94 The American Geologist. Feb. i89o 
reached its present amount formerly filled the space now 
occupied by water. 
The forces of undermining and falling in of overhanging 
cliffs can now be seen in operation, and the sea carries of!' the 
<*«»tritus as it did when Harrington sound was first formed. 
One can see to-day in the power of active erosion abundant 
cause for the excavation of Harrington sound without calling 
to its help a subsidence of the platform on which the islands 
rest. It may be said that the narrow entrance is not large enough 
to serve as a channel of exit, but we may conclude that the 
amount of water which enters Harrington sound is but a part 
of that which daily flows into that basin. There are subter- 
ranean water courses, notably those near Paynter's vale, 
through which water passes in great quantities. These are 
possibly agents of erosion as well as of transportation of mat- 
ter which has been eroded. It is said that the entrance to 
Harrington sound at "the Flats" is slowly filling, but it is not 
impossible that the amount of detritus conveyed through this 
channel is greatly in excess of that left in the passage. 
The "cave theory" of the formation of the present outlines 
of the Bermudas may be stated as follows : 
The modification of the coast line of the islands and the for- 
mation of the sounds is due to erosion by rain or by the 
sea. The sea is the most active agent, undermining the cliffs, 
forming caves or grottos, the roofs of which fall in from time 
to time, while the debris is in part carried off in solution or 
mechanical suspension by tides or currents. 
In support of this theor}'' there may be instanced observa- 
tion of the erosion indicated at the bases of the high abrupt 
cliffs and the caves found everywhere in the existing islands. 
The depth of Harrington sound is adduced as an evidence 
of subsidence. It is asked if the sea could excavate the floor at a 
depth of fifty feet below its present level. I am not sure that it can 
not or that instances of eroded caves in harder rock than coral 
limestone of an equal depth might be adduced. But even this 
depth does not disprove erosion or prove subsidence. I am 
not unwilling to believe that there were depressions in the 
great Bermuda which were never filled by rock, or sand before 
it hardened into rock. The nucleus of Harrington sound 
when erosion began may have been one of those. Indeed my 
