ON CAVES. 81 
rocky shore, or filled with heaped-up pebbles, if it opened on to 
a shingle beach. By its form and by its contents we could 
generally make a shrewd guess whether it was a sea-cave or 
not. We should ask whether the parts where the cave 
expands are those on which the sea would act with greatest 
force and efficiency, or whether the shape could be better 
explained by reference to torrents coming in the other 
way. We should examine the contents to see whether in 
their character or arrangement they indicated the action of 
the in-rushing water, or whether they are such as could never 
have survived the scour of tidal and wind-driven waves. 
When we have to inquire into the origin of caves in inland 
cliffs and on mountain-sides, now far above the sea, where 
many of the traces above-described may have been long 
removed by denudation, there are further tests to be applied. 
There we should have regard to their manner of occurrence 
and their place in the physical geography of the neighbour- 
hood. A sea-cave does not necessarily, or even commonly 
occur in the line of drainage from the uplands, but in the 
higher cliffs and headlands between the valleys that run down 
to the sea. Whereas the caves due to subterranean water- 
courses lie in the lines of drainage; and the caves due to 
sub-aérial waste coincide in distribution with the outcrop of 
the beds that readily lend themselves to that kind of 
weathering. 
Moreover, allowing for the possibility of unequal elevation 
of different parts of a coast-line, we can still generally find 
sufficient evidence to show whether the rock in which the 
cave occurs forms part of an old sea-cliff or of an escarpment.* 
We must remember also that during the formation of a 
sea-cave the base of the cliff is being swept by the sea. 
Sometimes an inland stream washes the base of a rock in 
which a watercourse cave has its outfall, but generally in the 
case of inland-formed caves a vast mass of talus is being 
formed along the base of the cliff in which the cave occurs. 
The scour of floods may keep the mouth open, but as the 
water is being drained off to other and lower levels, this sweep- 
ing of the cave mouth ceases, and the cave deposits show 
interbedded fallen rock and transported earth and stones, and 
often the remains of animals. 
As a general statement we may say that a typical sea-cave 
runs into a cliff which rises vertically from the level of the 
* Cf. Whitaker, Q.J.G.S. vol. xxiii. c. 186, p. 265. Geol. Mag. vol. iVe 
1867, pp. 457, 443, 
