y 
OBSERVATIONS ON SHELL-HEAPS AND SHELL-BEDS. 309 
tion of sand from tributary streams, none of which carry sand, 
but are asa rule, muddy. The quantities of sand brought in by 
the tide must be enormous, the action being in operation twice in 
the twenty-four hours, and it is evident there is ordinarily a large 
balance in favour of the flood-tide as compared with the ebb, in 
its silt-carrying capability. That the great bulk of the tidal 
water enters by an undercurrent would seem to be indicated by 
the well known phenomenon of the swelling of the tide before the 
ebb ceases to run out; even during a freshet this action continues, 
and in the case of the Clarence, the ebb has been observed to 
run out for three hours after the tide begins to swell. The scour- 
ing effect of the flood-tide would thus be intensified by the weight 
of the superincumbent outgoing current, the sand brought in being 
deposited at slack water, and the ebb (being mainly an upper 
current) not to any extent carrying it back again ; consequently, 
much of the sand brought in by the flood-tide would remain, 
gradually creeping upwards by repetition of this action until per- 
manent shoals were formed. 
Cook’s River is an instance near at hand which exemplifies this 
theory, being a muddy stream, the lower portion of which past 
Tempe becomes sandy. The flats above the Cook’s River dam, 
which are below the level of spring-tides, show numerous beds of 
shell, the occurrence of which no doubt suggested the idea of 
oyster-culture, which the late Mr. Thomas Holt so signally failed 
in, not realizing that conditions had so essentially altered that 
any such attempt was utterly hopeless. 
On the flats below Tempe, shell-beds have been worked for many 
years for the production of lime: these beds are in places eighteen 
‘inches in depth, resting on clay bottom, and covered by about a 
foot of sandy silt. These flats are now only covered by spring- 
tides, the diminished rise of tides being due to the obstruction of 
the tidal flow by the formation of sand shoals in the lower reaches 
of the river; this sand it is evident must be brought in by the 
tide, as the contribution of sand by the river itself is insignificant, 
the flood deposit being clay mud, These shell-beds afford evidence 
