A Chronology of the Explosive Eruptions of Kilauea 
Howard A. Powers 1 
INTRODUCTION 
The explosive eruptions of Kilauea, both 
prehistoric and historic, have been much dis¬ 
cussed in the literature on Hawaiian volcanoes. 
Some early general observations attributed all 
of the surface ash deposits to the known explo¬ 
sive eruption of 1790 (Dana, 1891: 42-45; 
Jaggar, 1921: 114-118); others recognized evi¬ 
dence of several different prehistoric explosions 
(Hitchcock, 1911: 166-169; Sidney Powers, 
1916). Later studies, with better exposures in 
artificial cuts on the windward rim of the crater, 
demonstrated a number of long intervals of 
quiet between several eruptions (Finch, 1925; 
Stone, 1926). This paper adds another, still 
more detailed, chapter to the accumulated 
knowledge, but much remains to be learned and 
recorded before complete understanding of the 
explosive phases of the volcano is attained. 
Many of the thoughts expressed herein are the 
result of discussion and exchange of ideas in 
the field with many coworkers, whose contri¬ 
butions are gratefully accepted and acknowl¬ 
edged. T. A. Jaggar, R. H. Finch, E. G. Win¬ 
gate, A. E. Jones, all formerly with the Hawaiian 
Volcano Observatory; J. E. Doerr, Jr., former 
naturalist, Hawaii National Park; H. S. Palmer 
of the University of Hawaii; C. K. Wentworth 
of the Honolulu Board of Water Supply; and 
G. A. Macdonald of the United States Geologi¬ 
cal Survey have all contributed to the accumu¬ 
lation and consideration of material which is 
presented in this paper. 
The erratic original deposition of the pyro- 
clastics of Kilauea, controlled by combinations 
of atmospheric elements and directed explo¬ 
1 Naturalist, Hawaii National Park; since July 1, 
Seismologist, Hawaiian Volcano Observatory, U. S. 
Geological Survey. Published by permission of the 
Director, National Park Service, Department of the 
Interior. Manuscript received December 23, 1947. 
sions, has been amply discussed by Wentworth 
(1938: chapter 3). Very few individual beds 
of material could be traced with assurance as 
horizon markers, even for the short distance 
around the circumference of the crater, because 
of this erratic deposition. Also, removal of 
material by wind and water erosion during and 
between eruptions has been tremendous. As a 
net result of these two factors, there is no cross 
section in the area which contains a complete 
representation of all the explosion deposits. 
Successful analysis of the stratigraphy and 
chronology of these deposits thus depends as 
much on appraising an erosion surface in the 
section as upon correctly describing the beds 
of pyroclastic materials. In this study, great 
emphasis has been placed on recognition of 
characteristics which can be used to distinguish 
unconformities and erosion surfaces produced 
during time intervals of perhaps a few hours 
between explosions from those produced dur¬ 
ing time intervals measured in years or tens of 
years between eruptions. The following charac¬ 
teristics of the present desert surface have been 
used to identify similar long-exposed surfaces 
in cross sections of the deposits: surface oxida¬ 
tion, encrustation, desert varnish on fragments, 
concentration of coarser particles, thinly lam¬ 
inated dust deposits, and truncation of pre¬ 
viously consolidated layers (Plate IB and 1C). 
Humus and other remains of plant growth are 
self-evident where they exist. 
On each of eight segments of the crater rim, 
a complete composite section of the pyroclastic 
deposits was built up by compilation from a 
number of cross sections which were sufficiently 
close together so that correlations between them 
could be made with confidence. Particular at¬ 
tention was paid to recording depositional 
[ 278 } 
