Eruptions of Kilauea—POWERS 
291 
ing of at least 2 6 separate explosive eruptions 
of Kilauea. Each eruption appears to have been 
caused either by phreatic or by magmatic ex¬ 
plosions; no single eruption seems to require 
both phreatic and magmatic explosions to ac¬ 
count for the lithologic constituents in its de¬ 
posit. 
The Uwekahuna formation seems to have 
been deposited on the floor and outer slopes 
of an earlier caldera by at least five magmatic 
eruptions separated by appreciable time inter¬ 
vals. 
The Keanakakoi formation seems to represent 
at least 10 phreatic and 11 magmatic eruptions 
which have occurred since the last major over¬ 
flows of lava which form the present crater rim. 
Details of distribution of the Keanakakoi 
deposits promise to be useful in working out 
the details of the late structural history of 
Kilauea. For example, the slumped rim blocks 
in the northeast corner carry the thickest deposit 
from eruption number 2-K, suggesting early 
collapse of these blocks; and most of the de¬ 
posits on Byron’s Ledge are not appreciably 
thicker than those on the adjacent southeast 
rim, which is 100 feet higher. Perhaps this 
indicates a late date for the collapse of Byron’s 
Ledge. 
The study of the early Keanakakoi magmatic 
explosion deposits has a direct bearing on the 
problem of the contribution of the Kilauea 
summit crater to the older Pahala ash formation 
of Hilina Pali and other areas 10 miles or more 
distant from the summit crater. 
The chronological table of explosive erup¬ 
tions may conceivably be useful as an actual 
time scale when coupled with results from 
studies of rate of reforestation. The most mature 
forest east and north of Kilauea Iki apparently 
was never killed by any explosive eruption dur¬ 
ing the Keanakakoi time and thus may repre¬ 
sent uninterrupted encroachment from the slope 
of Mauna Loa as much as 5 miles distant across 
the surface of the last rim overflow of Kilauea. 
A less mature forest extending from the crater 
rim to about a mile northeast of the rim may 
represent the total development of vegetation 
since the seventeenth eruption, which probably 
killed all vegetation to that distance. The growth 
immediately on the northeast rim may represent 
the recovery from a partial kill of vegetation 
in 1790. 
The general picture sketched in by this study 
suggests these long-period fluctuations in the 
intensity of lava pressure: 
(1) pre-Uwekahuna, high-pressure, dome¬ 
building phase 
(2) low-pressure collapse and explosive 
eruptions of Uwekahuna tuff 
(3) high-pressure, post-Uwekahuna resump¬ 
tion of dome-building 
Plate 4A. Keanakakoi pyroclastics on the outer slope of the southwest Kilauea rim. Deposits of the fif¬ 
teenth, seventeenth, eighteenth, and 1790 eruptions mantle an erosion surface which truncates beds of the 
fourteenth to the tenth eruptions. The erosion surface marked by the hammer truncates the vitric ash of erup¬ 
tions earlier than the tenth. Details of the section in upper left center show in Plate 4B, and details of the 
upper beds at the head of the re-entrant behind the figure show in Plate 4C. 
PLATE 4B. Remnants of two layers of 1790 gravels, separated by a depositional break, lie on a desert sur¬ 
face truncating the stony pisolite of the eighteenth eruption. Erosion surfaces have been emphasized by re¬ 
touching on the negative. Beds from the seventeenth eruption reach 8 inches in thickness at the right of the 
section and lie on the eroded surface of beds from the fifteenth eruption, the lowest layer which is continuous 
across the face of the section. A lens of pumice from the thirteenth eruption (marked by pocket knife at left) 
lies between remnant patches of beds from the fourteenth and twelfth eruptions. The vitric deposits from the 
third and fourth eruptions lie beneath the horizontal desert surface (not retouched), which extends across 
the photograph just below the grass clump on the right. 
PLATE 4C. Gravel of 1790, from 4 to 6 inches thick, lying on the eroded layer of stony pisolitic mud 
from the eighteenth eruption, originally 3 inches thick in this section. Erosion occurred after consolidation of 
the pisolitic mud, as it truncates the bedding within the layer, and has cut the bed to a remnant 1 inch thick 
in the photograph. Beneath the stony pisolite layer is a desert surface eroded on deposits from the seven¬ 
teenth eruption about 10 inches thick including the basal layer of coarsest fragments. 
Photographs by H. A. Powers, September, 1947, at locality U. 
