530 W. M. DAVIS SUBSIDENCE OF REEF-ENCIECLED ISLANDS 



3 miles long and 15 or 20 fathoms deep, lies a mile farther on the west. 



Gardiner is the only student of coral reefs who has described Rotnma 

 in detail. His accounts permit one to suppose that the eccentric position 

 of the volcanic island with respect to the bank is due to the recent unsym- 

 metrical addition of new cones to an older island that was more centrally 

 situated. It may indeed be by reason of the "broad lava streams'^ which 

 "can be traced to the sea" from several craters, the appearance of which 

 led to the belief that "they have not been long inactive" (1898, a, 438), 

 that "the coast is fairly even w^ith a complete absence of the 'long points 

 and deep fiord-like bays' which, according to Dana, would on a volcanic 

 island give indubitable evidence of subsidence" (499). Some of the re- 

 cent ash cones are clift to a height of 700 or 900 feet, this being a natural 

 result of the absence of reefs around their young shoreline ; but they now 

 have narrow fringing reefs; one of these, a few yards wide, drops off ab- 

 ruptly into 20 fathoms of water (440). As to the fringing reefs, it may 

 be agreed that they have "been formed entirely at the present level" by 

 outgrowth; but the submarine bank and its 15-fathom rim, together with 

 the plunging cliffs, strongly suggest that rapid submergence has taken 

 place here as well as in the area of the isolated banks. 



Inasmuch as, with the exception of Tutuila, described above, there is 

 no other known example of a submerged barrier reef so good as Rotuma 

 in the vast extent of the open Pacific, its occurrence in close association 

 with the chief group of Pacific submarine banks has an almost demon- 

 strative value for the subsidence theory; for, as Darwin said, "If an old 

 barrier reef were destroyed and submerged, and new reefs became attached 

 to the land, these would necessarily at first belong to the fringing class." 

 It is interesting to note that the young naturalist went on to say that, 

 fringing reefs being all colored red on his maps as indicating stationary 

 or rising coasts, examples of exceptional fringing reefs, formed as above 

 suggested, would be given the same color, "although the coast was sink- 

 ing" ; but he added, "I have no reason to believe that from this source of 

 error any coast has been wrongly colored with respect to movement indi- 

 cated" (124). With reference to Rotuma, it is later said in explanation 

 of the colors on the chart of reef distribution : "From the chart in Du- 

 perrey's atlas, I thought this island was encircled [that is, by a barrier 

 reef], and had colored it blue; but the Chev. Dillon assures me that the 

 reef is only a shore or fringing one, red" (162). One may imagine the 

 pleasure that Darwin would have felt on learning more fully the facts 

 about this curious island and the numerous "drowned atolls" with which 

 it is associated, and thus recognizing that he might restore the blue color 

 originally assigned to Rotuma. One may, indeed, here quote what Geikie 



