312 
K M 
FIG. 7. Section structure under northwest rift zone. 
resents the arrival times recorded at Barbers 
Point; the lower set, the arrival times recorded 
at Kualapuu. 
Because of the shortness of the shot line, 
interpretation had to rely on second and third 
arrivals. For these later arrivals correlation from 
one seismogram to the next was sought by look- 
ing for similarity in wave forms as well as lin- 
ear relations on the travel-time curve. 
Interpretation began by assuming that the 
first layer below the water layer had a velocity 
of 3.0 km/sec. Then the 4.96 km/sec line of 
the Kualapuu recording was paired with the 
4.97 km/sec line of Barbers Point. The 6.9 
km/sec line was paired with the 6.65 km/sec 
line. The 8.88 km/sec line was considered to be 
arrivals from the Mohorovicic discontinuity; no 
comparable line on the Barbers Point recording 
could be found to match this. The 17.16 km/sec 
line is probably a reflection from one of the 
layers. 
Inasmuch as the layers with velocities of 4.97 
km/sec and 6.8 km/sec showed no dip or very 
little dip, it was assumed that at least under line 
BRAVO the Mohorovicic discontinuity was hori- 
zontal. This assumption is justified even from 
gravity data, which showed that the BRAVO 
series had run along the strike of the flank of 
the volcano. This assumption results in a veloc- 
ity of 8.8 km/sec for the mantle. 
Calculations from the travel-time data yielded 
the following: 
PACIFIC SCIENCE, Vol. XIX, July 1965 
Layer 
Velocity 
(km/sec) 
Km to Top of Layer 
Dip 
Water 
1.5 
0 
0° 
a 
3.0 
.45 (average) 
b 
4.97 
2.6 
0° 
c 
6.8 
10 
1.1' 
d 
8.8 
21 
0° 
The section structure deduced from the travel- 
time curves from the BRAVO series is given in 
Figure 9. 
THE STRUCTURE OF KOOLAU VOLCANO 
FROM COMPOSITE DATA 
When the section structures of the caldera, 
northwest rift zone, and southern flank were 
fitted together, the over-all structure of Koolau 
Volcano shown in Figure 10 was obtained. 
Three layers were detected in the northwest 
rift zone. The unusual feature of a high veloc- 
ity layer, 7.6-77 km/sec, at very shallow depths, 
should be noted. 
An explanation of this phenomenon is that 
differentiation in the magma chamber under the 
rift zone separated the magma into a 57-6.1 
km/sec layer and a 7.6-77 km/sec layer. Fur- 
ther investigation with longer refraction tra- 
verses should be carried out over the rift zone 
to determine whether there is another layer 
with higher velocity under the 7.6-77 km/sec 
layer. 
From interpretations of the reflected phase 
of the GASHOUSE series, Adams and Furumoto 
? I 
(p. 296 in this issue) proposed a magma 
chamber at a depth between 3 and 4 km. This 
agrees well with the depth to the 7.6-77 
km/sec layer which, when extended to the cal- 
W KM E 
Fig. 8. Travel time plots of BRAVO series. 
