EARTHQUAKES—BULLEN 323 
tical direction. In contrast, the velocities below the crust depend 
very largely on the depth alone, and the variation with depth is 
much more steady. 
An early triumph in seismology was the revelation that the earth 
has a large core sharply separated from the part outside, called the 
mantle. The existence of a central core had been suspected for the 
last century, but was not established until Oldham supplied the nec- 
essary evidence from seismology in 1906. In 1913, Gutenberg (then 
in Gottingen, now of California) made a famous calculation which 
gave the depth of the core boundary as 1,800 miles, a figure that Jef- 
freys with the use of statistical theory showed in 1939 to be accu- 
rate within 3 miles or so. 
The discovery of the existence of the central core came about this 
way. Consider P waves issuing from a large hypothetical earth- 
quake with focus at the South Pole. Then it transpires that the 
waves would be strongly recorded at all stations in the Southern Hem- 
isphere, and, as well, in the Northern Hemisphere as far north as 
latitude 15°, i.e., up to the latitude of Guatemala. But between the 
latitudes of Guatemala (15° N.) and Winnipeg (52° N.), there would 
be a “shadow zone” in which the regular P waves would be much 
less prominent. Then at the latitude of Winnipeg, the P waves would 
come in strongly again and be well recorded from there to the North 
Pole. The whole of the United States would thus be inside the 
shadow zone for the particular earthquake. Similar shadow zones 
occur with all earthquakes large enough to be recorded on the opposite 
side of the earth, the location of the shadow zone depending, of 
course, on the location of the focus. 
On examination it was seen that the only possible explanation of 
the shadow is the presence of a central core. The regular P rays 
which emerge at epicentral distances up to 105° lie entirely in the 
mantle. Rays which emerge at 105° just graze the boundary of the 
core. Slightly steeper P rays strike the boundary at an angle and 
the shadow is produced through these rays being bent or “refracted” 
sharply downward at the boundary, with the result that they do not 
emerge at the surface until at distances beyond 142°. The phenom- 
enon is similar to the refraction of light rays at the boundary which 
separates a water surface from the air and which causes a straight 
stick partly immersed in the water to be apparently bent at the 
surface. 
NoS waves have ever been detected in the earth’s central core. This 
is part of the evidence that shows the central core, for most of its 
volume, to be in a fluid or molten state. The complete evidence for 
this conclusion includes measurements of tidal deformation of the 
solid earth, and astronomical observations that enable movements of 
the earth’s poles to be determined. These measurements, together 
