teeth in compensation for the backward evolution of Us feet. Scott has been especially investigating the upper Miocene horses; I quote from the MSS. he has kindly lent me, in which he proposes to remove also the classi- cal Anchitherium of Cuvier. He says : " These American genera, Mesohippus and Miohippus may confidently be regarded as important members of the equine stem, while Anchitherium (of Europe) from present information would appear to belong to an abortive side branch leading to no permanent results." Scott has also discovered an important intermediate form link- ing Miohippus with Protohippus. The Palaeotheres have not been found in America. y6 ^Hh. IN NORTH AMERICA. 39 The Tapir line has been traced by Cope anil myself back to Systcmodon of the VVahsatch, and Isectolophus of the Briilger and Uinta. These forms have simple premolars, but bear the most striking resemblance to the Tapirs in the molars both above and below. All previous attempts to determine the Miocene representatives of the Tapirs have been erroneous. Wortman and Earle have just published an account of two lower Miocene species of true Tapirs, which, both in foot and tooth structure, definitely carry the -American Tapir line up to the middle Miocene, where it is again lost sight of. These species belong to the genus Protapirus, which Filhol has found in the Ol '(jocene of France, thus adding an important geological parallel. The Wahsatch Tapirs were a little larger than the Horses or Hyracotheres which were about the size of a fox, and much smaller than the ancestral Titdotheres. Another family of snail, slender perissodactyls were most nearly allied to the Lophiodons of Europe of any American forms. ' These are the Hclaletidae, distinguished by feet tending to monodactylism, and narrow hoofs like those of the deer ; even in the Wahsatch Heptodon the lateral toes are quite short and raised off the ground. The molars, like those of the lophiodons of Europe, are intermediate between those of the Tapir and the Rhinoceros, but both teeth and feet preclude our uniting the.sc forms either with the Tapirs or with the Hyrachyus family, as Cope has done. The Bridger successor is Helaletes, which Marsh mistakenly supposed was an ancestral Tapir, and the integrity of this line is now firmly establishei' by the dis- covery of the Miocene Colodon. This is described by Marsh as a successor of Helaletes, and Wortman and Earle have just published an account of the teeth and feet, showing that Colodon is widely separated from the contemporary Tapirs, and is the last member of the Heptodon-Helaletes line. The Rhinoceroses of America comprised the true Acera- theriinae and Diceratheriinae, and what may be called the pseudo-rhinoceroses, the Hyracodons and Amynodons; all these forms present the true Rhinoceros molar pattern, but they diverge most widely in the structure of the anterior teeth and of the feet. The Hyracodons first appear in the numerous and 97 1 40 THE RISE OF THE MAMMALIA diversified Hyrachyus of the Bridger, some of which exhibited rudimentary horns upon the back part of the nasals (Colono- ceras) ; they retained a full set of equal-sized incisors and canines, and acquired a horse type of skull, skeleton, .ind locomotion. Scott has well named them the 'cursorial rhi- noceroses.' Colonoceras prob.ibly did not, as Marsh has sug- gested, branch off into Dicerather'um, for the horns of this true rhinoceros are developed at the ends of the nasals ; the Hyra- chyinae sent off as a side branch the deer-like Triplopus of the Washakie, and terminated in the Hyracodons of the lower Miocone. The Amynodons, at the time of their discovery by Marsh, were naturally supposed to be the long-sought Eocene rhinoceroses, but I have shown that no Amynodon can fill this r61e. Garman s discovery of the skull of the remarkable Miocene Metamynodon tended to confirm my views, and I have now to report the discovery of many skulls and a nearly complete skeleton by tne American Museum Expedition. This proves that the Amynodoniidae were remarkable side forms. In wide contrast with the true rhinoceroses, the upper and lower canines d^ivelop into huge, partly recurved tusks, like those of the boar. As in Elasmotherium, the premolars become greatly reduced, and the molars tend to hypsodontism. The lower molars are long and narrow, like those of the anomalous Cadurcotherium of the Oligocene of Europe — it is thus rendered probable that Cadurcotherium is not a sloth, as Filhol has suggested, but is an aberrant rhinoceros, related to, if not identical with, the Amynodons. The hypsodontism in some Metamynodoii teeth is accompanied by a partial loss of enamel. To complete the aberrant character of this family, we find that it has four equal-sized and completely func- tional toes in the forefoot, like those of the Titanotheres, not with the fifth toe reduced as in the contemporary Acera- theria. The true Rhinoceroses, we remember, are distinguished by the entire loss of upper canines. Wortman has just reported finding rudimentary upper canines in both the milk and permanent dentitions of the older Miocene species. The IN NORTH AMERICA. 41 true rhinoceroses suddenly appear in the lower Miocene of America and Oligocene of Europe ; we have not yet traced them back. In a collection of lower Miocene t.kulls recently obtained for the American Museum we find that the premolars are still very simple. In fhe higher Oreodon Beds all traces of the superior canine are lost, and the premolars have become more like the molars. As the origia of the rhinoceroses still remains a mystery, so their later evolution needs clearing up. The American series suddenly terminate in the huge, hornless forms of the upper Miocene. I find there is still no unanimity of opinion in Europe as to the phyletic relationships of the Miocene, Pliocene and existing species. Succession of t'e Artiodactyls. . The Eocene Artiodactyl phylogeny is still far behind that of the perissodactyls, but the Miocene and Pliocene succession has been worked up with great success and clearness by Cope and Scott. The latter says in a recent paper : " All the great groups of Artiodactyla are seen to arise independently from the Buno-Selenodonta which forms as it were a lake, from which several streams, flowing partly in parallel partly in divergent directions, are deriveci." The Elotheriidae appear in Parahyus of the Bridger and Achaenodon of the Washakie, and terminate in the middle Miocene in the gigantic Elotherium ramosum, an animal with a skull three feet long, both the jaws and skull being armed with long branching processes. The true bunodont pigs and peccaries have not yet been found lower than the White River. Scott has traced the Oreodons back to Protoreodon of the top of the Eocene. The aberrant Agriochoeridae, he believes, were doubtfully connected with the true Oreodons by a lower Eocene stem form. The true Oreodons, which exister' in great herds in the lower Miocene, have been divided by Cope and Scott into three parallel lines extending into the Loup Fork, namely, the large Merycochoerus, the medium-sized and more p.imitive IMerychius L.nd the small, highly-specialized Pith- ecistes. 99 42 THE RISE OF THE MAMMALIA The Tragulines are represented by Leptomcryx, Hyper- tragulus and Hypisodus. Lcptomeryx is believed to be a side member of the main family. Here I may speak of the recent discovery of the characters of the Protoceratidae, a new family with a remarkable ensemble of characters. In 189! Marsh described the female skull of Protoceras with a small pair of parietal protuberances. The male skull was found in 1892. It is armed not only by upper canine tusks, but by four pairs of cranial jirotuberances, two of which might be dignified by the name of osseous horns ; it thus presents the armature of an Uintathorium upon a small scale. Besidi s parietal and two pairs of fr ntal protuberances, there are a pair of most exceptional maxillary plates. The fore foot is like that of Tragulus, while the hind foot is didactyl like the deer. We can at present form no idea of its affinities. The oldest American Artiodactyl certainly known is the tritubercular Pantolestes of the VVahsatch. Cope believes the line of American Llamas may have sprung from this, and have been continued through Homacodon of the Bridger. The first undoubted cameloid is Leptotragulus of the Uinta, a comparatively recent discovery. It has strikingly reduced feet for such an early form. Poebrotherium«»)f the White River and John Day has quite the proportions of the living llama ; thence the line passes into Protolabis of the Deep River and John Day. Scott believes that these forms are undoubtedly related to both the camels and llamas, and that in the Loup Fork, perhaps in the two species of Proca- melus, the division occurs, P. angustidens passing into the camels, and P. occidentalis into the llamas. The Pliocene Homocamelus, Holomeniscus and ICschatius, Scott believes may represent a highly specialized side line of camels ; while Pliauchenia, still imperfectly known, may belong on the llama side. The deer represented by Cosoryx and Blastomeryx are, so far as we know, not of American origin, for they first appear in the Upper Miocene at Loup P'ork. LX A'ORTII AMEK/CA. The Ancylopoda, 43 The order Ancylopoda Cope presents the most signal exception to the law of correlation. It is only quite recently that Kilhol, Forsyth Major and Dcpcret have brou^'ht tof^ether the slotli-like phalanges with the ungulate type of teeth of the Chalicotheriidae. Siiue 1825, when Cuvier described the phalanges from ICppelshcini as those of a ^ panqolin gii^nntcsquf,* referring to their deep clefts, and 1833, when Kaup named the teeth, these stnvtures were always considered distinct. It is probable that Moropus and other sujjposed Sloths described by Marsh from our Miocene also belong in this cxception^Al order. As now restored by ' .)1 and myself, this remark- able Chalicotherium had a g, ess clumsy than the Sloth, and something between a huge cat and a hoofed animal; it com- bined the skull of a primitive ungulate with the molars of an eocene titanothere, for the premolars are simple. The limbs, wrist and ankle bones are chiefly ungulate and peris.sodactyl. In viewing this combination of characters, the first question to settle is which set of characters is secondaiy and adaptive. I agree with Deperet, as against I'Mlhol who regards this as an aberrant edentate, that the unguiculate characters are sec- ondary; but I do not believ<_ it is very near the perisso- dactyla. It seems to have sprung rather from the primitive ungulate stem before it had parted with its unguiculate char- acters. Perhaps it came off from the VVahsatcii Mcniscothe- rium, a member of the Condylarthra, which ic very closely resembles in its skull and molar structure and in its dental curve. Marsh, by the way, has just added to our knowledge of this little Wah.satch genus by describing its fore and hind feet, which are more primitive than tho.se of Phenacodus or Hyra.x. While the Creodonta were imitating all modern car- nivores, is it not possible that the Condylarthra gave off a sloth-like form for fossorial and semi-arboreal habits .' Last summer while this problem was being discussed, we we'-'- brought to face with the exact counterpart of Chalicotherium. which may be called a claivcd odd-toed form, by the surprising discovery of a hind foot, which represents a clawed even-toed •WiPVP 44 THE RISE OF THE MAMMALIA animal. This was found by the American Museum party in the Protoceras beds of South Dakota, and has been named Artionyx. This foot has a truly Artiodactyl tarsus and meta- tarsus like that of the pigs or oreodons. Yet it possesses five toes terminating in large unclcft claws. It has been suggested by Wortman and myself that it represents an Artionychinc (even-clawed) division and that Chalicotherium represents a Perissonyv-hinc (odd-clawed) division of the Ancylopoda ; in other words, that a double parallelism exists with the Ungu- lata. Another explanation may be that these genera are highly specialized Artiodactyla and Perissodactyla respectively; Scott has made the ingenious suggestion, tending to support this theory, that the Artionyx foot is the long unknown foot of the aberrant oreodont Agriochoerus of Leidy. This summer will probably determine the truth of this suggestion, for two parties are hunting in the beds in which Agriochoerus and Artionyx occur. Thus an immense number of problems still await solution, and demand the generous cooperation of European and Amer- ican specialists in the use of similar methods of research, in the prompt publication of descriptions and figures, and in the free use of museum collections. I may be pard(ftied for calling general attention to the service which the palaeontological department of the American Museum is trying to render in the immediate publication of stratigraphical and descriptive tables of western horizons and localities. The Factors of Evolution. A few words in conclusion upon the impressions which a study of the rise of the mammalia gives as to the factors of organic evolution. I refer also to recent papers by Cope, Scott and myself. The evolution of a family like the Titanotheres presents an uninterrupted march in one direction. While apparently prosperous and attaining a great size, it was really passing into a great corral of inadaptation to the grasses which were introduced in the middle Miocene. So with other families and lesser lines, extinction came in at the end of a term of >iSi>i^... ■ I i ^u i ,ui...MJW ly i imiW^- . jLI-J /.\' XOKTJl A.yfKRfCA. 45 development and hi};h specialization. With other families no causes for extinction can be assigned, as in the Ini)pin}; off of the smaller Miocene perissodactyls. The point is that ii certain trend of develoi)ment is taken leading to an adaptive or inadaptive final issue — hut extinction or survival of the fittest seems to exert little intluence tit route. The changes en route lead us to believe cither in predes- tination — a kind of internal perfecting tendency, or in kineto- genesis. I'or the trend of evolution is not tlie happy resultant of many trials, but is heralded in structures of the same form all the world over and in age after age, by similar minute changes advancing irresistil)ly from inutility to utility. It i.s an absolutely definite and lawful progression. The infinite number of contemporary developing, degenerating and station- ary characters preclude the possibility of fortuity. There is some law introc'-Jng and regulating each of these variations, as in the variations of imlividual growth. The limits of variation seem to lie partly in what I have called the 'potential of evolution.' As the oiisperni or fertilized ovum is the potential adult, so the ICocene mola'" is the poten- tial Miocene molar. We have seen that the variations of the horse and rhinoceros molars, apparently so diverse, are n.-ally uniform, — is not this evidence that the stem peris.sodactyl had these variations //?/tv///rt//)', waiting to be called forth by certain stimuli } This capacity of similar development under certain stimuli is part of the law of mammalian evolution, but this does not decide the crucial point whether the stimulus is spon- taneous in the germ or inherited from the parent. I incline to the latter opinion. Columbia College, August 3, 1893. 10]