376 
NATURE 
[Frsruary 18, 1897 
which is formed by a dead coral reef, constituted almost wholly 
of two species, one a massive Porites, and the other Heliopora 
cerulea, For a great part of the day this floor lies bare and dry, 
the frayed ends of the Heliopora standing like broken reeds, 6 
inches above its surface, and the great clumps of Porites forming 
a series of stepping-stones of equal height. Neither of these 
corals stands long exposure to the air; on Funafuti they require 
constant submergence, and we are thus led to regard their upper 
surface as marking what was at one time the level of low tide in 
the swamp ; but since the present level of low tide is below the 
level thus indicated, some change must have occurred in the 
level of low tides. Not necessarily an elevation of the reef : 
Darwin has admirably discussed this explanation, and it is quite 
conceivable that some change in local conditions, such as the 
exclusion of the sea by the growth of the hurricane beach, may 
ABC. Mangrove swamp. 
BE OF ; 
(Rxtx] e//opora reer 
Consolidated cora/ 
VAY \¥orming part of the floor 
of the swamp. 
Sand. 
5,0 Clinker Freld of 
P0"9)} cora/ Fragments. 
Fic. 4.—Corner of Funafuti, showing Mangrove Swamp and Heliopora Reef. 
have produced a local alteration in the height of the tides. The 
swamp communicates with the sea by pits in its floor, which 
enter subterranean channels running seawards. These passages 
are so narrow that the tide rises and falls in the swamp much 
more slowly than in the open sea. To determine whether any 
change of level has taken place, it thus becomes necessary to 
compare the highest and lowest water level of the swamp with 
that of the sea or of the lagoon. I accordingly levelled across 
the island from the lagoon to the sea, crossing the swamp on the 
way, and found that the high-water level at spring tides is 1 foot 
10 inches below high-water (spring tides) of the lagoon, so that 
given free access of the sea, the Heliopora reef would be covered 
I foot 10 inches deeper than at present ; but it is now submerged 
from 10 inches to 2 feet 2 inches at high water springs, and 
NO. 1425, VOL. 55] 
would accordingly be submerged from 2 feet 8 inches to q feet, 
with free access of the sea. The range of spring tides is at least 
6 feet, as I learn from Lieut. Dawson, but I am not quite 
sure that an extreme range of g feet 8 inches has not been ob- 
served. Taking, however, the smaller number, it becomes 
clear that for a considerable part of the day, the reef 
would be exposed to the 
air. It is not likely 
that under these con- 
ditions the corals would 
continue to live, and, 
I think, therefore, that 
the reef must have un- 
dergone some slight ele- 
vation, to the amount, 
ISLET FORMED CHIEFLY OF SANQ 
Fic. 5 
SECTION THROUGH THE /SLET OF PAVA, FUNAFUT!, 
Ocal, Horizontal :- ¢ fathoms to O-4 inch. Vertical :-6 feet to Or inch, 
PLATFORM OF HARO CORAL BRECC/A 
CLIFF 
7OCC/A 
TP COMingsS 
OUTER EOGE 
OF REEF 
perhaps, of 4 feet. This 
conclusion is in accord- 
ance with Dana’s view, 
and is supported by ob- 
servations on some other 
features of the island, 
such, for example, as the 
occurrence of an inter- 
rupted line of low cliffs, 
sometimes passing into a series of pinnacles, generally about 4 
feet in height, as measured from low-water level. In the annexed 
section (Fig. 5) the cliffs are further from the land than is 
usually the case. These cliffs consist of a consolidated breccia 
of coral fragments, and are now in process of denudation, as is 
LEVEL OF 
LOW TIDE 
