PuntaAltaissituatedaboutthirtymileshigheruponthenorthernsideofthissamebay：itconsistsofasmallplain，betweentwentyandthirtyfeetinheight，cutoffontheshorebyalineoflowcliffsaboutamileinlength，representedinFigure15withitsverticalscalenecessarilyexaggerated。ThelowerbedAismoreextensivethantheupperones;itconsistsofstratifiedgravelorconglomerate，cementedbycalcareo-
arenaceousmatter，andisdividedbycurvilinearlayersofpinkishmarl，ofwhichsomearepreciselyliketosca-rock，andsomemoresandy。Thebedsarecurvilinear，owingtotheactionofcurrents，anddipindifferentdirections;theyincludeanextraordinarynumberofbonesofgiganticmammifersandmanyshells。Thepebblesareofconsiderablesize，andareofhardsandstone，andofquartz，likethatoftheVentana：therearealsoafewwell-roundedmassesoftosca-rock。
ThesecondbedBisaboutfifteenfeetinthickness，buttowardsbothextremitiesofthecliffnotincludedinthediagramiteitherthinsoutanddiesaway，orpassesinsensiblyintoanoverlyingbedofgravel。Itconsistsofred，toughclayeymud，withminutelinearcavities;itismarkedwithfainthorizontalshadesofcolour;itincludesafewpebbles，andrarelyaminuteparticleofshell：inonespot，thedermalarmourandafewbonesofaDasypoidquadrupedwereembeddedinit：itfillsupfurrowsintheunderlyinggravel。Withtheexceptionofthefewpebblesandparticlesofshells，thisbedresemblesthetruePampeanmud;butitstillmorecloselyresemblestheclayeyflatsmentionedintheFirstChapter
separatingthesuccessivelyrisingparallelrangesofsand-dunes。
ThebedCisofstratifiedgravel，likethelowestone;itfillsupfurrowsintheunderlyingredmud，andissometimesinterstratifiedwithit，andsometimesinsensiblypassesintoit;astheredmudthinsout，thisuppergravelthickens。Shellsaremorenumerousinitthaninthelowergravel;
butthebones，thoughsomearestillpresent，arelessnumerous。Inonepart，however，wherethisgravelandtheredmudpassedintoeachother，I
foundseveralbonesandatolerablyperfectheadoftheMegatherium。SomeofthelargeVolutas，thoughembeddedinthegravel-bedC，werefilledwiththeredmud，includinggreatnumbersofthelittlerecentPaludestrinaaustralis。ThesethreelowerbedsarecoveredbyanunconformablemantleD
ofstratifiedsandyearth，includingmanypebblesofquartz，pumiceandphonolite，landandsea-shells。
M。d’OrbignyhasbeensoobligingastonameformethetwentyspeciesofMolluscaembeddedinthetwogravelbeds：theyconsistof：——
1。Volutellaangulata，d’Orbigny，“Voyage“Mollusq。andPal。
2。VolutaBraziliana，Sol3。OlicancilleriaBraziliensisd’Orbigny。
4。Olicancilleriaauricularia，d’Orbigny。
5。Olivinapuelchana，d’Orbigny。
6。Buccinanopscochlidium，d’Orbigny。
7。Buccinanopsglobulosum，d’Orbigny。
8。Colombellasertulariarum，d’Orbigny。
9。TrochusPatagonicus，andvar。ofditto，d’Orbigny。
10。PaludestrinaAustralis，d’Orbigny。
11。FissurellaPatagonica，d’Orbigny。
12。Crepidulamuricata，Lam。
13。Venuspurpurata，Lam。
14。Venusrostrata，Phillippi。
15。MytilusDarwinianus，d’Orbigny。
16。Nuculasemiornata，d’Orbigny。
17。CarditaPatagonica，d’Orbigny。
18。CorbulaPatagonica？，d’Orbigny。
19。Pectentethuelchus，d’Orbigny。
20。Ostreapuelchana，d’Orbigny。
21。AlivingspeciesofBalanus。
22and23。AnAstraeandencrustingFlustra，apparentlyidenticalwithspeciesnowlivinginthebay。
Alltheseshellsnowliveonthiscoast，andmostoftheminthissamebay。
Iwasalsostruckwiththefact，thattheproportionalnumbersofthedifferentkindsappearedtobethesamewiththosenowcastuponthebeach：inbothcasesspecimensofVoluta，Crepidula，Venus，andTrochusarethemostabundant。FourorfiveofthespeciesarethesamewiththeupraisedshellsonthePampasnearBuenosAyres。Allthespecimenshaveaveryancientandbleachedappearance，anddonotemit，whenheated，ananimalodour：someofthemarechangedthroughoutintoawhite，soft，fibroussubstance;othershavethespacebetweentheexternalwalls，eitherhollow，orfilledupwithcrystallinecarbonateoflime。ABulinus，mentionedintheIntroductiontothe“FossilMammalia“inthe“ZoologyoftheVoyageofthe’Beagle’“hassomuchfresheranappearance，thanthemarinespecies，thatIsuspectitmusthavefallenamongsttheothers，andbeencollectedbymistake。
Theremainsoftheextinctmammiferousanimals，fromthetwogravelbedshavebeendescribedbyProfessorOweninthe“ZoologyoftheVoyageofthe’Beagle’：“theyconsistof，1st，onenearlyperfectheadandthreefragmentsofheadsoftheMegatheriumCuvierii;2nd，alowerjawofMegalonyxJeffersonii;3rd，lowerjawofMylodonDarwinii;4th，fragmentsofaheadofsomegiganticEdentalquadruped;5th，analmostentireskeletonofthegreatScelidotheriumleptocephalum，withmostofthebones，includingthehead，vertebrae，ribs，someoftheextremitiestotheclaw-
bone，andeven，asremarkedbyProfessorOwen，theknee-cap，allnearlyintheirproperrelativepositions;6th，fragmentsofthejawandaseparatetoothofaToxodon，belongingeithertoT。Platensis，ortoasecondspecieslatelydiscoverednearBuenosAyres;7th，atoothofEquuscurvidens;8th，toothofaPachyderm，closelyalliedtoPalaeotherium，ofwhichpartsoftheheadhavebeenlatelysentfromBuenosAyrestotheBritishMuseum;inallprobabilitythispachydermisidenticalwiththeMacraucheniaPatagonicafromPortS。Julian，hereaftertobereferredto。
Lastly，and9thly，inacliffoftheredclayeybedB，therewasadoublepiece，aboutthreefeetlongandtwowide，ofthebonyarmourofalargeDasypoidquadruped，withthetwosidespressednearlyclosetogether：asthecliffisnowrapidlywashingaway，thisfossilprobablywaslatelymuchmoreperfect;frombetweenitsdoubled-upsides，Iextractedthemiddleandungualphalanges，unitedtogether，ofoneofthefeet，andlikewiseaseparatephalanx：henceoneormoreofthelimbsmusthavebeenattachedtothedermalcase，whenitwasembedded。Besidestheseseveralremainsinadistinguishablecondition，therewereverymanysinglebones：thegreaternumberwereembeddedinaspace200yardssquare。ThepreponderanceoftheEdentalquadrupedsisremarkable;asis，incontrastwiththebedsofMonteHermoso，theabsenceofRodents。Mostofthebonesarenowinasoftandfriablecondition，and，liketheshells，donotemitwhenburntananimalodour。Thedecayedstateofthebonesmaybepartlyowingtotheirlateexposuretotheairandtidal-waves。Barnacles，Serpulae，andcorallinesareattachedtomanyofthebones，butIneglectedtoobservewhetherthesemightnothavegrownonthemsincebeingexposedtothepresenttidalactionAfterhavingpackedupmyspecimensatBahiaBlanca，thispointoccurredtome，andInotedit;butforgotitonmyreturn，untiltheremainshadbeencleanedandoiled：myattentionhasbeenlatelycalledtothesubjectbysomeremarksbyM。d’Orbigny。;butIbelievethatsomeofthebarnaclesmusthavegrownontheScelidotherium，soonafterbeingdeposited，andbeforebeingWHOLLYcoveredupbythegravel。Besidestheremainsintheconditionheredescribed，Ifoundonesinglefragmentofboneverymuchrolled，andasblackasjet，soasperfectlytoresemblesomeoftheremainsfromMonteHermoso。
Verymanyoftheboneshadbeenbroken，abraded，androlled，beforebeingembedded。Others，evensomeofthoseincludedinthecoarsestpartsofthethenowhardconglomerate，stillretainalltheirminutestprominencesperfectlypreserved;sothatIconcludethattheyprobablywereprotectedbyskin，flesh，orligaments，whilstbeingcoveredup。InthecaseoftheScelidotherium，itisquitecertainthatthewholeskeletonwasheldtogetherbyitsligaments，whendepositedinthegravelinwhichIfoundit。Somecervicalvertebraeandahumerusofcorrespondingsizelaysoclosetogether，asdidsomeribsandthebonesofaleg，thatIthoughtthattheymustoriginallyhavebelongedtotwoskeletons，andnothavebeenwashedinsingle;butasremainswerehereverynumerous，Iwillnotlaymuchstressonthesetwocases。WehavejustseenthatthearmouroftheDasypoidquadrupedwascertainlyembeddedtogetherwithsomeofthebonesofthefeet。
ProfessorEhrenberghasexaminedformespecimensofthefinermatterfromincontactwiththesemammiferousremains：hefindsinthemtwoPolygastrica，decidedlymarineforms;andsixPhytolitharia，ofwhichoneisprobablymarine，andtheotherseitheroffresh-waterorterrestrialorigin。“MonatsberichtenderAkad。zuBerlin“April1845。Thelistconsistsof：——
POLYGASTRICA。
Gallionellasulcata。
Stauropteraaspera？fragm。
PHYTOLITHARIA。
Lithasteriscustuberculatus。
LithostylidiumClepsammidium。
Lithostylidiumquadratum。
Lithostylidiumrude。
Lithostylidiumunidentatum。
Spongolithisacicularis。
OnlyoneoftheseeightmicroscopicalbodiesiscommontotheninefromMonteHermoso：butfiveofthemareincommonwiththosefromthePampeanmudonthebanksoftheParana。Thepresenceofanyfresh-waterinfusoria，consideringthearidityofthesurroundingcountry，ishereremarkable：themostprobableexplanationappearstobe，thatthesemicroscopicalorganismswerewashedoutoftheadjoininggreatPampeanformationduringitsdenudation，andafterwardsredeposited。
Wewillnowseewhatconclusionsmaybedrawnfromthefactsabovedetailed。Itiscertainthatthegravel-bedsandintermediateredmudweredepositedwithintheperiod，whenexistingspeciesofMolluscaheldtoeachothernearlythesamerelativeproportionsastheydoonthepresentcoast。
Thesebeds，fromthenumberoflittoralspecies，musthavebeenaccumulatedinshallowwater;butnot，judgingfromthestratificationofthegravelandthelayersofmarl，onabeach。Fromthemannerinwhichtheredclayfillsupfurrowsintheunderlyinggravel，andisinsomepartsitselffurrowedbytheoverlyinggravel，whilstinotherpartsiteitherinsensiblypassesinto，oralternateswith，thisuppergravel，wemayinferseverallocalchangesinthecurrents，perhapscausedbyslightchanges，upordown，intheleveloftheland。Bytheelevationofthesebeds，towhichperiodthealluvialmantlewithpumice-pebbles，landandsea-shellsbelongs，theplainofPuntaAlta，fromtwentytothirtyfeetinheight，wasformed。Inthisneighbourhoodthereareotherandhighersea-formedplainsandlinesofcliffsinthePampeanformationwornbythedenudingactionofthewavesatdifferentlevels。Hencewecaneasilyunderstandthepresenceofroundedmassesoftosca-rockinthislowestplain;andlikewise，asthecliffsatMonteHermosowiththeirmammiferousremainsstandatahigherlevel，thepresenceoftheonemuch-rolledfragmentofbonewhichwasasblackasjet：possiblysomefewoftheothermuch-rolledbonesmayhavebeensimilarlyderived，thoughIsawonlytheonefragment，inthesameconditionwiththosefromMonteHermoso。M。d’OrbignyhassuggestedthatallthesemammiferousremainsmayhavebeenwashedoutofthePampeanformation，andafterwardsredepositedtogetherwiththerecentshells。
“Voyage“Part。Geolog。page49。Undoubtedlyitisamarvellousfactthatthesenumerousgiganticquadrupeds，belonging，withtheexceptionoftheEquuscurvidens，tosevenextinctgenera，andone，namely，theToxodon，notfallingintoanyexistingfamily，shouldhaveco-existedwithMollusca，allofwhicharestilllivingspecies;butanalogousfactshavebeenobservedinNorthAmericaandinEurope。Inthefirstplace，itshouldnotbeoverlooked，thatmostoftheco-embeddedshellshaveamoreancientandalteredappearancethanthebones。Inthesecondplace，isitprobablethatnumerousbonesnothardenedbysilexoranyothermineral，couldhaveretainedtheirdelicateprominencesandsurfacesperfectiftheyhadbeenwashedoutofonedeposit，andre-embeddedinanother：——thislaterdepositbeingformedoflarge，hardpebbles，arrangedbytheactionofcurrentsorbreakersinshallowwaterintovariouslycurvedandinclinedlayers？Theboneswhicharenowinsoperfectastateofpreservation，must，I
conceive，havebeenfreshandsoundwhenembedded，andprobablywereprotectedbyskin，flesh，orligaments。TheskeletonoftheScelidotheriumindisputablywasdepositedentire：shallwesaythatwhenheldtogetherbyitsmatrixitwaswashedoutofanoldgravel-bedtotallyunlikeincharactertothePampeanformation，andre-embeddedinanothergravel-bed，composedIspeakaftercarefulcomparisonofexactlythesamekindofpebbles，inthesamekindofcement？IwilllaynostressonthetwocasesofseveralribsandbonesoftheextremitieshavingAPPARENTLYbeenembeddedintheirproperrelativeposition：butwillanyonebesoboldastoaffirmthatitispossible，thatapieceofthethintessellatedarmourofaDasypoidquadruped，atleastthreefeetlongandtwoinwidth，andnowsotenderthatIwasunablewiththeutmostcaretoextractafragmentmorethantwoorthreeinchessquare，couldhavebeenwashedoutofonebed，andre-embeddedinanother，togetherwithsomeofthesmallbonesofthefeet，withouthavingbeendashedintoatoms？WemustthenwhollyrejectM。
d’Orbigny’ssupposition，andadmitascertain，thattheScelidotheriumandthelargeDasypoidquadruped，andashighlyprobable，thattheToxodon，Megatherium，etc。someofthebonesofwhichareperfectlypreserved，wereembeddedforthefirsttime，andinafreshcondition，inthestratainwhichtheywerefoundentombed。Thesegiganticquadrupeds，therefore，thoughbelongingtoextinctgeneraandfamilies，coexistedwiththetwentyabove-enumeratedMollusca，thebarnacleandtwocorals，stilllivingonthiscoast。Fromtherolledfragmentofblackbone，andfromtheplainofPuntaAltabeinglowerthanthatofMonteHermoso，Iconcludethatthecoarsesub-littoraldepositsofPuntaAlta，areofsubsequentorigintothePampeanmudofMonteHermoso;andthebedsatthislatterplace，aswehaveseen，areprobablyofsubsequentorigintothehightosca-plainroundtheSierraVentana：weshall，however，return，attheendofthischapter，totheconsiderationoftheseseveralstagesinthegreatPampeanformation。
BUENOSAYRESTOST。FEBAJADA，INENTRERIOS。
ForsomedistancenorthwardofBuenosAyres，theescarpmentofthePampeanformationdoesnotapproachveryneartothePlata，anditisconcealedbyvegetation：butinsectionsonthebanksoftheRiosLuxan，Areco，andArrecifes，IobservedbothpaleanddarkreddishPampeanmud，withsmall，whitishconcretionsoftosca;atalltheseplacesmammiferousremainshavebeenfound。InthecliffsontheParana，atSanNicolas，thePampeanmudcontainsbutlittletosca;hereM。d’OrbignyfoundtheremainsoftworodentsCtenomysBonariensisandKerodonantiquusandthejawofaCanis：
whenontheriverIcouldclearlydistinguishinthisfinelineofcliffs，“horizontallinesofvariationbothintintandcompactness。”Iquotethesewordsfrommynote-book，aswrittendownonthespot，onaccountofthegeneralabsenceofstratificationinthePampeanformationhavingbeeninsistedonbyM。d’Orbignyasaproofofthediluvialoriginofthisgreatdeposit。Theplainnorthwardofthispointisverylevel，butwithsomedepressionsandlakes;IestimateditsheightatfromfortytosixtyfeetabovetheParana。AttheA。MediothebrightredPampeanmudcontainsscarcelyanytosca-rock;whilstatashortdistancethestreamofthePabon，formsacascade，abouttwentyfeetinheight，overacavernousmassoftwovarietiesoftosca-rock;ofwhichoneisverycompactandsemi-
crystalline，withseamsofcrystallisedcarbonateoflime：similarcompactvarietiesaremetwithontheSalidilloandSeco。TheabsoluteidentityI
speakafteracomparisonofmyspecimensbetweensomeofthesevarieties，andthosefromTapalguen，andfromtheridgesouthofBahiaBlanca，adistanceof400milesoflatitude，isverystriking。
AtRosariothereisbutlittletosca-rock：nearthisplaceIfirstnoticedattheedgeoftherivertracesofanunderlyingformation，which，twenty-
fivemileshigherupintheestanciaofGorodona，consistsofapaleyellowishclay，aboundingwithconcretionarycylindersofaferruginoussandstone。Thisbed，whichisprobablytheequivalentoftheoldertertiarymarinestrata，immediatelytobedescribedinEntreRios，onlyjustrisesabovetheleveloftheParanawhenlow。TherestofthecliffatGorodona，isformedofredPampeanmud，with，inthelowerpart，manyconcretionsoftosca，somestalacti-formed，andwithonlyafewintheupperpart：attheheightofsixfeetabovetheriver，twogiganticskeletonsoftheMastodonAndiumwerehereembedded;theirboneswerescatteredafewfeetapart，butmanyofthemstillheldtheirproperrelativepositions：theyweremuchdecayedandassoftascheese，sothatevenoneofthegreatmolarteethfellintopiecesinmyhand。WehereseethatthePampeandepositcontainsmammiferousremainsclosetoitsbase。OnthebanksoftheCarcarana，afewmilesdistant，thelowestbedvisiblewaspalePampeanmud，withmassesoftosca-rock，inoneofwhichIfoundamuchdecayedtoothoftheMastodon：
abovethisbed，therewasathinlayeralmostcomposedofsmallconcretionsofwhitetosca，outofwhichIextractedawellpreserved，butslightlybrokentoothofToxodonPlatensis：abovethistherewasanunusualbedofverysoftimpuresandstone。InthisneighbourhoodInoticedmanysingleembeddedbones，andIheardofothershavingbeenfoundinsoperfectastatethattheywerelongusedasgate-posts：theJesuitFalknerfoundherethedermalarmourofsomegiganticEdentalquadruped。
InsomeoftheredmudscrapedfromatoothofoneoftheMastodonsatGorodona，ProfessorEhrenbergfindssevenPolygastricaandthirteenPhytolitharia，allofthem，Ibelieve，withtwoexceptions，alreadyknownspecies。“Monatsberichtenderkonig。Akad。zuBerlin“April1845。Thelistconsistsof：——
POLYGASTRICA。
Campylodiscusclypeus。
Coscinodiscussubtilis。
Coscinodiscusal。sp。
Eunotia。
Gallionellagranulata。
Himantidiumgracile。
Pinnulariaborealis。
Ofthesetwenty，thepreponderatingnumberareoffresh-waterorigin;onlytwospeciesofCoscinodiscusandaSpongolithisshowthedirectinfluenceofthesea;thereforeProfessorEhrenbergarrivesattheimportantconclusionthatthedepositmusthavebeenofbrackish-waterorigin。OfthethirteenPhytolitharia，ninearemetwithinthetwodepositsinBahiaBlanca，wherethereisevidencefromtwootherspeciesofPolygastricathatthebedswereaccumulatedinbrackishwater。Thetracesofcoral，sponges，andPolythalamia，foundbyDr。Carpenterinthetosca-rockofwhichImustobservethegreaternumberofspecimenswerefromtheupperbedsinthesouthernpartsoftheformation，apparentlyshowamorepurelymarineorigin。
AtST。FEBAJADA，inEntreRios，thecliffs，estimatedatbetweensixtyandseventyfeetinheight，exposeaninterestingsection：thelowerhalfconsistsoftertiarystratawithmarineshells，andtheupperhalfofthePampeanformation。Thelowestbedisanobliquelylaminated，blackish，induratedmud，withdistincttracesofvegetableremains。M。d’Orbigny“Voyage“Part。Geolog。page37，hasgivenadetaileddescriptionofthissection，butashedoesnotmentionthislowestbed，itmayhavebeenconcealedwhenhewastherebytheriver。Thereisaconsiderablediscrepancybetweenhisdescriptionandmine，whichIcanonlyaccountforbythebedsthemselvesvaryingconsiderablyinshortdistances。Abovethisthereisathickbedofyellowishsandyclay，withmuchcrystallisedgypsumandmanyshellsofOstreae，Pectens，andArcae：abovethistheregenerallycomesanarenaceouscrystallinelimestone，butthereissometimesinterposedabed，abouttwelvefeetthick，ofdarkgreen，soapyclay，weatheringintosmallangularfragments。Thelimestone，wherepurest，iswhite，highlycrystalline，andfullofcavities：itincludessmallpebblesofquartz，brokenshells，teethofsharks，andsometimes，asIwasinformed，largebones：itoftencontainssomuchsandastopassintoacalcareoussandstone，andinsuchpartsthegreatOstreaPatagonicachieflyabounds。CaptainSulivan，R。N。hasgivenmeaspecimenofthisshell，whichhefoundinthecliffsatPointCerrito，betweentwentyandthirtymilesabovetheBajada。Intheupperpart，thelimestonealternateswithlayersoffinewhitesand。TheshellsincludedinthesebedshavebeennamedformebyM。d’Orbigny：theyconsistof：——
1。OstreaPatagonica，d’Orbigny，“Voyage“Part。Pal。
2。OstreaAlvarezii，d’Orbigny，“Voyage“Part。Pal。
3。PectenParanensis，d’Orbigny，“Voyage“Part。Pal。
4。PectenDarwinianus，d’Orbigny，“Voyage“Part。Pal。
5。VenusMunsterii，d’Orbigny，“Voyage“Pal。
6。ArcaBonplandiana，d’Orbigny，“Voyage“Pal。
7。CardiumPlatense，d’Orbigny，“Voyage“Pal。
8。Tellina，probablynov。species，buttooimperfectfordescription。
PHYTOLITHARIA。
Lithasteriscustuberculatus。
Lithodontiumbursa。
Lithodontiumfurcatum。
Lithodontiumrostratum。
LithostylidiumAmphiodon。
LithostylidiumClepsammidium。
LithostylidiumHamus。
Lithostylidiumpolyedrum。
Lithostylidiumquadratum。
Lithostylidiumrude。
LithostylidiumSerra。
Lithostylidiumunidentatum。
SpongolithisFustis。
Thesespeciesareallextinct：thesixfirstwerefoundbyM。d’OrbignyandmyselfintheformationsoftheRioNegro，S。Josef，andotherpartsofPatagonia;andtherefore，asfirstobservedbyM。d’Orbigny，thesebedscertainlybelongtothegreatPatagonianformation，whichwillbedescribedintheensuingchapter，andwhichweshallseemustbeconsideredasaveryancienttertiaryone。NorthoftheBajada，M。d’Orbignyfound，inbedswhichheconsidersaslyingbeneaththestrataheredescribed，remainsofaToxodon，whichhehasnamedasadistinctspeciesfromtheT。PlatensisofthePampeanformation。MuchsilicifiedwoodisfoundonthebanksoftheParanaandlikewiseontheUruguay，andIwasinformedthattheycomeoutoftheselowerbeds;fourspecimenscollectedbymyselfaredicotyledonous。
Theupperhalfofthecliff，toathicknessofaboutthirtyfeet，consistsofPampeanmud，ofwhichthelowerpartispale-coloured，andtheupperpartofabrighterred，withsomeirregularlayersofanarenaceousvarietyoftosca，andafewsmallconcretionsoftheordinarykind。Closeabovethemarinelimestone，thereisathinstratumwithaconcretionaryoutlineofwhitehardtosca-rockormarl，whichmaybeconsideredeitherastheuppermostbedoftheinferiordeposits，orthelowestofthePampeanformation;atonetimeIconsideredthisbedasmarkingapassagebetweenthetwoformations：butIhavesincebecomeconvincedthatIwasdeceivedonthispoint。InthesectionontheParana，Ididnotfindanymammiferousremains;butattwomilesdistanceontheA。TapasatributaryoftheConchitas，theywereextremelynumerousinalowcliffofredPampeanmudwithsmallconcretions，preciselyliketheupperbedontheParana。Mostofthebonesweresolitaryandmuchdecayed;butIsawthedermalarmourofagiganticEdentalquadruped，formingacaldron-likehollow，fourorfivefeetindiameter，outofwhich，asIwasinformed，thealmostentireskeletonhadbeenlatelyremoved。IfoundsingleteethoftheMastodonAndium，ToxodonPlatensis，andEquuscurvidens，neartoeachother。Asthislattertoothapproachescloselytothatofthecommonhorse，Ipaidparticularattentiontoitstrueembedment，forIdidnotatthattimeknowthattherewasasimilartoothhiddeninthematrixwiththeothermammiferousremainsfromPuntaAlta。Itisaninterestingcircumstance，thatProfessorOwenfindsthattheteethofthishorseapproachmorecloselyintheirpeculiarcurvaturetoafossilspecimenbroughtbyMr。
LyellfromNorthAmerica，thantothoseofanyotherspeciesofEquus。
Lyell“TravelsinNorthAmerica“volume1page164and“ProceedingsofGeologicalSociety“volume4page39。
Theunderlyingmarinetertiarystrataextendoverawidearea：IwasassuredthattheycanbetracedinravinesinaneastandwestlineacrossEntreRiostotheUruguay，adistanceofabout135miles。InaS。E。
directionIheardoftheirexistenceattheheadoftheR。Nankay;andatP。GordainBandaOriental，adistanceof170miles，Ifoundthesamelimestone，containingthesamefossilshells，lyingataboutthesamelevelabovetheriverasatSt。Fe。Inasoutherlydirection，thesebedssinkinheight，foratanotherP。GordainEntreRios，thelimestoneisseenatamuchlessheight;andtherecanbelittledoubtthattheyellowishsandyclay，onalevelwiththeriver，betweentheCarcaranaandS。Nicholas，belongstothissameformation;asperhapsdothebedsofsandatBuenosAyres，whichlieatthebottomofthePampeanformation，aboutsixtyfeetbeneaththesurfaceofthePlata。Thesoutherlydeclinationofthesebedsmayperhapsbedue，nottounequalelevation，buttotheoriginalformofthebottomofthesea，slopingfromlandsituatedtothenorth;forthatlandexistedatnogreatdistance，wehaveevidenceinthevegetableremainsinthelowestbedatSt。Fe;andinthesilicifiedwoodandinthebonesofToxodonParanensis，foundaccordingtoM。d’Orbignyinstilllowerstrata。
BANDAORIENTAL。
ThisprovinceliesonthenorthernsideofthePlata，andeastwardoftheUruguay：ithasagentleundulatorysurface，withabasisofprimaryrocks;
andisinmostpartscoveredupwithanunstratifiedmass，ofnogreatthickness，ofreddishPampeanmud。Intheeasternhalf，nearMaldonado，thisdepositismorearenaceousthaninthePampas，itcontainsmanythoughsmallconcretionsofmarlortosca-rock，andothersofhighlyferruginoussandstone;inonesection，onlyafewyardsindepth，itrestedonstratifiedsand。NearMonteVideothisdepositinsomespotsappearstobeofgreaterthickness;andtheremainsoftheGlyptodonandotherextinctmammifershavebeenfoundinit。Inthelonglineofcliffs，betweenfiftyandsixtyfeetinheight，calledtheBarrancasdeS。Gregorio，whichextendwestwardoftheRioS。Lucia，thelowerhalfisformedofcoarsesandofquartzandfeldsparwithoutmica，likethatnowcastuponthebeachnearMaldonado;andtheupperhalfofPampeanmud，varyingincolourandcontaininghoneycombedveinsofsoftcalcareousmatterandsmallconcretionsoftosca-rockarrangedinlines，andlikewiseafewpebblesofquartz。Thisdepositfillsuphollowsandfurrowsintheunderlyingsand;
appearingasifwaterchargedwithmudhadinvadedasandybeach。Thesecliffsextendfarwestward，andatadistanceofsixtymiles，nearColoniadelSacramiento，IfoundthePampeandepositrestinginsomeplacesonthissand，andinothersontheprimaryrocks：betweenthesandandthereddishmud，thereappearedtobeinterposed，butthesectionwasnotaverygoodone，athinbedofshellsofanexistingMytilus，stillpartiallyretainingtheircolour。ThePampeanformationinBandaOrientalmightreadilybemistakenforanalluvialdeposit：comparedwiththatofthePampas，itisoftenmoresandy，andcontainssmallfragmentsofquartz;theconcretionsaremuchsmaller，andtherearenoextensivemassesoftosca-rock。
Intheextremewesternpartsofthisprovince，betweentheUruguayandalinedrawnfromColoniatotheR。PerdidoatributaryoftheR。Negro，theformationsarefarmorecomplicated。Besidesprimaryrocks，wemeetwithextensivetractsandmanyflat-topped，horizontallystratified，cliff-
bounded，isolatedhillsoftertiarystrata，varyingextraordinarilyinmineralogicalnature，someidenticalwiththeoldmarinebedsofSt。FeBajada，andsomewiththoseofthemuchmorerecentPampeanformation。
Thereare，also，extensiveLOWtractsofcountrycoveredwithadepositcontainingmammiferousremains，preciselylikethatjustdescribedinthemoreeasternpartsoftheprovince。Althoughfromthesmoothandunbrokenstateofthecountry，Ineverobtainedasectionofthislatterdepositclosetothefootofthehighertertiaryhills，yetIhavenottheleastdoubtthatitisofquitesubsequentorigin;havingbeendepositedaftertheseahadwornthetertiarystrataintothecliff-boundedhills。Thislaterformation，whichiscertainlytheequivalentofthatofthePampas，iswellseeninthevalleysintheestanciaofBerquelo，nearMercedes;ithereconsistsofreddishearth，fullofroundedgrainsofquartz，andwithsomesmallconcretionsoftosca-rockarrangedinhorizontallines，soasperfectlytoresemble，exceptincontainingalittlecalcareousmatter，theformationintheeasternpartsofBandaOriental，inEntreRios，andatotherplaces：inthisestanciatheskeletonofagreatEdentalquadrupedwasfound。InthevalleyoftheSarandis，atthedistanceofonlyafewmiles，thisdeposithasasomewhatdifferentcharacter，beingwhiter，softer，finer-grained，andfulloflittlecavities，andconsequentlyoflittlespecificgravity;nordoesitcontainanyconcretionsorcalcareousmatter：Ihereprocuredahead，whichwhenfirstdiscoveredmusthavebeenquiteperfect，oftheToxodonPlatensis，anotherofaMylodonThisheadwasatfirstconsideredbyProfessorOweninthe“Zoologyofthe’Beagle’s’Voyage“asbelongingtoadistinctgenus，namely，Glossotherium。，perhapsM。Darwinii，andalargepieceofdermalarmour，differingfromthatoftheGlyptodonclavipes。Thesebonesareremarkablefromtheirextraordinarilyfreshappearance;whenheldoveralampofspiritsofwine，theygiveoutastrongodourandburnwithasmallflame;
Mr。T。Reekshasbeensokindastoanalysesomeofthefragments，andhefindsthattheycontainabout7percentofanimalmatter，and8percentofwater。Liebig“ChemistryofAgriculture“page194statesthatfreshdrybonescontainfrom32to33percentofdrygelatine。SeealsoDr。Daubeny，in“EdinburghNewPhilosophicalJournal“volume37page293。
Theoldertertiarystrata，formingthehigherisolatedhillsandextensivetractsofcountry，vary，asIhavesaid，extraordinarilyincomposition：
withinthedistanceofafewmiles，Isometimespassedovercrystallinelimestonewithagate，calcareoustuffs，andmarlyrocks，allpassingintoeachother，——redandpalemudwithconcretionsoftosca-rock，quitelikethePampeanformation，——calcareousconglomeratesandsandstones，——brightredsandstonespassingeitherintoredconglomerate，orintowhitesandstone，——hardsiliceoussandstones，jasperyandchalcedonicrocks，andnumerousothersubordinatevarieties。Iwasunabletomarkouttherelationsofallthesestrata，andwilldescribeonlyafewdistinctsections：——inthecliffsbetweenP。GordaontheUruguayandtheA。deVivoras，theupperbediscrystallinecellularlimestoneoftenpassingintocalcareoussandstone，withimpressionsofsomeofthesameshellsasatSt。
FeBajada;atP。Gorda，thislimestoneisinterstratifiedwithandrestson，whitesand，whichcoversabedaboutthirtyfeetthickofpale-colouredclay，withmanyshellsofthegreatOstreaPatagonicaInmy“Journal“page1711stedition，IhavehastilyandinaccuratelystatedthatthePampeanmud，whichisfoundovertheeasternpartofB。Oriental，liesOVERthelimestoneatP。Gorda;Ishouldhavesaidthattherewasreasontoinferthatitwasasubsequentorsuperiordeposit。：beneaththis，intheverticalcliff，nearlyonalevelwiththeriver，thereisabedofredmudabsolutelylikethePampeandeposit，withnumerousoftenlargeconcretionsofperfectlycharacterisedwhite，compacttosca-rock。AtthemouthoftheVivoras，theriverflowsoverapalecavernoustosca-rock，quitelikethatinthePampas，andthisAPPEAREDtounderliethecrystallinelimestone;butthesectionwasnotunequivocallikethatatP。Gorda。Thesebedsnowformonlyanarrowandmuchdenudedstripofland;buttheymustoncehaveextendedmuchfurther;foronthenextstream，southoftheS。Juan，CaptainSulivan，R。N。foundalittlecliff，onlyjustabovethesurfaceoftheriver，withnumerousshellsoftheVenusMunsterii，D’Orbigny，——oneofthespeciesoccurringatSt。Fe，andofwhichtherearecastsatP。Gorda：
thelineofcliffsofthesubsequentlydepositedtruePampeanmud，extendfromColoniatowithinhalfamileofthisspot，andnodoubtoncecoveredupthisdenudedmarinestratum。AgainatColonia，aFrenchmanfound，indiggingthefoundationsofahouse，agreatmassoftheOstreaPatagonicaofwhichIsawmanyfragments，packedtogetherjustbeneaththesurface，anddirectlysuperimposedonthegneiss。Thesesectionsareimportant：M。
d’OrbignyisunwillingtobelievethatbedsofthesamenaturewiththePampeanformationeverunderlietheancientmarinetertiarystrata;andI
wasasmuchsurprisedatitashecouldhavebeen;buttheverticalcliffatP。Gordaallowedofnomistake，andImustbepermittedtoaffirm，thatafterhavingexaminedthecountryfromtheColoradotoSt。FeBajada，I
couldnotbedeceivedinthemineralogicalcharacterofthePampeandeposit。
Moreover，inaprecipitouspartoftheravineofLasBocas，aredsandstoneisdistinctlyseentooverlieathickbedofpalemud，alsoquitelikethePampeanformation，aboundingwithconcretionsoftruetosca-rock。Thissandstoneextendsovermanymilesofcountry：itisasredasthebrightestvolcanicscoriae;itsometimespassesintoacoarseredconglomeratecomposedoftheunderlyingprimaryrocks;andoftenpassesintoasoftwhitesandstonewithredstreaks。AttheCaleradelosHuerfanos，onlyaquarterofamilesouthofwhereIfirstmetwiththeredsandstone，thecrystallinewhitelimestoneisquarried：asthisbedistheuppermost，andasitoftenpassesintocalcareoussandstone，interstratifiedwithpuresand;andastheredsandstonelikewisepassesintosoftwhitesandstone，andisalsotheuppermostbed，Ibelievethatthesetwobeds，thoughsodifferent，areequivalents。Afewleaguessouthwardofthesetwoplaces，oneachsideofthelowprimaryrangeofS。Juan，therearesomeflat-topped，cliff-bounded，separatelittlehills，verysimilartothosefringingtheprimaryrangesinthegreatplainsouthofBuenosAyres：theyarecomposed——
1st，ofcalcareoustuffwithmanyparticlesofquartz，sometimespassingintoacoarseconglomerate;2nd，ofastoneundistinguishableontheclosestinspectionfromthecompactervarietiesoftosca-rock;and3rd，ofsemi-crystallinelimestone，includingnodulesofagate：thesethreevarietiespassinsensiblyintoeachother，andastheyformtheuppermoststratuminthisdistrict，Ibelievethatthey，also，aretheequivalentsofthepurecrystallinelimestone，andoftheredandwhitesandstonesandconglomerates。
BetweenthesepointsandMercedesontheRioNegro，therearescarcelyanygoodsections，theroadpassingoverlimestone，tosca-rock，calcareousandbrightredsandstones，andnearthesourceoftheSanSalvadoroverawideextentofjasperyrocks，withmuchmilkyagate，likethatinthelimestonenearSanJuan。IntheestanciaofBerquelo，theseparate，flat-topped，cliff-boundedhillsareratherhigherthanintheotherpartsofthecountry;theyrangeinaN。E。andS。W。direction;theiruppermostbedsconsistofthesamebrightredsandstone，passingsometimesintoaconglomerate，andinthelowerpartintosoftwhitesandstone，andevenintoloosesand：beneaththissandstone，Isawintwoplaceslayersofcalcareousandmarlyrocks，andinoneplaceredPampean-likeearth;atthebaseofthesesections，therewasahard，stratified，whitesandstone，withchalcedoniclayers。NearMercedes，bedsofthesamenatureandapparentlyofthesameage，areassociatedwithcompact，white，crystallinelimestone，includingmuchbotryoidalagate，andsingularmasses，likeporcelain，butreallycomposedofacalcareo-siliceouspaste。Insinkingwellsinthisdistrictthechalcedonicstrataseemtobethelowest。Beds，suchastheredescribed，occuroverthewholeofthisneighbourhood;buttwentymilesfurtheruptheR。Negro，inthecliffsofPerika，whichareaboutfiftyfeetinheight，theupperbedisaprettilyvariegatedchalcedony，mingledwithapurewhitetallowylimestone;beneaththisthereisaconglomerateofquartzandgranite;beneaththismanysandstones，somehighlycalcareous;andthewholelowertwo-thirdsofthecliffconsistsofearthycalcareousbedsofvariousdegreesofpurity，withonelayerofreddishPampean-likemud。
Whenexaminingtheagates，thechalcedonicandjasperyrocks，someofthelimestones，andeventhebrightredsandstones，Iwasforciblystruckwiththeirresemblancetodepositsformedintheneighbourhoodofvolcanicaction。InowfindthatM。Isabelle，inhis“VoyageaBuenosAyres，“hasdescribedcloselysimilarbedsonItaquyandIbicuywhichentertheUruguaysomewaynorthoftheR。Negroandthesebedsincludefragmentsofreddecomposedtruescoriaehardenedbyzeolite，andofblackretinite：wehavethenheregoodevidenceofvolcanicactionduringourtertiaryperiod。
Stillfurthernorth，nearS。Anna，wheretheParanamakesaremarkablebend，M。Bonplandfoundsomesingularamygdaloidalrocks，whichperhapsmaybelongtothissameepoch。M。d’Orbigny“Voyage“Part。Geolog。page29I
mayremarkthat，judgingfromthesizeandwell-roundedconditionoftheblocksofrockintheabove-describedconglomerates，massesofprimaryformationprobablyexistedatthistertiaryperiodabovewater：thereis，also，accordingtoM。Isabelle，muchconglomeratefurthernorth，atSalto。
>FromwhateversourceandthroughwhatevermeansthegreatPampeanformationoriginated，weherehave，Imustrepeat，unequivocalevidenceofasimilaractionataperiodbeforethatofthedepositionofthemarinetertiarystratawithextinctshells，atSantaFeandP。Gorda。Duringalsothedepositionofthesestrata，wehaveintheintercalatedlayersofredPampean-likemudandtosca-rock，andinthepassagenearS。Juanofthesemi-crystallinelimestoneswithagateintotoscaundistinguishablefromthatofthePampas，evidenceofthesameaction，thoughcontinuedonlyatintervalsandinafeeblemanner。Wehavefurtherseenthatinthisdistrict，ataperiodnotonlysubsequenttothedepositionofthetertiarystrata，buttotheirupheavementandmostextensivedenudation，truePampeanmudwithitsusualcharactersandincludingmammiferousremains，wasdepositedroundandbetweenthehillsorisletsformedofthesetertiarystrata，andoverthewholeeasternandlowprimarydistrictsofBandaOriental。
EARTHYMASS，WITHEXTINCTMAMMIFEROUSREMAINS，OVERTHEPORPHYRITICGRAVEL
ATS。JULIAN，LATITUDE49DEGREES14’S。INPATAGONIA。
FIGURE16。SECTIONOFTHELOWESTPLAINATPORTS。JULIAN。
Sectionthroughbedsfromtoptobottom：A，B，C，D，E，F。
AA。Superficialbedofreddishearth，withtheremainsoftheMacrauchenia，andwithrecentsea-shellsonthesurface。
B。Gravelofporphyriticrocks。
C。andD。Pumiceousmudstone——Ancienttertiaryformation。
E。andF。Sandstoneandargillaceousbeds——Ancienttertiaryformation。
Thiscase，thoughnotcomingstrictlyunderthePampeanformation，maybeconvenientlygivenhere。Onthesouthsideoftheharbour，thereisanearlylevelplainmentionedintheFirstChapteraboutsevenmileslong，andthreeorfourmileswide，estimatedatninetyfeetinheight，andborderedbyperpendicularcliffs，ofwhichasectionisrepresentedinFigure16。
Theloweroldtertiarystratatobedescribedinthenextchapterarecoveredbytheusualgravelbed;andthisbyanirregularearthy，sometimessandymass，seldommorethantwoorthreefeetinthickness，exceptwhereitfillsupfurrowsorgullieswornnotonlythroughtheunderlyinggravel，buteventhroughtheuppertertiarybeds。Thisearthymassisofapalereddishcolour，likethelesspurevarietiesofPampeanmudinBandaOriental;itincludessmallcalcareousconcretions，likethoseoftosca-
rockbutmorearenaceous，andotherconcretionsofagreenish，induratedargillaceoussubstance：afewpebbles，also，fromtheunderlyinggravel-bedarealsoincludedinit，andthesebeingoccasionallyarrangedinhorizontallines，showthatthemassisofsub-aqueousorigin。Onthesurfaceandembeddedinthesuperficialparts，therearenumerousshells，partiallyretainingtheircolours，ofthreeorfourofthenowcommonestlittoralspecies。NearthebottomofonedeepfurrowrepresentedinFigure16，filledupwiththisearthydeposit，IfoundalargepartoftheskeletonoftheMacraucheniaPatachonica——agiganticandmostextraordinarypachyderm，allied，accordingtoProfessorOwen，tothePalaeotherium，butwithaffinitiestotheRuminants，especiallytotheAmericandivisionoftheCamelidae。Severalofthevertebraeinachain，andnearlyallthebonesofoneofthelimbs，eventothesmallestbonesofthefoot，wereembeddedintheirproperrelativepositions：hencetheskeletonwascertainlyunitedbyitsfleshorligaments，whenenvelopedinthemud。Thisearthymass，withitsconcretionsandmammiferousremains，fillingupfurrowsintheunderlyinggravel，certainlypresentsaverystrikingresemblancetosomeofthesectionsforinstance，atP。AltainB。Blanca，orattheBarrancasdeS。GregoriointhePampeanformation;butImustbelievethatthisresemblanceisonlyaccidental。Isuspectthatthemudwhichatthepresentdayisaccumulatingindeepandnarrowgulliesattheheadoftheharbour，would，afterelevation，presentaverysimilarappearance。ThesouthernmostpartofthetruePampeanformation，namely，ontheColorado，lies560milesoflatitudenorthofthispoint。InthesucceedingchapterIshallhavetorefertoagreatdepositofextinctmammiferousremains，latelydiscoveredbyCaptainSulivan，R。N。atapointstillfurthersouth，namely，attheR。Gallegos;theiragemustatpresentremaindoubtful。
WithrespecttotheageoftheMacrauchenia，theshellsonthesurfaceprovethatthemassinwhichtheskeletonwasenvelopedhasbeenelevatedabovetheseawithintherecentperiod：Ididnotseeanyoftheshellsembeddedatasufficientdepthtoassuremethoughitbehighlyprobable
thatthewholethicknessofthemasswascontemporaneouswiththeseINDIVIDUALSPECIMENS。ThattheMacrauchenialivedsubsequentlytothespreadingoutofthegravelonthisplainiscertain;andthatthisgravel，attheheightofninetyfeet，wasspreadoutlongaftertheexistenceofrecentshells，isscarcelylesscertain。For，itwasshownintheFirstChapter，thatthislineofcoasthasbeenupheavedwithremarkableequability，andthatoveravastspacebothnorthandsouthofS。Julian，recentspeciesofshellsarestrewedonorembeddedinthesurfaceofthe250feetplain，andofthe350feetplainuptoaheightof400feet。Thesewidestep-formedplainshavebeenformedbythedenudingactionofthecoast-wavesontheoldtertiarystrata;andtherefore，whenthesurfaceofthe350feetplain，withtheshellsonit，firstroseabovethelevelofthesea，the250feetplaindidnotexist，anditsformation，aswellasthespreadingoutofthegravelonitssummit，musthavetakenplacesubsequently。Soalsothedenudationandthegravel-coveringofthe90feetplainmusthavetakenplacesubsequentlytotheelevationofthe250feetplain，onwhichrecentshellsarealsostrewed。HencetherecannotbeanydoubtthattheMacrauchenia，whichcertainlywasentombedinafreshstate，andwhichmusthavebeenaliveafterthespreadingoutofthegravelonthe90feetplain，existed，notonlysubsequentlytotheupraisedshellsonthesurfaceofthe250feetplain，butalsotothoseonthe350to400feetplain：theseshells，eightinnumbernamely，threespeciesofMytilus，twoofPatella，oneFusus，Voluta，andBalanus，areundoubtedlyrecentspecies，andarethecommonestkindsnowlivingonthiscoast。AtPuntaAltainB。Blanca，Iremarkedhowmarvellousitwas，thattheToxodon，amammifersounliketoallknowngenera，shouldhaveco-existedwithtwenty-
threestilllivingmarineanimals;andnowwefindthattheMacrauchenia，aquadrupedonlyalittlelessanomalousthantheToxodon，alsoco-existedwitheightotherstillexistingMollusca：itshould，moreover，beborneinmind，thatatoothofapachydermatousanimalwasfoundwiththeotherremainsatPuntaAlta，whichProfessorOwenthinksalmostcertainlybelongedtotheMacrauchenia。
Mr。LyellhasarrivedatahighlyimportantconclusionwithrespecttotheageoftheNorthAmericanextinctmammifersmanyofwhicharecloselyalliedto，andevenidenticalwith，thoseofthePampeanformation，namely，thattheylivedsubsequentlytotheperiodwhenerraticbouldersweretransportedbytheagencyoffloatingiceintemperatelatitudes。
“GeologicalProceedings“volume4page36。NowinthevalleyoftheSantaCruz，onlyfiftymilesoflatitudesouthofthespotwheretheMacraucheniawasentombed，vastnumbersofgigantic，angularboulders，whichmusthavebeentransportedfromtheCordilleraonicebergs，liestrewedontheplain，attheheightof1，400feetabovethelevelofthesea。Inascendingtothislevel，severalstep-formedplainsmustbecrossed，allofwhichhavenecessarilyrequiredlongtimefortheirformation;hencethelowestorninetyfeetplain，withitssuperficialbedcontainingtheremainsoftheMacrauchenia，musthavebeenformedverylongsubsequentlytotheperiodwhenthe1，400feetplainwasbeneaththesea，andbouldersweredroppedonitfromfloatingmassesofice。Itmustnotbeinferredfromtheseremarks，thattheice-actionceasedinSouthAmericaatthiscomparativelyancientperiod;forinTierradelFuegoboulderswereprobablytransportedcontemporaneouslywith，ifnotsubsequentlyto，theformationoftheninetyfeetplainatS。Julian，andatotherpartsofthecoastofPatagonia。Mr。
Lyell’sconclusion，therefore，isthusfarconfirmedinthesouthernhemisphere;anditisthemoreimportant，asoneisnaturallytemptedtoadmitsosimpleanexplanation，thatitwastheice-periodthatcausedtheextinctionofthenumerousgreatmammiferswhichsolatelyswarmedoverthetwoAmericas。
SUMMARYANDCONCLUDINGREMARKSONTHEPAMPEANFORMATION。
Oneofitsmoststrikingfeaturesisitsgreatextent;IpassedcontinuouslyoveritfromtheColoradotoSt。FeBajada，adistanceof500
geographicalmiles;andM。d’Orbignytraceditfor250milesfurthernorth。
InthelatitudeofthePlata，Iexaminedthisformationatintervalsoveraneastandwestlineof300milesfromMaldonadototheR。Carcarana;andM。d’Orbignybelievesitextends100milesfurtherinland：fromMr。
Caldcleugh’stravels，however，Ishouldhavethoughtthatithadextended，southoftheCordoveserange，tonearMendoza，andImayaddthatIheardofgreatboneshavingbeenfoundhighuptheR。Quinto。HencetheareaofthePampeanformation，asremarkedbyM。d’Orbigny，isprobablyatleastequaltothatofFrance，andperhapstwiceorthriceasgreat。Inabasin，surroundedbygravel-cliffataheightofnearlythreethousandfeet，southofMendoza，thereis，asdescribedintheThirdChapter，adepositverylikethePampean，interstratifiedwithothermatter;andagainatS。
Julian’s，inPatagonia，560milessouthoftheColorado，asmallirregularbedofanearlysimilarnaturecontains，aswehavejustseen，mammiferousremains。IntheprovincesofMoxosandChiquitos1，000milesnorthwardofthePampas，andinBolivia，ataheightof4，000metres，M。d’Orbignyhasdescribedsimilardeposits，whichhebelievestohavebeenformedbythesameagencycontemporaneouslywiththePampeanformation。Consideringtheimmensedistancesbetweentheseseveralpoints，andtheirdifferentheights，itappearstomeinfinitelymoreprobable，thatthissimilarityhasresultednotfromcontemporaneousnessoforigin，butfromthesimilarityoftherockyframeworkofthecontinent：itisknownthatinBrazilanimmenseareaconsistsofgneissicrocks，andweshallhereaftersee，overhowgreatalengththeplutonicrocksoftheCordillera，theoverlyingpurpleporphyries，andthetrachyticejections，arealmostidenticalinnature。
ThreetheoriesontheoriginofthePampeanformationhavebeenpropounded：——First，thatofagreatdebaclebyM。d’Orbigny;thisseemsfoundedchieflyontheabsenceofstratification，andonthenumberofembeddedremainsofterrestrialquadrupeds。AlthoughthePampeanformationlikesomanyargillaceousdepositsisnotdividedintodistinctandseparatestrata，yetwehaveseenthatinonegoodsectionitwasstripedwithhorizontalzonesofcolour，andthatinseveralspecifiedplacestheupperandlowerpartsdiffered，notonlyconsiderablyincolour，butgreatlyinconstitution。InthesouthernpartofthePampastheuppermasstoacertainextentstratifiedgenerallyconsistsofhardtosca-rock，andthelowerpartofredPampeanmud，oftenitselfdividedintotwoormoremasses，varyingincolourandinthequantityofincludedcalcareousmatter。InWesternBandaOriental，bedsofasimilarnature，butofagreaterage，conformablyunderlieandareintercalatedwiththeregularlystratifiedtertiaryformation。Asageneralrule，themarlyconcretionsarearrangedinhorizontallines，sometimesunitedintoirregularstrata：
surely，ifthemudhadbeentumultuouslydepositedinmass，theincludedcalcareousmatterwouldhavesegregateditselfirregularly，andnotintonodulesarrangedinhorizontallines，oneabovetheotherandoftenfarapart：thisarrangementappearstometoprovethatmud，differingslightlyincomposition，wassuccessivelyandquietlydeposited。Onthetheoryofadebacle，aprodigiousamountofmud，withoutasinglepebble，issupposedtohavebeenborneoverthewidesurfaceofthePampas，whenunderwater：
ontheotherhand，overthewholeofPatagonia，thesameoranotherdebacleissupposedtohavebornenothingbutgravel，——thegravelandthefinemudintheneighbourhoodoftheRiosNegroandColoradohavingbeenbornetoanequaldistancefromtheCordillera，orimaginedlineofdisturbance：
assuredlydirectlyoppositeeffectsoughtnottobeattributedtothesameagency。Where，again，couldamassoffinesediment，chargedwithcalcareousmatterinafitstateforchemicalsegregation，andinquantitysufficienttocoveranareaatleast750mileslong，and400milesbroad，toadepthoffromtwentytothirtyfeettoahundredfeet，havebeenaccumulated，readytobetransportedbythesupposeddebacle？Tomyminditislittleshortofdemonstration，thatagreatlapseoftimewasnecessaryfortheproductionanddepositionoftheenormousamountofmudlikematterformingthePampas;norshouldIhavenoticedthetheoryofadebacle，haditnotbeenadducedbyanaturalistsoeminentasM。d’Orbigny。
Asecondtheory，firstsuggested，Ibelieve，bySirW。Parish，isthatthePampeanformationwasthrowndownonlowandmarshyplainsbytheriversofthiscountrybeforetheyassumedtheirpresentcourses。Theappearanceandcompositionofthedeposit，themannerinwhichitslopesupandroundtheprimaryranges，thenatureoftheunderlyingmarinebeds，theestuaryandsea-shellsonthesurface，theoverlyingsandstonebedsatM。Hermoso，areallquiteopposedtothisview。NordoIbelievethatthereisasingleinstanceofaskeletonofoneoftheextinctmammifershavingbeenfoundinanuprightposition，asifithadbeenmired。
Thethirdtheory，ofthetruthofwhichIcannotentertainthesmallestdoubt，isthatthePampeanformationwasslowlyaccumulatedatthemouthoftheformerestuaryofthePlataandintheseaadjoiningit。Ihavecometothisconclusionfromthereasonsassignedagainstthetwoforegoingtheories，andfromsimplegeographicalconsiderations。FromthenumerousshellsoftheAzaralabiatalyinglooseonthesurfaceoftheplains，andnearBuenosAyresembeddedinthetosca-rock，weknowthatthisformationnotonlywasformerlycoveredby，butthattheuppermostpartsweredepositedin，thebrackishwateroftheancientLaPlata。SouthwardandseawardofBuenosAyres，theplainswereupheavedfromunderwaterinhabitedbytruemarineshells。WefurtherknowfromProfessorEhrenberg’sexaminationofthetwentymicroscopicalorganismsinthemudroundthetoothoftheMastodonhighupthecourseoftheParana，thatthebottom-
mostpartofthisformationwasofbrackish-waterorigin。Asimilarconclusionmustbeextendedtothebedsoflikecomposition，attheleveloftheseaandunderit，atM。HermosoinBahiaBlanca。Dr。Carpenterfindsthatthehardervarietiesoftosca-rock，collectedchieflytothesouth，containmarinespongoidbodies，minutefragmentsofshells，corals，andPolythalamia;theseperhapsmayhavebeendriftedinwardsbythetides，fromthemoreopenpartsofthesea。Theabsenceofshells，throughoutthisdeposit，withtheexceptionoftheuppermostlayersnearBuenosAyres，isaremarkablefact：canitbeexplainedbythebrackishconditionofthewater，orbythedeepmudatthebottom？Ihavestatedthatboththereddishmudandtheconcretionsoftosca-rockareoftenpenetratedbyminute，linearcavities，suchasfrequentlymaybeobservedinfresh-watercalcareousdeposits：——weretheyproducedbytheburrowingofsmallworms？
OnlyonthisviewofthePampeanformationhavingbeenofestuaryorigin，cantheextraordinarynumberspresentlytobealludedtooftheembeddedmammiferousremainsbeexplained。Itisalmostsuperfluoustogivethenumerouscasesforinstance，inSumatra;Lyell“Principles“volume3page325sixthedition，ofthecarcassesofanimalshavingbeenwashedouttoseabyswollenrivers;butImayrefertoarecentaccountbyMr。
Bettington“AsiaticSociety“1845June21st，ofoxen，deer，andbearsbeingcarriedintotheGulfofCambray;seealsotheaccountinmy“Journal“2ndeditionpage133，ofthenumbersofanimalsdrownedinthePlataduringthegreat，oftenrecurrent，droughts。
Withrespecttothefirstoriginofthereddishmud，Iwillonlyremark，thattheenormousareaofBrazilconsistsinchiefpartofgneissicandothergraniticrocks，whichhavesuffereddecomposition，andbeenconvertedintoared，gritty，argillaceousmass，toagreaterdepththaninanyothercountrywhichIhaveseen。Themixtureofroundedgrains，andevenofsmallfragmentsandpebblesofquartz，inthePampeanmudofBandaOriental，isevidentlyduetotheneighbouringandunderlyingprimaryrocks。TheestuarymudwasdriftedduringthePampeanperiodinamuchmoresoutherlycourse，owingprobablytotheeastandwestprimaryridgessouthofthePlatanothavingbeenthenelevated，thanthemudofthePlataatpresentis;foritwasformerlydepositedasfarsouthastheColorado。Thequantityofcalcareousmatterinthisformation，especiallyinthoselargedistrictswherethewholemasspassesintotosca-rock，isverygreat：Ihavealreadyremarkedonthecloseresemblanceinexternalandmicroscopicalappearance，betweenthistosca-rockandthestrataatCoquimbo，whichhavecertainlyresultedfromthedecayandattritionofrecentshells：Idarenot，however，extendthisconclusiontothecalcareousrocksofthePampas，moreespeciallyastheunderlyingtertiarystratainwesternBandaOrientalshowthatatthatperiodtherewasacopiousemissionofcarbonateoflime，inconnectionwithvolcanicaction。Imayadd，thattherearenearlysimilarsuperficialcalcareousbedsatKingGeorge’sSoundinAustralia;andtheseundoubtedlyhavebeenformedbythedisintegrationofmarineremainssee“VolcanicIslands“etc。page144。Thereis，however，somethingveryremarkableinthefrequencyofsuperficial，thinbedsofearthycalcareousmatter，indistrictswherethesurroundingrocksarenotcalcareous。MajorCharters，inaPaperreadbeforetheGeographicalSocietyApril13，1840
andabstractedinthe“Athenaeum“page317，statesthatthisisthecaseinpartsofMexico，andthathehasobservedsimilarappearancesinmanypartsofSouthAfrica。ThecircumstanceoftheuppermoststratumroundtheraggedSierraVentana，consistingofcalcareousormarlymatter，withoutanycoveringofalluvialmatter，strikesmeasverysingular，inwhatevermannerweviewthedepositionandelevationofthePampeanformation。
ThePampeanformation，judgingfromitssimilarcomposition，andfromtheapparentabsolutespecificidentityofsomeofitsmammiferousremains，andfromthegenericresemblanceofothers，belongsoveritsvastarea——
throughoutBandaOriental，EntreRios，andthewideextentofthePampasasfarsouthastheColorado，——tothesamegeologicalepoch。Themammiferousremainsoccuratalldepthsfromthetoptothebottomofthedeposit;andImayaddthatnowhereinthePampasisthereanyappearanceofmuchsuperficialdenudation：someboneswhichIfoundneartheGuardiadelMontewereembeddedclosetothesurface;andthisappearstohavebeenthecasewithmanyofthosediscoveredinBandaOriental：ontheMatanzas，twentymilessouthofBuenosAyres，aGlyptodonwasembeddedfivefeetbeneaththesurface;numerousremainswerefoundbyS。Muniz，nearLuxan，atanaveragedepthofeighteenfeet;inBuenosAyresaskeletonwasdisinterredatsixtyfeetdepth，andontheParanaIhavedescribedtwoskeletonsoftheMastodononlyfiveorsixfeetabovetheverybaseofthedeposit。Withrespecttotheageofthisformation，asjudgedofbytheordinarystandardoftheexistenceofMollusca，theonlyevidencewithinthelimitsofthetruePampaswhichisatalltrustworthy，isaffordedbythestilllivingAzaralabiatabeingembeddedintosca-rocknearBuenosAyres。AtPuntaAlta，however，wehaveseenthatseveraloftheextinctmammifers，mostcharacteristicofthePampeanformation，co-existedwithtwentyspeciesofMollusca，abarnacleandtwocorals，allstilllivingonthissamecoast;——
forwhenwerememberthattheshellshaveamoreancientappearancethanthebones;thatmanyofthebones，thoughembeddedinacoarseconglomerate，areperfectlypreserved;thatalmostallthepartsoftheskeletonoftheScelidotherium，eventotheknee-cap，werelyingintheirproperrelativepositions;andthatalargepieceofthefragiledermalarmourofaDasypoidquadruped，connectedwithsomeofthebonesofthefoot，hadbeenentombedinaconditionallowingthetwosidestobedoubledtogether，itmustassuredlybeadmittedthatthesemammiferousremainswereembeddedinafreshstate，andthereforethatthelivinganimalsco-existedwiththeco-embeddedshells。Moreover，theMacraucheniaPatachonicaofwhich，accordingtoProfessorOwen，remainsalsooccurinthePampasofBuenosAyres，andatPuntaAltahasbeenshownbysatisfactoryevidenceofanotherkind，tohavelivedontheplainsofPatagonialongaftertheperiodwhentheadjoiningseawasfirsttenantedbyitspresentcommonestmolluscousanimals。Wemust，therefore，concludethatthePampeanformationbelongs，intheordinarygeologicalsenseoftheword，totheRecentPeriod。M。d’Orbignybelieves“Voyage“Part。Geolog。page81，thatthisformation，though“tresvoisinedelanotre，estneanmoinsdebeaucoupanterieureanotrecreation。”
AtSt。FeBajada，thePampeanestuaryformation，withitsmammiferousremains，conformablyoverliesthemarinetertiarystrata，whichasfirstshownbyM。d’OrbignyarecontemporaneouswiththoseofPatagonia，andwhich，asweshallhereaftersee，belongtoaveryancienttertiarystage。
Whenexaminingthejunctionbetweenthesetwoformations，Ithoughtthattheconcretionarylayerofmarlmarkedapassagebetweenthemarineandestuarystages。M。d’Orbignydisputesthisviewasgiveninmy“Journal“，andIadmitthatitiserroneous，thoughinsomedegreeexcusable，fromtheirconformabilityandfrombothaboundingwithcalcareousmatter。Itwould，indeed，havebeenagreatanomalyiftherehadbeenatruepassagebetweenadepositcontemporaneouswithexistingspeciesofmollusca，andoneinwhichallthemolluscaappeartobeextinct。NorthwardofSantaFe，M。d’Orbignymetwithferruginoussandstones，marlyrocks，andotherbeds，whichheconsidersasadistinctandlowerformation;buttheevidencethattheyarenotpartsofthesamewithanalteredmineralogicalcharacter，doesnotappeartomequitesatisfactory。
InWesternBandaOriental，whilethemarinetertiarystratawereaccumulating，therewerevolcaniceruptions，muchsilexandlimewereprecipitatedfromsolution，coarseconglomerateswereformed，beingderivedprobablyfromadjoiningland，andlayersofredmudandmarlyrocks，likethoseofthePampeanformation，wereoccasionallydeposited。ThetruePampeandeposit，withmammiferousremains，insteadofasatSantaFeoverlyingconformablythetertiarystrata，ishereseenatalowerlevelfoldingroundandbetweentheflat-topped，cliff-boundedhills，formedbyaupheavalanddenudationofthesesametertiarystrata。Theupheaval，havingoccurredhereearlierthanatSantaFe，maybenaturallyaccountedforbythecontemporaneousvolcanicaction。AttheBarrancasdeS。Gregorio，thePampeandeposit，aswehaveseen，overliesandfillsupfurrowsincoarsesand，preciselylikethatnowaccumulatingontheshoresnearthemouthofthePlata。Icanhardlybelievethatthislooseandcoarsesandiscontemporaneouswiththeoldtertiaryandoftencrystallinestrataofthemorewesternpartsoftheprovince;andaminducedtosuspectthatitisofsubsequentorigin。IfthatsectionnearColoniacouldbeimplicitlytrusted，inwhich，ataheightofonlyfifteenfeetabovethePlata，abedoffresh-lookingmussels，ofanexistinglittoralspecies，appearedtoliebetweenthesandandthePampeanmud，IshouldconcludethatBandaOrientalmusthavestood，whenthecoarsesandwasaccumulating，atonlyalittlebelowitspresentlevel，andhadthensubsided，allowingtheestuaryPampeanmudtocoverfarandwideitssurfaceuptoaheightofsomehundredfeet;andthatafterthissubsidencetheprovincehadbeenupliftedtoitspresentlevel。
InWesternBandaOriental，weknow，fromtwounequivocalsectionsthatthereisamass，absolutelyundistinguishablefromthetruePampeandeposit，beneaththeoldtertiarystrata。Thisinferiormassmustbeverymuchmoreancientthantheupperdepositwithitsmammiferousremains，foritliesbeneaththetertiarystratainwhichalltheshellsareextinct。
Nevertheless，theloweranduppermasses，aswellassomeintermediatelayers，aresosimilarinmineralogicalcharacter，thatIcannotdoubtthattheyareallofestuaryorigin，andhavebeenderivedfromthesamegreatsource。Atfirstitappearedtomeextremelyimprobable，thatmudofthesamenatureshouldhavebeendepositedonnearlythesamespot，duringanimmenselapseoftime，namely，fromaperiodequivalentperhapstotheEoceneofEuropetothatofthePampeanformation。Butas，attheverycommencementofthePampeanperiod，ifnotatastillearlierperiod，theSierraVentanaformedaboundarytothesouth，——theCordilleraortheplainsinfrontofthemtothewest，——thewholeprovinceofCorrientesprobablytothenorth，for，accordingtoM。d’Orbigny，itisnotcoveredbythePampeandeposit，——andBrazil，asknownbytheremainsinthecaves，tothenorth-east;andasagain，duringtheoldertertiaryperiod，landalreadyexistedinWesternBandaOrientalandnearSt。FeBajada，asmaybeinferredfromthevegetabledebris，fromthequantitiesofsilicifiedwood，andfromtheremainsofaToxodonfound，accordingtoM。d’Orbigny，instilllowerstrata，wemayconclude，thatatthisancientperiodagreatexpanseofwaterwassurroundedbythesamerockyframeworkwhichnowboundstheplainsofPampeanformation。Thishavingbeenthecase，thecircumstanceofsedimentofthesamenaturehavingbeendepositedinthesameareaduringanimmenselapseoftime，thoughhighlyremarkable，doesnotappearincredible。
TheelevationofthePampas，atleastofthesouthernparts，hasbeenslowandinterruptedbyseveralperiodsofrest，asmaybeinferredfromtheplains，cliffs，andlinesofsand-duneswithshellsandpumice-pebbles
standingatdifferentheights。Ibelieve，also，thatthePampeanmudcontinuedtobedeposited，afterpartsofthisformationhadalreadybeenelevated，inthesamemannerasmudwouldcontinuetobedepositedintheestuaryofthePlata，ifthemud-banksonitsshoreswerenowupliftedandchangedintoplains：Ibelieveinthisfromtheimprobabilityofsomanyskeletonsandboneshavingbeenaccumulatedatonespot，whereM。Hermosonowstands，atadepthofbetweeneighthundredandonethousandfeet，andatavastdistancefromanylandexceptsmallrockyislets，——asmusthavebeenthecase，ifthehightosca-plainroundtheVentanaandadjoiningSierras，hadnotbeenalreadyupliftedandconvertedintoland，supportingmammiferousanimals。AtPuntaAltawehavegoodevidencethatthegravel-
strata，whichcertainlybelongtothetruePampeanperiod，wereaccumulatedaftertheelevationinthatneighbourhoodofthemainpartofthePampeandeposit，whencetheroundedmassesoftosca-rockwerederived，andthatrolledfragmentofblackboneinthesamepeculiarconditionwiththeremainsatMonteHermoso。
ThenumberofthemammiferousremainsembeddedinthePampasis，asIhaveremarked，wonderful：itshouldbeborneinmindthattheyhavealmostexclusivelybeenfoundinthecliffsandsteepbanksofrivers，andthat，untillately，theyexcitednoattentionamongsttheinhabitants：IamfirmlyconvincedthatadeeptrenchcouldnotbecutinanylineacrossthePampas，withoutintersectingtheremainsofsomequadruped。ItisdifficulttoformanopinioninwhatpartofthePampastheyaremostnumerous;inalimitedspottheycouldnotwellhavebeenmorenumerousthantheywereatP。Alta;thenumber，however，latelyfoundbySenorF。Muniz，nearLuxan，inacentralspotinthePampas，isextraordinarilygreat：attheendofthischapterIwillgivealistofallthelocalitiesatwhichIhaveheardofremainshavingbeendiscovered。Veryfrequentlytheremainsconsistofalmostperfectskeletons;butthereare，also，numeroussinglebones，asforinstanceatSt。Fe。Theirstateofpreservationvariesmuch，evenwhenembeddedneareachother：IsawnoneotherssoperfectlypreservedastheheadsoftheToxodonandMylodonfromthewhitesoftearthybedontheSarandisinBandaOriental。ItisremarkablethatintwolimitedsectionsI
foundnolessthanfiveteethseparatelyembedded，andIheardofteethhavingbeensimilarlyfoundinotherparts：maywesupposethattheskeletonsorheadswereforalongtimegentlydriftedbycurrentsoverthesoftmuddybottom，andthattheteethoccasionally，hereandthere，droppedout？
Itmaybenaturallyasked，wheredidthesenumerousanimalslive？FromtheremarkablediscoveriesofMM。LundandClausen，itappearsthatsomeofthespeciesfoundinthePampasinhabitedthehighlandsofBrazil：theMastodonAndiumisembeddedatgreatheightsintheCordillerafromnorthoftheequatortoatleastasfarsouthasTarijaHumboldtstatesthattheMastodonhasbeendiscoveredinNewGranada：ithasbeenfoundinQuito。
WhenatLima，IsawatoothofaMastodoninthepossessionofDonM。
Rivero，foundatPlayaChicaontheMaranon，neartheGuallaga。EveryonehasheardofthenumerousremainsofMastodoninBolivia。;andasthereisnohigherland，therecanbelittledoubtthatthisMastodonmusthavelivedontheplainsandvalleysofthatgreatrange。Thesecountries，however，appeartoofardistantforthehabitationoftheindividualsentombedinthePampas：wemustprobablylooktonearerpoints，forinstancetotheprovinceofCorrientes，which，asalreadyremarked，issaidnottobecoveredbythePampeanformation，andmaytherefore，attheperiodofitsdeposition，haveexistedasdryland。IhavealreadygivenmyreasonsforbelievingthattheanimalsembeddedatM。HermosoandatP。
AltainBahiaBlanca，livedonadjoiningland，formedofpartsofthealreadyelevatedPampeandeposit。Withrespecttothefoodofthesemanygreatextinctquadrupeds，Iwillnotrepeatthefactsgiveninmy“Journal“
secondeditionpage85，showingthatthereisnocorrelationbetweentheluxurianceofthevegetationofacountryandthesizeofitsmammiferousinhabitants。Idonotdoubtthatlargeanimalscouldnowexist，asfarastheamount，notkind，ofvegetationisconcerned，onthesterileplainsofBahiaBlancaandoftheR。Negro，aswellasontheequally，ifnotmoresterileplainsofSouthernAfrica。Theclimate，however，mayperhapshavesomewhatdeterioratedsincethemammifersembeddedatBahiaBlancalivedthere;forwemustnotinfer，fromthecontinuedexistenceofthesameshellsonthepresentcoasts，thattherehasbeennochangeinclimate;forseveraloftheseshellsnowrangenorthwardalongtheshoresofBrazil，wherethemostluxuriantvegetationflourishesunderatropicaltemperature。Withrespecttotheextinction，whichatfirstfillsthemindwithastonishment，ofthemanygreatandsmallmammifersofthisperiod，I
mayalsorefertotheworkabovecitedsecondeditionpage173，inwhichIhaveendeavouredtoshow，thathoweverunablewemaybetoexplaintheprecisecause，weoughtnotproperlytofeelmoresurprisedataspeciesbecomingextinctthanatonebeingrare;andyetweareaccustomedtoviewtherarityofanyparticularspeciesasanordinaryevent，notrequiringanyextraordinaryagency。