Icannotconceiveanyexistingreptilebeingconvertedintoamammal。FromhomologiesIshouldlookatitascertainthatallmammalshaddescendedfromsomesingleprogenitor。Whatitsnaturewas，itisimpossibletospeculate。Morelike，probably，theOrnithorhynchusorEchidnathananyknownform;astheseanimalscombinereptiliancharacters（andinalessdegreebirdcharacter）withmammalian。Wemustimaginesomeformasintermediate，asisLepidosirennow，betweenreptilesandfish，betweenmammalsandbirdsontheonehand（fortheyretainlongerthesameembryologicalcharacter）andreptilesontheotherhand。Withrespecttoamammalnotbeingdevelopedonanyisland，besideswantoftimeforsoprodigiousadevelopment，theremusthavearrivedontheislandthenecessaryandpeculiarprogenitor，havingacharacterliketheembryoofamammal;andnotanALREADYDEVELOPEDreptile，birdorfish。
Wemightgivetoabirdthehabitsofamammal，butinheritancewouldretainalmostforeternitysomeofthebird—likestructure，andpreventanewcreaturerankingasatruemammal。
Ihaveoftenspeculatedonantiquityofislands，butnotwithyourprecision，oratallunderthepointofviewofNaturalSelectionNOT
havingdonewhatmighthavebeenanticipated。TheargumentoflittoralMioceneshellsattheCanaryIslandsisnewtome。Iwasdeeplyimpressed（fromtheamountofthedenudation）[withthe]antiquityofSt。Helena，anditsageagreeswiththepeculiarityoftheflora。WithrespecttobatsatNewZealand（N。B。TherearetwoorthreeEuropeanbatsinMadeira，andI
thinkintheCanaryIslands）nothavinggivenrisetoagroupofnon—volantbats，itis，nowyouputthecase，surprising;moreespeciallyasthegenusofbatsinNewZealandisverypeculiar，andthereforehasprobablybeenlongintroduced，andtheynowspeakofCretaceanfossilsthere。Butthefirstnecessarystephastobeshown，namely，ofabattakingtofeedontheground，oranyhow，andanywhere，exceptintheair。IamboundtoconfessIdoknowonesinglesuchfact，viz。ofanIndianspecieskillingfrogs。Observe，thatinmywretchedPolarBearcase，Idoshowthefirststepbywhichconversionintoawhale\"wouldbeeasy，\"\"wouldoffernodifficulty\"！！Sowithseals，Iknowofnofactshowinganytheleastincipientvariationofsealsfeedingontheshore。Moreover，sealswandermuch;Isearchedinvain，andcouldnotfindONEcaseofanyspeciesofsealconfinedtoanyislands。Andhencewandererswouldbeapttocrosswithindividualsundergoinganychangeonanisland，asinthecaseoflandbirdsofMadeiraandBermuda。Thesameremarkapplieseventobats，astheyfrequentlycometoBermudafromthemainland，thoughabout600milesdistant。WithrespecttotheAmblyrhynchusoftheGalapagos，onemayinferasprobable，frommarinehabitsbeingsorarewithSaurians，andfromtheterrestrialspeciesbeingconfinedtoafewcentralislets，thatitsprogenitorfirstarrivedattheGalapagos;fromwhatcountryitisimpossibletosay，asitsaffinityIbelieveisnotverycleartoanyknownspecies。Theoffspringoftheterrestrialspecieswasprobablyrenderedmarine。NowinthiscaseIdonotpretendIcanshowvariationinhabits;