CORAL REEFS AND SUBMARINE BANKS 391 
Macclesfield bank, which is instanced as typical of its class, and 
inquire particularly as to the validity of the assumptions that are 
necessary in order to explain it by the glacial-control processes. 
The assumptions appear to be about as follows: 
If a volcanic island about the size of Hawaii stood still long 
enough in preglacial time, it would be reduced to low relief and - 
would be surrounded by shoals of its own detritus mingled with 
reef deposits. If any defending reefs were killed by the chilled 
and lowered glacial ocean, the low island would be attacked by the 
waves; and, if the attack endured long enough, the island, still 
stationary, would be reduced to a smooth platform having the form 
of a very flat cone. Such a platform would have about the area 
of the Macclesfield bank. ‘Then as the sea finally warmed and rose 
again the bank would be more or less rimmed with a reef and 
aggraded with wave-swept sediments. This succession of supposi- 
tions is, like many another, easily conceivable; but it is so full of 
improbabilities that it cannot command acceptance, as the following 
considerations will show. 
The instability of the Australasian Archipelago.—The improba- 
bility that wave attack was effective in truncating preglacial islands 
has already been sufficiently set forth; some uncertainty as to the 
lowering of the glacial ocean by so much as 35 or 40 fathoms will be 
pointed out below in the section on extra-tropical submarine 
banks; we may here inquire whether the region of the Macclesfield 
bank can be fairly regarded as one of long-enduring stability. No 
inquiry based on geological evidence regarding the region concerned 
appears to have been made on this point. Stability seems to have 
been assumed in order that an even platform might be abraded and 
a smooth bank formed by subsequent aggradation. If this assump- 
tion be temporarily set aside, the geological relations of the bank 
may be inquired into impartially. The general situation of the 
bank in a deep sea on the border of a large continent is not one that 
would be chosen, on general principles, as presumably stable. The 
prevailing opinion as to the instability of continental borders is 
summarized by Chamberlin and Salisbury, who suggest that 
mutual crowding and crumpling may take place, accompanied by 
fracture and slipping, where the edges of elevated continental 
