Whilepassingacrosstheseinterminable-lookingplains，theeyerestswithpleasureonasmallflower，whichexistsinsuchnumbersastogiveitsownhuetothegroundOnebroadbandofyellowstretchesacrossourpathOnlookingattheflowerswhichformedthisgoldencarpet，wesaweveryvarietyofthatcolor，fromthepalestlemontotherichestorangeCrossingahundredyardsofthis，wecameuponanotherbroadbandofthesameflower，butblue，andthiscolorisvariedfromthelightesttinttodarkblue，andevenpurple。
Ihadbeforeobservedthesameflowerpossessingdifferentcolorsindifferentpartsofthecountry，andonceagreatnumberofliver-coloredflowers，whichelsewherewereyellowEventhecolorofthebirdschangedwiththedistrictwepassedthrough；butneverbeforedidIseesuchamarkedchangeasfromyellowtoblue，repeatedagainandagainonthesameplainAnotherbeautifulplantattractedmyattentionsostronglyontheseplainsthatIdismountedtoexamineit。
TomygreatdelightIfoundittobeanoldhomeacquaintance，aspeciesofDrosera，closelyresemblingourownsundew`DroseraAnglia’。
Theflower-stalkneverattainsaheightofmorethantwoorthreeinches，andtheleavesarecoveredwithreddishhairs，eachofwhichhasadropofclammyfluidatitstip，makingthewholeappearasifspangledoverwithsmalldiamondsInoticeditfirstinthemorning，andimaginedtheappearancewascausedbythesunshiningondropsofdew；
but，asitcontinuedtomaintainitsbrilliancyduringtheheatoftheday，Iproceededtoinvestigatethecauseofitsbeauty，andfoundthatthepointsofthehairsexudedpureliquid，in，apparently，capsulesofclear，glutinousmatterTheywerethuslikedewdropspreservedfromevaporationTheclammyfluidisintendedtoentrapinsects，which，dyingontheleaf，probablyyieldnutrimenttotheplant。
DuringourseconddayonthisextensiveplainIsufferedfrommytwenty-seventhattackoffever，atapartwherenosurface-waterwastobefoundWeneverthoughtitnecessarytocarrywaterwithusinthisregion；andnow，whenIwasquiteunabletomoveon，mymensoonfoundwatertoallaymyburningthirstbydiggingwithsticksafewfeetbeneaththesurfaceWehadthusanopportunityofobservingthestateoftheseremarkableplainsatdifferentseasonsoftheyear。
Nextdaywepursuedourway，andonthe8thofJunewefordedtheLotembwatotheN。WofDilolo，andregainedourformerpath。
TheLotembwahereisaboutamilewide，aboutthreefeetdeep，andfullofthelotus，papyrus，arum，mat-rushes，andotheraquaticplants。
Ididnotobservethecourseinwhichthewaterflowedwhilecrossing；
but，havingnoticedbeforethattheLotembwaontheothersideoftheLakeDiloloflowedinasoutherlydirection，IsupposedthatthiswassimplyaprolongationofthesameriverbeyondDilolo，andthatitroseinthislargemarsh，whichwehadnotseeninourprogresstotheN。WButwhenwecametotheSouthernLotembwa，wewereinformedbyShakatwalathattheriverwehadcrossedflowedinanoppositedirectionnotintoDilolo，butintotheKasai。
Thisphenomenonofariverrunninginoppositedirectionsstruckevenhismindasstrange；and，thoughIdidnotobservethecurrent，simplyfromtakingitforgrantedthatitwastowardthelake，Ihavenodoubtthathisassertion，corroboratedasitwasbyothers，iscorrect，andthattheDiloloisactuallythewatershedbetweentheriversystemsthatflowtotheeastandwest。
Iwouldhavereturnedinordertoexaminemorecarefullythismostinterestingpoint，but，havinghadmylowerextremitieschilledincrossingtheNorthernLotembwa，Iwasseizedwithvomitingofblood，and，besides，sawnoreasontodoubtthenativetestimony。
ThedistancebetweenDiloloandthevalleysleadingtothatoftheKasaiisnotmorethanfifteenmiles，andtheplainsbetweenareperfectlylevel；
and，hadIreturned，IshouldonlyhavefoundthatthislittlelakeDilolo，bygivingaportiontotheKasaiandanothertotheZambesi，distributesitswaterstotheAtlanticandIndianOceansIstatethefactexactlyasitopenedtomyownmind，foritwasonlynowthatIapprehendedthetrueformoftheriversystemsandcontinentIhadseenthevariousriversofthiscountryonthewesternsideflowingfromthesubtendingridgesintothecentre，andhadreceivedinformationfromnativesandArabsthatmostoftheriversontheeasternsideofthesamegreatregiontookasomewhatsimilarcoursefromanelevatedridgethere，andthatallunitedintwomaindrains，theoneflowingtothenorthandtheothertothesouth，andthatthenortherndrainfounditswayoutbytheCongotothewest，andthesouthernbytheZambesitotheeast。
Iwasthusonthewatershed，orhighestpointofthesetwogreatsystems，butstillnotmorethan4000feetabovethelevelofthesea，and1000feetlowerthanthetopofthewesternridgewehadalreadycrossed；
yet，insteadofloftysnow-cladmountainsappearingtoverifytheconjecturesofthespeculative，wehadextensiveplains，overwhichonemaytravelamonthwithoutseeinganythinghigherthananant-hilloratreeIwasnotthenawarethatanyoneelsehaddiscoveredtheelevatedtroughformofthecentreofAfrica。
Ihadobservedthattheoldschistoserocksonthesidesdippedintowardthecentreofthecountry，andtheirstrikenearlycorrespondedwiththemajoraxisofthecontinent；andalsothatwherethelatereruptedtraprockshadbeenspreadoutintabularmassesoverthecentralplateau，theyhadborneangularfragmentsoftheolderrocksintheirsubstance；butthepartialgeneralizationwhichtheobservationsledtowas，thatgreatvolcanicactionhadtakenplaceinancienttimes，somewhatinthesamewayitdoesnow，atdistancesofnotmorethanthreehundredmilesfromthesea，andthatthisigneousaction，extendingalongbothsidesofthecontinent，hadtiltedupthelateralrocksinthemannertheyarenowseentolieThegreaterenergyandmoreextendedrangeofigneousactioninthoseveryremoteperiodswhenAfricawasformed，embracingalltheflanks，impartedtoititspresentverysimpleliteraloutlineThiswasthelengthtowhichIhadcome。
Thetraprocks，whichnowconstitutethe“fillingup“ofthegreatvalley，werealwaysapuzzletometillfavoredwithSirRoderickMurchison’sexplanationoftheoriginalformofthecontinent，forthenIcouldseeclearlywhythesetraprocks，whichstilllieinaperfectlyhorizontalpositiononextensiveareas，heldintheirsubstanceangularfragments，containingalgaeoftheoldschists，whichformthebottomoftheoriginallacustrinebasin：thetraps，inburstingthrough，hadbrokenthemoffandpreservedthem。
Thereare，besides，rangesofhillsinthecentralparts，composedofclayandsandstoneschists，withtheripplemarkdistinct，inwhichnofossilsappear；butastheyareusuallytiltedawayfromthemassesofhorizontaltrap，itisprobablethattheytoowereaportionoftheoriginalbottom，andfossilsmayyetbefoundinthem。*
*AfterdwellinguponthegeologicalstructureoftheCapeColonyasdevelopedbyMrABain，andtheexistenceinveryremoteperiodsoflacustrineconditionsinthecentralpartofSouthAfrica，asprovedbyfresh-waterandterrestrialfossils，SirRoderickMurchisonthuswrites：
“SuchasSouthAfricaisnow，suchhavebeenhermainfeaturesduringcountlesspastagesanteriortothecreationofthehumanrace；
fortheoldrockswhichformherouterfringeunquestionablycircledroundaninteriormarshyorlacustrinecountry，inwhichtheDicynodonflourished，atatimewhennotasingleanimalwassimilartoanylivingthingwhichnowinhabitsthesurfaceofourglobeThepresentcentralandmeridianzoneofwaters，whetherlakesormarshes，extendingfromLakeTchadtoLake’Ngami，withhippopotamiontheirbanks，arethereforebutthegreatmodernresidualgeographicalphenomenaofthoseofamesozoicageThedifferences，however，betweenthegeologicalpastofAfricaandherpresentstateareenormous。
Sincethatprimevaltime，thelandshavebeenmuchelevatedabovethesea-leveleruptiverockspiercinginpartsthroughthem；
deeprentsanddefileshavebeensuddenlyformedinthesubtendingridgesthroughwhichsomeriversescapeoutward。
“Travelerswilleventuallyascertainwhetherthebasin-shapedstructure，whichishereannouncedashavingbeenthegreatfeatureofthemostancient，asitisoftheactualgeographyofSouthAfricai。e。，fromprimevaltimestothepresentday，does，ordoesnot，extendintoNorthernAfricaLookingatthatmuchbroaderportionofthecontinent，wehavesomereasontosurmisethatthehighermountainsalsoform，inageneralsense，itsflanksonly。”President’sAddress，RoyalGeographicalSociety，1852，pcxxiii。
ThecharacteristicsoftherainyseasoninthiswonderfullyhumidregionmayaccountinsomemeasurefortheperiodicalfloodsoftheZambesi，andperhapstheNileTherainsseemtofollowthecourseofthesun，fortheyfallinOctoberandNovember，whenthesunpassesoverthiszoneonhiswaysouthOnreachingthetropicofCapricorninDecember，itisdry；
andDecemberandJanuaryarethemonthsinwhichinjuriousdroughtsaremostdreadednearthattropicfromKolobengtoLinyanti。
AshereturnsagaintothenorthinFebruary，March，andApril，wehavethegreatrainsoftheyear；andtheplains，whichinOctoberandNovemberwerewellmoistened，andimbibedrainlikesponges，nowbecomesupersaturated，andpourforththosefloodsofclearwaterwhichinundatethebanksoftheZambesiSomewhatthesamephenomenonprobablycausestheperiodicalinundationsoftheNileThetworiversriseinthesameregion；butthereisadifferenceintheperiodofflood，possiblyfromtheirbeingonoppositesidesoftheequator。
ThewatersoftheNilearesaidtobecometurbidinJune；
andthefloodattainsitsgreatestheightinAugust，ortheperiodwhenwemaysupposethesupersaturationtooccurThesubjectisworthytheinvestigationofthosewhomayexaminetheregionbetweentheequatorand10DegS。；fortheNiledoesnotshowmuchincreasewhenthesunisatitsfarthestpointnorth，ortropicofCancer，butatthetimeofitsreturningtotheequator，exactlyasintheothercasewhenheisonCapricorn，andtheZambesiisaffected。*
*Theaboveisfrommyownobservation，togetherwithinformationderivedfromthePortugueseintheinteriorofAngola；andImayaddthattheresultofmanyyears’observationbyMessrsGabrielandBrandatLoanda，onthewestcoast，isinaccordancetherewith。
Itrainstherebetweenthe1stand30thofNovember，butJanuaryandDecemberareusuallybothwarmanddry。
Theheavierrainscommenceaboutthe1stofFebruary，andlastuntilthe15thofMayThennorainfallsbetweenthe20thofMayandthe1stofNovemberTherainaveragesfrom12to15inchesperannumIn1852itwas12。034inches；
in1853，15。473inchesAlthoughIhadnomeansofmeasuringtheamountofrainwhichfellinLonda，Ifeelcertainthattheannualquantityexceedsverymuchthatwhichfallsonthecoast，becauseforalongtimewenoticedthateverydawnwasmarkedbyadelugingshower，whichbeganwithoutwarning-dropsorthunder。
Iobservedthattherainceasedsuddenlyonthe28thofApril，andthelesserrainscommencedaboutafortnightbeforethebeginningofNovember。
FrominformationderivedfromArabsofZanzibar，whomImetatNalieleinthemiddleofthecountry，theregiontotheeastofthepartsofLondaoverwhichwehavetraveledresemblestheminitsconformation。
Theyreportswampysteppes，someofwhichhavenotrees，wheretheinhabitantsusegrass，andstalksofnativecorn，forfuel。
Alargeshallowlakeisalsopointedoutinthatdirection，namedTanganyenka，whichrequiresthreedaysforcrossingincanoes。
ItisconnectedwithanothernamedKalagweGarague？，farthernorth，andmaybetheNyanjaoftheMaravimFromthislakeisderived，bynumeroussmallstreams，theRiverLoapula，theeasternbranchoftheZambesi，which，comingfromtheN。E。，flowspastthetownofCazembe。
ThesouthernendofthislakeistendaysnortheastofthetownofCazembe；
andasthatisprobablymorethanfivedaysfromShinte，wecannothavebeennearertoitthan150milesProbablythislakeisthewatershedbetweentheZambesiandtheNile，asLakeDiloloisthatbetweentheLeebaandKasaiBut，howeverthismaybe，thephenomenaoftherainyseasonshowthatitisnotnecessarytoassumetheexistenceofhighsnowymountainsuntilwegetreliableinformation。
This，itistobehoped，willbeoneoftheresultsoftheresearchesofCaptainBurtoninhispresentjourney。
TheoriginalvalleyformationofthecontinentdeterminedthenorthernandsoutherncourseoftheZambesiinthecentre，andalsooftheancientriverwhichonceflowedfromtheLinyantibasintotheOrangeRiverItalsogavedirectiontothesouthernandnorthernflowoftheKasaiandtheNileWefindthatbetweenthelatitudes，say6Degand12DegS。，fromwhich，inallprobability，theheadwatersofthoseriversdiverge，thereisasortofelevatedpartitioninthegreatlongitudinalvalleyPresumingonthecorrectnessofthenativeinformation，whichplacesthehumidregiontowhichtheNileandZambesiprobablyowetheiroriginwithinthelatitudesindicated，whydoessomuchmorerainfalltherethaninthesamelatitudesnorthoftheequator？WhydoesDarfurnotgiverisetogreatrivers，likeLondaandthecountryeastofit？TheprevailingwindsintheoceanoppositetheterritorypointedoutaresaidtobefromtheN。EandS。E。