82 J. D. Dana—Oceanie Coral Island Subsidence. 
even to a distance of two thousand miles from Hawaii, or as: 
far as from New York to Salt Lake City; and how much 
farther is unknown, as the line of coral islands here passes the 
boundary of the coral reef seas, or the region where coral 
records are possible. 2 
ther ranges of submerged summits are shown to extend 
through the whole central Pacific, even where not a rocky peak 
remains above the surface; for all the coral islands from the 
eastern Paumotus to Wakes Island, near long. 170° E. and lat. 
19° N., north of the Ralick and Radack (or Marshall) groups, 
are in linear ranges: and they have, along with the equally 
linear ranges of high islands just south, a nearly uniform trend, 
curving into northwest and north-northwest at the western ex- 
tremity. The coral islands consequently cap the summits of 
linear ranges of elevations, and all these linear ranges together 
constitute a grand chain of heights, the whole over five 
thousand miles in length. Thus, the coral islands are records 
of the earth’s submarine orography, as well as of slow changes 
of level in the ocean’s bottom. 
This coral island subsidence is an example of one of the 
great secular movements of the earth’s crust. The axis of the 
subsiding area* has a length of more than six thousand miles 
scoquel to one-quarter of the circumference of the globe; and 
the breadth, reckoning only from the Sandwich Islands to the 
Friendly Group (or to Tongatabu) is over twenty-five hundred 
miles, thus equalling the width of the North American con- 
tinent. A movement of such extent, involving so large a part 
of the earth’s crust, could not have been a local change of level, 
but one in which the whole sphere was concerned as a unit ; 
for all parts, whether participating or not, must have in some 
way been in sympathy. with it. 
his subsidence was in progress, in all probability, during 
the Glacial era, the thickness of the reefs proving that in their 
origin they run back through a very long age, if not also into 
the Tertiary. It was a downward movement for the tropical 
Pacific, and perhaps for the warmer latitudes of all the oceanic 
areas, while the more northern continental lands, or at least 
those of North America, were making their upward movement, 
preparatory to or during that era of ice. 
he subsidence connected with the origin of coral islands 
and barrier reefs in the Pacific has been shownt+ to have 
amounted to several thousands of feet, perhaps full ten thou- 
sand. And, it may be here repeated that, although this sounds 
large, the change of level is not greater than the elevation whi 
* The position of this area is stated on page 328 of the volume on Corals and 
Coral Islands. 
~ + Ibid, p. 329. 
