134 
PACIFIC SCIENCE, Vol. XX, January 1966 
The subsidence at Kapoho in 1924 was lim- 
ited to the fault trough, a zone about 0.4 mile 
wide. Kalapana also lies in a graben, between 
two faults, and it appears probable that the sink- 
ing in 1868 was restricted largely, if not wholly, 
to the graben. However, other areas, not within 
grabens, also have sunk. Just south of Cape 
Kumukahi, the eastern point of the island, pre- 
vious to I960 Hawaiian fishpond walls could be 
seen submerged 4-6 ft deep in the ocean. ( They 
were buried in part by the lava flow of I960.) 
Judging from similar walls elsewhere, they give 
evidence of a sinking of at least 5 ft. The area 
is directly on the active east rift zone of Kilauea 
volcano, and it might be supposed that the sink- 
ing was localized, due to a change in volume 
within the eruptive core of the rift zone, a 
change s’milar to that which produces the swell- 
ing and shrinking of the rift zone before and 
after eruptions (Macdonald and Eaton, 1964: 
101). However, evidence of sinking exists also 
at Keaau, 14.2 miles northwest of Cape Kumu- 
kahi (Fig. 7). There, according to Eldon S. 
English and other members of the Shipman 
family, long-time owners of the property, coco- 
nut stumps are submerged below sea level and 
are invisible even at the lowest tides. The bases 
of the stumps are buried in sand, and it is dif- 
ficult to determine just how deep the former 
ground level now lies below sea level; and also, 
of course, its original height above sea level is 
not known. Therefore, the extent of sinking is 
uncertain, but it must amount to several feet. 
The recency of sinking also is uncertain, but it 
appears unlikely that the coconut stumps could 
have survived more than a few centuries. The 
Fig. 9. Unnamed marine shelf projecting northward from the shore of Kapu-wai cove, Honaunau Bay, at 
low tide, June 8, 1964. Paper rolls mark some of the concavities made by ancient Hawaiians. (City of Refuge 
negative 1006.) 
