— ie- 

 ls it possible tliat a professing geoloijist, is ignorant of 

 tlie important diü'ercnccs in circuinstanccs and rcsulting in- 

 ferences, bctwccn the causcs of underinining and overtlow- 

 ing ! lu the formcr the certain iuference is that tlie Icvel of 

 the land so undermined is higher , higher perhaps by thousands 

 of feet, than that of the sea, whilst an overflowing action or 

 process denotes, that the level of the sea is, at the time of 

 producing that effect, higher than that of the land. That Mr. 

 Darwin had seen the key to the phenomenon of //old cocoa- 

 //uut trees being undermined and Mling" but, was unable to 

 rccognise it , appears from his having observed as folio ws. // It 

 // seems from some old charts (1) that the long island to wind- 

 // vvard (the S. E", isle) was formerly separated by wide chan- 

 //nels into several islets, and this fact is likewise indicated by 

 //the Icss age of the trees in certain positions" (3) page 548. 

 We have said that this observation is a key to the under- 

 mining process, along the lagoon shores. Thus whilst these 



(1) Had Mr. Darwin exaniined those upfilled cliannels Avith sufficiënt 

 iittcntion , he would have conclnded, that they were filled long before 

 Europeans passcd eastward by Cape of Good Hope , and secing that these 

 places are alraost exclusively occnpied ■\vith the low crooked groMÏng „spe- 

 cies of teak" (Pitch teak) and the few coconut trees interspcrsed not ri- 

 sii'.g to half tlie hcight of those on the ori^zinal islets, that the vessel on 

 board whicli the chart was drawn passcd the islets at snch a distance, that 

 these low trees were not visiblé, and consequently the spaccs were f)rc- 

 sumcd to be openiiigs bctwcen the islets. Mr. Darwin probably rcctivcd 

 dircctly, or through captain Eitzuoy his knowledge of the chart iii qncs- 

 lion from Mr. Lkisk, who had seen in niy hnnd, whilst talking ujjoii the 

 subject with Mr. J. B. Guay a copy, which I had made from the origiiial. 

 That oiiginal had no date, bnt indicaiions are observable of the time being 

 early iu the last contury. Indeed even in the present time, a vessel 

 passing at not loss thuu ten miles distanco, would observo only the ori- 

 ginal islets and as sucli separated 0'oni cacli other. (J. C. R.). 



(2) Not „Icss age," but stuntcd growth ; there as elsewherc a propor- 

 lioa is dying of old age. These channels I)cing gencrally filled up wilh 

 coarsc debris , the coconut tree (which requircs to have fine sand about 

 jl's roots) does not thcrcin grow at all vigorously, nor producc more 

 thau u vcry few vcry small uuts ia t/hc course of the ycar. 



