49 GEOGNOSY. 

Headlands project from the land because for the most part they consist 
of rock which has been better able to withstand the shock of the 
breakers. Bays and creeks, on the other hand, have been cut by the 
waves out of less durable materials. Again, by the sinking of land, 
ranges of hills have become capes and headlands, while the valleys have 
passed into the condition of bays, inlets, or fjords. By the uprise of the 
sea-bottom, tracts of low alluvial ground have been added to the land. 
Hence speculations as to the history of the elevation of the 
land, based merely upon inferences from the form of coast-lines as 
expressed upon ordinary maps, are apt to be of little value. To be 
of real service, they demand a careful scrutiny of the actual coast- 
lines, and an amount of geological investigation which would require 
long and patient toil for its accomplishment. . 
Passing from the mere external form of the land to the com- 
position and structure of its materials, we may begin by considering 
the general density of the entire globe, computed from observations 
and compared with that of the outer and accessible portion of the 
planet. Reference has already been made to the comparative density 
of the earth among the other members of the solar system. In 
inquiries regarding the history of our globe, the density of the whole 
mass of the planet as compared with water—the standard to which 
the specific gravities of terrestrial bodies are referred—is a question 
of prime importance. Various methods have been employed for 
determining the earth’s density. The deflection of the plumb-line on 
either side of a mountain of known structure and density, the time of 
oscillation of the pendulum at great heights, at the sea-level, and in 
deep mines, the comparative force of gravitation as measured by the 
torsion balance, have each been tried with the following various results: 
Plumb-line experiments on Schiehallien (Maskelyne and Playfair) 
gave as the mean density of the earth . 4°713 
Do. on Arthur’s Seat, Edinburgh (James) . h 5 ‘ . orale 
Pendulum experiments on Mont Cenis (Carlini and Giulio) \ » 4°950 
Do. in Harton coal-pit, Newcastle (Airy) . ‘ : . 6°565 
Torsion balance experiments (Cavendish, 1798). : ‘ ; . 5°480 
Do. do. (Reich, 1838) : ; Z : . 9°49 
Do. do. (Baily, 1843) ; ; ‘ F . 5:°660 
Do. do. (Cornu and Baille, 1872, 3) . : . 9°50-5°56 
Though these observations are somewhat discrepant, we may feel 
satisfied that the globe has a mean density neither much more nor much - 
less than 5:5; that is to say, it is five and a half times heavier than 
one of the same dimensions formed of pure water. Now the average 
density of the materials which compose the accessible portions of the 
earth is between 2°5 and 3; so that the mean density of the whole 
globe is about twice as much as that of its outer part. We might 
therefore infer that the inside consists of much heavier materials. 
than the outside, and consequently that the mass of the planet must 
contain at least two dissimilar portions—an exterior lighter crust or 
vind, and an interior heavier nucleus, But the effect of pressure 
