— 21 — 



plenty of materials. Moreover the tidal undulations produce 

 the same effects on those coasts respectivel}^, the warm super- 

 undulation descendiug on the western sides of the oceans, and 

 the sub-undulations rising cold from the ocean depths on their 

 eastern sides. 



XVIII. j>The xi'sual direction of the island groups in the central parts 

 »of the Pacific i'i N. W. and S. E. This must be noticed, because sub- 

 »tcrranean disturbaiices are known to follow the coast lines of the land. 

 » Commencing on the shores of America, there are abundant proofs that 

 »the greater part has been elevated within the recent peviad, but as co- 

 »ral reefs do not occur there, ie is uot immediitely connected with our 

 » present subject. Immediately adjoining that continent there 'n an exfent 

 sof ocean remarkably free from islands and where of conrse, there exists 

 »iio postible indication of change of level. We then come to a N. W. 

 »b. W. liae, dividiag the open sea from one strewed with lagoon islands 

 »imd incluoing the twQ beautiful groups of enciicled island.«, the Society and 

 » Georgian Archipelagoes, whereia no marks of elcvation exist. This great 

 »band, having a length of more than four thoiisand miles by six hunderd 

 »broad, must according to our view be an area of subsidence. We will 

 » at present pass over the spaca of ocean immediately adjoining it, and 

 » procsed to the chaiu of islands including the New HebriJes, Salomon 

 »and New Ireland; any one who examines the charts of the separate 

 »islan(ls in the Pacific, wiil be struck wich the absence of all di-itant or 

 »enciicling reefs ronnd these groups, yet it is known that coral occurs 

 »abnndantly close in shore. Here then, acording to the theory, there 

 »are no proofs of siibbidence, and in conforuiity to this, we find in the 

 »works of PoRSTER, Lbsson, Labillardière, QüOT.and Benket, con- 

 »stant alluj'ion to the masscs of elevated coral. These islands form 

 »therefore a well determined band of clevaiion; between it and the great 

 » area of subsi lence there is a hroad space of sea, wLich we passed over 

 »iQ gding from the foimer to the latter, irregnlarly scattered with iilets 

 » of all classes, some with proofs of recent elevation and raerely fringed 

 » by reefs, others enciicled , and some lagoon islands. One of the latter 

 sis dcsciibed by captaio Cook as a grand circle of breakers without a 

 »8ingle spot of land. In this C;i83 we maybeüeve that an ordiiiary lagoon 

 j>island has been recently submerged. On the otherhand, there are proofs 

 » of other lagoon islands having been lifted up, several yards above the 

 »lcvel of the sea, but which still rcthin a pool of salt water in their 

 »C3ntrcs. These fac;s show an iiregular ac,i)n in the snbterranean for- 

 »ces, and when we remeniber that the space lies directly between the 

 »well marked area of elevalion and the enormous one of silsilence, al- 

 • ternate and irrcgular movement secms ahnest probable." 



