384 
PACIFIC SCIENCE, Vol. XIX, July 1965 
waii if a density of 2.0 g/cc were assumed for 
the material above sea level, rather than a den- 
sity of 2.3 g/cc. As can be seen, the effect of 
reducing the density is to raise the anomaly 
values and to give a positive correlation with 
surface elevation. A density of 2.3 g/cc, there- 
fore, appears to be more correct, as it yields 
anomalies which show no correlation with local 
changes in surface elevation. 
The densities of the materials in the sub- 
surface cannot be sampled directly but must 
be established indirectly. One of the best 
sources of indirect evidence as to the densities 
present is seismic information. In addition to 
the seismic work carried out on the Hawaiian 
Ridge by the Hawaii Institute of Geophysics 
and reported by Furumoto et al. (p. 306 in 
this issue) and Adams and Furumoto (p. 296 
in this issue), explosion seismic refraction meas- 
urements have been made in the waters adja- 
cent to the Hawaiian Islands by Raitt (1956), 
Gaskell and Swallow (1953), Shor (I960), 
Shor and Pollard (1964), and Western Geo- 
physical Company (unpublished). In addition, 
Jones (1935) and Eaton (1962) have used 
observatory earthquake information to study 
the velocity structure on the Hawaiian Ridge. 
Although details vary, the seismic results pre- 
sent an amazingly consistent picture of the 
over-all velocity structure of the Hawaiian 
Ridge and the area around it. 
Over the shallow waters on the Ridge and 
on the islands away from volcanic pipes and 
rift zones, the velocity structure consists of 2-3 
km of material with a velocity lying between 
2.9 and 4.0 km/sec, 6-8 km of material with 
a velocity between 4.5 and 5.2 km/sec, and 
4-7 km of material with a velocity between 
6.4 and 7.2 km/sec, with the depth to Moho 
ranging from 14 to 16 km. The only significant 
deviation from this picture is seen in the re- 
sults obtained by Furumoto et al. (see p. 306 
in this issue), which show a much thicker 
lower crustal layer and a Moho depth of 20+ 
km. However, this depth does not appear to be 
representative of the Ridge as a whole, which 
appears to be 14-16 km. This does not imply 
that the 20-km depth south of Oahu is incor- 
rect, for, as is seen in the gravity map of Figure 
1, a thicker crust there is substantiated by a 
pronounced gravity minimum over the area 
that cannot be related to bathymetry or to a 
thick section of sediments. 
A drastically different picture is found from 
seismic work carried out over the volcanic plugs 
or along major rift zones. Here velocities in 
excess of 7.0 km/sec— usually in the range of 
7. 5-8.0 km/sec— are found at depths from 2 to 
7 km below sea level. This material is sometimes 
overlain by material whose velocity is about 
6.0 km/sec. Examples of this can be seen in 
the papers by Furumoto et al. and Adams and 
Furumoto in this issue. These results are simi- 
lar to those obtained elsewhere in the Hawaiian 
area by other investigators. 
In the case of Hawaiian volcanic rocks, spe- 
cial care must be taken in converting velocity 
information to density information since the 
relations which exist are considerably different 
from those commonly encountered in conti- 
nental-type igneous rocks. These differences re- 
sult primarily from three factors: (1) the ex- 
ceptionally high grain density of the Hawaiian 
rocks, ( 2 ) the very large porosities which exist, 
and ( 3 ) the presence of glass. 
Manghnani and Woollard (p. 291 in this 
issue) summarize laboratory velocity measure- 
ments of a number of Hawaiian rocks at sur- 
face temperatures and pressures. They find that 
the velocity of the basalts is controlled by the 
amount of glass and olivine present as well as 
by the physical structure of the rock. When 
using these laboratory measurements to relate 
seismic velocities observed in the field densities, 
it is important to correct for changes in envi- 
ronmental conditions. In addition to the effect 
of the difference in ambient temperature and 
pressure, there is also the effect of interstitial 
water pressure on the very porous lavas. The 
effect of interstitial water pressure, however, is 
closely related to permeability, and these work- 
ers have found a definite correlation between 
seismic velocity in flow material and apparent 
porosity which appears to be related to differ- 
ences in permeability. 
In clastic sediments, the hydrostatic pressure 
of the water in the pore spaces has been shown 
by various investigators to result in a decrease 
in velocity with increasing pore water pressure. 
This effect in sediments apparently results from 
