Atnight，astrangerarrivedatthehouseofDonBenito，andaskedpermissiontosleepthere。Hesaidhehadbeenwanderingaboutthemountainsforseventeendays，havinglosthisway。HestartedfromGuasco，andbeingaccustomedtotravellingintheCordillera，didnotexpectanydifficultyinfollowingthetracktoCopiapo；buthesoonbecameinvolvedinalabyrinthofmountains，whencehecouldnotescape。Someofhismuleshadfallenoverprecipices，andhehadbeeningreatdistress。Hischiefdifficultyarosefromnotknowingwheretofindwaterinthelowercountry，sothathewasobligedtokeepborderingthecentralranges。
　　Wereturneddownthevalley，andonthe22ndreachedthetownofCopiapo。
　　Thelowerpartofthevalleyisbroad，formingafineplainlikethatofQuillota。Thetowncoversaconsiderablespaceofground，eachhousepossessingagarden：butitisanuncomfortableplace，andthedwellingsarepoorlyfurnished。Everyoneseemsbentontheoneobjectofmakingmoney，andthenmigratingasquicklyaspossible。Alltheinhabitantsaremoreorlessdirectlyconcernedwithmines；andminesandoresarethesolesubjectsofconversation。Necessariesofallsortsareextremelydear；asthedistancefromthetowntotheportiseighteenleagues，andthelandcarriageveryexpensive。Afowlcostsfiveorsixshillings；meatisnearlyasdearasinEngland；firewood，orrathersticks，arebroughtondonkeysfromadistanceoftwoandthreedays’journeywithintheCordillera；andpasturageforanimalsisashillingaday：allthisforSouthAmericaiswonderfullyexorbitant。
　　June26th——IhiredaguideandeightmulestotakemeintotheCordillerabyadifferentlinefrommylastexcursion。Asthecountrywasutterlydesert，wetookacargoandahalfofbarleymixedwithchoppedstraw。
　　Abouttwoleaguesabovethetownabroadvalleycalledthe“Despoblado。”
　　oruninhabited，branchesofffromthatonebywhichwehadarrived。Althoughavalleyofthegrandestdimensions，andleadingtoapassacrosstheCordillera，yetitiscompletelydry，exceptingperhapsforafewdaysduringsomeveryrainywinter。Thesidesofthecrumblingmountainswerefurrowedbyscarcelyanyravines；andthebottomofthemainvalley，filledwithshingle，wassmoothandnearlylevel。Noconsiderabletorrentcouldeverhavefloweddownthisbedofshingle；forifithad，agreatcliff-boundedchannel，asinallthesouthernvalleys，wouldassuredlyhavebeenformed。Ifeellittledoubtthatthisvalley，aswellasthosementionedbytravellersinPeru，wereleftinthestatewenowseethembythewavesofthesea，asthelandslowlyrose。Iobservedinoneplace，wheretheDespobladowasjoinedbyaravinewhichinalmostanyotherchainwouldhavebeencalledagrandvalley，thatitsbed，thoughcomposedmerelyofsandandgravel，washigherthanthatofitstributary。Amererivuletofwater，inthecourseofanhour，wouldhavecutachannelforitself；butitwasevidentthatageshadpassedaway，andnosuchrivulethaddrainedthisgreattributary。Itwascurioustobeholdthemachinery，ifsuchatermmaybeused，forthedrainage，all，withthelasttriflingexception，perfect，yetwithoutanysignsofaction。Everyonemusthaveremarkedhowmud-banks，leftbytheretiringtide，imitateinminiatureacountrywithhillanddale；andherewehavetheoriginalmodelinrock，formedasthecontinentroseduringthesecularretirementoftheocean，insteadofduringtheebbingandflowingofthetides。Ifashowerofrainfallsonthemud-bank，whenleftdry，itdeepensthealready-formedshallowlinesofexcavation；
　　andsoitiswiththerainofsuccessivecenturiesonthebankofrockandsoil，whichwecallacontinent。
　　Werodeonafteritwasdark，tillwereachedasideravinewithasmallwell，called“Aguaamarga。”Thewaterdeserveditsname，forbesidesbeingsalineitwasmostoffensivelyputridandbitter；sothatwecouldnotforceourselvestodrinkeitherteaormate。IsupposethedistancefromtheriverofCopiapotothisspotwasatleasttwenty-fiveorthirtyEnglishmiles；inthewholespacetherewasnotasingledropofwater，thecountrydeservingthenameofdesertinthestrictestsense。YetabouthalfwaywepassedsomeoldIndianruinsnearPuntaGorda：Inoticedalsoinfrontofsomeofthevalleys，whichbranchofffromtheDespoblado，twopilesofstonesplacedalittlewayapart，anddirectedsoastopointupthemouthsofthesesmallvalleys。Mycompanionsknewnothingaboutthem，andonlyansweredmyqueriesbytheirimperturbable“quiensabe？”
　　IobservedIndianruinsinseveralpartsoftheCordillera：themostperfectwhichIsaw，weretheRuinasdeTambillos，intheUspallataPass。
　　Smallsquareroomsweretherehuddledtogetherinseparategroups：someofthedoorwayswereyetstanding；theywereformedbyacrossslabofstoneonlyaboutthreefeethigh。UlloahasremarkedonthelownessofthedoorsintheancientPeruviandwellings。Thesehouses，whenperfect，musthavebeencapableofcontainingaconsiderablenumberofpersons。
　　Traditionsays，thattheywereusedashalting-placesfortheIncas，whentheycrossedthemountains。TracesofIndianhabitationshavebeendiscoveredinmanyotherparts，whereitdoesnotappearprobablethattheywereusedasmereresting-places，butyetwherethelandisasutterlyunfitforanykindofcultivation，asitisneartheTambillosorattheIncasBridge，orinthePortilloPass，atallwhichplacesIsawruins。IntheravineofJajuel，nearAconcagua，wherethereisnopass，Iheardofremainsofhousessituatedatagreatheight，whereitisextremelycoldandsterile。
　　AtfirstIimaginedthatthesebuildingshadbeenplacesofrefuge，builtbytheIndiansonthefirstarrivaloftheSpaniards；butIhavesincebeeninclinedtospeculateontheprobabilityofasmallchangeofclimate。
　　InthisnorthernpartofChile，withintheCordillera，oldIndianhousesaresaidtobeespeciallynumerous：bydiggingamongsttheruins，bitsofwoollenarticles，instrumentsofpreciousmetals，andheadsofIndiancorn，arenotunfrequentlydiscovered：anarrow-headmadeofagate，andofpreciselythesameformwiththosenowusedinTierradelFuego，wasgivenme。IamawarethatthePeruvianIndiansnowfrequentlyinhabitmostloftyandbleaksituations；butatCopiapoIwasassuredbymenwhohadspenttheirlivesintravellingthroughtheAndes，thattherewereverymanymuchisimasbuildingsatheightssogreatasalmosttoborderupontheperpetualsnow，andinpartswherethereexistnopasses，andwherethelandproducesabsolutelynothing，andwhatisstillmoreextraordinary，wherethereisnowater。Neverthelessitistheopinionofthepeopleofthecountryalthoughtheyaremuchpuzzledbythecircumstance，that，fromtheappearanceofthehouses，theIndiansmusthaveusedthemasplacesofresidence。Inthisvalley，atPuntaGorda，theremainsconsistedofsevenoreightsquarelittlerooms，whichwereofasimilarformwiththoseatTambillos，butbuiltchieflyofmud，whichthepresentinhabitantscannot，eitherhereor，accordingtoUlloa，inPeru，imitateindurability。Theyweresituatedinthemostconspicuousanddefencelessposition，atthebottomoftheflatbroadvalley。Therewasnowaternearerthanthreeorfourleagues，andthatonlyinverysmallquantity，andbad：thesoilwasabsolutelysterile；Ilookedinvainevenforalichenadheringtotherocks。Atthepresentday，withtheadvantageofbeastsofburden，amine，unlessitwereveryrich，couldscarcelybeworkedherewithprofit。YettheIndiansformerlychoseitasaplaceofresidence！Ifatthepresenttimetwoorthreeshowersofrainweretofallannually，insteadofone，asnowisthecaseduringasmanyyears，asmallrillofwaterwouldprobablybeformedinthisgreatvalley；andthen，byirrigationwhichwasformerlysowellunderstoodbytheIndians，thesoilwouldeasilyberenderedsufficientlyproductivetosupportafewfamilies。
　　IhaveconvincingproofsthatthispartofthecontinentofSouthAmericahasbeenelevatednearthecoastatleastfrom400to500，andinsomepartsfrom1000to1300feet，sincetheepochofexistingshells；andfurtherinlandtherisepossiblymayhavebeengreater。AsthepeculiarlyaridcharacteroftheclimateisevidentlyaconsequenceoftheheightoftheCordillera，wemayfeelalmostsurethatbeforethelaterelevations，theatmospherecouldnothavebeensocompletelydrainedofitsmoistureasitnowis；andastherisehasbeengradual，sowouldhavebeenthechangeinclimate。Onthisnotionofachangeofclimatesincethebuildingswereinhabited，theruinsmustbeofextremeantiquity，butIdonotthinktheirpreservationundertheChilianclimateanygreatdifficulty。WemustalsoadmitonthisnotionandthisperhapsisagreaterdifficultythatmanhasinhabitedSouthAmericaforanimmenselylongperiod，inasmuchasanychangeofclimateeffectedbytheelevationofthelandmusthavebeenextremelygradual。AtValparaiso，withinthelast220years，therisehasbeensomewhatlessthan19feet：atLimaasea-beachhascertainlybeenupheavedfrom80to90feet，withintheIndo-humanperiod：butsuchsmallelevationscouldhavehadlittlepowerindeflectingthemoisture-bringingatmosphericcurrents。Dr。Lund，however，foundhumanskeletonsinthecavesofBrazil，theappearanceofwhichinducedhimtobelievethattheIndianracehasexistedduringavastlapseoftimeinSouthAmerica。
　　WhenatLima，Iconversedonthesesubjects[3]withMr。Gill，acivilengineer，whohadseenmuchoftheinteriorcountry。
　　Hetoldmethataconjectureofachangeofclimatehadsometimescrossedhismind；butthathethoughtthatthegreaterportionofland，nowincapableofcultivation，butcoveredwithIndianruins，hadbeenreducedtothisstatebythewater-conduits，whichtheIndiansformerlyconstructedonsowonderfulascale，havingbeeninjuredbyneglectandbysubterraneanmovements。Imayheremention，thatthePeruviansactuallycarriedtheirirrigatingstreamsintunnelsthroughhillsofsolidrock。Mr。Gilltoldme，hehadbeenemployedprofessionallytoexamineone：hefoundthepassagelow，narrow，crooked，andnotofuniformbreadth，butofveryconsiderablelength。Isitnotmostwonderfulthatmenshouldhaveattemptedsuchoperations，withouttheuseofironorgunpowder？Mr。Gillalsomentionedtomeamostinteresting，and，asfarasIamaware，quiteunparalleledcase，ofasubterraneandisturbancehavingchangedthedrainageofacountry。TravellingfromCasmatoHuaraznotveryfardistantfromLima，hefoundaplaincoveredwithruinsandmarksofancientcultivationbutnowquitebarren。Nearitwasthedrycourseofaconsiderableriver，whencethewaterforirrigationhadformerlybeenconducted。Therewasnothingintheappearanceofthewater-coursetoindicatethattheriverhadnotflowedthereafewyearspreviously；insomeparts，bedsofsandandgravelwerespreadout；inothers，thesolidrockhadbeenwornintoabroadchannel，whichinonespotwasabout40yardsinbreadthand8feetdeep。Itisself-evidentthatapersonfollowingupthecourseofastream，willalwaysascendatagreaterorlessinclination：Mr。Gill，therefore，wasmuchastonished，whenwalkingupthebedofthisancientriver，tofindhimselfsuddenlygoingdownhill。Heimaginedthatthedownwardslopehadafallofabout40or50feetperpendicular。Weherehaveunequivocalevidencethataridgehadbeenupliftedrightacrosstheoldbedofastream。Fromthemomenttheriver-coursewasthusarched，thewatermustnecessarilyhavebeenthrownback，andanewchannelformed。Fromthatmoment，also，theneighbouringplainmusthavelostitsfertilizingstream，andbecomeadesert。
　　June27th——Wesetoutearlyinthemorning，andbymiddayreachedtheravineofPaypote，wherethereisatinyrillofwater，withalittlevegetation，andevenafewalgarrobatrees，akindofmimosa。Fromhavingfire-wood，asmelting-furnacehadformerlybeenbuilthere：wefoundasolitarymaninchargeofit，whosesoleemploymentwashuntingguanacos。
　　Atnightitfrozesharply；buthavingplentyofwoodforourfire，wekeptourselveswarm。
　　28th——Wecontinuedgraduallyascending，andthevalleynowchangedintoaravine。Duringthedaywesawseveralguanacos，andthetrackoftheclosely-alliedspecies，theVicuna：thislatteranimalispre-eminentlyalpineinitshabits；itseldomdescendsmuchbelowthelimitofperpetualsnow，andthereforehauntsevenamoreloftyandsterilesituationthantheguanaco。Theonlyotheranimalwhichwesawinanynumberwasasmallfox：Isupposethisanimalpreysonthemiceandothersmallrodents，which，aslongasthereistheleastvegetation，subsistinconsiderablenumbersinverydesertplaces。InPatagonia，evenonthebordersofthesalinas，whereadropoffreshwatercanneverbefound，exceptingdew，theselittleanimalsswarm。Nexttolizards，miceappeartobeabletosupportexistenceonthesmallestanddriestportionsoftheearth——evenonisletsinthemidstofgreatoceans。
　　Thesceneonallsidesshoweddesolation，brightenedandmadepalpablebyaclear，uncloudedsky。Foratimesuchsceneryissublime，butthisfeelingcannotlast，andthenitbecomesuninteresting。Webivouackedatthefootofthe“primeralinea。”orthefirstlineofthepartitionofwaters。Thestreams，however，ontheeastsidedonotflowtotheAtlantic，butintoanelevateddistrict，inthemiddleofwhichthereisalargesaline，orsaltlake；thusformingalittleCaspianSeaattheheight，perhaps，oftenthousandfeet。Whereweslept，thereweresomeconsiderablepatchesofsnow，buttheydonotremainthroughouttheyear。Thewindsintheseloftyregionsobeyveryregularlawseverydayafreshbreezeblowsupthevalley，andatnight，anhourortwoaftersunset，theairfromthecoldregionsabovedescendsasthroughafunnel。Thisnightitblewagaleofwind，andthetemperaturemusthavebeenconsiderablybelowthefreezing-point，forwaterinavesselsoonbecameablockofice。