18 
ine calcareous organisms, these rocks have been consolidated super- 
ficially into dense travertine (“cap-stone”) in many places by alter- 
nate solution and precipitation of carbonate of lime. As a result of 
this mode of origin concretionary structure is very wide spread. 
Travertine formation has been very irregular, or else solution chan- 
nels have been frequent, or both factors in development have been 
operative. As a result extraordinary “nigger-heads” and “teeth” of 
limestone have been formed or left amongst 1 Lie less solid sand. 
These are well exhibited hi the railway cuttings near the Show Ground 
and in the river bank near the boat-sheds at Peppermint Grove. 
Plant roots have formed channels for percolating water and 
may, perhaps, have contributed organic solvents during life or dur- 
ing decay, so that the segregation of calcium carbonate has been 
directed, in its first stages at all events, by root distribution. The 
structure lines so initiated have been extended and enlarged, produc- 
ing the arborescent rods of limestone which form so striking a fea- 
ture of the railway cuttings between Oottesloe Beach and North Fre- 
mantle. 
False bedding has been extensively developed during the forma- 
tion of those rocks and gives rise to very striking features in the 
topography produced by them (for example, near the Mount Lyell 
Chemical Works at Rocky Bay). 
In places [e.g, 7 North Fremantle) the superficial crust of lime- 
stone is sufficiently pure to be burnt for lime, while at various places 
in the Metropolitan area the subjacent calcareous sandstones have 
supplied rather inferior building stone and road metal. 
(k.) Next in succession we come to the actual “Shore Line” of the 
Indian Ocean. This is composed mostly of sandy beaches backed 
by actively moving sand dunes, which, in places, are encroaching 
very seriously upon residential and industrial areas. Extensive rocky 
promontories are scarcely existent, but small headlands of coastal 
limestone alternate with the sand beaches. A well-defined wave-cut 
platform of limestone is recognisable in places at a level such that it 
is extensively exposed at low water. The significance of this platform 
is being discussed in another paper by Mr. J. L. Somerville. 
(/.) In the immediate neighbourhood of Fremantle the shallow 
waters of “Gage Roads” enclosed between the shore and the Iiottnest 
Island to Rockingham Bay reefs constitutes another physiographic 
element of no mean importance. 
(m.) The zone of islands and reefs extending through Carnac 
arid Garden Island and, perhaps, Rottnest Island, forms the most 
westerly of the physiographic elements included within the scope 
of the present paper. The author has not had an opportunity of 
examining them personally. 
It is obvious, of course, that a method of subdivision like that 
attempted above is to some extent a matter of convenience. The 
various elements shade off into one another in many instances so 
