276 
sumably could also represent magmatic differ- 
entiates of an original tholeiitic magma. Alter- 
nate explanations are: (a) there are regional 
variations in the chemical composition of the 
mantle, (b) the composition of basalt is a 
function of partial melting of the mantle, or 
(c) of the depth at which the magma is gener- 
ated. In connection with this hypothesis, the 
laboratory investigations of Yoder and Tilley 
(1962) suggest that the composition of the 
magma can be a function of pressure and tem- 
perature, and that alkalic basalt represents a 
deeper, higher pressure environment than does 
tholeiitic basalt. 
To investigate this problem the writer and 
Gordon A. Macdonald started a program of 
integrated geophysical and geological studies in 
the Hawaiian Islands and Samoa in 1963 under 
the sponsorship of the National Science Foun- 
dation. Although this,, study has not been com- 
pleted, preliminary results which will be pre- 
sented in this report suggest that magmatic dif- 
ferentiation best explains the relations noted in 
Hawaii and probably Samoa also. On other 
islands, such as the Line Islands group, gravity 
data suggest that the primary magma extruded 
was alkalic, but there is no clue as to the depth 
of origin. The only actual data bearing on this 
point are the relations noted in Japan by Kuno 
(1959), who found that the alkalic basalts are 
associated with zones of deep-focus earthquakes 
and the tholeiitic basalts with shallow-focus 
earthquakes. If there is a relation between earth- 
quake focal depths and magma generation, and 
there does appear to be such a relation in Ha- 
waii with magma supplying present active vents 
originating from about 60 km, Kuno’s observa- 
tions and the Hawaiian data both appear to 
support the findings of Yoder and Tilley. With 
regard to partial melting Engel and Engel 
( 1964), in an excellent review study of oceanic 
basalts, note that this implies a mantle that is 
even more deficient in radiogenic elements than 
is tholeiitic basalt. These investigators, on the 
basis of their comprehensive study of basalts in 
all oceanic areas, conclude that deep-lying oce- 
anic basalts in general are tholeiitic; that they 
are all remarkably similar in composition; and 
that, since their volume is about a thousand 
times greater than that of alkalic basalts, the 
PACIFIC SCIENCE, Vol. XIX, July 1965 
alkalic basalts are probably magmatic differen- 
tiates of tholeiitic magma. 
Thus, the geologic evidence for the composi- 
tion of the mantle from studies of basalts is not ij 
conclusive, and even may be said to be contra- 
dictory. The writers approach to the problem 
has been through geophysical studies. As tho- 
leiitic basalt has a higher density than alkalic 
basalt, there should be a real difference in the 
gravity field associated with primary volcanic 
pipes supplying the different lavas. On the j 
assumption that these pipes were filled with 
essentially undifferentiated mantle material in 
the early stages of intrusion, when the bulk of 
the lavas was being rapidly extruded and the 
bulk of the volcanic pile was being developed, 
as noted recently off Iceland, and that these 
pipes were sealed off at the top soon after the 
external pressure was sufficienteto- develop in- 
termittent volcanism with subsequent eruption 
via flank rifts developed by doming and frac- |j 
turing of the volcanic pile or through secondary 
vents, the primary pipes should contain a sub- 
stantial sample of the primary magma with lit- 
tle or no differentiation at shallow depth. Cer- 
tainly the field relations and geophysical meas- 
urements on Oahu suggest such a history. On 
the island of Oahu the two primary pipes whose 
caldera He on opposite sides of the island at or 
dose to sea level are marked by local gravity 
anomalies of about +110 mgal (+310 mgal 
absolute Bouguer). Later vents, such as Koko 
Head and Diamond Head, have no appreciable 
gravity effect, although the bulk density of the 
surface flow material is only 2.3 gm/cc. The 
rift flows forming the Koolau Mountains, rep- 
resenting an Intermediate stage in the vol- 
canic history of the island, are marked by inter- 
mediate gravity anomaly values of about +50 
mgal Somewhat similar conditions are noted 
on the island of Hawaii, where there is still 
active volcanism. All of the primary pipes, such 
as Mauna Kea, Mauna Loa, Kohala, Hualalai, 
and the present active pipe at Kilauea, are 
marked by local gravity highs of about +100 
mgal. All stages of eruptive history are repre- 
sented. Mauna Kea is dormant; Mauna Loa, 
the principal source of lava in the recent past, 
erupts only periodically and mostly through 
flank fissures; and Kilauea represents an active, 
