Tothesouththisbasin-likeplaincontracts，andrisingscarcelyperceptiblywithasmoothsurface，passesthrougharemarkablelevelgapinthemountains，formingatrueland-strait，andcalledtheAngostura。Itthenimmediatelyexpandsintoasecondbasin-formedplain：thisagaintothesouthcontractsintoanotherland-strait，andexpandsintoathirdbasin，which，however，fallssuddenlyinlevelaboutfortyfeet。Thisthirdbasin，tothesouth，likewisecontractsintoastrait，andthenagainopensintothegreatplainofSanFernando，stretchingsofarsouththatthesnowypeaksofthedistantCordilleraareseenrisingaboveitshorizonasabovethesea。Theseplains，neartheCordillera，aregenerallyformedofathickstratifiedmassofshingle(TheplainofSanFernandohas，accordingtoMM。MeyenandGay“Reise“etc。Th。1ss。295and298，neartheCordillera，anupperstep-formedplainofclay，onthesurfaceofwhichtheyfoundnumerousblocksofrocks，fromtwotothreefeetlong，eitherlyingsingleorpiledinheaps，butallarrangedinnearlystraightlines。);inotherparts，ofaredsandyclay，oftenwithanadmixtureofpumiceousmatter。Althoughthesebasinsareconnectedtogetherlikeanecklace，inanorthandsouthline，bysmoothland-straits，thestreamswhichdrainthemdonotallflownorthandsouth，butmostlywestward，throughbreacheswornintheboundingmountains;andinthecaseofthesecondbasin，orthatofRancagua，therearetwodistinctbreaches。Eachbasin，moreover，isnotdrainedsingly;thus，togivethemoststrikinginstance，butnottheonlyone，inproceedingsouthwardovertheplainofRancagua，wefirstfindthewaterflowingnorthwardtoandthroughthenorthernland-strait;then，withoutcrossinganymarkedridgeorwatershed，weseeitflowingsouth-westwardtowardsthenorthernoneofthetwobreachesinthewesternmountainousboundary;andlastly，againwithoutanyridge，itflowstowardsthesouthernbreachinthesesamemountains。Hencethesurfaceofthisonebasin-likeplain，appearingtotheeyesolevel，hasbeenmodelledwithgreatnicety，sothatthedrainage，withoutanyconspicuouswatersheds，isdirectedtowardsthreeopeningsintheencirclingmountains。((ItappearsfromCaptainHerbert’saccountoftheDiluviumoftheHimalaya，“GleaningsofScience“Calcuttavolume2page164，thatpreciselysimilarremarksapplytothedrainageoftheplainsorvalleysbetweenthosegreatmountains。)Thestreamsflowingfromthesouthernbasin-likeplains，afterpassingthroughthebreachestothewest，uniteandformtheriverRapel，whichentersthePacificnearNavidad。I
followedthesouthernmostbranchofthisriver，andfoundthatthebasinorplainofSanFernandoiscontinuouslyandsmoothlyunitedwiththoseplains，whichweredescribedintheSecondChapter，asbeingwornnearthecoastintosuccessivecave-eatenescarpments，andstillnearertothecoast，asbeingstrewedwithupraisedrecentmarineremains。
Imighthavegivendescriptionsofnumerousotherplainsofthesamegeneralform，someatthefootoftheCordillera，somenearthecoast，andsomehalfwaybetweenthesepoints。Iwillalludeonlytooneother，namely，theplainofUspallata，lyingontheeasternoroppositesideoftheCordillera，betweenthatgreatrangeandtheparallellowerrangeofUspallata。AccordingtoMiers，itssurfaceis6，000feetabovethelevelofthesea：itisfromtentofifteenmilesinwidth，andissaidtoextendwithanunbrokensurfacefor180milesnorthwards：itisdrainedbytworiverspassingthroughbreachesinthemountainstotheeast。OnthebanksoftheRiverMendozaitisseentobecomposedofagreataccumulationofstratifiedshingle，estimatedat400feetinthickness。Ingeneralappearance，andinnumerouspointsofstructure，thisplaincloselyresemblesthoseofChile。
Theoriginandmannerofformationofthethickbedsofgravel，sandyclay，volcanicdetritus，andcalcareoustuff，composingthesebasin-likeplains，isveryimportant;because，asweshallpresentlyshow，theysendarmsorfringesfarupthemainvalleysoftheCordillera。Manyoftheinhabitantsbelievethattheseplainswereonceoccupiedbylakes，suddenlydrained;
butIconceivethatthenumberoftheseparatebreachesatnearlythesamelevelinthemountainssurroundingthemquiteprecludesthisidea。HadnotsuchdistinguishednaturalistsasMM。MeyenandGaystatedtheirbeliefthatthesedepositswereleftbygreatdebaclesrushingdownfromtheCordillera，Ishouldnothavenoticedaview，whichappearstomefrommanyreasonsimprobableinthehighestdegree——namely，fromthevastaccumulationofWELL-ROUNDEDPEBBLES——theirfrequentstratificationwithlayersofsand——theoverlyingbedsofcalcareoustuff——thissamesubstancecoatingandunitingthefragmentsofrockonthehummocksintheplainofSantiago——andlastlyevenfromtheworn，rounded，andmuchdenudedstateofthesehummocks，andoftheheadlandswhichprojectfromthesurroundingmountains。Ontheotherhand，theseseveralcircumstances，aswellasthecontinuousunionofthebasinsatthefootoftheCordillera，withthegreatplainoftheRioRapelwhichstillretainsthemarksofsea-actionatvariouslevels，andtheirgeneralsimilarityinformandcompositionwiththemanyplainsnearthecoast，whichareeithersimilarlymarkedorarestrewedwithupraisedmarineremains，fullyconvincemethatthemountainsboundingthesebasin-plainswerebreached，theirislet-likeprojectingrocksworn，andtheloosestratifieddetritusformingtheirnowlevelsurfacesdeposited，bythesea，asthelandslowlyemerged。Itishardlypossibletostatetoostronglytheperfectresemblanceinoutlinebetweenthesebasin-like，long，andnarrowplainsofChile(especiallywhenintheearlymorningthemistshanginglowrepresentedwater)，andthecreeksandfiordsnowintersectingthesouthernandwesternshoresofthecontinent。
Wecanonthisviewofthesea，whenthelandstoodlower，havinglongandtranquillyoccupiedthespacesbetweenthemountain-ranges，understandhowtheboundariesoftheseparatebasinswerebreachedinmorethanoneplace;
forweseethatthisisthegeneralcharacteroftheinlandbaysandchannelsofTierradelFuego;wethere，also，seeinthesawingactionofthetides，whichflowwithgreatforceinthecrosschannels，apowersufficienttokeepthebreachesopenasthelandemerged。Wecanfurtherseethatthewaveswouldnaturallyleavethesmoothbottomofeachgreatbayorchannel，asitbecameslowlyconvertedintoland，gentlyinclinedtoasmanypointsasthereweremouths，throughwhichtheseafinallyretreated，thusformingsomanywatersheds，withoutanymarkedridges，onanearlylevelsurface。Theabsenceofmarineremainsinthesehighinlandplainscannotbeproperlyadducedasanobjectiontotheirmarineorigin：
forwemayconclude，fromshellsnotbeingfoundinthegreatshinglebedsofPatagonia，thoughcopiouslystrewedontheirsurfaces，andfrommanyotheranalogousfacts，thatsuchdepositsareeminentlyunfavourablefortheembedmentofsuchremains;andwithrespecttoshellsnotbeingfoundstrewedonthesurfaceofthesebasin-likeplains，itwasshowninthelastchapterthatremainsthusexposedintimedecayanddisappear。
(FIGURE13。SECTIONOFTHEPLAINATTHEEASTERNFOOTOFTHECHILEAN
CORDILLERA。
>FromCordillera(left)throughTalus-plainandLevelsurface，2，700feetabovesea，toGravelterraces(right)。)
IobservedsomeappearancesontheplainsattheeasternandoppositefootoftheCordillerawhichareworthnotice，asshowingthattheseatherelongactedatnearlythesamelevelasonthebasin-plainsofChile。Themountainsonthiseasternsideareexceedinglyabrupt;theyriseoutofasmooth，talus-like，verygentle，slope，fromfivetotenmilesinwidth(asrepresentedinFigure13)，entirelycomposedofperfectlyroundedpebbles，oftenwhite-washedwithanaluminoussubstancelikedecomposedfeldspar。
Thisslopingplainortalusblendsintoaperfectlyflatspaceafewmilesinwidth，composedofreddishimpureclay，withsmallcalcareousconcretionsasinthePampeandeposit，——offinewhitesandwithsmallpebblesinlayers，——andoftheabove-mentionedwhitealuminousearth，allinterstratifiedtogether。ThisflatspacerunsasfarasMendoza，thirtymilesnorthward，andstandsprobablyataboutthesameheight，namely，2，700feet(PentlandandMiers)abovethesea。Totheeastitisboundedbyanescarpment，eightyfeetinheight，runningformanymilesnorthandsouth，andcomposedofperfectlyroundpebbles，andloose，white-washed，orembeddedinthealuminousearth：behindthisescarpmentthereisasecondandsimilaroneofgravel。NorthwardofMendoza，theseescarpmentsbecomebrokenandquiteobliterated;anditdoesnotappearthattheyeverenclosedalake-likearea：Iconclude，therefore，thattheywereformedbythesea，whenitreachedthefootoftheCordillera，likethesimilarescarpmentsoccurringatsomanypointsonthecoastsofChileandPatagonia。
Thetalus-likeplainslopesupwithasmoothsurfaceintothegreatdryvalleysoftheCordillera。OneachhandofthePortillovalley，themountainsareformedofredgranite，mica-slate，andbasalt，whichallhavesufferedatrulyastonishingamountofdenudation;thegravelinthevalley，aswellasonthetalus-likeplaininfrontofit，iscomposedoftheserocks;butatthemouthofthevalley，inthemiddle(heightprobablyaboutthreethousandfivehundredfeetabovethesea)，afewsmallisolatedhillocksofseveralvarietiesofporphyryproject，roundwhich，onallsides，smoothandoftenwhite-washedpebblesofthesesameporphyries，totheexclusionofallothers，extendtoacircumscribeddistance。Now，itisdifficulttoconceiveanyotheragency，exceptthequietandlong-continuedactionoftheseaonthesehillocks，whichcouldhaveroundedandwhitewashedthefragmentsofporphyry，andcausedthemtoradiatefromsuchsmallandquiteinsignificantcentres，inthemidstofthatvaststreamofstoneswhichhasdescendedfromthemainCordillera。
SLOPINGTERRACESOFGRAVELINTHEVALLEYSOFTHECORDILLERA。
(FIGURE14。GROUND-PLANOFABIFURCATINGVALLEYINTHECORDILLERA，borderedbysmooth，slopinggravel-fringes(AA)，wornalongthecourseoftheriverintocliffs。)
AllthemainvalleysonbothflanksoftheChileanCordillerahaveformerlyhad，orstillhave，theirbottomsfilleduptoaconsiderablethicknessbyamassofrudelystratifiedshingle。InCentralChilethegreaterpartofthismasshasbeenremovedbythetorrents;cliff-boundedfringes，moreorlesscontinuous，beingleftatcorrespondingheightsonbothsidesofthevalleys。Thesefringes，orastheymaybecalledterraces，haveasmoothsurface，andasthevalleysrise，theygentlyrisewiththem：hencetheyareeasilyirrigated，andaffordgreatfacilitiesfortheconstructionoftheroads。Fromtheiruniformity，theygivearemarkablecharactertothesceneryofthesegrand，wild，brokenvalleys。Inwidth，thefringesvarymuch，sometimesbeingonlybroadenoughfortheroads，andsometimesexpandingintonarrowplains。Theirsurfaces，besidesgentlyrisingupthevalley，areslightlyinclinedtowardsitscentreinsuchamannerastoshowthatthewholebottommustoncehavebeenfilledupwithasmoothandslightlyconcavemass，asstillarethedryunfurrowedvalleysofNorthernChile。Wheretwovalleysuniteintoone，theseterracesareparticularlywellexhibited，asisrepresentedinFigure14。Thethicknessofthegravelformingthesefringes，onarudeaverage，maybesaidtovaryfromthirtytosixtyoreightyfeet;butnearthemouthsofthevalleysitwasinseveralplacesfromtwotothreehundredfeet。Theamountofmatterremovedbythetorrentshasbeenimmense;yetinthelowerpartsofthevalleystheterraceshaveseldombeenentirelywornawayoneitherside，norhasthesolidunderlyingrockbeenreached：higherupthevalleys，theterraceshavefrequentlybeenremovedononeortheotherside，andsometimesonbothsides;butinthislattercasetheyreappearafterashortintervalontheline，whichtheywouldhaveheldhadtheybeenunbroken。Wherethesolidrockhasbeenreached，ithasbeencutintodeepandnarrowgorges。
Stillhigherupthevalleys，theterracesgraduallybecomemoreandmorebroken，narrower，andlessthick，until，ataheightoffromseventoninethousandfeet，theybecomelost，andblendedwiththepilesoffallendetritus。
Icarefullyexaminedinmanyplacesthestateofthegravel，andalmosteverywherefoundthepebblesequallyandperfectlyrounded，occasionallywithgreatblocksofrock，andgenerallydistinctlystratified，oftenwithpartingseamsofsand。Thepebblesweresometimescoatedwithawhitealuminous，andlessfrequentlywithacalcareous，crust。Atgreatheightsupthevalleysthepebblesbecomelessrounded;andastheterracesbecomeobliterated，thewholemasspassesintothenatureofordinarydetritus。I
wasrepeatedlystruckwiththegreatdifferencebetweenthisdetritushighupthevalleys，andthegraveloftheterraceslowdown，namely，inthegreaternumberofthequiteangularfragmentsinthedetritus，——intheunequaldegreetowhichtheotherfragmentshavebeenrounded，——inthequantityofassociatedearth，——intheabsenceofstratification，——andintheirregularityoftheuppersurfaces。Thisdifferencewaslikewisewellshownatpointslowdownthevalleys，whereprecipitousravines，cuttingthroughmountainsofhighlycolouredrock，havethrowndownwide，fan-
shapedaccumulationsofdetritusontheterraces：insuchcases，thelineofseparationbetweenthedetritusandtheterracecouldbepointedouttowithinaninchortwo;thedetritusconsistingentirelyofangularandonlypartiallyroundedfragmentsoftheadjoiningcolouredrocks;thestratifiedshingle(asIascertainedbycloseinspection，especiallyinonecase，inthevalleyoftheRiverMendoza)containingonlyasmallproportionofthesefragments，andthosefewwellrounded。
Iparticularlyattendedtotheappearanceoftheterraceswherethevalleysmadeabruptandconsiderablebends，butIcouldperceivenodifferenceintheirstructure：theyfollowedthebendswiththeirusualnearlyequableinclination。Iobserved，also，inseveralvalleys，thatwhereverlargeblocksofanyrockbecamenumerous，eitheronthesurfaceoftheterraceorembeddedinit，thisrocksoonappearedhigherupinsitu：thusIhavenoticedblocksofporphyry，ofandesiticsyenite，ofporphyryandofsyenite，alternatelybecomingnumerous，andineachcasesucceededbymountainsthusconstituted。Thereis，however，oneremarkableexceptiontothisrule;foralongthevalleyoftheCachapual，M。Gayfoundnumerouslargeblocksofwhitegranite，whichdoesnotoccurintheneighbourhood。I
observedtheseblocks，aswellasothersofandesiticsyenite(notoccurringhereinsitu)，nearthebathsofCauquenesataheightofbetweentwoandthreehundredfeetabovetheriver，andthereforequiteabovetheterraceorfringewhichbordersthatriver;somemilesupthevalleystherewereotherblocksataboutthesameheight。Ialsonoticed，atalessheight，justabovetheterrace，blocksofporphyries(apparentlynotfoundintheimmediatelyimpendingmountains)，arrangedinrudelines，asonasea-beach。Alltheseblockswererounded，andthoughlarge，notgigantic，likethetrueerraticbouldersofPatagoniaandFuegia。M。Gaystatesthatthegranitedoesnotoccurinsituwithinadistanceoftwentyleagues(“AnnalesdesScienceNat。“1seriestome28。M。Gay，asIwasinformed，penetratedtheCordillerabythegreatobliquevalleyofLosCupressos，andnotbythemostdirectline。);Isuspect，forseveralreasons，thatitwillultimatelybefoundatamuchlessdistance，thoughcertainlynotintheimmediateneighbourhood。ThebouldersfoundbyMM。MeyenandGayontheupperplainofSanFernando(mentionedinapreviousnote)probablybelongtothissameclassofphenomena。
ThesefringesofstratifiedgraveloccuralongallthegreatvalleysoftheCordillera，aswellasalongtheirmainbranches;theyarestrikinglydevelopedinthevalleysoftheMaypu，Mendoza，Aconcagua，Cachapual，andaccordingtoMeyen，intheTinguirica。(“Reise“etc。Th。1s。302。)Inthevalleys，however，ofNorthernChile，andinsomeontheeasternflankoftheCordillera，asinthePortilloValley，wherestreamshaveneverflowed，orarequiteinsignificantinvolume，thepresenceofamassofstratifiedgravelcanbeinferredonlyfromthesmoothslightlyconcaveformofthebottom。Onenaturallyseeksforsomeexplanationofsogeneralandstrikingaphenomenon;thatthematterformingthefringesalongthevalleys，orstillfillinguptheirentirebeds，hasnotfallenfromtheadjoiningmountainslikecommondetritus，isevidentfromthecompletecontrastineveryrespectbetweenthegravelandthepilesofdetritus，whetherseenhighupthevalleysontheirsides，orlowdowninfrontofthemoreprecipitousravines;thatthematterhasnotbeendepositedbydebacles，evenifwecouldbelieveindebacleshavingrusheddownEVERYvalley，andalltheirbranches，eastwardandwestwardfromthecentralpinnaclesoftheCordillera，wemustadmitfromthefollowingreasons，——fromthedistinctstratificationofthemass，——itssmoothuppersurface，——thewell-roundedandsometimesencrustedstateofthepebbles，sodifferentfromtheloosedebrisonthemountains，——andespeciallyfromtheterracespreservingtheiruniforminclinationroundthemostabruptbends。Tosupposethatasthelandnowstands，theriversdepositedtheshinglealongthecourseofeveryvalley，andalltheirmainbranches，appearstomepreposterous，seeingthatthesesameriversnotonlyarenowremovingandhaveremovedmuchofthisdeposit，butareeverywheretendingtocutdeepandnarrowgorgesinthehardunderlyingrocks。
Ihavestatedthatthesefringesofgravel，theoriginofwhichareinexplicableonthenotionofdebaclesorofordinaryalluvialaction，aredirectlycontinuouswiththesimilarly-composedbasin-likeplainsatthefootoftheCordillera，which，fromtheseveralreasonsbeforeassigned，I
cannotdoubtweremodelledbytheagencyofthesea。NowifwesupposethattheseaformerlyoccupiedthevalleysoftheChileanCordillera，inpreciselythesamemannerasitnowdoesinthemoresouthernpartsofthecontinent，wheredeepwindingcreekspenetrateintotheveryheartof，andinthecaseofObstructionSoundquitethrough，thisgreatrange;andifwesupposethatthemountainswereupraisedinthesameslowmannerastheeasternandwesterncoastshavebeenupraisedwithintherecentperiod，thentheoriginandformationofthesesloping，terrace-likefringesofgravelcanbesimplyexplained。Foreverypartofthebottomofeachvalleywill，onthisview，havelongstoodattheheadofaseacreek，intowhichthethenexistingtorrentswillhavedeliveredfragmentsofrocks，where，bytheactionofthetides，theywillhavebeenrolled，sometimesencrusted，rudelystratified，andthewholesurfacelevelledbytheblendingtogetherofthesuccessivebeachlines。(Slopingterracesofpreciselysimilarstructurehavebeendescribedbyme“PhilosophicalTransactions“1839page58，inthevalleysofLochaberinScotland，where，athigherlevels，theparallelroadsofGlenRoyshowthemarksofthelongandquietresidenceofthesea。IhavenodoubtthattheseslopingterraceswouldhavebeenpresentinthevalleysofmostoftheEuropeanranges，hadnoteverytraceofthem，andallwrecksofsea-action，beensweptawaybytheglacierswhichhavesinceoccupiedthem。Ihaveshownthatthisisthecasewiththemountains(“LondonandEdinburghPhilosophicalJournal“
volume21page187)ofNorthWales。)Asthelandrose，thetorrentsineveryvalleywillhavetendedtohaveremovedthematterwhichjustbeforehadbeenarrestedon，ornear，thebeach-lines;thetorrents，also，havingcontinuedtogaininforcebythecontinuedelevationincreasingtheirtotaldescentfromtheirsourcestothesea。ThisslowrisingoftheCordillera，whichexplainssowelltheotherwiseinexplicableoriginandstructureoftheterraces，judgingfromallknownanalogies，willprobablyhavebeeninterruptedbymanyperiodsofrest;butweoughtnottoexpecttofindanyevidenceoftheseperiodsinthestructureofthegravel-
terraces：for，asthewavesattheheadsofdeepcreekshavelittleerosivepower，sotheonlyeffectoftheseahavinglongremainedatthesamelevelwillbethattheupperpartsofthecreekswillhavebecomefilledupatsuchperiodstothelevelofthewaterwithgravelandsand;andthatafterwardstheriverswillhavethrowndownonthefilled-uppartsatalusofsimilarmatter，ofwhichtheinclination(asattheheadofapartiallyfilled-uplake)willhavebeendeterminedbythesupplyofdetritus，andtheforceofthestream。(Ihaveattemptedtoexplainthisprocessinamoredetailedmanner，inalettertoMr。Maclaren，publishedinthe“EdinburghNewPhilosophicalJournal“volume35page288。)Hence，afterthefinalconversionofthecreeksintovalleys，almosttheonlydifferenceintheterracesatthosepointsatwhichtheseastoodlong，willbeasomewhatmoregentleinclination，withriver-worninsteadofsea-worndetritusonthesurface。
Iknowofonlyonedifficultyontheforegoingview，namely，thefar-
transportedblocksofrockhighonthesidesofthevalleyoftheCachapual：Iwillnotattemptanyexplanationofthisphenomenon，butImaystatemybeliefthatamountain-ridgeneartheBathsofCauqueneshasbeenupraisedlongsubsequentlytoalltheotherrangesintheneighbourhood，andthatwhenthiswaseffectedthewholefaceofthecountrymusthavebeengreatlyaltered。Inthecourseofages，moreover，inthisandothervalleys，eventsmayhaveoccurredlike，butevenonagranderscalethan，thatdescribedbyMolina，whenaslipduringtheearthquakeof1762bankedupfortendaysthegreatRiverLontue，whichthenburstingitsbarrier“inundatedthewholecountry，“anddoubtlesstransportedmanygreatfragmentsofrock。(“CompendiodelaHist。”etc。etc。tome1page30。M。
Brongniart，inhisreportonM。Gay’slabours“AnnalesdesSciences“1833，considersthatthebouldersintheCachapualbelongtothesameclasswiththeerraticbouldersofEurope。AstheblockswhichIsawarenotgigantic，andespeciallyastheyarenotangular，andastheyhavenotbeentransportedfairlyacrosslowspacesorwidevalleys，Iamunwillingtoclassthemwiththosewhich，bothinthenorthernandsouthernhemisphere“GeologicalTransactions“volume6page415，havebeentransportedbyice。
ItistobehopedthatwhenM。Gay’slong-continuedandadmirablelaboursinChilearepublished，morelightwillbethrownonthissubject。However，thebouldersmayhavebeenprimarilytransported;thefinalpositionofthoseofporphyry，whichhavebeendescribedasarrangedatthefootofthemountaininrudelines，Icannotdoubt，hasbeenduetotheactionofwavesonabeach。ThevalleyoftheCachapual，inthepartwherethebouldersoccur，burststhroughthehighridgeofCauquenes，whichrunsparallelto，butatsomedistancefrom，theCordillera。Thisridgehasbeensubjectedtoexcessiveviolence;trachyticlavahasburstfromit，andhotspringsyetflowatitsbase。SeeingtheenormousamountofdenudationofsolidrockintheupperandmuchbroaderpartsofthisvalleywhereitenterstheCordillera，andseeingtowhatextenttheridgeofCauquenesnowprotectsthegreatrange，Icouldnothelpbelieving(asalludedtointhetext)
thatthisridgewithitstrachyticeruptionshadbeenthrownupatamuchlaterperiodthantheCordillera。Ifthishasbeenthecase，theboulders，afterhavingbeentransportedtoalowlevelbythetorrents(whichexhibitineveryvalleyproofsoftheirpowerofmovinggreatfragments)，mayhavebeenraiseduptotheirpresentheight，withthelandonwhichtheyrested。)Finally，notwithstandingthisonecaseofdifficulty，Icannotentertainanydoubt，thattheseterrace-likefringes，whicharecontinuouslyunitedwiththebasin-shapedplainsatthefootoftheCordillera，havebeenformedbythearrestmentofriver-bornedetritusatsuccessivelevels，inthesamemannerasweseenowtakingplaceattheheadsofallthosemany，deep，windingfiordsintersectingthesoutherncoasts。Tomymind，thishasbeenoneofthemostimportantconclusionstowhichmyobservationsonthegeologyofSouthAmericahaveledme;forwethuslearnthatoneofthegrandestandmostsymmetricalmountain-chainsintheworld，withitsseveralparallellines，hasbeentogetherupliftedinmassbetweensevenandninethousandfeet，inthesamegradualmannerashavetheeasternandwesterncoastswithintherecentperiod。(Idonotwishtoaffirmthatallthelineshavebeenupliftedquiteequally;slightdifferencesintheelevationwouldleavenoperceptibleeffectontheterraces。Itmay，however，beinferred，perhapswithoneexception，thatsincetheperiodwhentheseaoccupiedthesevalleys，theseveralrangeshavenotbeendislocatedbyGREATandABRUPTfaultsorupheavals;forifsuchhadoccurred，theterracesofgravelatthesepointswouldnothavebeencontinuous。TheoneexceptionisatthelowerendofaplainintheValledelYeso(abranchoftheMaypu)，where，atagreatheight，theterracesandvalleyappeartohavebeenbrokenthroughbyalineofupheaval，ofwhichtheevidenceisplainintheadjoiningmountains;thisdislocation，perhaps，occurredAFTERTHEELEVATIONofthispartofthevalleyabovethelevelofthesea。Thevalleyhereisalmostblockedupbyapileaboutonethousandfeetinthickness，formed，asfarasIcouldjudge，fromthreesides，entirely，oratleastinchiefpart，ofgravelanddetritus。Onthesouthside，theriverhascutquitethroughthismass;onthenorthernside，andontheverysummit，deepravines，paralleltothelineofthevalley，areworn，asifthedrainagefromthevalleyabovehadpassedbythesetwolinesbeforefollowingitspresentcourse。)
FORMATIONOFVALLEYS。
ThebulkofsolidrockwhichhasbeenremovedinthelowerpartsofthevalleysoftheCordillerahasbeenenormous。ItisonlybyreflectingonsuchcasesasthatofthegravelbedsofPatagonia，coveringsomanythousandsquareleaguesofsurface，andwhich，ifheapedintoaridge，wouldformamountain-rangealmostequaltotheCordillera，thattheamountofdenudationbecomescredible。Thevalleyswithinthisrangeoftenfollowanticlinalbutrarelysynclinallines;thatis，thestrataonthetwosidesmoreoftendipfromthelineofvalleythantowardsit。Ontheflanksoftherange，thevalleysmostfrequentlyrunneitheralonganticlinalnorsynclinalaxes，butalonglinesofflexureorfaults：thatis，thestrataonbothsidesdipinthesamedirection，butwithdifferent，thoughoftenonlyslightlydifferent，inclinations。AsmostofthenearlyparallelridgeswhichtogetherformtheCordillerarunapproximatelynorthandsouth，theeastandwestvalleyscrosstheminzig-zaglines，burstingthroughthepointswherethestratahavebeenleastinclined。Nodoubtthegreaterpartofthedenudationwasaffectedattheperiodswhentidal-
creeksoccupiedthevalleys，andwhentheouterflanksofthemountainswereexposedtothefullforceofanopenocean。Ihavealreadyalludedtothepowerofthetidalactioninthechannelsconnectinggreatbays;andI
mayherementionthatoneofthesurveyingvesselsinachannelofthiskind，thoughundersail，waswhirledroundandroundbytheforceofthecurrent。Weshallhereaftersee，thatofthetwomainridgesformingtheChileanCordillera，theeasternandloftiestoneowesthegreaterpartofitsANGULARupheavaltoaperiodsubsequenttotheelevationofthewesternridge;anditislikewiseprobablethatmanyoftheotherparallelridgeshavebeenangularlyupheavedatdifferentperiods;consequentlymanypartsofthesurfacesofthesemountainsmustformerlyhavebeenexposedtothefullforceofthewaves，which，iftheCordillerawerenowsunkintothesea，wouldbeprotectedbyparallelchainsofislands。Thetorrentsinthevalleyscertainlyhavegreatpowerinwearingtherocks;ascouldbetoldbythedullrattlingsoundofthemanyfragmentsnightanddayhurryingdownwards;andaswasattestedbythevastsizeofcertainfragments，whichIwasassuredhadbeencarriedonwardsduringfloods;yetwehaveseeninthelowerpartsofthevalleys，thatthetorrentshaveseldomremovedallthesea-checkedshingleformingtheterraces，andhavehadtimesincethelastelevationinmassonlytocutintheunderlyingrocks，gorges，deepandnarrow，butquiteinsignificantindimensionscomparedwiththeentirewidthanddepthofthevalleys。
AlongtheshoresofthePacific，Ineverceasedduringmymanyandlongexcursionstofeelastonishedatseeingeveryvalley，ravine，andevenlittleinequalityofsurface，bothinthehardgraniticandsofttertiarydistricts，retainingtheexactoutline，whichtheyhadwhenthesealefttheirsurfacescoatedwithorganicremains。Whentheseremainsshallhavedecayed，therewillbescarcelyanydifferenceinappearancebetweenthislineofcoast-landandmostothercountries，whichweareaccustomedtobelievehaveassumedtheirpresentfeatureschieflythroughtheagencyoftheweatherandfresh-waterstreams。Intheoldgraniticdistricts，nodoubtitwouldberashtoattributeallthemodificationsofoutlineexclusivelytothesea-action;forwhocansayhowoftenthislatelysubmergedcoastmaynotpreviouslyhaveexistedasland，wornbyrunningstreamsandwashedbyrain？Thissourceofdoubt，however，doesnotapplytothedistrictssuperficiallyformedofthemoderntertiarydeposits。Thevalleyswornbythesea，throughthesofterformations，bothontheAtlanticandPacificsidesofthecontinent，aregenerallybroad，winding，andflat-bottomed：theonlydistrictofthisnaturenowpenetratedbyarmsofthesea，istheislandofChiloe。
Finally，theconclusionatwhichIhavearrived，withrespecttotherelativepowersofrainandseawaterontheland，is，thatthelatterisfarthemostefficientagent，andthatitschieftendencyistowidenthevalleys;whilsttorrentsandriverstendtodeepenthem，andtoremovethewreckofthesea’sdestroyingaction。Asthewaveshavemorepower，themoreopenandexposedthespacemaybe，sowilltheyalwaystendtowidenmoreandmorethemouthsofvalleyscomparedwiththeirupperparts：hence，doubtless，itis，thatmostvalleysexpandattheirmouths，——thatpart，atwhichtheriversflowinginthem，generallyhavetheleastwearingpower。
Whenreflectingontheactionoftheseaonthelandatformerlevels，theeffectofthegreatwaves，whichgenerallyaccompanyearthquakes，mustnotbeoverlooked：fewyearspasswithoutasevereearthquakeoccurringonsomepartofthewestcoastofSouthAmerica;andthewavesthuscausedhavegreatpower。AtConcepcion，aftertheshockof1835，Isawlargeslabsofsandstone，oneofwhichwassixfeetlong，threeinbreadth，andtwointhickness，thrownhighuponthebeach;andfromthenatureofthemarineanimalsstilladheringtoit，itmusthavebeentornupfromaconsiderabledepth。Ontheotherhand，atCallao，therecoil-waveoftheearthquakeof1746carriedgreatmassesofbrickwork，betweenthreeandfourfeetsquare，somewayoutseaward。Duringthecourseofages，theeffectthusproducedateachsuccessivelevel，cannothavebeensmall;andinsomeofthetertiarydepositsonthislineofcoast，Iobservedgreatbouldersofgraniteandotherneighbouringrocks，embeddedinfinesedimentarylayers，thetransportalofwhich，exceptbythemeansofearthquake-waves，alwaysappearedtomeinexplicable。
SUPERFICIALSALINEDEPOSITS。
Thissubjectmaybehereconvenientlytreatedof：Iwillbeginwiththemostinterestingcase，namely，thesuperficialsalinebedsnearIquiqueinPeru。Theporphyriticmountainsonthecoastriseabruptlytoaheightofbetweenonethousandninehundredandthreethousandfeet：betweentheirsummitsandaninlandplain，onwhichthecelebrateddepositofnitrateofsodalies，thereisahighundulatorydistrict，coveredbyaremarkablesuperficialsaliferouscrust，chieflycomposedofcommonsalt，eitherinwhite，hard，opaquenodules，ormingledwithsand，inthislattercaseformingacompactsandstone。Thissaliferoussuperficialcrustextendsfromtheedgeofthecoast-escarpment，overthewholefaceofthecountry;butneverattains，asIamassuredbyMr。Bollaert(longresidenthere)anygreatthickness。Althoughaveryslightshowerfallsonlyatintervalsofmanyyears，yetsmallfunnel-shapedcavitiesshowthatthesalthasbeeninsomepartsdissolved。(Itissingularhowslowly，accordingtotheobservationsofM。Cordieronthesalt-mountainofCardonainSpain“Ann。
desMines，TranslationofGeolog。Mem。”byDelaBechepage60，saltisdissolved，wheretheamountofrainissupposedtobeasmuchas31。4ofaninchintheyear。Itiscalculatedthatonlyfivefeetinthicknessisdissolvedinthecourseofacentury。)InseveralplacesIsawlargepatchesofsand，quitemoist，owingtothequantityofmuriateoflime(asascertainedbyMr。T。Reeks)containedinthem。Fromthecompactsalt-
cementedsandbeingeitherred，purplish，oryellow，accordingtothecolouroftherockystrataonwhichitrested，Iimaginedthatthissubstancehadprobablybeenderivedthroughcommonalluvialactionfromthelayersofsaltwhichoccurinterstratifiedinthesurroundingmountains(“JournalofResearches“page444firstedition。)：butfromtheinterestingdetailsgivenbyM。d’Orbigny，andfromfindingonafreshexaminationofthisagglomeratedsand，thatitisnotirregularlycemented，butconsistsofthinlayersofsandofdifferenttintsofcolour，alternatingwithexcessivelyfineparallellayersofsalt，Iconcludethatitisnotofalluvialorigin。M。d’OrbignyobservedanalogoussalinebedsextendingfromCobijaforfivedegreesoflatitudenorthward，andatheightsvaryingfromsixhundredtoninehundredfeet(“Voyage“etc。page102。M。d’Orbignyfoundthisdepositintersected，inmanyplaces，bydeepravines，inwhichtherewasnosalt。Streamsmustonce，thoughhistoricallyunknown，haveflowedinthem;andM。d’Orbignyarguesfromthepresenceofundissolvedsaltoverthewholesurroundingcountry，thatthestreamsmusthavearisenfromrainorsnowhavingfallen，notintheadjoiningcountry，butonthenowaridCordillera。Imayremark，thatfromhavingobservedruinsofIndianbuildingsinabsolutelysterilepartsoftheChilianCordillera(“Journal“2ndeditionpage357)，Iamledtobelievethattheclimate，atatimewhenIndianmaninhabitedthispartofthecontinent，wasinsomeslightdegreemorehumidthanitisatpresent。)：fromfindingrecentsea-
shellsstrewedonthesesaliferousbeds，andunderthem，greatwell-roundedblocks，exactlylikethoseontheexistingbeach，hebelievesthatthesalt，whichisinvariablysuperficial，hasbeenleftbytheevaporationofthesea-water。Thissameconclusionmust，Inowbelieve，beextendedtothesuperficialsaliferousbedsofIquique，thoughtheystandaboutthreethousandfeetabovethelevelofthesea。
Associatedwiththesaltinthesuperficialbeds，therearenumerous，thin，horizontallayersofimpure，dirty-white，friable，gypseousandcalcareoustuffs。Thegypseousbedsareveryremarkable，fromaboundingwith，soassometimestobealmostcomposedof，irregularconcretions，fromthesizeofaneggtothatofaman’shead，ofveryhard，compact，heavygypsum，intheformofanhydrite。Thisgypsumcontainssomeforeignparticlesofstone;itisstained，judgingfromitsactionwithborax，withiron，anditexhalesastrongaluminousodour。Thesurfacesoftheconcretionsaremarkedbysharp，radiating，orbifurcatingridges，asiftheyhadbeen(butnotreally)corroded：internallytheyarepenetratedbybranchingveins(likethoseofcalcareoussparintheseptariaoftheLondonclay)ofpurewhiteanhydrite。Theseveinsmightnaturallyhavebeenthoughttohavebeenformedbysubsequentinfiltration，hadnoteachlittleembeddedfragmentofrockbeenlikewiseedgedinaveryremarkablemannerbyanarrowborderofthesamewhiteanhydrite：thisshowsthattheveinsmusthavebeenformedbyaprocessofsegregation，andnotofinfiltration。SomeofthelittleincludedandCRACKEDfragmentsofforeignrockarepenetratedbytheanhydrite，andportionshaveevidentlybeenthusmechanicallydisplaced：atSt。Helena，Iobservedthatcalcareousmatter，depositedbyrainwater，alsohadthepowertoseparatesmallfragmentsofrockfromthelargermasses。(“VolcanicIslands“etc。page87。)Ibelievethesuperficialgypseousdepositiswidelyextended：IreceivedspecimensofitfromPisagua，fortymilesnorthofIquique，andlikewisefromArica，whereitcoatsalayerofpuresalt。M。d’OrbignyfoundatCobijaabedofclay，lyingaboveamassofupraisedrecentshells，whichwassaturatedwithsulphateofsoda，andincludedthinlayersoffibrousgypsum。(“VoyageGeolog。”etc。page95。)Thesewidelyextended，superficial，bedsofsaltandgypsum，appeartomeaninterestinggeologicalphenomenon，whichcouldbepresentedonlyunderaverydryclimate。
Theplainorbasin，onthebordersofwhichthefamousbedofnitrateofsodalies，issituatedatthedistanceofaboutthirtymilesfromthesea，beingseparatedfromitbythesaliferousdistrictjustdescribed。Itstandsataheightof3，300feet;itssurfaceislevel，andsomeleaguesinwidth;itextendsfortymilesnorthward，andhasatotallength(asIwasinformedbyMr。BelfordWilson，theConsul-GeneralatLima)of420miles。
Inawellneartheworks，thirty-sixyardsindepth，sand，earth，andalittlegravelwerefound：inanotherwell，nearAlmonte，fiftyyardsdeep，thewholeconsisted，accordingtoMr。Blake，ofclay，includingalayerofsandtwofeetthick，whichrestedonfinegravel，andthisoncoarsegravel，withlargeroundedfragmentsofrock。(Seeanadmirablepaper“GeologicalandMiscellaneousNoticesofTarapaca“in“Silliman’sAmericanJournal“volume44page1。)Inmanypartsofthisnowutterlydesertplain，rushesandlargeprostratetreesinahardenedstate，apparentlyMimosas，arefoundburied，atadepthfromthreetosixfeet;accordingtoMr。
Blake，theyhaveallfallentothesouth-west。Thebedofnitrateofsodaissaidtoextendforfortytofiftyleaguesalongthewesternmarginoftheplain，butisnotfoundinitscentralparts：itisfromtwotothreefeetinthickness，andissohardthatitisgenerallyblastedwithgunpowder;itslopesgentlyupwardsfromtheedgeoftheplaintobetweentenandthirtyfeetaboveitslevel。Itrestsonsandinwhich，itissaid，vegetableremainsandbrokenshellshavebeenfound;shellshavealsobeenfound，accordingtoMr。Blake，bothonandinthenitrateofsoda。Itiscoveredbyasuperficialmassofsand，containingnodulesofcommonsalt，and，asIwasassuredbyaminer，muchsoftgypseousmatter，preciselylikethatinthesuperficialcrustalreadydescribed：certainlythiscrust，withitscharacteristicconcretionsofanhydrite，comesclosedowntotheedgeoftheplain。
Thenitrateofsodavariesinpurityindifferentparts，andoftencontainsnodulesofcommonsalt。AccordingtoMr。Blake，theproportionofnitrateofsodavariesfrom20to75percent。AnanalysisbyMr。A。Hayes，ofanaveragespecimen，gave：——
NitrateofSoda……64。98
SulphateofSoda……3。00
ChlorideofSoda……28。69
IodicSalts……0。63
ShellsandMarl……2。60
99。90
The“mother-water“atsomeoftherefineriesisveryrichiniodicsalts，andissupposedtocontainmuchmuriateoflime。(“LiteraryGazette“1841
page475。)Inanunrefinedspecimenbroughthomebymyself，Mr。T。Reekshasascertainedthatthemuriateoflimeisveryabundant。Withrespecttotheoriginofthissalinemass，fromthemannerinwhichthegentlyinclined，compactbedfollowsforsomanymilesthesinuousmarginoftheplain，therecanbenodoubtthatitwasdepositedfromasheetofwater：
fromthefragmentsofembeddedshells，fromtheabundantiodicsalts，fromthesuperficialsaliferouscrustoccurringatahigherlevelandbeingprobablyofmarineorigin，andfromtheplainresemblinginformthoseofChileandthatofUspallata，therecanbelittledoubtthatthissheetofwaterwas，atleastoriginally，connectedwiththesea。(Fromanofficialdocument，shownmebyMr。BelfordWilson，itappearsthatthefirstexportofnitrateofsodatoEuropewasinJuly1830，onFrenchaccount，inaBritishship：——
Inyear，theentireexportwasinQuintals。
1830……17，300
1831……40，885
1832……51，400
1833……91，335
1834……149，538
TheSpanishquintalnearlyequals100Englishpounds。)
THIN，SUPERFICIAL，SALINEINCRUSTATIONS。
ThesesalineincrustationsarecommoninmanypartsofAmerica：HumboldtmetwiththemonthetablelandofMexico，andtheJesuitFalknerandotherauthorsstatethattheyoccuratintervalsoverthevastplainsextendingfromthemouthofthePlatatoRiojaandCatamarca。(Azara“Travels“volume1page55，considersthattheParanaistheeasternboundaryofthesaliferousregion;butIheardof“salitrales“intheProvinceofEntreRios。)Henceitisthatduringdroughts，mostofthestreamsinthePampasaresaline。InowheremetwiththeseincrustationssoabundantlyasnearBahiaBlanca：squaremilesofthemud-flats，whichnearthatplaceareraisedonlyafewfeetabovethesea，justenoughtoprotectthemfrombeingoverflowed，appear，afterdryweather，whiterthanthegroundafterthethickesthoar-frost。Afterrainthesaltsdisappear，andeverypuddleofwaterbecomeshighlysaline;asthesurfacedries，thecapillaryactiondrawsthemoistureuppiecesofbrokenearth，deadsticks，andtuftsofgrass，wherethesalteffloresces。Theincrustation，wherethickest，doesnotexceedaquarterofaninch。M。Parchappehasanalysedit(M。d’Orbigny“Voyage“etc。Part。Hist。tome1page664。);andfindsthatthespecimenscollectedattheextremeheadofthelowplain，neartheRiverManuello，consistof93percentofsulphateofsoda，and7ofcommonsalt;whilstthespecimenstakenclosetothecoastcontainonly63percentofthesulphate，and37ofthemuriateofsoda。Thisremarkablefact，togetherwithourknowledgethatthewholeofthislowmuddyplainhasbeencoveredbytheseawithintherecentperiod，mustleadtothesuspicionthatthecommonsalt，bysomeunknownprocess，becomesintimechangedintothesulphate。Friable，calcareousmatterishereabundant，andthecaseoftheapparentdoubledecompositionoftheshellsandsaltonSanLorenzo，shouldnotbeforgotten。
Thesalineincrustations，nearBahiaBlanca，arenotconfinedto，thoughmostabundanton，thelowmuddyflats;forInoticedsomeonacalcareousplainbetweenthirtyandfortyfeetabovethesea，andevenalittleoccursinstillhighervalleys。LowalluvialtractsinthevalleysoftheRiversNegroandColoradoarealsoencrusted，andinthelattervalleysuchspacesappearedtobeoccasionallyoverflowedbytheriver。IobservedsalineincrustationsinsomeofthevalleysofSouthernPatagonia。AtPortDesirealow，flat，muddyvalleywasthicklyincrustedbysalts，whichonanalysisbyMr。T。Reeks，arefoundtoconsistofamixtureofsulphateandmuriateofsoda，withcarbonateoflimeandearthymatter。Onthewesternsideofthecontinent，thesoutherncoastsaremuchtoohumidforthisphenomenon;
butinNorthernChileIagainmetwithsimilarincrustations。Onthehardenedmud，inpartsofthebroad，flat-bottomedvalleyofCopiapo，thesalinematterencruststhegroundtothethicknessofsomeinches：
specimens，sentbyMr。BingleytoApothecaries’Hallforanalysis，weresaidtoconsistofcarbonateandsulphateofsoda。MuchsulphateofsodaisfoundinthedesertofAtacama。InallpartsofSouthAmerica，thesalineincrustationsoccurmostfrequentlyonlowdampsurfacesofmud，wheretheclimateisratherdry;andtheselowsurfaceshave，inalmosteverycase，beenupraisedabovethelevelofthesea，withintherecentperiod。
SALT-LAKESOFPATAGONIAANDLAPLATA。
Salinas，ornaturalsalt-lakes，occurinvariousformationsontheeasternsideofthecontinent，——intheargillaceo-calcareousdepositofthePampas，inthesandstoneoftheRioNegro，wheretheyareverynumerous，inthepumiceousandotherbedsofthePatagoniantertiaryformation，andinsmallprimarydistrictsinthemidstofthislatterformation。PortS。Julianisthemostsoutherlypoint(latitude49degreesto50degrees)atwhichsalinasareknowntooccur。(AccordingtoAzara“Travels“volume1page56，therearesalt-lakesasfarnorthasChaco(latitude25degrees)，onthebanksoftheVermejo。Thesalt-lakesofSiberiaappear(Pallas“Travels“
EnglishTranslationvolume1page284)tooccurinverysimilardepressionstothoseofPatagonia。)Thedepressions，inwhichthesesalt-lakeslie，arefromafewfeettosixtymetres，asassertedbyM。d’Orbigny，belowthesurfaceofthesurroundingplains(“VoyageGeolog。”page63。);and，accordingtothissameauthor，neartheRioNegrotheyalltrend，eitherintheN。E。andS。W。orinE。andW。lines，coincidentwiththegeneralslopeoftheplain。Thesedepressionsintheplaingenerallyhaveonesidelowerthantheothers，buttherearenooutletsfordrainage。Underalessdryclimate，anoutletwouldsoonhavebeenformed，andthesaltwashedaway。
Thesalinasoccuratdifferentelevationsabovethesea;theyareoftenseveralleaguesindiameter;theyaregenerallyveryshallow，butthereisadeeponeinaquartz-rockformationnearC。Blanco。Inthewetseason，thewhole，orapart，ofthesaltisdissolved，beingredepositedduringthesucceedingdryseason。Atthisperiodtheappearanceofthesnow-whiteexpanseofsaltcrystallisedingreatcubes，isverystriking。Inalargesalina，northwardoftheRioNegro，thesaltatthebottom，duringthewholeyear，isbetweentwoandthreefeetinthickness。
Thesaltrestsalmostalwaysonathickbedofblackmuddysand，whichisfetid，probablyfromthedecayoftheburrowingwormsinhabitingit。
(ProfessorEhrenbergexaminedsomeofthismuddysand，butwasunabletofindinitanyinfusoria。)Inasalina，situatedaboutfifteenmilesabovethetownofElCarmenontheRioNegro，andthreeorfourmilesfromthebanksofthatriver，Iobservedthatthisblackmudrestedongravelwithacalcareousmatrix，similartothatspreadoverthewholesurroundingplains：atPortS。Julianthemud，also，restedonthegravel：hencethedepressionsmusthavebeenformedanteriorlyto，orcontemporaneouslywith，thespreadingoutofthegravel。IwasinformedthatonesmallsalinaoccursinanalluvialplainwithinthevalleyoftheRioNegro，andthereforeitsoriginmustbesubsequenttotheexcavationofthatvalley。
WhenIvisitedthesalina，fifteenmilesabovethetown，thesaltwasbeginningtocrystallise，andonthemuddybottomtherewerelyingmanycrystals，generallyplacedcrosswaysofsulphateofsoda(asascertainedbyMr。Reeks)，andembeddedinthemud，numerouscrystalsofsulphateoflime，fromonetothreeinchesinlength：M。d’Orbignystatesthatsomeofthesecrystalsareacicularandmorethanevennineinchesinlength(“VoyageGeolog。”page64。);othersaremacledandofgreatpurity：thoseIfoundallcontainedsomesandintheircentres。Astheblackandfetidsandoverliesthegravel，andthatoverliestheregulartertiarystrata，Ithinktherecanbenodoubtthattheseremarkablecrystalsofsulphateoflimehavebeendepositedfromthewatersofthelake。Theinhabitantscallthecrystalsofselenite，thepadredelsal，andthoseofthesulphateofsoda，themadredelsal;theyassuredmethatbotharefoundunderthesamecircumstancesinseveraloftheneighbouringsalinas;andthatthesulphateofsodaisannuallydissolved，andisalwayscrystallisedbeforethecommonsaltonthemuddybottom。(Thisiswhatmighthavebeenexpected;forM。
Ballardasserts“Acad。desSciences“October7，1844，thatsulphateofsodaisprecipitatedfromsolutionmorereadilyfromwatercontainingmuriateofsodainexcess，thanfrompurewater。)Theassociationofgypsumandsaltinthiscase，aswellasinthesuperficialdepositsofIquique，appearstomeinteresting，consideringhowgenerallythesesubstancesareassociatedintheolderstratifiedformations。
Mr。ReekshasanalysedformesomeofthesaltfromthesalinaneartheRioNegro;hefindsitcomposedentirelyofchlorideofsodium，withtheexceptionof0。26ofsulphateoflimeandof0。22ofearthymatter：therearenotracesofiodicsalts。SomesaltfromthesalinaChiquitos，inthePampeanformation，isequallypure。Itisasingularfact，thatthesaltfromthesesalinasdoesnotservesowellforpreservingmeat，assea-saltfromtheCapedeVerdeIslands;andamerchantatBuenosAyrestoldmethatheconsidereditas50percentlessvaluable。ThepurityofthePatagoniansalt，orabsencefromitofthoseothersalinebodiesfoundinallsea-
water，istheonlyassignablecauseforthisinferiority;aconclusionwhichissupportedbythefactlatelyascertained，thatthosesaltsanswerbestforpreservingcheesewhichcontainmostofthedeliquescentchlorides。(“HorticulturalandAgriculturalGazette“1845page93。)(ItwouldprobablywellanswerforthemerchantsofBuenosAyres(consideringthegreatconsumptionthereofsaltforpreservingmeat)toimportthedeliquescentchloridestomixwiththesaltfromthesalinas：Imaycallattentiontothefact，thatatIquique，alargequantityofmuriateoflime，leftintheMOTHER-WATERduringtherefinementofthenitrateofsoda，isannuallythrownaway。)
Withrespecttotheoriginofthesaltinthesalinas，theforegoinganalysisseemsopposedtotheviewentertainedbyM。d’Orbignyandothers，andwhichseemssoprobableconsideringtherecentelevationofthislineofcoast，namely，thatitisduetotheevaporationofsea-waterandtothedrainagefromthesurroundingstrataimpregnatedwithsea-salt。Iwasinformed(Iknownotwhetheraccurately)thatonthenorthernsideofthesalinaontheRioNegro，thereisasmallbrinespringwhichflowsatalltimesoftheyear：ifthisbeso，thesaltinthiscaseatleast，probablyisofsubterraneanorigin。Itatfirstappearsverysingularthatfreshwatercanoftenbeprocuredinwells，andissometimesfoundinsmalllakes，quiteclosetothesesalinas。(SirW。Parishstates“BuenosAyres“
etc。pages122and170，thatthisisthecasenearthegreatsalinaswestwardoftheS。Ventana。IhaveseensimilarstatementsinanancientMS。JournallatelypublishedbyS。Angelis。AtIquique，wherethesurfaceissothicklyencrustedwithsalinematter，Itastedwateronlyslightlybrackish，procuredinawellthirty-sixyardsdeep;buthereonefeelslesssurpriseatitspresence，aspurewatermightpercolateundergroundfromthenotverydistantCordillera。)Iamnotawarethatthisfactbearsparticularlyontheoriginofthesalt;butperhapsitisratheropposedtotheviewofthesalthavingbeenwashedoutofthesurroundingsuperficialstrata，butnottoitshavingbeentheresidueofsea-water，leftindepressionsasthelandwasslowlyelevated。
CHAPTERIV。ONTHEFORMATIONSOFTHEPAMPAS。
Mineralogicalconstitution。
Microscopicalstructure。
BuenosAyres，shellsembeddedintosca-rock。
BuenosAyrestotheColorado。
SanVentana。
BahiaBlanca;M。Hermoso，bonesandinfusoriaof;P。Alta，shells，bones，andinfusoriaof;co-existenceoftherecentshellsandextinctmammifers。
BuenosAyrestoSantaFe。
SkeletonsofMastodon。
Infusoria。
Inferiormarinetertiarystrata，theirage。
Horse’stooth。
BANDAORIENTAL。
SuperficialPampeanformation。
Inferiortertiarystrata，variationof，connectedwithvolcanicaction;
MacraucheniaPatachonicaatSanJulianinPatagonia，ageof，subsequenttolivingmolluscaandtotheerraticblockperiod。
SUMMARY。
AreaofPampeanformation。
Theoriesoforigin。
Sourceofsediment。
Estuaryorigin。
Contemporaneouswithexistingmollusca。
Relationstounderlyingtertiarystrata。
Ancientdepositofestuaryorigin。
ElevationandsuccessivedepositionofthePampeanformation。
Numberandstateoftheremainsofmammifers;theirhabitation，food，extinction，andrange。
Conclusion。
LocalitiesinPampasatwhichmammiferousremainshavebeenfound。
ThePampeanformationishighlyinterestingfromitsvastextent，itsdisputedorigin，andfromthenumberofextinctgiganticmammifersembeddedinit。Ithasuponthewholeaveryuniformcharacter：consistingofamoreorlessdullreddish，slightlyindurated，argillaceousearthormud，often，butnotalways，includinginhorizontallinesconcretionsofmarl，andfrequentlypassingintoacompactmarlyrock。Themud，whereverIexaminedit，evenclosetotheconcretions，didnotcontainanycarbonateoflime。
Theconcretionsaregenerallynodular，sometimesroughexternally，sometimesstalactiformed;theyareofacompactstructure，butoftenpenetrated(aswellasthemud)byhair-likeserpentinecavities，andoccasionallywithirregularfissuresintheircentres，linedwithminutecrystalsofcarbonateoflime;theyareofwhite，brown，orpalepinkishtints，oftenmarkedbyblackdendriticmanganeseoriron;theyareeitherdarkerorlightertintedthanthesurroundingmass;theycontainmuchcarbonateoflime，butexhaleastrongaluminousodour，andleave，whendissolvedinacids，alargebutvaryingresidue，ofwhichthegreaterpartconsistsofsand。Theseconcretionsoftenuniteintoirregularstrata;andoververylargetractsofcountry，theentiremassconsistsofahard，butgenerallycavernousmarlyrock：someofthevarietiesmightbecalledcalcareoustuffs。
Dr。Carpenterhaskindlyexaminedunderthemicroscope，slicedandpolishedspecimensoftheseconcretions，andofthesolidmarl-rock，collectedinvariousplacesbetweentheColoradoandSantaFeBajada。Inthegreaternumber，Dr。Carpenterfindsthatthewholesubstancepresentsatolerablyuniformamorphouscharacter，butwithtracesofincipientcrystallinemetamorphosis;inotherspecimenshefindsmicroscopicallyminuteroundedconcretionsofanamorphoussubstance(resemblinginsizethoseinooliticrocks，butnothavingaconcentricstructure)，unitedbyacementwhichisoftencrystalline。Insome，Dr。Carpentercanperceivedistincttracesofshells，corals，Polythalamia，andrarelyofspongoidbodies。Forthesakeofcomparison，IsentDr。Carpenterspecimensofthecalcareousrock，formedchieflyoffragmentsofrecentshells，fromCoquimboinChile：inoneofthesespecimens，Dr。Carpenterfinds，besidesthelargerfragments，microscopicalparticlesofshells，andavaryingquantityofopaqueamorphousmatter;inanotherspecimenfromthesamebed，hefindsthewholecomposedoftheamorphousmatter，withlayersshowingindicationsofanincipientcrystallinemetamorphosis：hencetheselatterspecimens，bothinexternalappearanceandinmicroscopicalstructure，closelyresemblethoseofthePampas。Dr。Carpenterinformsmethatitiswellknownthatchemicalprecipitationthrowsdowncarbonateoflimeintheopaqueamorphousstate;
andheisinclinedtobelievethatthelong-continuedattritionofacalcareousbodyinastateofcrystallineorsemi-crystallineaggregation(as，forinstance，intheordinaryshellsofMollusca，which，whensliced，aretransparent)mayyieldthesameresult。FromtheintimaterelationsbetweenalltheCoquimbospecimens，Icanhardlydoubtthattheamorphouscarbonateoflimeinthemhasresultedfromtheattritionanddecayofthelargerfragmentsofshell：whethertheamorphousmatterinthemarlyrocksofthePampashaslikewisethusoriginated，itwouldbehazardoustoconjecture。
Forconvenience’sake，Iwillcallthemarlyrockbythenamegiventoitbytheinhabitants，namely，Tosca-rock;andthereddishargillaceousearth，Pampeanmud。Thislattersubstance，Imaymention，hasbeenexaminedformebyProfessorEhrenberg，andtheresultofhisexaminationwillbegivenundertheproperlocalities。
Iwillcommencemydescriptionsatacentralspot，namely，atBuenosAyres，andthenceproceedfirstsouthwardtotheextremelimitofthedeposit，andafterwardsnorthward。TheplainonwhichBuenosAyresstandsisfromthirtytofortyfeetinheight。ThePampeanmudishereofaratherpalecolour，andincludessmallnearlywhitenodules，andotherirregularstrataofanunusuallyarenaceousvarietyoftosca-rock。Inawellatthedepthofseventyfeet，accordingtoIgnatioNunez，muchtosca-rockwasmetwith，andatseveralpoints，atonehundredfeetdeep，bedsofsandhavebeenfound。
IhavealreadygivenalistoftherecentmarineandestuaryshellsfoundinmanypartsonthesurfacenearBuenosAyres，asfarasthreeorfourleaguesfromthePlata。SpecimensfromnearEnsenada，givenmebySirW。
Parish，wheretherockisquarriedjustbeneaththesurfaceoftheplain，consistofbrokenbivalves，cementedbyandconvertedintowhitecrystallinecarbonateoflime。Ihavealreadyalluded，inthefirstchapter，toaspecimen(alsogivenmebySirW。Parish)fromtheA。delTristan，inwhichshells，resemblingineveryrespecttheAzaralabiata，d’Orbigny，asfarastheirwornconditionpermitsofcomparison，areembeddedinareddish，softish，somewhatarenaceousmarlyrock：aftercarefulcomparison，withtheaidofamicroscopeandacids，IcanperceivenodifferencebetweenthebasisofthisrockandthespecimenscollectedbymeinmanypartsofthePampas。Ihavealsostated，ontheauthorityofSirW。Parish，thatnorthwardofBuenosAyres，onthehighestpartsoftheplain，aboutfortyfeetabovethePlata，andtwoorthreemilesfromit，numerousshellsoftheAzaralabiata(andIbelieveofVenussinuosa)occurembeddedinastratifiedearthymass，includingsmallmarlyconcretions，andsaidtobepreciselylikethegreatPampeandeposit。HencewemayconcludethatthemudofthePampascontinuedtobedepositedtowithintheperiodofthisexistingestuaryshell。Althoughthisformationisofsuchimmenseextent，Iknowofnootherinstanceofthepresenceofshellsinit。
BUENOSAYRESTOTHERIOCOLORADO。
Withtheexceptionofafewmetamorphicridges，thecountrybetweenthesetwopoints，adistanceof400geographicalmiles，belongstothePampeanformation，andinthesouthernpartisgenerallyformedoftheharderandmorecalcareousvarieties。Iwillbrieflydescribemyroute：abouttwenty-
fivemilesS。S。W。ofthecapital，inawellfortyyardsindepth，theupperpart，and，asIwasassured，theentirethickness，wasformedofdarkredPampeanmudwithoutconcretions。NorthoftheRiverSalado，therearemanylakes;andonthebanksofone(neartheGuardia)therewasalittlecliffsimilarlycomposed，butincludingmanynodularandstalactiformconcretions：Ifoundherealargepieceoftessellatedarmour，likethatoftheGlyptodon，andmanyfragmentsofbones。ThecliffsontheSaladoconsistofpale-colouredPampeanmud，includingandpassingintogreatmassesoftosca-rock：hereaskeletonoftheMegatheriumandthebonesofotherextinctquadrupeds(seethelistattheendofthischapter)werefound。Largequantitiesofcrystallisedgypsum(ofwhichspecimensweregivenme)occurinthecliffsofthisriver;andlikewise(asIwasassuredbyMr。Lumb)inthePampeanmudontheRiverChuelo，sevenleaguesfromBuenosAyres：ImentionthisbecauseM。d’OrbignylayssomestressonthesupposedabsenceofthismineralinthePampeanformation。
SouthwardoftheSaladothecountryislowandswampy，withtosca-rockappearingatlongintervalsatthesurface。Onthebanks，however，oftheTapalguen(sixtymilessouthoftheSalado)thereisalargeextentoftosca-rock，somehighlycompactandevensemi-crystalline，overlyingpalePampeanmudwiththeusualconcretions。Thirtymilesfurthersouth，thesmallquartz-ridgeofTapalguenisfringedonitsnorthernandsouthernflank，bylittle，narrow，flat-toppedhillsoftosca-rock，whichstandhigherthanthesurroundingplain。BetweenthisridgeandtheSierraofGuitru-gueyu，adistanceofsixtymiles，thecountryisswampy，withthetosca-rockappearingonlyinfourorfivespots：thissierra，preciselylikethatofTapalguen，isborderedbyhorizontal，oftencliff-bounded，littlehillsoftosca-rock，higherthanthesurroundingplain。Here，also，anewappearancewaspresentedinsomeextensiveandlevelbanksofalluviumordetritusoftheneighbouringmetamorphicrocks;butIneglectedtoobservewhetheritwasstratifiedornot。BetweenGuitru-gueyuandtheSierraVentana，Icrossedadryplainoftosca-rockhigherthanthecountryhithertopassedover，andwithsmallpiecesofdenudedtablelandofthesameformation，standingstillhigher。
ThemarlyorcalcareousbedsnotonlycomeupnearlyhorizontallytothenorthernandsouthernfootofthegreatquartzosemountainsoftheSierraVentana，butinterfoldbetweentheparallelranges。Thesuperficialbeds(forInowhereobtainedsectionsmorethantwentyfeetdeep)retain，evenclosetothemountains，theirusualcharacter：theuppermostlayer，however，inoneplaceincludedpebblesofquartz，andrestedonamassofdetritusofthesamerock。Attheveryfootofthemountains，thereweresomefewpilesofquartzandtosca-rockdetritus，includingland-shells;
butatthedistanceofonlyhalfamilefromtheselofty，jagged，andbatteredmountains，Icouldnot，tomygreatsurprise，findontheboundlesssurfaceofthecalcareousplainevenasinglepebble。Quartz-
pebbles，however，ofconsiderablesizehaveatsomeperiodbeentransportedtoadistanceofbetweenfortyandfiftymilestotheshoresofBahiaBlanca。(Schmidtmeyer“TravelsinChile“page150，statesthathefirstnoticedonthePampas，verysmallbitsofredgranite，whenfiftymilesdistantfromthesouthernextremityofthemountainsofCordova，whichprojectontheplain，likeareefintothesea。)
ThehighestpeakoftheSt。Ventanais，byCaptainFitzroy’smeasurement，3，340feet，andthecalcareousplainatitsfoot(fromobservationstakenbysomeSpanishofficers)840feetabovethesea-level。(“LaPlata“etc。bySirW。Parishpage146。)Ontheflanksofthemountains，ataheightofthreehundredorfourhundredfeetabovetheplain，therewereafewsmallpatchesofconglomerateandbreccia，firmlycementedbyferruginousmattertotheabruptandbatteredfaceofthequartz——tracesbeingthusexhibitedofancientsea-action。Thehighplainroundthisrangesinksquiteinsensiblytotheeyeonallsides，excepttothenorth，whereitssurfaceisbrokenintolowcliffs。RoundtheSierrasTapalguen，Guitru-gueyu，andbetweenthelatterandtheVentanawehaveseen(andshallhereafterseeroundsomehillsinBandaOriental)，thatthetosca-rockformslow，flat-
topped，cliff-boundedhills，higherthanthesurroundingplainsofsimilarcomposition。Fromthehorizontalstratificationandfromtheappearanceofthebrokencliffs，thegreaterheightofthePampeanformationroundtheseprimaryhillsoughtnottobealtogetherorinchiefpartattributedtotheseseveralpointshavingbeenupliftedmoreenergeticallythanthesurroundingcountry，buttotheargillaceo-calcareousmudhavingcollectedroundthem，whentheyexistedasisletsorsubmarinerocks，atagreaterheight，thanatthebottomoftheadjoiningopensea;——thecliffshavingbeensubsequentlywornduringtheelevationofthewholecountryinmass。
SouthwardoftheVentana，theplainextendsfartherthantheeyecanrange;
itssurfaceisnotverylevel，havingslightdepressionswithnodrainageexits;itisgenerallycoveredbyafewfeetinthicknessofsandyearth;
andinsomeplaces，accordingtoM。Parchappe，bybedsofclaytwoyardsthick。(M。d’Orbigny“Voyage“PartGeolog。pages47，48。)OnthebanksoftheSauce，fourleaguesS。E。oftheVentana，thereisanimperfectsectionabouttwohundredfeetinheight，displayingintheupperparttosca-rockandinthelowerpartredPampeanmud。AtthesettlementofBahiaBlanca，theuppermostplainiscomposedofverycompact，stratifiedtosca-rock，containingroundedgrainsofquartzdistinguishablebythenakedeye：thelowerplain，onwhichthefortressstands，isdescribedbyM。Parchappeascomposedofsolidtosca-rock(Ibid。);butthesectionswhichIexaminedappearedmorelikearedepositedmassofthisrock，withsmallpebblesandfragmentsofquartz。IshallimmediatelyreturntotheimportantsectionsontheshoresofBahiaBlanca。Twentymilessouthwardofthisplace，thereisaremarkableridgeextendingW。byN。andE。byS。formedofsmall，separate，flat-topped，steep-sidedhills，risingbetweenonehundredandtwohundredfeetabovethePampeanplainatitssouthernbase，whichplainisalittlelowerthanthattothenorth。Theuppermoststratuminthisridgeconsistsofpale，highlycalcareous，compacttosca-rock，resting(asseeninoneplace)onreddishPampeanmud，andthisagainonapalerkind：
atthefootoftheridge，thereisawellinreddishclayormud。IhaveseennootherinstanceofachainofhillsbelongingtothePampeanformation;andasthestratashownosignsofdisturbance，andasthedirectionoftheridgeisthesamewiththatcommontoallthemetamorphiclinesinthiswholearea，IsuspectthatthePampeansedimenthasinthisinstancebeenaccumulatedonandoveraridgeofhardrocks，insteadof，asinthecaseoftheabove-mentionedSierras，roundtheirsubmarineflanks。
Southofthislittlechainoftosca-rock，aplainofPampeanmuddeclinestowardsthebanksoftheColorado：inthemiddleawellhasbeenduginredPampeanmud，coveredbytwofeetofwhite，softish，highlycalcareoustosca-rock，overwhichliessandwithsmallpebblesthreefeetinthickness——thefirstappearanceofthatvastshingleformationdescribedintheFirstChapter。InthefirstsectionaftercrossingtheColorado，anoldtertiaryformation，namely，theRioNegrosandstone(tobedescribedinthenextchapter)，ismetwith：butfromtheaccountsgivenmebytheGauchos，IbelievethatatthemouthoftheColoradothePampeanformationextendsalittlefurthersouthwards。
BAHIABLANCA。
Toreturntotheshoresofthisbay。AtMonteHermosothereisagoodsection，aboutonehundredfeetinheight，offourdistinctstrata，appearingtotheeyehorizontal，butthickeningalittletowardstheN。W。
Theuppermostbed，abouttwentyfeetinthickness，consistsofobliquelylaminated，softsandstone，includingmanypebblesofquartz，andfallingatthesurfaceintoloosesand。Thesecondbed，onlysixinchesthick，isahard，dark-colouredsandstone。Thethirdbedispale-colouredPampeanmud;
andthefourthisofthesamenature，butdarkercoloured，includinginitslowerparthorizontallayersandlinesofconcretionsofnotverycompactpinkishtosca-rock。Thebottomofthesea，Imayremark，toadistanceofseveralmilesfromtheshore，andtoadepthofbetweensixtyandonehundredfeet，wasfoundbytheanchorstobecomposedoftosca-rockandreddishPampeanmud。ProfessorEhrenberghasexaminedformespecimensofthetwolowerbeds，andfindsinthemthreePolygastricaandsixPhytolitharia。
(Thefollowinglistisgiveninthe“Monatsberichtenderkonig。Akad。zuBerlin“April1845：——
POLYGASTRICA。
Fragilariarhabdosoma。
Gallionelladistans。
Pinnularia？
PHYTOLITHARIA。
LithodontiumBursa。
Lithodontiumfurcatum。
Lithostylidiumexesum。
Lithostylidiumrude。
LithostylidiumSerra。
SpongolithisFustis？)
Ofthese，onlyone(SpongolithisFustis？)isamarineform;fiveofthemareidenticalwithmicroscopicalstructuresofbrackish-waterorigin，hereaftertobementioned，whichformacentralpointinthePampeanformation。Inthesetwobeds，especiallyinthelowerone，bonesofextinctmammifers，someembeddedintheirproperrelativepositionsandotherssingle，areverynumerousinasmallextentofthecliffs。Theseremainsconsistof，first，theheadofCtenomysantiquus，alliedtothelivingCtenomysBraziliensis;secondly，afragmentoftheremainsofarodent;
thirdly，molarteethandotherbonesofalargerodent，closelyalliedto，butdistinctfrom，theexistingspeciesofHydrochoerus，andthereforeprobablyaninhabitantoffreshwater;fourthandfifthly，portionsofvertebrae，limbs，ribs，andotherbonesoftworodents;sixthly，bonesoftheextremitiesofsomegreatmegatheroidquadruped。(See“FossilMammalia“
page109byProfessorOwen，inthe“ZoologyoftheVoyageofthe’Beagle’;“
andCataloguepage36ofFossilRemainsinMuseumofRoyalCollegeofSurgeons。)Thenumberoftheremainsofrodentsgivestothiscollectionapeculiarcharacter，comparedwiththosefoundinanyotherlocality。Allthesebonesarecompactandheavy;manyofthemarestainedred，withtheirsurfacespolished;someofthesmalleronesareasblackasjet。
MonteHermosoisbetweenfiftyandsixtymilesdistantinaS。E。linefromtheVentana，withtheintermediatecountrygentlyrisingtowardsit，andallconsistingofthePampeanformation。Whatrelation，then，dothesebeds，attheleveloftheseaandunderit，beartothoseontheflanksoftheVentana，attheheightof840feet，andontheflanksoftheotherneighbouringsierras，which，fromthereasonsalreadyassigned，donotappeartoowetheirgreaterheighttounequalelevation？Whenthetosca-
rockwasaccumulatingroundtheVentana，andwhen，withtheexceptionofafewsmallruggedprimaryislands，thewholewidesurroundingplainsmusthavebeenunderwater，werethestrataatMonteHermosodepositingatthebottomofagreatopensea，betweeneighthundredandonethousandfeetindepth？Imuchdoubtthis;forifso，thealmostperfectcarcassesoftheseveralsmallrodents，theremainsofwhicharesoverynumerousinsolimitedaspace，musthavebeendriftedtothisspotfromthedistanceofmanyhundredmiles。Itappearstomefarmoreprobable，thatduringthePampeanperiodthiswholeareahadcommencedslowlyrising(andinthecliffs，atseveraldifferentheightswehaveproofsofthelandhavingbeenexposedtosea-actionatseverallevels)，andthattractsoflandhadthusbeenformedofPampeansedimentroundtheVentanaandtheotherprimaryranges，onwhichtheseveralrodentsandotherquadrupedslived，andthatastream(inwhichperhapstheextinctaquaticHydrochoeruslived)driftedtheirbodiesintotheadjoiningsea，intowhichthePampeanmudcontinuedtobepouredfromthenorth。Asthelandcontinuedtorise，itappearsthatthissourceofsedimentwascutoff;andinitsplacesandandpebbleswerebornedownbystrongercurrents，andconformablydepositedoverthePampeanstrata。
(FIGURE15。SECTIONOFBEDSWITHRECENTSHELLSANDEXTINCTMAMMIFERS，AT
PUNTAALTAINBAHIABLANCA。(Showingbedsfrombottomtotop：A，B，C，D。))
