Anda di halaman 1dari 46

Composition

and Origin of
Hawaiian Lavas

GORDON A . MACDONALD
Hawaii Institute of Geophysics, University of Hawaii, Honolulu

ABSTRACT

New chemical analyses for major elements in 76 Hawaiian rocks are pre-
sented and bring the total number of such modern analyses to about 470.
Many determinations of minor elements also are becoming available. Hawaiian
petrology is discussed against this total background. The three major rock
suites, tholeiitic, alkalic, and nephelinic, are chemically intergradational. The
main mass of the volcanoes is tholeiitic, followed by a relatively small volume
(generally less than 1 percent) of alkalic lavas; the two types of lavas are
interbedded in a thin transitional zone. The nephelinic lavas are separated
from the others by a long time interval that is marked by a profound erosional
unconformity.
Variations within the rock suites are largely the result of crystal differen-
tiation. All three rock suites probably are derived from a single type of parent
magma, which varies slightly from one volcanic center to another, of olivine
tholeiite composition. Crystallization of this magma in shallow magma cham-
bers leads to eruptible magmas of tholeiitic composition. In the last stages of
volcanism, consolidation of the upper part of the magma body leads to crys-
tallization at deeper levels under higher pressure and to production of alkalic
magmas. Finally, crystallization at depths of several tens of kilometers pro-
duces nephelinic magmas that are erupted after a long period of volcanic
quiescence.

477

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
478 STUDIES IN VOLCANOLOGY

CONTENTS
Introduction -478
Hawaiian petrographic suites 479
Ultramafic inclusions 483
The chemical analyses 486
Discussion of chemical data 497
Origin and diversification of Hawaiian magmas 509
The parent magma 509
Origin of the alkalic and nephelinic suites 512
Diversification within suites 516
Conclusions 518
Acknowledgments ... 519
References cited .519
Figure
1. Alkali: silica diagram of Hawaiian rocks 481
2. Abundance of titania in relation to soda in Hawaiian rocks 504
3. Abundance of P 2 O s in relation to N a 2 0 in Hawaiian rocks 504
4. AFM diagram of all Hawaiian volcanic rocks 506
5. Von Wolff diagram of all Hawaiian volcanic rocks 507
6. Von Wolff diagram showing trend lines of the Hawaiian rock suites 508
7. Alkali: silica diagram showing variational trends of Hawaiian rock suites
and fractionation trends found in the laboratory 514
Table
1. Lavas of the upper member of the Waianae volcano, Oahu 488
2. Lavas of West Maui and Molokai 490
3. Lavas of Mauna Kea and Kohala 493
4. Lavas of Hualalai volcano 495
5. Lavas of the Honolulu volcanic series, Oahu, and the Kiekie volcanic
series, Niihau 496
6. Miscellaneous tholeiitic lavas 498
7. Rocks of the Koloa volcanic series, Kauai 500
8. Average compositions of Hawaiian lavas 502
9. Derivation of other alkalic rock types from alkalic olivine basalt 517

INTRODUCTION
It is fitting that this paper should appear in a volume dedicated to Prof.
Howel Williams. My interest in volcanoes and their products was first aroused
33 years ago in Prof. Williams' fascinating seminars on volcanology, and my
work on Hawaiian volcanoes and rocks over the past 29 years is a direct
outgrowth of that interest.
In the William Smith Lecture before the Geological Society of London,
in 1952, Prof. Williams pointed out that the science of volcanology is under-
going a change from an early descriptive stage to an interpretive stage (Wil-
liams, 1954). This is conspicuously true of Hawaiian volcanology. When the
systematic and continuous studies of the active volcanoes were begun by
T. A. Jaggar and his associates in 1911, a voluminous record of activity
already existed, but it was purely descriptive. Nothing was known of the

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 479

internal structure and mechanism of the volcanoes. Since then, the geophysi-
cal and geochemical observations of the Hawaiian Volcano Observatory staff,
the geologic study and mapping of the islands by H. T. Stearns and his asso-
ciates of the U.S. Geological Survey, the petrographic studies of H . A. Powers,
Horace Winchell, and others, and recently the geophysical work of the Vol-
cano Observatory and the Hawaii Institute of Geophysics, have brought us
to the stage where we are beginning to understand the internal working of
the volcanoes. Geophysical evidence is emerging on the depth of origin of
the magma and on its accumulation in a shallow reservoir, until it is finally
released at the surface (Eaton and Murata, 1960). Knowledge of the chem-
istry of the lavas has increased to the point where we can begin to use it
statistically, with confidence that serious gaps in it are very unlikely. Further-
more, as is absolutely essential to the study of petrogenesis, the petrography
can be closely related to the structure and history of the volcanoes. We still
have a long way to go to a complete understanding of Hawaiian volcanoes,
but the progress is gratifying!
The following pages present additional petrographic data that will con-
tribute to the ultimate understanding. There now exist more than 470 modern
analyses of major chemical elements in Hawaiian lavas. More than 350 of
these analyses have been made since 1940. An earlier study (Macdonald and
Katsura, 1964) was concentrated on the early-stage tholeiitic lavas. The
present work focuses instead on the late-stage lavas, though some additional
analyses of tholeiitic rocks also are given. The work has been financed by
grant number GP-1958, from the National Science Foundation to the Hawaii
Institute of Geophysics. In Tokyo, 98 new chemical analyses were made at
the Japan Analytical Chemistry Research Institute. For comparison, eight of
the same powdered samples were analyzed also in U.S. Geological Survey
Laboratory in Denver. The results are described elsewhere (Macdonald and
Powers, in prep.), but it may be said here that the two sets of analyses are
in reasonably good agreement. Analyses of late lavas of Haleakala will be
given elsewhere (Macdonald and Powers, in press), and are not included here.

HAWAIIAN PETROGRAPHIC SUITES

The nature of the suites of rocks that are recognized in the Hawaiian
Islands and their stratigraphic and structural relationships have been discussed
elsewhere (Macdonald and Katsura, 1964) and will be only briefly re-
viewed here. The additional work reported in the present paper has merely
strengthened the former conclusions.
The geologic mapping of the islands has led to the recognition of a suc-
cession of stages in Hawaiian volcanism (Stearns, 1940a; 1946, p. 1 7 - 2 2 ) .
Some uncertainty still exists about the subsea history, and deep drilling is
badly needed to explore the composition and structure of the lower portions
of the volcanoes. Above sea level, however, the stages are as follows:

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
480 STUDIES IN VOLCANOLOGY

( 1 ) A shield-building stage, in which frequent eruptions of very fluid lava


give rise to widespread thin lava flows. Explosive activity is very minor and
pyroclastic rocks form less than 1 percent of the whole. The flows build a
broadly rounded shield volcano. Toward the end of this stage, the summit of
the shield usually, though perhaps not always, collapses to form a caldera.
The lack of explosion in this stage results f r o m both the low viscosity of the
magma and a low gas content, probably generally less than 1 percent by weight
(Macdonald, 1963b).
( 2 ) A caldera-filling stage, in which flows of fluid lava accumulate within
the caldera and gradually fill it. Because they are confined within the depres-
sion, many of these flows become thick and massive; and because their thick-
ness causes them to cool slowly, the gas bubbles tend to escape from the
main part of the flow and make the flows dense. It should be emphasized,
however, that during this period, eruptions continue on the outer slopes of
the shield and form thin vesicular flows that are indistinguishable f r o m the
earlier precaldera flows. Toward the end of the period, at least on some of
the volcanoes, eruptions become less frequent and somewhat more explosive.
( 3 ) A postcaldera stage, during which a relatively thin cap is built over
the top of the shield and hides the caldera. The reduction in tempo of vol-
canism, already detectable at the end of the previous stage, becomes more
marked. The lava flows are shorter and thicker than in the earlier stages, and
in extreme cases pile u p over the vent to form volcanic domes. Many of the
eruptions are moderately explosive, partly because of greater viscosity of the
magma, but also apparently because of increased gas content.
This stage constitutes the end of what can be considered the principal
period of Hawaiian volcanism. It is followed by a long interval of volcanic
quiet, during which waves cut sea cliffs hundreds, or even thousands, of feet
high, and streams cut canyons thousands of feet deep into the volcanoes.
( 4 ) A posterosional stage, during which renewed volcanism pours flows
over the deeply eroded surface. The flows are fluid and far-spreading, but
are commonly thick and dense because of ponding in the valleys. Eruptions
commonly are moderately explosive because of a moderately high gas con-
tent in the magma.
The petrographic suites can be closely correlated with the stages of vol-
canism. In 1935, Powers recognized a suite of "primitive" lavas, followed by
a suite of "differentiated" lavas; and in 1950, Tilley termed these the tholeiitic
and alkalic suites. The names have become firmly established. I quite agree
with Chayes (1966) as to the inappropriateness of the term "tholeiitic" for
these rocks, but it has become too generally used to b e easily uprooted, and
since it is well understood in its present context, it will continue to be used
in this paper.
Tholeiitic rocks have been defined as those in which a reaction relation-
ship exists between Ca-poor pyroxene and magnesian olivine, whereas the
reaction relationship was presumed to be absent in the alkalic rocks (Tilley,
1950). However, the reaction relationship is often difficult to recognize, and

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 481

moreover the existence of an olivine reaction relationship in alkalic rocks


also has been demonstrated (Schairer and Yoder, 1960). The relative merits
of mineralogical and chemical methods of distinguishing between the tholeiitic
and alkalic suites have been discussed elsewhere (Macdonald and Katsura,
1964). Although mineralogical criteria are very useful where chemical data
are not available, the best criterion appears to me to be the ratio of silica to
alkalies in the rock, as determined by chemical analysis and plotted on an
alkali:silica diagram such as Figure 1 of the present paper. Most rocks can
unhesitatingly be assigned to one suite or the other. However, it should again
be emphasized that there is a complete chemical gradation between the two
suites.
The rocks of the tholeiitic suite in Hawaii include tholeiite, olivine tholeiite,
picrite-basalt of oceanite type, and one small body of rhyodacite (Macdonald
and Katsura, 1964), as well as small segregation veinlets and patches of iron-
rich basalt, pegmatoid, and granophyre (Kuno and others, 1957). Those of
the alkalic suite include alkalic olivine basalt, hawaiite, mugearite, soda

% S¡0 2

Figure 1. Alkali ¡silica diagram of Hawaiian rocks. Solid circles, tholeiitic rocks; open
circles, rocks of the alkalic suite; crosses, posterosional rocks of the nephelinic suite;
G, granophyre associated with the Palolo quartz diabase; open triangles, low-K tholeiites
of the deep Pacific basin (Engel and others, 1965); K, weighted average of first phase of
1959 eruption of Kilauea (Murata and Richter, 1966); P, pyrolite (Green and Ringwood,
1963); E, "eclogite" inclusions from Salt Lake tuff (Yoder and Tilley, 1962, p. 482;
Macdonald and Katsura, 1964, p. 123). The pairs of points connected by lines represent
analyses of the same rock powder by two different laboratories.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
482 STUDIES IN VOLCANOLOGY

trachyte, and picrite-basalt of ankaramite type, as well as rocks transitional


from mugearite to trachyte that have been termed benmoreite by Tilley, Yoder,
and Schairer (1965). Oceanites belonging to the alkalic suite have not been
found in the Hawaiian Islands, though they are fairly common in some other
regions, such as Samoa. The nephelinic suite includes nephelinite, melilite
nephelinite, basanite, alkalic olivine basalt, and small amounts of ankaratrite
and picrite-basalt of mimosite type. The rock names are used herein as de-
fined in previous papers (Macdonald, 1949a, p. 1544; 1960, p. 172-175;
Macdonald and Katsura, 1964, p. 84-86).
The rocks of the shield-building stage of Hawaiian volcanism are wholly
tholeiitic. Most contain considerably more K , 0 than the deep ocean basalts
described by Engel, Engel, and Havens (1965), though such low-K rocks
are by no means unknown among subaerial Hawaiian lavas. They are par-
ticularly abundant in Kohala volcano, where all of the 14 analyzed sam-
ples of the shield-building lavas (Pololu volcanic series) contain less than
0.3 percent K 2 0, and 11 of them contain less than 0.2 percent (Macdonald
and Katsura, 1964, p. 117). It would be of much interest to determine whether
low-K rocks increase in abundance at increasing depths below sea level within
the volcanoes, that is, whether there is a general increase in the abundance
of K (and related elements) throughout the life of the volcano. Submarine
basalts from the lower flank of Kilauea that are described by Moore (1965,
p. 44) all resemble the tholeiites exposed above sea level, but they represent
flows recently erupted onto the present carapace of the volcano and indicate
nothing about the nature of lavas that were erupted in early stages of growth
of the volcano.
Alkalic rocks appear during the caldera-filling stage. In a thin zone that
appears to be only a few flows thick, rocks of definitely alkalic and tholeiitic
character are interbedded, and some of the associated rocks are chemically
gradational in type. In the Waianae volcano of Oahu the interbedding is
actually of greater extent than formerly was apparent, because one of the
previously reported chemical analyses (specimen C-47, Macdonald and Kat-
sura, 1964, p. 116, 126) actually is an alkalic basalt, not tholeiitic basalt,
as stated in the locality description. The postcaldera stage consists almost
wholly of alkalic rocks, though recent work indicates that there may be an
occasional recurrence of tholeiitic lava into rather late stages of the volcanic
history (see section on Chemical Analyses).
The posterosional rocks of Kauai, Niihau, Oahu, and West Maui belong
to the nephelinic suite. A small posterosional flow that rests on alluvium in
Kolekole Pass, in the Waianae Range of Oahu (Stearns, 1940b, p. 47; Mac-
donald, 1940, p. 88) is too much weathered for chemical analysis, but to
judge from the microscopic sections it is an alkalic olivine basalt. Thin-section
examination and partial chemical analysis of the lava of the small Kalaupapa
shield volcano, which was built against the base of the great erosional sea
cliff of East Molokai (Stearns and Macdonald, 1947, p. 25, 109), indicate
that it also is alkalic basalt. Many of the alkalic olivine basalts of the neph-

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 483

elinic suite are markedly undersaturated with silica and contain several percent
of nepheline in the norm, although none is detectable in the mode. These
rocks may be referred to as basanitoids, or as linosaites (Winchell, 1947,
p. 28).
The caps of alkalic rocks constitute only a very small proportion of the
total volume of the volcanoes. Estimated proportions for the different vol-
canoes range from 0.1 to 3 percent, and average about 1 percent (Macdonald,
1963a). The proportion of rocks of the nephelinic suite is even smaller and
ranges from a mere trace, or none at all, to a fraction of 1 percent.

ULTRAMAFIC INCLUSIONS

Nodules of peridotite and related rocks are common in lavas of the alkalic
and nephelinic suites in Hawaii, but, as elsewhere in the world, they are rare
in the tholeiitic lavas (Forbes and Kuno, 1965). A detailed study of the
inclusions and their host rocks has recently been completed by R. W. White,
of the University of California, and another study by E. D. Jackson, of the
U.S. Geological Survey, is in progress. White (1966) finds a close correlation
between the types of inclusions and the types of host rocks that indicates a
genetic relationship between the two. For the most part, they cannot be
chance xenoliths that were picked up by unrelated rising magma. For details
on the texture and mineralogy of the inclusions the reader is referred to the
paper by White (1966) and forthcoming reports by Jackson. Only a few
features, of importance in relation to the discussion of genesis of the magma
types, are summarized here.
The inclusions in tholeiitic rocks are dunite, which commonly contains a
little feldspar. They have been found in both historic and prehistoric lavas
of both Mauna Loa and Kilauea, but they are conspicuously unusual. Chrome
diopside has been reported by Ross, Foster, and Myers (1954) in inclusions
that were collected by me from the 1950 lava flow of Mauna Loa, but subse-
quent search by both White and me has failed to discover any clinopyroxene-
bearing inclusions in that flow. The textures of the dunites suggest that they
are formed by the accumulation of olivine crystals that were sinking from
overlying magma. There is no evidence of disequilibrium between the inclu-
sions and the surrounding magma. The dunites are commonly associated with
much more widespread inclusions of fine-grained olivine gabbro that are
characterized by a very open texture that resembles the diktytaxitic texture
of some basaltic lava flows. Similar texture is found in small gabbro intrusions
that were emplaced at high levels in the older, dissected Hawaiian volcanoes
(Macdonald, 1942, p. 328-331). It suggests that the gabbro inclusions, and
presumably also the associated dunite inclusions, were derived from tholeiitic
masses that crystallized at shallow depths within the volcano.
The inclusions in the rocks of the alkalic suite are found by White (1966,
p. 261) to be mostly dunite, wehrlite, and gabbro. The same types had pre-

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
484 STUDIES IN VOLCANOLOGY

viously been recognized in the 1801 flow of Hualalai volcano by Cross (1915,
p. 34-35), Macdonald (1949b, p. 76), and Richter and Murata (1961, p.
217). The 1801 lava is a typical alkalic olivine basalt, an analysis of which
is given in Table 4 of the present paper. Untold thousands of ultrabasic inclu-
sions, up to at least 27 inches in diameter, are present in the flow (Richter
and Murata, 1961). Except for their large size and tremendous number, the
inclusions in it are typical of those in rocks of the alkalic suite in general.
They show a complete gradation from dunite and wehrlite to gabbro, and
rarely to anorthosite. Some are quite massive, but others show a distinct,
or even a marked, layering that resembles layering found in many big gabbroic
intrusions, such as the Skaergaard and Stillwater intrusions (Jackson, 1961;
Hess, 1960; Wager and Deer, 1939). It appears probable that the inclusions
in the 1801 lava, and by implication the inclusions of other rocks of the
alkalic suite, were derived from intrusive bodies at depth. Studies by Roedder
(1963) have shown the presence of tiny inclusions of carbon dioxide in the
olivine crystals of the dunite in the 1801 lava. The gas bubbles are under a
pressure which indicates to Roedder that the enclosing olivine was formed
at a depth from 8 to 16 km below the surface. Presumably the intrusive bodies
from which the inclusions were derived, lie at depths in that range, and the
alkalic basalt magma which carried the inclusions to the surface must have
come from at least that depth. Furthermore, the magma must have risen from
that depth with great rapidity in order to overcome the sinking tendency of
the large, heavy inclusions. Oxburgh (1964, p. 16) indicates that the rate
of rise of basalt magma with a viscosity of 3000 poises must have exceeded
22 m/hour to carry upward a 6-inch xenolith 0.7 gm/cc denser than itself.
The viscosity of tholeiitic magma reaching the surface in Hawaii is on the
order of 3000 poises (Macdonald, 1963b), and the evidence suggests that the
alkalic basalts, such as the lava of 1801, are less viscous than the tholeiitic
basalts.
Both Richter and Murata (1961) and White (1966, p. 306) find some
indication of disequilibrium and resulting reaction between the inclusions and
the surrounding magma, but it was not extreme. Possibly the time available
from the beginning of rise of the magma from the level of origin of the inclu-
sions to its final consolidation on the surface was insufficient to allow much
reaction, but the indicated depth of formation of the inclusions is not actually
very great and the physical environment there is not greatly different from
that at the surface, so that disequilibrium due to change of physical condi-
tions probably was small.
The inclusions in rocks of the nephelinic suite are predominantly lherzolite
with more orthopyroxene than clinopyroxene that grades into dunite (White,
1966, p. 258, 303). The lherzolite generally is completely massive with
no evidence of layering. The textures suggest metamorphic recrystallization
(White, 1966, p. 263). The high content of alumina in the orthopyroxenes,
which ranges from 1.0 to 5.4 percent and averages 3.0 percent (White, 1966,
p. 278-281) suggests formation of the minerals under moderately high pres-

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 485

sure (Forbes and Kuno, 1965, p. 176), possibly at depths in the range from
30 to 60 km (Boyd and England, 1963). White finds abundant evidence of
reaction between the inclusions and the enclosing magma that includes reduc-
tion of the jadeite component of diopside and that indicates adjustment to a
changed physical environment. The lherzolite inclusions suggest that the
nephelinic rocks originate at a greater depth than do those of the alkalic suite.
The rare inclusions of wehrlite in nephelinic rocks may have been picked up
at shallow depths, where the magma came in contact with consolidated intru-
sive masses that were genetically unrelated to it.
Some of the nodules of lherzolite in the nephelinic rocks are as much as
8 inches in diameter, and again the magma that contains them must have
risen rapidly from the depths at which they were picked up.
Some of the inclusions in both the alkalic and the nephelinic rocks show
marked crushing, undulatory extinction, deformation lamellae, and kink band-
ing in the olivine and pyroxene, but these features do not necessarily indicate
more than minor shearing due to movements of a few feet or tens of feet in
the volcanic hearth.
Inclusions of garnet peridotite and garnet pyroxenite ("eclogite") are
abundant in nephelinite lava and tuff at Salt Lake Crater and the adjacent
Aliamanu and Makalapa Craters on Oahu and are associated with abundant
websterite and lherzolite (Winchell, 1947, p. 13, 26; White, 1966, p. 250).
They have been found nowhere else in Hawaii. The inclusions are dominantly
of two types, though all gradations occur between them. One type consists
of about 75 percent clinopyroxene, 15 percent garnet, and 10 percent olivine;
the other type consists of about 75 percent olivine, 15 percent orthopyroxene,
and 10 percent clinopyroxene (Jackson, 1966). Although, as Jackson points
out, many of the inclusions cannot be chemically equivalent to Hawaiian lavas,
two published analyses of the garnet pyroxenite (Yoder and Tilley, 1962,
p. 482; Macdonald and Katsura, 1964, p. 123) correspond quite closely to
olivine tholeiite approaching oceanite. Both contain abundant normative
hypersthene, and chemically they belong to the tholeiitic suite (Fig. 1). They
are low in potassium. Their modal mineral facies indicates formation at a
pressure above that of the basalt-eclogite transformation, although recent
work (Kushiro and Yoder, 1965, p. 94) suggests that this may not repre-
sent any great depth in the oceanic mantle — certainly not necessarily more
than the 60-kilometer depth that is indicated for the origin of tholeiitic magma
by earthquakes at Kilauea (Eaton and Murata, 1960).
The garnet pyroxenites demonstrate that material with the chemical com-
position of olivine tholeiite is available at depths as great as those from which
came the inclusions in the nephelinic magmas. It appears quite possible that
they represent the composition of the liquid fraction that was produced by
partial melting of the mantle under the pressure and temperature conditions
in the zone of magma generation beneath the Hawaiian volcanoes, and that
the associated lherzolitic fragments represent the portion of the mantle that
remains unmelted.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
486 STUDIES IN VOLCANOLOGY

THE CHEMICAL ANALYSES


The Waianae volcano is the older of the two major volcanoes that make
up the island of Oahu. Table 1 presents 15 new analyses of lavas of the upper
member of the Waianae volcanic series (Stearns and Vaksvik, 1935, p. 67;
Macdonald, 1940, p. 64). All belong to the postcaldera stage. With analyses
previously reported of the lavas of the lower (precaldera) and middle (caldera-
filling) members (Macdonald and Katsura, 1964, Tables 1 and 2), they
represent as nearly complete and continuous a stratigraphic section as it is
possible to obtain of any Hawaiian volcano in natural exposures.
The analyses in Table 1 are arranged in ascending stratigraphic order. All
are hawaiites. The appearance of hypersthene in the norms is at least partly
the result of the oxidized state of the iron. The rocks are the freshest that
were obtainable in the natural exposures. Although some are very slightly
weathered, none are much so, as is demonstrated by the moderate water
contents. The oxidation is largely an original feature. As noted by Tilley,
Yoder and Schairer (1965, p. 80), the alkalic lavas are typically more oxi-
dized than the tholeiitic lavas, and the degree of oxidation increases with
rising silica and alkali content as is demonstrated by the Fe 2 0 3 :Fe0 ratios
in Table 8. The hypersthene in the upper Waianae rocks is wholly normative
— none has been identified in the mode, and all the modal pyroxene appears
to be normal augite.
The first 13 columns of Table 1 represent the sequence of lavas that com-
prise the main portion of the upper member of the Waianae volcanic series.
Columns 14 and 15 are analyses of lavas from very late cones on the southern
slope of the volcano. Although they are not separated from the rest of the
upper member by any profound erosional unconformity, these cones and
flows are distinctly later than most of the Waianae volcanics.
If we consider the first 13 columns of the table, it will be seen that there
is a general upward increase in the abundance of silica and alkalies, and the
uppermost flows (columns 11, 12, and 13) approach mugearite in composi-
tion. The lavas of the late cones (columns 14 and 15) are more mafic and
poorer in alkalies than the earlier lavas and are transitional from hawaiite
toward alkalic basalt. Indeed, the rock of another of the late cones, Puu
Kapolei, is a typical alkalic olivine basalt (Macdonald and Katsura, 1964,
p. 116, column 7). As already pointed out, the small posterosional flow in
Kolekole Pass also is alkalic olivine basalt.
A dozen new analyses of Haleakala lavas are presented in another paper
(Macdonald and Powers, in press). The lavas of the Kula volcanic series
(postcaldera) are predominantly hawaiites, though some mugearites, alkalic
basalts, and ankaramites are present (Macdonald, 1942, p. 279, 289; Mac-
donald and Powers, 1946). Lavas sampled along the Halemau'u Trail, up
the west end of Haleakala Crater, show the same upward enrichment in silica
and alkalies as is noted in the upper member of the Waianae volcanic series.
The lavas of the later Hana volcanic series are separated in many places from

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 487

those of the Kula series by a great erosional unconformity (Stearns, 1942).


They also include alkalic olivine basalts, hawaiites, and ankaramites, but
basalts are decidedly predominant in most areas and most of the hawaiites
are close to basalt in chemical composition. Thus, the latest lavas of Halea-
kala, like those of the Waianae volcano, show a tendency to revert to a more
mafic composition. Most of the late lavas of Haleakala, particularly those
of the Hana series, are notably undersaturated in silica. Many of them con-
tain in excess of 8 percent nepheline, and some as much as 16 percent, in
the norm. No modal nepheline has been detected, however. Much, if not all,
of the silica deficiency probably is in the pyroxenes.
Table 2 presents new analyses of lavas of West Maui and Molokai. The
West Maui rocks are slightly weathered, as is indicated by the high state of
oxidation of the iron and the high water content, but they are the freshest
that could be found. The Molokai rocks are fresher because of the drier
climate. Specimen C-153, column 5, is the only hawaiite yet found on West
Maui. It confirms the fact that in volcanoes of the "Kohala type" an occa-
sional flow of hawaiite may be present with the predominant mugearites, just
as the predominant hawaiites of volcanoes of the "Haleakala type" are accom-
panied by a few mugearites (Macdonald and Katsura, 1964, p. 83, 97). One
of the rocks from East Molokai (column 9) is an unusually mafic mugearite
and has a remarkably high content of P.O-,. It resembles in composition a
"trachyandesite" from the posterosional Koloa volcanic series on Kauai.
Probably the most important feature of the analyses in table 2 is that sev-
eral of the rocks (columns 1-4, 6, and 11) fall in the range between mugearite
and trachyte. (These are the "benmoreites" of Tilley, Yoder, and Schairer,
1965.) In Hawaii, trachytes are not more abundant than rocks intermediate
between them and basalt, as supposed by Chayes (1963). Rather, there is a
gradual decrease in the abundance of types more and more removed from
basalt.
In Table 3, specimen C-203 (column 1) is a tholeiite from the Hamakua
volcanic series and is stratigraphically above the part of the section that, at
the mouth of Laupahoehoe Gulch, 3.5 miles to the northwest, contains inter-
bedded tholeiitic and alkalic basalts (Macdonald and Katsura, 1964, p. 97).
It thus extends the upward known range of the tholeiitic basalts. Specimen
C-205 (column 4) is from a thick stubby flow associated with a dome just
southeast of Ookala. It was considered in the field to belong to the later
Laupahoehoe volcanic series and to be, probably, a hawaiite. Chemically, it
falls almost exactly on the line between alkalic and tholeiitic basalt! The other
rocks are typical members of the alkalic series. Those of Kohala Volcano
(columns 10-12) are all mugearites.
The lavas of Hualalai volcano, listed in Table 4, all are typical alkalic
olivine basalts. The volcano is so young that it is almost wholly undissected
by erosion, and since all of the samples came from the present surface, all
are from approximately the same stratigraphic level. This probably is a large
factor in explaining the very small amount of compositional variation among

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
488 STUDIES IN VOLCANOLOGY

MH H» M H — , 1 vi Tf 00 oo VO 00 ov
ov es io N m h >o *—< 00 VI Ov Vi ^ vo O "I vi O VO TH
- Ov
r-
vi Tf -«t oo oo Ov en o O es o o d VOes OÍ en Ov vO c4 OS en
Ü o< —
i i
Ti- oo th in N o ' es ve VOVI en Vi en vi Tt- • VO O O
en
VCI VI P>; TT; TF o ow-i ve
<s o Vi
O eñ d O
vo
Ov
es —
i i
t^ vs
00 Vi o
TH
- pi DO t ^ oí m OS es Ov oí 06 d

i i
Ö Tf rt
es m P-Î m v i u-i . ov Ov Ov 00 <N< 00 VO oo en vo • es o es o Vi
OO N M « ^ O Í os ve r» es VOvo ""t " O os Ov t-; so r-; Ov
I OÍ » 4 t t ö o en d d
o vi ov ; en ^H O vi es' T—Id
en T—1
O
VOw vo en 00 OS en r- • o en O Tt-
00 t"*; en Cv¡ VI VO TTOVl t- es VO es vi • OV en 1—1 t-; Ov
oí rí \o "ñ en ^ es o O en O o o es 00 o : en en O r—i es"
3 o T
— <en es
o00 V, en 00 O 00 00 C-l es O <D
1—I
oo en 00 vi OS en es T—1oes_ vi • os es
T-Î O o en o o O O T—1r-' f-í
1 en t—i
; en en t-í
ó O

os vO vi Ov r- o\ Tt 00 r-H 00 vo 00 vo es es en es • O O
ON es ^ ^ R-^ VI 00 vO en Ov es VI i vo o • vo O vi
FH oô vi oo en r-- T-Î O d en o o o Ó —i vi O : Tf vi
o en es
Ó
oo SO _ V) 00 00 en oen es en O • O o O
00 Ov en Os en Tf • Tf
i » -h
»1 »1 -st i-í Ö o en o d O d 1-H K NO' ; >ñ Tf
Ü O T
— <en —,1
T—1
t- tH rt 00 00 n N O en vo r~ oo 00 rt 00 • O Os VO en
r- TT r-; 0\ 0\ -H Tt; os eS -H vo en *—i wv oen »—< O VO 00
o vo
t vo vi r-' vi o¿ en o o Tf
" O o d oo
* « r-' vi en O en ^H*
O o cu en OÍ :
« ü
^ ^ » ^ O lo in o Tt- r~ — vo vi <r> w os VI VO O vo O en i—t
O 00 o\ 0\ ^o m OOVOVI so , ii rí- en

VO es 00 *—i Ov o
• ^ Tt \o if oo ri o © Tf d d T O
f
IS ov es oo 1 t-: VO d en O d o
O tH o B oí en
%
m ^ m M N vi •O OI Vi en vo o vo vqen Ov oo oo O os en es
oo r-; v¡ « o\ m
vo 'i oo ^ oo m O v vo Ov vo Ov O
w v • Vi " f r —i
T f rt Ö © © o o
o 00 o¿ ;
en *—i t-í vi -H es" o en

N m r^ » ^t Tt os O en 00 VI VO00 es o O
vo vi en vi en es © VOen 00 00 O vo o so . Tf vo
r-í o¿ vi ^ o¿ T—1Ö o en ^H O d o Ov en so 1 VO
Tj- T-Í o
« t Ov m o\ N O 00 o Vi O 00 Vi •<t vo VO • o es o so es en
vi vo Ov VI en Ov t • Ov Ov t-; o VO Ov vo
d o Oo o o¿ ^ oí ; M-' —i es' d en T-l
o en
SO en 00 00 OS VI o\ Ov en VO t- VO es 00 VOSO • es o O
Os OO O O t-; O en VI r- Ov Vi en r- Ov Tf • os ov
^t oô \û ^ « m Od Oo o 1-H K en oí ; vi vi vo
RJ- H o i-H
ri en es ** OO CMeS Os r- VO o Tí os oes vi O ^ ^en es en
00 Ov Ov OS -H VI Ov Tt VI Ov Vi vi
r- vi tí- oo r- en © O en o o d
o enH'oes
00 — en -H vi es" es'
Tf —1
t f- en Ov en Vi w e- vi oo Ov O o OO o Ov so vo
vi oo Ov —
M so vv en Ov Ti- r- en vi • V
Ov es en es Tt es
vo
^ —* tj- o\ vs oo en — O © o o 8 O o
en es VO TÍ es" o vi en

Ç «W
O co utí «4ai
-1 O
«MdM
H

<
o. o " 2 - ° " o â Q ° « o o o o " d ' 9 in .o e «
CO Y ^ A> Y "3 «M MÄ J o « mü
K C h f c S û Z ^ ï I H i - S

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — COMPOSITION AND ORIGIN OF HAWAIIAN LAVAS 489

Ë
a
v Ia
SH Ë 3
o
I ¿2
a- es &
t- £
NO o
00 U >
aes ao< es'
& u
00
•a 3 3
¡3
•a 3
3
£
£ P*
so
NO o
v-l •O en
aes
cl 3
>
m
/-> cl
ON
•a
CU
o "
CL *4-l
Ph S O
Ë R •§
o u u
oo oo vi « «a
^h ro ci voo u
G
t X! N VÍ •a a O
> a u
¡3 es
D. 10 3
3¿4 O, *3
m \o CL NO es ¡3 Ë
aes 3
3
z xi Oh
-H Cl o Ë h
53 o.a o.
ra o c & * o. •O £) J3 U <D
oo ci Si "O
ai ^ a à Ë S uË
•S z Q. O tü u Ë
Ë Ë
(S 1 * £ o u
o u o o S1 g< '35
« M 60 E "§ "3 Ck u
£ ëo •a es S
'C
ON O (S >M .Si,
s i U
tu
00 ON © o •u
r-' ci 2 B. g SI
es c
s a c CI es i
Ë ."2 o
<u
•o .3
<L>
Ë O f
•C
es
1
•a N & Ë x
3 ... •3 O ^
•a> g u -g « Ê
•a aN (U •o
•o
m •o 'I £
< 3 •3 „es V -l
>n
3 3 3
ta c
X —
tO 'S
N
O
d T0)3
o •g S
u
"S
-t-»u
"S o. .a o
>> <s
N "O <u
<u < o (u
U
-S » Ë <u «M -Si «
cï o vi NO •a 2 - 1 » N -S is O N v-1 v-1
c\ ro »n c« 'S T3 r^ m •O
o es
p) M ^ û o Cl Ccl1
m C
Cl es
. o ¿3 "S jî
S >
P, O
a.
•C
•a
£
es
"S
-S -g .^
vn m
«J ••o
H
—3
r-' ci es .2 •s g •a Ë 3 S
"H, >2 3M ais o * * Ë Ë
MH
.. J3
<D <u o u a> ü S e
n H c] O > :a es
O cl o
NO 00 ci
g n S "S es es
S u ^ s % I %
CS es es es
& £ S &
es es
«

eS es

X K w a
CS J 3 ça X
a « 2 X X
*-t ÍS o rf »1 « t- OO ON O ^
00 00
ON
o r- r- r- r- f» f- t'- co NO
<c I U Ü U Ü o o o o o o o o o o
O
N Wlo' \¿ I s 00 o H N n ^
a s o, Ë
es J3

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
490 STUDIES IN VOLCANOLOGY

,-1 SO «•N
so o
S •• en
M es t- O
» P N l O ^ i n N T t t m t t- Vi NO es es 5 Or r
Os *en
"! f
i o M\ m c i « « 1d N ' 0 0 d es O o vi es O es t Os vi ©
U f "- SÌ es

S t c o ^ v i n n N H i n i f i oes O es os O Os vi Vl • rH o SO O
00 •fr es Os • OOVl es en
^ h h rñ rn h n / i n o o «M O o Os es' r-' Os —
t1 : o o O
rH
Os

vi rfr os vs 00 os en en O O 90 vi so VI oo •t O es •
f - OS in en es O 00 en Os es es O VI <n Os Os t», •
ü f - es' 00 vi O es O g O vi
en
00
»—i en en es r* *

Vl -t © Os en
os en oo en O
00 Os en es os
SO VI 00
VI r* en 00
-fr O OS
Vi S© t r- -fr vi < © o es es e os Os Ö vi K en es' »-i
O « r l en (

S M (S oo Tt a a T SO en O so es
S ^ 1 ^ 1
00 <S (S —I 00 «
es m so oo o- H (So rn Oss 00 S Os
m h N ¿ f i vi so es O O en f»' »H o O
(J m " »•H

O SO SO 00 © oo
00 ^t m so Os vi
Üvi oo t - f r m e n v i e S O O « — 1¡Í
r*
t> rir*
Vl

en SO
oo O VI es •<-i -el- oen S Vi 00 t'- 00 s es so • • . . ©
« ~ r- es en
es
en es en vs O
r» Os vi . . • ' es
o £ os Ö ri s-i SO en ri es' ri o Ö
8 es r-' es' vi ; ; ; ; en
VI

•fr Vi SO Vi OS O00 so es OS OS OS en O -«t 00 SO so o oos m o oo


[s Cl N w O Ol r- so vi SO SO Vi t» f; m ifl o\ oo ^
en •fr O vi os es O O ö ci ö ö os ri m m ri Os vi ->fr t-^ vi
it « 9s es es

fssc^-oenoomvi'^ooo^'ooi— so en vi O
so 00 Os
T i-- 't es° »-! to so ri o ri wi o © SO Ö en
(J »i « vi

00 v> OS es vi es
o r» VI o OS 00 in O S vi os
SO -fr en
SO es SO
00 1—1 r* :
: ; ; ; es
* •
so oö so' T—1© es" vi es' 1 M SO r-'
VI T
— < r-t

O fl
Vl eS
fl\ OOrtffi es Os O es o
eS SO © SO OS Vi es 00 en oo es V)
00 Mo
os o es Os §2
"7 so' 06 fi ri e»¡ es' O ö tH O o Os vi oó en
ÇJ Vi ~ os Y—1rr r H

Os es Vl r-
r- Os Î3 SO 00 Vl 00 00 en f* SO en
00 00 es 00 00 Vi oo • os
00 so en r-; . o
ví f 1 en
Vl <—1 •>fr — es o Ö o o es so r-' ;
rH VI

Is 5 «

o' m O O o O o
+ I
<D r\
A . W i-t «
«O
u
M(SVw pi
O« On o"

o" 9
-„ —
W « N « Ï5 ei £ u, Ä e «
C o u a c

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 491

t- n
P- rfn
wi VI "i

O so <s
<S O
in N N

< »—, I— f-0


so «n r-¡ ^ ©
« «i n w w

l; « » <H «
<N «-« \e i--
ifi « b »i

r- oo r-
n f3 ~ «© o\"
ie
m" ci ci

o
in « « s
ve' ci ts

-h -g- oo
ve o, o *-¡
«-< r-*

« » t
es <o\¿ mM
\¿

O ÍH 06 ***
« « m N v
»1 ri « « ó

n « « «
*© V> <-<

rt
t- O
© N N
W N
s

CI
O N
oo «oo
V0 CÍ «

0\ o «NintmoooiO'; (S
VO
•<t x-i
O O O U U U O Ü Ü Ü U o
>-< ci fi ^ vi v¿ oo' os o vi ci
"o fia S* I o

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
492 STUDIES IN VOLCANOLOGY

en vi en Os O t- VI 00 TÍ Os es es 00 o t> os Os O es
eS O s 00 Os es Ti; OS eS -H VO es so SO en o so es e-. en
00° vi vi so' 00
en vi - O C> ci ö ö Os en en ci ^H*
t en ^H TÍ
1
Ò T o<

O r-c- so r-H T—1 1 00 en so es VI Vi Tf Tf Tj- o es


00 O
00* vi vi SO mOS M en
Vi
os vs OS es o
ci « ö oo
en en 00 OS r^ Os
T-í o virn ci en es' o
TT V1 Tf

o O r- M en O T—( OO en es Os 00 so O O so
C4 00 o T-l Os I-; eS es SO I-; Vi VI es O
,-H fi SO en O
s¿ vi ci o o
o ci TÍ ci ci «
vs Tt

oCTSm O s en r- t'-
SO es
, 1 Os en so r-
— eS
SO

oo es 00 ci o
(S «ri O
o en 00 so'
vi vi
T-i Ö ö © O
SO
oo
<N V) f-
ci es os
Os V)
en o es'
Ó TT —
,1
1

o O s o Os 5 vs en
en O 00
so so SO TJ- TJ- es O
oo I-; Tf m t O es S 00 SO 00 es O o 00
(S Vl or-' oen so ci vi vi ci o O vi vi O Ö
SO so SO OS en v¡
so t-' ,-H ^H* es

vi »—i o w

o SO oo VI Os en en es 00 es VI oo V) O SO
O en r- s Vi t»
so »—i so es I-; os
ens TÍ Os vi V)
O so io <
(S SO vi o o ö o" so vi so' en
v-t Ö1 o es es ,-H
ò

S O es 0\ ci o so o r-- OS V)
oeS v-1 > Os Tf O SO
OS Os *-< so o SO o TJ- es CU I-- os es SO vi O
Ö so' TÍ so en so' vi ci o O o'
o o ci ci ci ^H w:
vi Tt —
' i
Cft

(S en os Vi es f- T* o
(S en o »—i so
so Os Vi s o es
es so Z
es
w Oen Vi r^ O Os
00 SO f-
Os so' so' so' en o ö o O TÍ en o'
O
TÍ o ens T
—<
^H

r» vs 00 00 as oo oo
00 en
"o
o 00 O oO Tt; 00 es
es o f- en so t c r- TT O
P-;
es
o oe> <S
es O O
en t-;
es SO TÍ O s vi ci o o o Os ^ vi O so en
i Tt T—1 Os es
ü

O en OS o •>fr SO 00
O
eS t-;s 00 Os oo os eS o
es
<s r- o
oo 00
O O en
es VI Vi
O s vi so en s¿ Ö Ö oo so' Ö ci ci ^ O
Ó T»- i •<t (S
"

r- f-; - en 00 oo oo so
SO oo SO so <s os es 00 en
o 00 t so
SO 00
Vi VI
es
o
TÍ o es
TF oo es
eS SO TÍ en O vi Ö ci o O O O O vi en vi
TT —
, 1 O es es w vi TÍ
Ö
TH
-

SO oo SO Os Os iH <s OS Vi H OS en r- Vi SO TJ- en es o o
O O en1 00 so so oo <S es r- V) 00 en ^ SO VI SO en vo_ SO rn
(S K M TÍ oo so" TH- ci O o' ö ci O ö o —
i i oó oô en
oI TJ-
<
1 o es
T—1
S c «M
>
o
a
O — i" J, « U S
ou O« O" OM O « O O O o O O e 'S
O, « « o
a
an

Cv M NN
ab

ne
or

CO co < fe fe
Z X E K H Ph S

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 493

NO "o es es O S o
<s m q aj ID .2 S NO •C
ci r-' vi <u (S

0 e •a M •a s
S « cO
o
o< o, ciiO Oi •c
3 3
es NO O oo s s S cO 3 C
O S
»H CT\
»-I 00 <è O •J •6 h-l .2s a
r- § J .2 =3 <u
ci -t—> « 3
3 « -i S
s a S M o •»
CO
•a ° > u
'S "S CO u
ON o o U >2
1s o co O c
3
sCO
Ü
O NO ON NO P-; "3 X g w M 0>
en en oV I > 2 s s
0
CS
3
5a •>»
CO
<U
i I
p-
o S
S 1
0 ra6 3
cO M
O ON m O o O S ¿ 5 3co
ON . I 00
— 00 NO S «s «
u
, I oó TI-" vi en O
— 3 o 3
S £ cO I S §
(U 2 hJ £cO a I
, o «
.y « I •S c "3
•8
O Ü « A
C4 oo VI <U jS CO
Cj en Ti* r» en » S, CO C
oo rn Tf es' es' £ Os
w .s? & M 0 g «
2 s
1 -S
CO O
C
O 1
e
o 5 'S
S s £ X 2 X
oTt- >o m NO o Tí x> 2 « A (U
3
r- ci en — I es NO en S O
&
ci NO P-' ^ O o o G •s? 3
Q 2 G c
-3 -,
co
Id o> O
g c(U l es'
ON 2"o *
om m'— NO NO en NO NO M ^
l à o
P M
Z NO O O o es O es en •<u o •S â 3 G
1

en NO es D O 3 O ..
w
O en 3
-, _ ° C
». fi .
o ju 4L
> J3 CO T3 O « O ci
U s i C 33
CO O •S K -2
"H s? •g s * >M s
id '3 o O 3 3
J >o O NO es es 00
o c¡ 00 es
vi
en G •o
cCO « a -S •S £
m o en O vi es" •o eO —<
£ OÑ u X I S .S s J O lu
oa CO A ^ O
W CO cO
3oo . UiS Z** J3 e c3f?
O | w * Û S
a BO M
^ Ü «
S
o O ON o O2 CO a1 « c OS £ 2
r- 00 ON en £3
Tt ...2 >
— 1 1
S* 2
00 ^ vi NO ^ c fcH ^
•C « 0 «fc-c 3 3 T3 "n
EL H S
5û û e«
& -sg s o g.2 .2 £o0 .2
.v
. O _
•o
S <s.
o > 3
C S —
3 « tí ® fl cO-
•s g
u
o 00 ON
oo
NO es 2O S t ^ s 2
NO V)
en en o en '3 c -g co « > o ?
NO ci O O vi es" S > « z a z
» g
co s I^
> O •a « °
"S S M o o ja • Sv ^ -Û
•S ¡3 a o 0 <u CO » ^ »J d -fcj
<u ai cO _ '3 _ -fio o 3 O" S2
O r^ en t VI Tj- TÍ-
(S t-; t-; 00 en
en ci ci es' vi N
G° > •
JS
e S
o >o
g s. ti -3 c
3
Q 3 .3
O A
O o 1 i .2, o, c
o <D l i 3 So
1-1 Í
8 ¡ 2 o C
O ^
¡3
•a o »«9
cO
« S .3 •O i>
cO •C <D 2 o S . S.2 3 S . Ë
O ON ON r- P- 3 _ o ~ M b' b' Ü
O —I NO u 1)
<3 » o u *3 a> a> co a> c a *
OÑ vi O M
<u .2 == c w C O :s« «o
•S CO G :a a u o60 U)
® a a« .5 S oU)
C
Ofe COo 3 Ü »Ô •r 2 A!
3 "3 O3 3 3
° 3 cO cO <i cO ? 2- §2 <2G g>
< s e a S S S
tí 0 cO JS 1-1
(U «H
o es' H -î NO 00 ON O M pi
es od (NoI vi o(N c)o O (S
O O
(VI
—I
CS (S <N
—I

U o u Ó Ó ô Ô Ó ÔÓ Ó
ri a
J3 "o 3 cO

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
494 STUDIES IN VOLCANOLOGY

the lavas. Specimen C-223 (column 6) is from the lower vent of the eruption
of 1801 (Stearns and Macdonald, 1946, p. 148). The analysis is consider-
ably higher in alumina than the analysis of the flow from the upper vent made
by H. S. Washington (1923, p. 102), but it is very close to a large number
of analyses of the upper flow made recently by the U.S. Geological Survey
in connection with the studies of inclusions in the-flow by E. D, Jackson.
It appears probable that Washington's value for alumina is too low.
Tlie analyses in Table 5 are all of posterosional lavas. Column 1 is of an
alkalic olivine basalt from one of the vents of the Koko fissure that crosses
the extreme eastern end of the island of Oahu. Like the flow from the south-
ern end of the same fissure, close to Hanauma Bay (Winchell, 1947, p. 30),
it contains normative nepheline, but there is none in the mode, Winchell
(1947, p. 28) referred to these rocks as linosaites, but they may also be
termed basanitoids. Column 2 is of a basanite from the Pali lava flow (Stearns
and Vaksvik, 1935, p. 116). The rock contains a little nepheline, but more
abundant plagioclase in both the norm and the mode. The dike of nephelinite
exposed nearby in the Pali (cliff), analyzed by J. H. Scoon (Yoder and Tilley,
1962, p. 362), is clearly not the feeder of this flow. The material analyzed
in columns 3 and 6 is primarily black glass. The higher water content, un-
doubtedly, is the result of hydration of the glass, which, however, appears
perfectly fresh.
Specimen C-168 (Table 5, column 7) is of an alkalic olivine basalt that is
very rich in phenocrysts of olivine, up to about 5 mm in diameter. It comes
closer to being an oceanite than any other Hawaiian alkalic basalt thus far
studied. Similar rocks are common in American Samoa.
Columns 1, 2, and 3 in Table 6 are analyses of tholeiites from the caldera-
filling member of the Koolau volcano (Kailua volcanic series of Stearns and
Vaksvik, 1935). The rocks have been altered, apparently by gases rising
through the caldera fill; the pyroxene is partly changed to chlorite with libera-
tion of silica, which has been redeposited as secondary chalcedony and quartz.
Zeolites also are present in places. The analyses indicate that, except for an
increase in water content, there has been little change in the bulk composition
of the rocks. All of the caldera-filling rocks of the Koolau volcano appear
to be tholeiitic basalts, which are chloritized to varying degrees. The "amphi-
bolitic dike rock" reported by Manghnani and WooHard (1965) is actually
a portion of a melilite nephelinite lava flow (the Training School flow) that
belongs to the posterosional Honolulu volcanic series. An "eclogite" described
by the same writers apparently came from Salt Lake Crater, a Honolulu series
vent on the western flank of the Koolau volcano, not from the area of the
Koolau caldera.
Column 4, Table 6, represents a tholeiite flow beneath the Salt Lake Tuff.
It apparently is part of an eroded ridge of the older Koolau rocks that are
buried by the tuff. The rock in column 5 is exposed at the base of Puu Kapu-
wai, one of the late cones of the Waianae volcano. It was mapped (Macdonald,
1940, PI. 6) as part of the upper member of the Waianae volcanic series.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 495

T A B L E 4 . LAVAS OF HUALALAI VOLCANO, HAWAII

1 2 3 4 5 6
Spec. no. C-218 C-219 C-220 C-221 C-222 C-223
Si0 2 46.78 45.63 46.50 46.88 46.54 46.12
AI 2 0 3 14.38 13.86 15.33 15.41 13.95 14.13
Fe 2 0 : I 4.55 3.25 3.12 3.85 3.16 3.00
FeO 8.13 10.18 10.27 9.29 9.97 10.12
MgO 8.53 10.01 8.05 7.15 9.40 9.34
CaO 11.51 10.39 9.64 11.00 10.74 10.77
Na 2 0 2.44 2.71 3.22 2.84 2.69 2.68
K2O 0.85 1.00 1.09 0.93 0.87 0.97
H 2 O+ 0.62 0.68 0.18 0.32 0.41 0.31
H 2 O- 0.22 0.12 0.12 0.12 0.24 0.16
Ti0 2 2.13 2.41 2.57 2.33 2.25 2.35
P2O5 0.24 0.31 0.35 0.31 0.29 0.29
MnO 0.16 0.18 0.17 0.16 0.16 0.17
Total 100.54 100.73 100.61 100.59 100.67 100.41
NORMS (CIPW)
or 5.00 6.12 6.67 5.56 5.56 5.56
ab 20.44 16.24 20.96 22.01 18.86 18.34
an 25.85 22.52 23.91 26.41 23.07 23.63
ne — 3.69 3.41 1.14 2.27 2.27
( wo 12.18 11.37 9.16 10.90 11.83 11.60
di en 8.20 7.20 5.40 6.60 7.30 7.10
( fs 3.04 3.43 3.30 3.70 3.83 3.83
en 1.90 — — — — —
hy j fs 0.66 — — — — —

fo 7.84 12.46 10.29 7.91 11.34 11.34


- Í fa 3.26 6.94 7.04 5.10 6.53 6.63
mt 6.50 4.64 4.41 5.57 4.64 4.41
il 4.10 4.56 4.86 4.41 4.26 4.41
ap 0.67 0.67 0.67 0.67 0.67 0.67
Analyst: Shiro Imai.
1. C-218. Alkalic olivine basalt; on new road to Kailua, 1.0 mile from junction with
belt road at Kainaliu, Kona. Hualalai volcanic series.
2. C-219. Alkalic olivine basalt; on new road to Kailua, 0.65 mile from same junction.
Hualalai volcanic series.
3. C-220. Alkalic olivine basalt; on highway leading N from Kailua near coast, 0.35
mile from junction with Palani road. Hualalai volcanic series.
4. C-221. Alkalic olivine basalt; on same road (road from Kailua to Honokahau), 0.45
mile from junction with Palani road. Hualalai volcanic series.
5. C-222. Alkalic olivine basalt; on same road, 2.95 miles from junction with Palani
road. Hualalai volcanic series.
6. C-223. Alkalic olivine basalt; lava flow of 1801, at lower edge of Puhiopele cone,
downslope from highway. Historic member of Hualalai volcanic series.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
496 STUDIES IN VOLCANOLOGY

TABLE 5 . LAVAS OF THE HONOLULU VOLCANIC SERIES, OAHU, AND THE KIEKIE VOLCANIC
SERIES, NIIHAU

1 2 3 4 5 6 7
Spec. no. C-164 C-163 C-196 C-195 C-165 C-197 C-168
Si0 2 44.85 41.68 41.42 38.92 39.55 36.00 44.67
AI2O3 13.26 11.24 13.47 12.25 11.01 10.67 15.17
FE2OS 2.36 5.94 3.89 5.33 4.79 6.43 1.98
FeO 10.60 8.38 8.63 7.88 9.19 9.57 10.39
MgO 10.82 12.84 8.25 12.95 13.60 11.73 11.65
CaO 10.91 11.30 13.64 13.21 13.07 12.78 11.10
Na 2 0 3.00 3.30 2.78 3.92 3.69 5.62 2.27
K2O 0.80 0.82 0.76 1.26 1.02 1.84 0.27
H20+ 0.75 0.34 0.91 0.33 0.56 1.15 1.30
H2O- 0.25 0.75 1.22 0.38 0.44 0.48 0.43
Ti0 2 2.13 2.57 2.14 2.62 2.50 2.70 1.01
PA 0.53 0.80 0.40 1.11 1.00 1.41 0.16
MnO 0.18 0.19 0.17 0.20 0.20 0.23 0.18
Total 100.44 100.15 99.93* 100.36 100.62 100.61 100.58
NORMS (CIPW)
or 5.00 5.00 4.45 — — 1.67
ab 13.10 10.48 3.14 — — — 13.10
an 20.29 13.34 21.96 11.95 10.29 — 30.58
ne 6.53 9.37 11.08 17.89 17.04 24.42 3.12
1c — — — 6.10 4.80 8.28 —

( wo 12.64 15.66 17.98 13.11 15.08 8.93 9.86


di I en 7.70 11.60 11.50 9.80 10.80 6.40 6.00
{ fs
fo
4.22
13.51
2.51
14.35
5.28
6.37
1.98
15.82
2.90
13.72
1.72
16.03
3.30
16.17
ol
ol Ji
fa 7.75 3.26 3.16 3.26 4.18 5.00 9.69
cs — — — 9.12 3.70 20.30 —

mt 3.48 8.58 5.57 7.66 6.96 9.28 3.02


il 4.10 4.86 4.10 5.02 4.71 5.17 1.98
ap 1.34 2.02 1.01 2.69 2.35 3.36 0.34
Analysts: columns 1, 2, 5, and 7, Hiroshi Asari; columns 3, 4, and 6, Y. Watanabe.
* Total includes 2.25% C0 2 .
1. C-164. Basanitoid; Kaupo lava flow, roadcut just E of entrance to Makapuu Beach
Park, Oahu. Honolulu volcanic series.
2. C-163. Basanite; roadcut on new Pali Highway on Kailua side of lower tunnel. Pali
lava flow, Honolulu volcanic series.
3. C-196. Black glassy nephelinite ash; Black Point ash, on Diamond Head Road, Hono-
lulu. Honolulu volcanic series.
4. C-195 Nephelinite; Upper Nuuanu flow, on Pali Highway just S of Wyllie St. junc-
tion, Honolulu. Honolulu volcanic series.
5. C-165. Melilite nephelinite; Castle flow, main face of Kapaa Quarry, near Kailua,
Oahu. Honolulu volcanic series.
6. C-197. Black glassy melilite nephelinite cinder; Tantalus cinder, quarry on Round
Top Drive, Honolulu. Honolulu volcanic series.
7. C-168. Alkalic olivine basalt; near Kiekie village, Niihau. Kiekie volcanic series.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 497

However, its relationships to other rocks are obscure, and it may belong to
the lower member. This interpretation is supported by the fact that its stron-
tium content, as determined by N. J. Hubbard, is lower than that of any other
analyzed rock of the middle and upper members. Columns 6 through 11 are
analyses of historic lavas and one very late prehistoric lava of Kilauea and
Mauna Loa volcanoes.
Table 7 lists new analyses of lavas of the nephelinic suite from the island
of Kauai. Most of them resemble earlier analyses of rocks of the same types
(Macdonald, Davis, and Cox, 1960, p. 110). Specimen 192, column 3, is
a coarse-grained intrusive rock corresponding in mineral composition most
closely to an ijolite. The rock appears to vary into an oligoclase gabbro
(kauaiite). It contains many inclusions of lherzolite, dunite, and a few of
orthopyroxenite (Macdonald and others, 1960, p. 74; White, 1966, p. 251,
locality 181). Sample 183 (column 10) is noteworthy because it represents
a type hitherto undescribed in the nephelinic series. Under the microscope
the rock is leucocratic, with a marked predominance of feldspar over pyroxene
and olivine. Although the norm contains more than 10 percent nepheline,
staining by Shand's (1939) method has revealed no nepheline in the mode.
Chemically, the rock most closely resembles a trachyandesite.

DISCUSSION OF CHEMICAL DATA

New average chemical compositions of the various types of Hawaiian lavas


are given in Table 8. No new analyses of oceanite, ankaratrite, and rhyodacite
have been made, but the old averages are included for convenience in com-
paring them with the other types. From inspection of the table, it will be
seen that silica, alumina, and alkalies increase steadily from alkalic olivine
basalt to trachyte. As compared to tholeiite, the alkalic olivine basalt is richer
in alumina and alkalies, but is poorer in silica. The feldspar-phyric rocks are
richer in silica and alumina than the other alkalic basalts, and indeed are
richer in alumina than the hawaiites. They are generally regarded as cumu-
lates that are enriched in feldspar phenocrysts from adjacent magma, but
this is far from certain. They may represent magma that is enriched in alumina
in some other way, in which the abundance of alumina, together with the
other appropriate constituents, and the physical conditions brought about the
early separation and long-continued growth of plagioclase crystals.
The nephelinic posterosional lavas display a marked increase in alkalies
combined with a decrease in silica and alumina, as compared with alkalic
olivine basalt. Except for the basanites, the nephelinic lavas also display an
increase in lime and total iron and a less marked average increase in magnesia.
The alkalic olivine basalts of the posterosional group, as compared with those
of the earlier period of volcanism, are poorer in alumina, slightly poorer in
silica and iron, slightly richer in lime, and notably richer in magnesia, but

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
498 STUDIES IN VOLCANOLOGY

vi vo en O 1 O ov o oo en so O It o\ VO O «1 o «
t o o\ oo vi n » 1*H rt ín IH VI SO VO oo « 00 « «
o VO Ol es O O o es' o O O s¿ h ^ ri rl
v> es es OÑ

00 es en O o\
v¡ 00 r- •o 0\ 00es es
t
f» es <— NO esesovmooooes
'tMmwOíOímíelO
en en vo' o es' O o O O O «Noivódímoil'
v» »—1 « N H

« » « N S B h f r t v O O ^ l í es O es 00 © en O 00
vo es V
00 o TT en •"fr t-~ 00
•rt'eneñr-^t-^OMOÓOfs'©© vi es oó vi
es 1—< P-' en o
•o —
'I »—I

O es es
^O ^—; ^—4 »•—; NO c-vl os —
' ; es es 00 O
en 0\
r- O O o Os
Os o Tt-
-< en Tf t-^ r-^ o es' O O O 1-i o ó es vi © ci
es 1-H
Vi *-H >-H

<
«o •<t es vo Ov O O -I •n Vi r» es ©eS<SO\vooo©rf
eS 00 «i en es en <1 t T o rí es
O es es OÑ OÑ O ri O o oo H M I« - 'es^ —
Oi l O n «« «
ó VI i —H
i

o vo 00 —
t o 00*—< en r-
en en t- en VO es O O O en
oI Os TT 00 vo es ©•t es 1 vo
Tf q VO en Vi
00 en •<t en en r-;
y—ien es oó o es O O vi i oó vi en en vo
o VI — es O VO

t^eS'3'OOViVOOv©OvOvf~-<J- esooenvien©Tr©vo
a i w v i i s h H M o o o H o d *! *! 9 "!
o H N Nesmesr i-c
i o o m rO
-< n

00
ov O Vi oo O T—1 v> t- vi VI 00 en vo vi 00 oo es »H I- O t-H o vooo
V) 00 00 VO Ov Tf VO vi VI es »-i vo -—i es en Os P- »H »—1 OS
es' es en oó es" O o o es' O o' Os Os es' oó en vi es' en vi
3 vi Os es

TI- «»OTtNOrt—i
»NOOUN m O l ^ «In o Vi en
Tj-ON-t-H oooomtNONOM
OS Tf en i
o
I N ^ < t « \ ¿ 5 v N C Í o 6 es' O O
Vi l-H
8i es es

oo en O r- r- •"t Vi
es i i es o rf
— VI VI vo O
00 oo
Os
VI en
en fO ©®\00©vi©vo©vi
oo oo it en m o t^ 1;
Ó vi rf
1—t vi vi vo os <N O ó es O f-cJi-ivioÓvOrHOÍi-i
es es

es en r--
vo Tf es vo o\
es Os VO es Ov © O Tt
1—1 es VO OV »h en **
es xt so oo oo o o\ O es
H (f| OO in N 1H Hso en
o en vi vi vo O es o
o VI —i rH N O O

« \e í » « n O » N W H O f
o\ TH o\ n m tn rt 00 flj H í¡ t rt
o t ií »i « o N o
O vo es
VI 00
00
es
VO Os
8 en
es o 00
00 Vi
V > ,-1 A O O oó vi es oó

i a
<U <ffi

Di
o"®"°"o °«o o o o cf >>
JS
Cti Ss fl) «) tai Cí ^ M <N W M o s

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
MACDONALD—COMPOSITION AND ORIGIN OF HAWAIIAN LAVAS 499

v© O t^
^
v¿ ^t O !
a

O O NO
Tt © a
•o
ci
3
O •d
X CD
vi r» «o
n « « 3
«X «*> ó O g
«1-1 «

W
Z
"> £ S
© en a
NO en o §
00
•o © ta
« o
— o o J
[x » x> Pi a ci
m i c
m en ©
•o x 3
£ ai •o 00
a a
m
— O
§ c
o
^ «o « ci C
•sf en © •8
3 ,3 C

On ©
H m
Z
O
X O f s
O O
O O
V)
S CO S 0
•s
Sn •<t
no r-
«
no
o P
•3 -J
C
rr Tt o « 3 m 3
fí a
a 3 1
s« M
•g «M
a S S Ï O
>>
en NO -H _ o c tu o
o « S
r-. vi o 'fi
« t «
o la 'S
G
a) ed ta «
g C
§ § .3 •feo
eS © C-
TF ~ NO
->t o S S * o c
•s S C «
«•2
O "« fa
* > .
«1 >.
.2-
u. »-< ^ co I«2 "- m
NO H
$
•a « .a *
NO TJ- O a « -3 > S js
t-' "fr « et £ J es •S .2 o 60
60 ON 60 O r-
S 00 O
S «S I®
I3 x¡ S M' C ON
•C «2
gZ £ §
2 « 83
... ... o ... «
.9 o
«
ss
"33 3
S 73
0 II
3 "oo "s.
JS J3 ! 2 >
H H
no'
i «—<
r^ •fr-
es
15 es• esi es es es•
<a o a O 0 0 U U Ó Ü O
Vi NO t— 00 On O H
S » fî

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
500 STUDIES IN VOLCANOLOGY

es Ov r-4 Tf es O O 00 VI
en <i O VI r- V Vi 00 en ov
r- 11 en
o r- es OS
o VO O VO
es o en 00 es
Ov ->t VI ©
vi es' eñ Os <—l rH es' Ó O © O O oO vi O en en Ov en' en SO
1 l-H es *—i

Tf en en os TÍ O vo Ov es
es vo O 00 o os 00
es Tf vi o ve vi en o
8 es rH
TH
-

O Ov es r- T í oov O
Tf Ov V i T-<
es
oov rH
Vi V
VI
O es
es vt íi
Vi vo ( • » 00 O O o O r- 00 r-
TÍ Tf 00 H-« es' rH o O es' O
VO en oe> t^ p —< vo
ve es en rji en
i O o VO 00 vo' 00
O TJ" v—' rH

T í t - - T í - < f r o v t - v i v o v o o v o o r - f - en es 00 •t en 00 O vn so Ov
en rn —¡ t-; O ve vo Ov ^h en en ov ^h o i -t en VI Ov Ov r- O Vl
O ^ 9¡ h pi ri O O Ó N o O O VO O eñ vo r-' —
H Tf Tf
O rH

e s o o - < í © e n v o v © t - . o s © o v r - vi 00 ov c- es 00 O 00 00 r»
f- es p es ov 00 oo es VI
3 2' m Á t i n r Í H o d r i o d Ó o t-í vo 00 o -* TÍ r-í en
TH
ü r-i rH

tí os Tf VO en en t í en Ov es en vo 00 vo VI en O r- 00 en
00 00 § O en 00 V I en Ov V i V I en r- VI en -í es VI r- 00 vo
3 en es' en Ov t í O es o Ó ó
t í rH
O 6o >
>
vi 00 Ov vo
rH
es' 00° es' Ov so
H
OH
«
3 vien»-HViesnenesooenc^ vo g 00 —1 en so o en es 00
5 os tí r» VI T
Vl os VI 00 VI o
f'i^Tí'^ooenenvqvit^^H Ov O r-' rH es' DC en en es' Tí
M O t ov ri N CN o ' O ci o © o\ Z rH T—1
TÍ rH

ve v e T f r f s - H © v o v e © f S o s T t - o i ^ r- en O ov 00 Vi O O VI 00
00 » O * « O » O N 1 »1 N W rt Ov es r-1 r- t-; vi es 00
H H ^ o ¡ N H n H d d m d d Ov tr-í es es' r¡ r-í es' en cj en
3 ^ i—I r—I < OV rH

00
r-; 00
es V O Vl o Vl 00 en VO o r- O en 00
p Ov vo O
es
r-< 00 vo VI 1—1 r-1 1 I-; vo p en
vo Tí o o en rH o OS ve rn* os eñ o' t-í r
H
Ov . 1 es

Tf t O es 00 vo vo en r- o es Vl en
•t es © ^ ^ t " ^ en r- Ov vi vi os Tf •"t
o o vo t í es es © —I OI O o O es' vi ve' SO vo' en o VO r-4*
O rH i—, i —H
T—1

-Henovvo00t--.voenTíenvo-H00 vi es oo 00 en O en VO
rt rj| r.I oo Ov 1 vi vi vo t-; oo <-<Ov os OV O 00 en
os Ov vi oo vi
en
en •-<©'©' es © o oí
t-i »h .
VO vo vo en es oó en
Ov r-H ^ T—1
O
* CA £ C3
en

§ +
<

o&
° O Q. ¿ o " o" 9
B5 a ffi af P x> a <u
0 al oi a Si •3 O

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 501

•<t o
SO 00

O SO "H
»o CS O

•a
x
S *
J3
en so vs
© vi en o3 Sf
SO Tf ci s le
c — ts c
o s o
I
ë * 1 o
.2. o
£ S1
' i l
aS n
2 O
co
J & •g
"S 8
& '3
•a
&
S? £ PL,
M 3«
o
u cû u OJS M
3 'S c
z £« fSe rt ° r3
Z
H
Pi Bù 5
-J o a xr¡ «
W u
ô T3
° § a «i M on
"
o
O so
00
a,
i?
«
>• pî ra ,o «8
H vi « a S
J I S o 35
n « s ! O,
fí ci -C I 2 a
£ o. S •a s g I o
« s? .s - « o
>3 .g j¡
O « g S 1es ta5
so' o „ •§ «s 5 S 5 i-l .S
<0
n¡ •o
ss
»
0
s. «o "

S £ 'O ts ""Ci m
-o
«a •s a g •a o
O rs oo os "S "S 5
cj oo o so fe «¡o «
« » ¿ N W S 2 s
SO § c
O 1 S •2
çua a
•'S w
° s ^ O
o W
s
o
Ifi W) « J3
ai u • ,
C4 <-« © - g s CO >>
a
>,
« t^. zn Svv
O « N
s s
U g <0
•a
*
J3

ja
"

£a <u
o,
3 00
•o a
oo eo
s
e« u o - . u
CS <u
•a -"S o •o
h lO N N OJ <u ¡§2,2
M 'o 'o 'S 'o ctu
^ SO en © £ 'S 'S 'H 'S
M VI •s
tj- r- vi ri 8 C .3 s
v> =3 w
Vi
M «S w Z& PQ8 M « m
S pu w PQ pCqD PQ PalQ f-E1C

"3
O Ü O O O U U Ü U Ü Ü
N m ^ "n se oo os o •-<
Ë r

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
502 STUDIES IN VOLCANOLOGY

ajiaijaqdau NO 00 OS NO NO. « o 00 ri r*
siüIiaH no © vi oo ri ci rt ,-i • r i ©
Ci r—< I rt
Tf(SVi©i-<C>t~r)CIOOM

•8
3};ui[aq<Ì3(si n' © o
I
13
•& SJISOUIIJAJ «Tf © vi rtoo m ri ri © ri © ©
u « rt
%

h i>. \e h pi e e o « « N
sjiiressg ^ - r i c i o i ^ - i o c i i - i r i o ©
^ I RH I—I

(piojnreseq)
o o i - ^ r j T t ^ T t i t ^ o s c s vir)
IJBSBq 5 N m a Ì H H N O N Ò o
3UIAI[0 OIJEiflY ^T 1—I T-l 1-H
© ci vi -fr r) rt r-J i/i ri
aiAqoBjj Bpog ©' ri (-' Tf ©' ©' ©'

so oo © so v> os oo r< t-, ri


sjisaouiusg V)

VO ONeS^HCnT-J-tt^HTJlT-JiS
sjueaSnj^ s—I vo' tí vo ci vo' vi ri ri ri ©
v>

o\ o \ o \ v o oo©rjviTj;t~<s
ajIIBMBH v i ^ r - ' - ^ - o Ö T t r t ' m o ö
1
JIBseq VO Vi H oo Ci oo t r)
ouAqd-j^dspiaj VO* ••fr oo vi as Ö Ö ©

}[BSBq •H r) Vi o © O <t N
3UIAIJ0 o i j B 3 [ j y
Vi rt Tf OÌ © ci T^ fi © ©
•t

•-< -h r) © vi on r- ri
N m ci ri ri © ri

3)U3[Oq) SUIAI[0 © © Vi Tj- r) Vi o© ©


r)
pus ajjpioqx O o\ ci 00 00 ri ri
n •fr
o
•B
O Vi vi CI © M r)
3JIUB330 NO oo ri © © ri
S n


M
fi Ir V
O
il o" ® o o o $ o
N
O O
«0
S

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 503

• «-< t-; oo en © es © vo vi en
od oó "i oo r>i v¿ t h »i "i PÍ

• <s <s f- « © oo en © • Vl eS ©
; ri « »i « « 6 -H ; » « N

• ve en r» «n c> oo o » P; r^ « e
; ©' t-' >n K m ci ri ci -i t-' vi ri

« rt t « * * 00 M <-; vo t»
irt N n -I t-' en ri •<t m oo' vi

« © vo in 9¡ vi n ©
en m © m
v¿ <•* «ri © m 0\ m vi vi «

\o m i-. e> • os es t-- in oo í t pi


vi ci OÑ vo ; e«i OÑ en en vi ; ->f <-î

C> t- >e PI
©
m
es
N
vo
© o o

r-; 00 00 © m r-; es
« « ri
f—t « , . , ** ci •-< ©'

es sO 00 os >n • r-- • r* • . os sq c-_


ri <ñ vi ri ri : ri ; ri ; ; h <t m
* —
t1

Os < © m en <s so >n es M >n f-


oo e«ñ © © V f» vi —t so «
n «s

es o\ © © so t oo en >n
« «es ©«
t»' ori SO ->fr •«f -J -rt ci
es es

« * VO «o Os 00 Tj- © oo O
so' ci
« ri Ci vô oo -fr SO vi *-¡

Oi w « » vo q es © es r»
« ri N N m rt SO so vi o
«i es

es es en © vo os ve •«»•
ri ri oo I— Os vrf ri it s¿ •>r "t ©
« es

r-_ »-< se c4 es m en
« m lei « ri M m en so o'
«S

» S js S & 5 £

O S £ M "3 8 S a §" o

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
504 STUDIES IN VOLCANOLOGY

they are not richer in alkalies. The presence of nepheline in the norm is due
to factors other than an increase of alkalies.
Titanium increases with the alkalies from alkalic olivine basalt to hawaiite,
but then decreases greatly in the mugearites and trachytes (Fig. 2). Its be-
havior is independent of that of silica, as shown by its increase in the neph-
elinic rocks. In all the rock types, except benmoreite (transitional mugearite-
trachyte), trachyte, and rhyodacite, including the tholeiites, titania is nearly
always present in excess of 1.75 percent, and thus conforms with Chayes's
(1964) generalization for basaltic rocks of oceanic regions.
Phosphorous behaves somewhat like titanium and increases from alkalic
olivine basalt to the nephelinic rocks on one hand and the hawaiites and
mugearites on the other, but again decreases in the trachytes. However, it
reaches its peak abundance in the mugearite group, at about 5 percent soda
(Fig. 3), whereas the maximum for titania falls at about 3.7 percent soda.
It will be noted in Figures 2 and 3 that there is a considerable scattering of
the points that represent abundance of both phosphorous and titania, as
related to soda.
Titanium tends to be concentrated in the pyroxenes, and its sparsity in the
felsic rock types may be attributed to removal of pyroxene, in the formation
of the felsic magmas. Crystals of gray to brownish-gray apatite, up to 0.1 mm
thick and 0.5 mm long, commonly are present in the mugearites, and the

Figure 2. Abundance of titania in rela- Figure 3. Abundance of P,0 5 in relation


tion to soda in Hawaiian rocks. The letter to N a , 0 in Hawaiian rocks. The letter E
E indicates the "eclogite" (garnet pyroxen- indicates the "eclogite" of Salt Lake Crater;
ite) of Salt Lake Crater; X's represent the X's represent the low-K tholeiites of the
low-potassium tholeiites of the deep Pacific deep Pacific basin (Engel and others,
basin (Engel and others, 1965). 1965).

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 505

abrupt decrease of phosphorous in the more felsic rocks probably results


from crystallization and separation of apatite at that stage. It is interesting
to note that very similar large, gray apatite crystals are present in the lavas
of the pre-Etna Trifoglietto Volcano on Sicily.
About 170 unpublished determinations of strontium in Hawaiian rocks by
N. J. Hubbard, together with about 30 others by Wagner and Mitchell (1953),
Turekian and Kulp (1956), Lessing and Catanzaro (1964), and Hamilton
(1965), demonstrate that strontium behaves much like phosphorous in Ha-
waiian rocks. It increases in abundance with the alkalies in the nephelinic
rocks, and in the alkalic rocks as far as the mugearites, but decreases again
(although the alkalies keep on increasing) in the benmoreites, and in the
extreme trachytes it reaches a level even lower than the strontium in the
picrite-basalts. At least part of the decrease in strontium in the most felsic
rocks probably results from separation of strontian apatite at the mugearite
stage, as does the decrease in phosphorous.
In contrast, both rubidium and zirconium appear to continue to increase
with the alkalies throughout the entire series of Hawaiian rocks. Larsen and
Gottfried (1960) find that barium also is more abundant in the alkalic rocks
than in the tholeiitic rocks. This is supported by the unpublished work of
Hubbard, which also confirms earlier indications that nickel is notably less
abundant in the alkalic rocks than in the tholeiitic ones and shows a strong
correlation in behavior with magnesium. Work by Schilling and Winchester
(1966) suggests that the rare-earth elements also are more abundant in the
alkalic basalts than in the tholeiites.
Lead, uranium, and thorium appear to increase in abundance with the
alkalies in the alkalic suite, and even more in the nephelinic suite, and reach
their maximum abundance in the nephelinites (Larsen and Gottfried, 1960;
Tatsumoto, 1966b).
In the alkali:silica diagram (Fig. 1) are plotted all the published analyses
of Hawaiian volcanic rocks, except for a few that have been shown to be
erroneous. The division between the tholeiitic and alkalic suites is conspicu-
ous, but it is also evident that there is a gradation in composition from olivine
tholeiites to alkalic olivine basalts in terms of alkali:silica ratio. The few rocks
of the alkalic suite that fall within the tholeiitic field are all ankaramites —
cumulate rocks, in which there has been a concentration of the mafic minerals
that are poor in alkalies. The garnet pyroxenite ("eclogite") inclusions from
Salt Lake Crater, Oahu, are distinctly lower in alkalies than are tholeiitic
lavas of the same silica content. On the other hand, the low-potassium tholei-
ites of the deep Pacific basin (Engel, Engel, and Havens, 1965, p. 720) plot
within the main group of tholeiites. Though their potassium content is less
than that of most Hawaiian tholeiites, this is compensated in the diagram by
the fact that their sodium content is somewhat greater than that of most
Hawaiian tholeiites of similar silica content.
In Hawaiian rocks, potassium and sodium increase together in a general
way, but the relative increase of potassium is much more rapid. From the

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
506 STUDIES IN VOLCANOLOGY

alkali-poor tholeiites of Kohala volcano to the mugearites, the sodium content


shows approximately a four-fold increase, whereas the potassium increases
approximately twenty-fold. Similarly, from average tholeiite to average alkalic
olivine basalt (Table 8), soda increases less than 1.4 times, whereas potassa
increases 2.5 times.
The field of the nephelinic rocks in Figure 1 is, for the most part, distinct
from the fields of the other suites, but merges with that of the alkalic suite.
Some alkalic basalts of the two suites are closely similar in chemical com-
position, although those of the nephelinic suite tend to be more undersaturated
in silica, as expressed by a greater content of normative nepheline.
Figure 4 is an AFM diagram, in which are plotted all unsuperseded analyses
of Hawaiian rocks. Like the similar, though less complete, diagrams pub-

6\

£ f „ %^
0

i\l'r %

Figure 4. AFM diagram of Hawaiian volcanic rocks. Solid dots, tholeiitic rocks;
large solid circle, average lava of the first phase of the 1959 eruption of Kilauea (Murata
and Richter, 1966); open circles, alkalic rocks; crosses, nephelinic rocks; V, iron-enriched
veinlet in tholeiite, Kilauea (Kuno and others, 1957); G, granophyre associated with
quartz dolerite, Palolo Quarry, Honolulu (Kuno and others, 1957); R, rhyodacite,
Waianae Range, Oahu (Macdonald and Katsura, 1964); T, trachyte; E, "eclogite" of
Salt Lake Crater; O, olivine phenocrysts in tholeiitic lava of the 1959 eruption of Kilauea
(Macdonald and Katsura, 1961). In the AFM diagram, A = N a , 0 + K 2 0, F = total
iron +MnO, and M = MgO.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 507

lished previously (Macdonald and Katsura, 1964, p. 107, 109), it indicates


a continuous and well-defined trend from olivine tholeiites through tholeiites
and rocks of the alkalic suite to soda trachyte that closely resembles the
trends of such differentiated masses as the Skaergaard, Dillsburg, and Tas-
manian dolerite intrusions. The lesser silica saturation of the Hawaiian trach-
ytes, as compared with the felsic differentiates of the above-mentioned bodies,
is not apparent on this type of diagram.
Figure 5 is a von Wolff diagram of all Hawaiian rocks. It shows more
clearly than the AFM diagram the separation into a tholeiitic trend that ex-
tends from oceanite to rhyodacite, and an alkalic trend from ankaramite to
soda trachyte (Macdonald and Katsura, 1964, p. 105-110). The two trends
intersect in the region of the oceanites. The rocks of the posterosional, neph-
elinic suite form a separate group that merges with the alkalic suite and that
shows two sharply divergent trends. One trend, which includes the alkalic

Figure 5. Von Wolff diagram of all Hawaiian volcanic rocks. Solid dots, tholeiitic
rocks; open circles, alkalic rocks; crosses, nephelinic (posterosional) rocks; G, granophyre
associated with quartz dolerite; P, pegmatoid associated with quartz dolerite (Kuno and
others, 1957); R, rhyodacite; T, trachyte; O, oceanite; N, melilite nephelinite. The large
solid circle represents the weighted average of the first phase of the 1959 eruption. In the
Von Wolff diagram, L = feldspar, Q = quartz, and M = mafic minerals.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
508 STUDIES IN VOLCANOLOGY

olivine basalts, basanitoids, and basanites, is close to and nearly parallel to


that of the main alkalic suite, but is less extensive. The other trend, which
represents the nephelinites, melilite nephelinites, and picrite-basalts of mimo-
site type (Macdonald, 1949a, p. 1548), projects from the region of the
oceanites in the general direction of the alkali corner of the diagram, repre-
senting a combined increase in alkali content and decrease in silication, with
relatively little change in the total content of mafic materials.
The low-K oceanic tholeiites of Engel, Engel, and Havens (1965), briefly
mentioned above, do not appear to fall on the variational trend of the Ha-
waiian lavas (Figs. 2, 3, and 6). Although some Hawaiian tholeiites, notably
• the lowest exposed lavas of the Kohala volcano, are similarly poor in potassium
(Macdonald and Katsura, 1964, p. 117), the Hawaiian rocks are notably

Figure 6. Von Wolff diagram showing variational trend lines of the Hawaiian rock
suites, as indicated in Fig. 5. The dashed line encloses the field of the oceanites. K,
weighted average of the first phase of the 1959 eruption of Kilauea (Murata and Richter,
1966); E, "eclogites" from Salt Lake Crater; AP, average peridotite (Nockolds, 1954);
P, pyrolite (Green and Ringwood, 1963); OL, olivine; 1, clinopyroxene from Salt Lake
"eclogite" (Yoder and Tilley, 1962); 2, augite from ankaramite of Haleakala (Washington
and Merwin, 1922); 3, hypersthene from Salt Lake "eclogite" (Yoder and Tilley, 1962);
4, average of four tholeiitic clinopyroxenes; 5, hypersthene from Uwekahuna tholeiitic
gabhro-picrite (Richter and Murata, 1961); 6, arbitrary mixture of olivine and Haleakala
augite. In the Von Wolff diagram, L = feldspar, Q = quartz, and M = mafic minerals.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 509

lower in alumina than the deep-ocean rocks and are somewhat lower in soda
at the same percentage of silica. There appears to be no convincing reason to
consider the low-K deep-ocean tholeiites to be ancestral to the Hawaiian
tholeiites.

ORIGIN AND DIVERSIFICATION OF


HAWAIIAN MAGMAS
The Parent Magma

There is general agreement that the tholeiitic rocks of Hawaii are derived
from magma generated in the upper mantle, probably at a depth of about
60 km (Eaton and Murata, 1960). Also, it has generally been considered
that the primary tholeiitic magma is essentially saturated in silica, because the
great majority of the rocks exposed above sea level are of that sort (Powers,
1955). Olivine phenocrysts are common in them, but are the result of early
crystallization of olivine, often greatly in excess of its stoichiometric ratio (the
so-called "Bowen-Anderson effect"), which has failed to react with the sur-
rounding liquid and become transformed into pyroxene before the consolida-
tion of the magma. If reaction were carried to completion, most of the visible
tholeiitic rocks of Hawaii would be olivine-free.
However, there is some evidence that suggests that the majority of the
visible tholeiitic lavas may not be representative of the parent tholeiitic magma,
as it originates in the mantle. Recently, there have been recognized masses
of very dense material with high-seismic velocities beneath the summit regions
of most of the Hawaiian volcanoes (Kinoshita, 1965; Kinoshita and Okamura,
1965; Adams and Furumoto, 1965). The nature of these masses is still un-
certain, but the most likely possibility is that they are bodies of cumulate rock
that was formed by the lagging behind of heavy phenocrysts in the rising
magma. In the volcanoes that have not progressed beyond the tholeiitic stage,
such cumulates should consist very largely of olivine. Murata and Richter
(1966, p. 8) have considered that the most olivine-rich lavas of the 1959
eruption of Kilauea resulted from the entrainment of olivine grains that were
eroded by unusually rapid-flowing magma from a cumulate mass within the
magma chamber of the volcano.
The mass of dense rock beneath the Koolau volcano of Oahu is about
6 km across and extends to an unknown depth, possibly all the way to the
mantle. Addition of such a mass of olivine to the lava flows of the Koolau
volcano would very considerably reduce their degree of silica saturation.
Furthermore, other evidence suggests that the olivine-rich rock at depth is not
restricted to these conspicuously dense plugs beneath the summits, but actually
is a good deal more extensive. Zones of high gravity extend along the rift
zones of the volcanoes (Kinoshita, 1965; Strange, Machesky, and Woollard,
1965), and in the Koolau volcano, seismic velocities of 6 km/sec or more
extend to within about 1.5 km of sea level beneath the northeast rift zone,

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
510 STUDIES IN VOLCANOLOGY

as compared with velocities from 4.2 to 4.6 km/sec for most of the mass of
the volcano (Furumoto, Thompson, and Woollard, 1965; Hill, 1966). It
seems a reasonable implication that the tholeiitic magma rising from depth
may contain considerably more olivine, both as suspended phenocrysts and
in solution, than the lavas that reach the surface.
From laboratory studies of melting relationships, Green and Ringwood
(1967, p. 166-168) conclude that the liquid derived from a mantle of gen-
eral peridotite or pyrolite (dunite plus basalt) composition under pressures
from 30 to 40 kb and a high degree of partial melting would be of tholeiitic
composition with 20 percent or more normative olivine in solution. O'Hara
(1965, p. 19-27), who uses a similar approach, also concludes that the pri-
mary magma must be an undersaturated olivine tholeiite, and that the silica-
saturated common tholeiites are the result of fractionation of olivine tholeiite
magma during its rise to the surface.
Both the very high rate of discharge of lava during the 1959 eruption of
Kilauea and the unusually high temperature of the lava reaching the surface —
about 1190° to 1200° C at the highest, or from 70° to 80° higher than the
temperatures measured in most eruptions •— suggests an unusually rapid rise
of magma from depth and, correlatively, a lesser than usual opportunity for
fractionation en route. The 1959 lava in Kilauea Iki Crater can be regarded
as closer than most of the surficial lavas to the composition of the primitive
magma at depth. Macdonald and Katsura (1961) assumed the average of
two analyses of the predominant olivine-rich lava of that eruption as the
probable composition of the parent magma. It was stated to contain about
18 percent normative olivine; but, as pointed out by Green and Ringwood
(1967, p. 110), the average value of alumina was wrongly reported as 12.9
instead of 12.1 percent, and a recalculated value for normative olivine is
16.7 percent. Very similar is a weighted average composition for the lava of
the first phase of the eruption that was published by Murata and Richter
(1966, p. 26). It contains 17 percent normative olivine. Both these averages
have been used by Green and Ringwood (1964, 1967) as the approximate
composition of the parent magma of the Hawaiian province. However, the
magma reaching the surface already contained phenocrysts of olivine, and
it is not unlikely that most of the erupting magma had already lost some
olivine during its rise from depth. The magma that was discharged most
rapidly and at the highest temperature (close to 1200° C) contained the
greatest amount of olivine, which reached from 27 to 30 percent.
The amount of olivine that can be contained in solution in a magma de-
pends, of course, on the temperature of the magma. Completely glassy Pele's
hair of the 1959 eruption contains 6.5 percent of normative olivine (Mac-
donald and Katsura, 1961, p. 362). Drever and Johnston (1958, p. 493,
495) regard rocks that contain up to 21 percent normative olivine as repre-
senting magmas that were once wholly liquid, and other rocks that contain
from 34 and 35 percent norm olivine they regard as probably once wholly
liquid. Richter and Murata (1966, p. 6), in discussing the rocks of the 1959

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 511

eruption of Kilauea, state: "Sample S-5, a pumice erupted on November 18,


contains approximately 30 percent large olivine phenocrysts, as much as 4 mm
in diameter, in the mode. The olivine phenocrysts are mostly euhedral; a few
exhibit incipient rounding of the crystal form." This description suggests
strongly that the olivine had crystallized from the enclosing liquid at slightly
higher temperature and was only beginning to react with the enclosing melt
at the time of eruption. Although some olivine may have been added from
higher levels by gravitative settling, it appears probable that the bulk com-
position of the rock represents fairly closely the composition of a wholly
liquid magma, and that tholeiitic magmas can contain at least 30 percent
olivine in solution.
Still another line of evidence suggests a parent magma that is rich in olivine.
In Figures 5 and 6, the variational trend lines of all of the Hawaiian suites
converge in the region of oceanite. This could be explained as the result of
formation of olivine-rich cumulates in the tholeiitic and alkalic suites, in both
of which the crystallization and movement of olivine appear to be very im-
portant factors in producing the observed compositional variations. However,
the plotted position of olivine lies far from the projection of the variational
trend of the nephelinites and melilite nephelinites, which nevertheless inter-
sects the other trends in the oceanite field. Judged wholly on the basis of the
trend lines in Figures 5 and 6, oceanite is the most likely single parent.
Thus, various lines of evidence seem to indicate that the primitive magma
of Hawaiian volcanism, derived by melting in the upper mantle, probably
has the chemical composition of olivine tholeiite that contains at least 20
percent of normative olivine in solution. Such magmas probably had tempera-
tures in excess of 1350° C (Tilley, Yoder, and Schairer, 1965, Fig. 1) at
depth, but as they rise, they soon cool to the crystallization temperature of
olivine, and phenocrysts of that mineral start to form. These phenocrysts may
settle faster than the rate of rise of the enclosing magma, which appears
usually to be quite slow from the zone of magma generation to the shallow
reservoir. During its usual relatively long pause in the shallow reservoir, the
extraction of most of the olivine is completed, so that the magmas reaching
the surface ordinarily are much depleted in olivine, and separation of olivine
in excess of its stoichiometric proportion commonly produces oversaturated,
quartz-normative tholeiites.
Other lines of evidence indicate that the parental magmas of the various
Hawaiian volcanoes are not completely uniform. The tholeiitic basalts vary
in composition from one volcanic center to another (Macdonald and Katsura,
1964, p. 103, 111). Differences between volcanoes also exist in the abun-
dance of minor elements, such as the isotopes of strontium (Powell and
DeLong, 1966; confirmed by unpublished work by N. J. Hubbard), and lead
(Tatsumoto, 1966a). These differences have been explained as due to minor
variations in the composition of the mantle in the source region of the magmas,
but Green and Ringwood (1967, p. 175-179) suggest instead, a process of
variable enrichment in certain "incompatible elements" that are derived by

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
512 STUDIES IN VOLCANOLOGY

extraction of the lowest melting components from the mantle wall rocks dur-
ing the slow ascent of the magma at intermediate depths.

Origin of the Alkalic and Nephelinic Suites


The time relationships of lava types in Hawaii strongly indicate derivation
of the alkalic suite from a tholeiitic parent. In 1935, H. A. Powers referred
to the rocks now known as the alkalic suite as "differentiated," and to those
of the present tholeiitic suite as "primitive." Powers (1935, 1955) also pointed
out, however, that an undersaturated magma, such as alkalic olivine basalt,
cannot be derived from a saturated magma, such as tholeiitic basalt, by any
probable process of crystal differentiation in a low-pressure environment,
because none of the minerals likely to separate from the magma contain more
silica than the magma.
Because of the apparent difficulties in deriving the alkalic rocks from
tholeiitic magma that is essentially saturated with silica, Kuno and his co-
workers (1957) and Yoder and Tilley (1962) assumed two independent
magmas, that formed by melting at different depths in the mantle, to be the
parents respectively of the Hawaiian tholeiitic and alkalic (including neph-
elinic) suites. In contrast, the volume and spacial relationships, together with
the chemical intergradation, have led most other workers in Hawaii to believe
that the alkalic rocks have been derived in some way from a tholeiitic parent
(Macdonald and Katsura, 1962, p. 194). Current work on lead isotopes by
M. Tatsumoto (personal commun., October, 1966) also indicates that the
Hawaiian tholeiitic and alkalic rocks come from a common source, whether
by fractional crystallization or by partial melting of similar mantle material.
In 1949, I attempted to demonstrate that Hawaiian alkalic rocks that range
in composition from hawaiite (andesine andesite) to trachyte could be de-
rived by crystal differentiation of a slightly undersaturated olivine basalt
magma having the average composition of all available analyses of Hawaiian
olivine-bearing basalt, including those rocks that were not normatively under-
saturated. This "average olivine basalt" magma was not labeled as tholeiitic
at that time, but although some alkalic olivine basalts were included in the
average, the great preponderance of analyses of tholeiitic rocks gave it a
distinctly tholeiitic character. Consequently, the calculations of the deriva-
tion of other rocks from that parent are applicable, with only slight modifica-
tions, to the derivation of the rocks from a purely olivine tholeiite parent.
The minerals that theoretically must be separated from the "average olivine
basalt" parent in order to leave rest magmas of hawaiitic to trachytic composi-
tion included a large proportion of hypersthene, which may have formed as
a result of a high-pressure environment (Larsen, 1940, p. 925; Macdonald,
1949a, p. 1575). Subtraction of material that contained a large proportion
of orthopyroxene from an ultrabasic magma also was suggested as a mechan-
ism for the derivation of the nephelinites and melilite nephelinites (Macdonald,
1949a, p. 1580-1584).

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 513

More recently, Murata (1960) has attempted to demonstrate the derivation


of alkalic basalts from undersaturated tholeiitic magma by crystal differentia-
tion that is controlled by the separation of clinopyroxene, though Yoder and
Tilley (1962, p. 416) consider that the results are not closely similar to
alkalic rocks.
Laboratory investigations at low pressures indicated the presence of a
thermal barrier which would prevent a magma from moving from the tholei-
itic to the alkalic field by crystal differentiation (Yoder and Tilley, 1962, p.
401). For a time it appeared that, despite such objections, small amounts of
weakly alkalic basalt actually had formed from tholeiitic magma during the
consolidation of the lava lake formed in Kilauea Iki Crater in 1959 (Mac-
donald and Katsura, 1961; Richter and Moore, 1966, p. 25); but reanalysis
of some of the samples suggests that the original analyses were in error (Peck,
D. L., personal commun., Nov. 14, 1966), and the formation of alkalic basalt
in the lava lake must now be regarded as very uncertain, if not disproven.
However, more recent laboratory work has shown that at higher pressures
the thermal barrier does not exist, and that at depth, crystal differentiation
can produce alkalic basalt magma from undersaturated tholeiitic magma.
Tilley, Yoder, and Schairer (1965, p. 79), Green and Ringwood (1964,
1967), and O'Hara (1965) have all demonstrated that alkalic olivine basalts
can be derived from olivine tholeiite that approaches oceanite in composition
by fractional crystallization that involves the separation of pyroxene (espe-
cially orthopyroxene) at moderately high pressures. This does not exclude
the possible independent origin of the two magma types by direct partial
melting of a peridotite or pyrolite mantle at different depths (Kuno and others,
1957; Yoder and Tilley, 1962), and indeed the latter probably is the origin
of the abundant alkalic olivine basalts in regions, such as the Hebridean
province, where alkalic olivine basalts equal or even greatly exceed the asso-
ciated olivine tholeiites in volume (Green and Ringwood, 1967, p. 170).
However, in provinces, such as the Hawaiian, where the rocks of the alkalic
suite are present in relatively small volume, their derivation from olivine
tholeiite magma appears the more probable. The evidence for the possibility
of pressure-controlled fractionation of olivine tholeiite to alkalic olivine basalt
has been discussed by the above-mentioned writers, and is not repeated here.
O'Hara (1965) has shown that all three suites — tholeiitic, alkalic, and
nephelinic — can be derived from a single parent magma of olivine tholeiite
composition by crystal differentiation. The separated minerals and the com-
position of the remaining liquid are governed by the pressure in the region
where the process is taking place. Differentiation of a relatively dry melt under
high pressure leads, according to O'Hara, to a potash-rich mafic magma that
may be further fractionated into leucitites and related rocks, none of which
are found in Hawaii. Under somewhat lower pressure, fractional crystalliza-
tion of the same original magma gives rise to a silica-poor liquid that, in turn,
may fractionate into nephelinite, melilite nephelinite, and related magmas.
Under lower pressure the same parent yields a submagma that gives rise to

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
514 STUDIES IN VOLCANOLOGY

alkalic olivine basalt magma, which in turn may yield the related rocks of the
Hawaiian alkalic suite. Finally, under low pressure that corresponds essen-
tially to surface conditions, the undersaturated olivine tholeiite parent pro-
duces silica-saturated tholeiitic basalts.
It is interesting to note that the fractionation trend found by Green and
Ringwood (1967, Fig. 8) for olivine tholeiite that contains about 20 percent
olivine (SiOo 46.95 percent, total alkalies 1.81 percent) under pressures of
13 to 18 kb, that correspond approximately to depths of 40 to 60 km, yields
a liquid product containing from 45.7 to 45.9 percent SiO, and 2.1 percent
alkalies. If further fractionation of this product followed the same trend or
one parallel to that of fractionation of the olivine basalt (OB, Fig. 7), it would
yield a liquid that resembles in alkali and silica content the junction area of the
Hawaiian alkalic and nephelinic suites. The fractionation trend of alkalic
olivine basalt found by Green and Ringwood (1967, Fig. 8) under similar
pressures (from 13 to 18 kb) is almost exactly parallel to the variation trend
of the Hawaiian nephelinic rocks, though the starting point is closer to the
silica-rich edge of the diagram (Fig. 7).

Figure 7. Alkali:silica diagram showing variational trends of Hawaiian rock suites


and fractionation trends found in the laboratory by Green and Ringwood (1967, Fig. 8).
The experimental rock systems of Green and Ringwood are indicated as follows: OT,
olivine tholeiite; AB, alkali olivine basalt; OB, olivine basalt (mixture of OT and AB);
the letters H, M, and L indicate the fractionation trends found respectively at high (13-18
kb), medium (9 kb) and low (atmospheric) pressures. KI indicates the cross-trend found
in the Kilauea Iki lava lake (Macdonald and Katsura, 1961), and P the similar cross-
trend in the Paiolo quartz diabase, both of which must have developed at low pressure.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 515

The low-pressure fractionation trend of the olivine tholeiite is more or less


parallel to the variation trend of the tholeiitic suite (Fig. 7), and the trend
of alkalic olivine basalt is almost exactly parallel to the variation trend of the
alkalic suite, suggesting that although the alkalic basalt magma was derived
from olivine tholeiite at fairly high pressures, differentiation of both the tho-
leiitic and alkalic suites took place under lower pressures at depths distinctly
less than 30 km.
The sequence of depths, from shallow for tholeiitic derivatives through
intermediate for the alkalic basalt magma and deep for the nephelinic magma,
is in agreement with the evidence of depth of origin of the ultrabasic inclu-
sions in the lavas, discussed on an earlier page. It also is harmonious with
what is known of the structure of Hawaiian volcanoes. The mechanism by
which the three principal Hawaiian rock suites are derived from olivine tholei-
ite magma may be somewhat as follows:
A magma reservoir at a depth of only 2 or 3 km has been demonstrated
for Kilauea (Eaton and Murata, 1960) and is indicated for Mauna Loa by
similar patterns of surface tilting that are associated with eruptions of that
volcano. By implication, shallow chambers are believed to have existed in
the other Hawaiian shield volcanoes, all of which, like Kilauea and Mauna
Loa, erupted tholeiitic lavas. The maintenance in a fluid condition of this
shallow chamber, which formed by melting within the base of the cone itself,
must demand frequent eruption, with a high rate of heat transport toward
the surface. As the frequency of eruption decreases in later stages (demon-
strated by weathered and eroded interflow surfaces), the shallow magma body
must begin to congeal from its top downward. Residual pockets of magma
may remain in its upper part for a time, but the level of the main mass of
magma, in which crystallization is taking place, becomes steadily deeper,
and the increased pressure leads to the separation of pyroxene in place of
olivine and the formation of alkalic olivine basalt magma. For a brief time,
tholeiitic magma from shallow pockets and alkalic magma from a deeper level
are erupted alternately, but soon the upper magma chamber becomes com-
pletely consolidated, and throughout the remaining period of volcanism,
alkalic lavas alone are erupted. To a considerable extent, however, the erup-
tive fissures are still governed by the rift structures that are established dur-
ing the main, tholeiitic stage of activity. If this scheme is correct, the large,
shallow magma chamber of the tholeiitic stage is absent in the later, alkalic
stage, though smaller shallow chambers filled with alkalic magma may exist.
In this connection it is interesting to note Rittmann's (1963, p. 798) con-
clusion that Etna, a volcano remarkably like Mauna Kea and other late-stage
Hawaiian volcanoes, both in its general form and in the composition of its
lavas, lacks any shallow magma chamber.
The final resurgence of volcanism, which produces nephelinic magmas
after a long period of volcanic quiescence and deep erosion, comes from still
greater depth. Not only are the pyroxenes of the inclusions brought up by
the magma aluminous, but also the pressure at the depth of origin of the

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
516 STUDIES IN VOLCANOLOGY

magma must have exceeded that of the gabbro-eclogite transition, at least


locally. The eruptive fissures bear little or no relationship to the rift zones
of the main period of volcanism and often cut sharply across them. Indeed,
they appear more likely to be related to fundamental crustal structures of the
Pacific Basin and to be essentially of tectonic origin.
Although the origin of the parent magmas of the alkalic and nephelinic
suites can be explained by crystal fractionation of a primary olivine tholeiite
parent, this explanation does not prove that such was the case, or that no
other processes have contributed to the change. Indeed, the relative rates of
increase of some minor elements in the alkalic rocks relative to the tholeiitic
rocks, and within the alkalic suite itself, appear to be incompatible with an
evolution of the rock series wholly by crystal differentiation. As already noted,
Green and Ringwood (1967) have appealed to absorption of minor elements
from the wall rocks along the path of ascent of the magma. The same process
might increase the content of other low-melting components, such as alkalies
and particularly potassium. Thermal diffusion and migration of volatile com-
pounds (not necessarily as a gas phase) are processes that must be going on
in magmas. Their effects are difficult to demonstrate and may be generally
unimportant, but until it is shown that they actually are unimportant, these
processes should not be ignored. They may play an important part in increas-
ing the amount of alkalies and some of the minor elements in the alkalic
magmas.

Diversification within Suites


All recent workers are agreed that the major variations within the tholeiitic
and alkalic suites in Hawaii are largely the result of crystal differentiation.
This is not to say, of course, that no other processes have any effect, but
simply that the movement of crystals in the magma is the dominant control.
The variations in the more mafic part of the tholeiitic suite are largely the
result of removal or addition of olivine, although movement of a small amount
of pyroxene also seems commonly to be involved (Macdonald, 1949a, p.
1576; Muir and Tilley, 1957, p. 253). In the more felsic part of the suite,
the effects of crystallization of pyroxene and feldspar become more important
(Macdonald and Katsura, 1964, p. 105, fig. 4). In the alkalic suite also, the
variations appear to be controlled by separation of pyroxene, olivine, and
feldspar. In Figure 6 the trend line for the alkalic suite projects through the
oceanite field directly toward an arbitrary mixture of olivine and clinopyroxene,
such as is found in the phenocrysts of the ankaramites.
Table 9, columns 1 and 2, gives the average composition of hawaiite as
compared to that of a theoretical rock obtained by subtracting augite of the
composition found in ankaramite of Haleakala (Washington and Merwin,
1922), olivine, calcic plagioclase, and magnetite from the average alkalic
olivine basalt magma in the proportions that are indicated in the explanation
of the table. The close agreement indicates that it is possible to derive ha-
waiite magma from alkalic olivine basalt magma by crystal differentiation.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 517

TABLE 9 . DERIVATION OF OTHER ALKALIC ROCK TYPES FROM ALKALIC OLIVINE BASALT

1 2 3 4

Si0 2 48.5 48.6 44.1 44.4


A1 2 0 3 16.0 16.0 12.1 12.3
Fe2Os \
12.1 12.1 12.5 13.2
FeO J
MgO 4.8 4.8 13.0 13.1
CaO 8.1 7.9 11.5 11.8
Na 2 0 4.2 4.0 1.9 1.8
K2O 1.5 1.4 .7 .5
Ti02 3.4 4.1 2.7 1.9
P2Os .7 .6 .3 .2
MnO .2 .3 .2 .1

1. Average hawaiite, recalculated to 100%.


2. Average alkalic olivine basalt, minus 18% augite, 7% olivine (Fo 70 ), 8% bytownite
(An 80 ), and 2% magnetite, recalculated to 100%.
3. Average ankaramite, recalculated to 100%.
4. Average alkalic olivine basalt, 50%, plus 23% augite, 15% olivine (Fo 70 ), 10%
bytownite (An 80 ), and 2% magnetite.

The same sort of calculations can be made for mugearite and trachyte with
similar results.
In the same table, columns 3 and 4 give the average composition of anka-
ramite compared with that of a rock derived by combining average alkalic oliv-
ine basalt with the same minerals removed in the formation of hawaiite, again
in the proportions indicated. The agreement between the theoretical and actual
compositions is not perfect, but is close enough to indicate that the anka-
ramites probably are essentially simple cumulate rocks. The titania content,
which is somewhat too high in the calculated hawaiite and too low in the cal-
culated ankaramite, suggests that the augite involved in the differentiation is
actually somewhat higher in titania than is indicated by the Haleakala analysis.
I have formerly suggested (Macdonald, 1949a, p. 1579) that the assimilation
of a small amount of limestone was involved in the formation of the ankara-
mites. Such assimilation certainly would make the thoretical derivation of these
rocks easier, but there is absolutely no other evidence to support the sugges-
tion, and it appears very unlikely that sedimentary limestone is available to be
assimilated. The possibility of assimilation of igneous carbonatites at depth
cannot be ruled out, but there is no independent evidence of the existence of
carbonatites in Hawaii. The calculations in Table 9 seem to indicate that
assimilation of any sort of limestone is unnecessary to derive the ankaramites.
Within the nephelinic suite there appear to be two distinct lines of variation
(Figs. 5 and 6). One trend, close to and parallel to that of the main alkalic

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
518 STUDIES IN VOLCANOLOGY

suite, is probably the result of movement of crystals of pyroxene and olivine,


and perhaps some feldspar, in the magma. The trend resembles the low-
pressure trend found by Green and Ringwood (1967) in alkalic olivine basalt
and is probably the result of fractionation at shallow depths. It includes the
alkalic olivine basalts of the nephelinic suite and also the basanitoids and
basanites. The other trend leads to the nephelinites and melilite nephelinites.
In Figures 5 and 6 it extends toward the bottom edge of the diagram. Pro-
jected in the opposite direction through the field of the oceanites, this trend
intersects the plotted position of the orthopyroxenes, and it may be largely
controlled by the separation of orthopyroxene in place of olivine in a high-
pressure environment. As already pointed out in Figure 7, this trend is parallel
to that found by Green and Ringwood for alkalic olivine basalt at moderate
pressures. Thus, crystal fractionation appears adequate to explain the varia-
tions within the alkalic and nephelinic suite also, although some enrichment
in alkalies and related elements by volatile transfer or diffusion is not excluded.

CONCLUSIONS
Current field and laboratory evidence thus appear to indicate that:
(1) The Hawaiian tholeiitic, alkalic, and nephelinic suites all were derived
from a single primitive magma type, olivine-rich tholeiite, that was formed
by partial melting of an ultrabasic mantle, which probably resembled pyrolite
in composition, at a depth of about 60 km.
(2) The primitive magma varied slightly from one volcano to another,
either as a result of minor inhomogeneity in the mantle, or acquisition of
varying small amounts of material from the wall rocks during its rise.
(3) Alkalic olivine basalt magma and nephelinic magma were formed
from olivine-rich tholeiite by fractionation at depths of about 30 and 50 km
respectively.
(4) The trends of differentiation within the individual suites were con-
trolled largely or entirely by fractional crystallization, mostly at depths less
than about 15 km, although the nephelinites probably resulted from frac-
tionation at greater depths.
(5) The silica-saturated tholeiites that predominate at the surface are not
the same as the primitive magma at depth, but have been depleted of olivine
at relatively shallow depths to form the masses of very dense rock shown by
geophysical studies to exist beneath the volcanoes.
(6) Such other processes as volatile transfer and thermal or pressure-
controlled diffusion are not wholly ruled out, however, and they should be
kept in mind until the importance of their effects is more definitely known.
(7) Although field evidence favors derivation of the alkalic suite from
an olivine tholeiitic parent magma, origin of the three suites by partial melting
at different depths in the mantle still remains a possibility from, the physico-
chemical standpoint.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 519

ACKNOWLEDGMENTS
I wish to thank Gary D. Stice and Floyd W. McCoy, Jr., of the Hawaii
Institute of Geophysics, for aid in collecting the samples of the upper member
of the Waianae volcanic series, Oahu; Howard A. Powers, of the U.S. Geo-
logical Survey, for the use of an unpublished analysis of the 1801 lava of
Hualalai volcano in calculating the average composition of the alkalic olivine
basalt in Table 8 and for arranging for the comparison analyses by the U.S.
Geological Survey; and Norman J. Hubbard, of the Hawaii Institute of Geo-
physics, for the use of unpublished minor-element analyses.

REFERENCES CITED
Adams, W. M., and Furumoto, A. S., 1965, A seismic refraction study of the
Koolau volcanic plug: Pacific Sci., v. 19, p. 296-305.
Boyd, F. R., and England, J. L., 1963, Pyroxene and associated minerals in the
crust and mantle; some effects of pressure on phase relations in the system
Mg0-Al 2 0 3 -Si0 2 : Carnegie Inst. Washington Year Book 62, p. 121-124.
Chayes, F., 1963, Relative abundance of intermediate members of the oceanic
basalt-trachyte association: Jour. Geophys. Research, v. 68, p. 1519-1534.
—.—. 1964, A petrographic distinction between Cenozoic volcanics in and around
the open oceans: Jour. Geophys. Research, v. 69, p. 1573-1588.
1966, Alkaline and subalkaline basalts: Am. Jour. Sci., v. 264, p. 128-145.
Cross, W., 1915, Lavas of Hawaii and their relations: U.S. Geol. Survey, Prof.
Paper 88, 97 p.
Drever, H. I., and Johnston, R., 1958, The petrology of picritic rocks in minor
intrusions—-a Hebridean group: Royal Soc. Edinburgh Trans., v. 63, pt. 3,
p. 459-499.
Eaton, J. P., and Murata, K. J., 1960, How volcanoes grow: Science, v. 132, no.
3432, p. 925-938.
Engel, A. E. J., Engel, C. G., and Havens, R. G., 1965, Chemical characteristics
of oceanic basalts and the upper mantle: Geol. Soc. America Bull., v. 76,
p. 719-734.
Forbes, R. B., and Kuno, H., 1965, The regional petrology of peridotite in-
clusions and basaltic host rocks: Internat. Union Geol. Sci., Upper Mantle
Symposium, New Delhi, 1964, p. 161-179.
Furumoto, A. S., Thompson, N. J., and Woollard, G. P., 1965, The structure of
Koolau volcano from seismic refraction studies: Pacific Sci., v. 19, p. 306-
314.
Green, D. H., and Ringwood, A. E., 1963, Mineral assemblages in a model
mantle composition: Jour. Geophys. Research, v. 68, p. 937-945.
1964, Fractionation of basalt magmas at high pressures: Nature, v. 201,
no. 4926, p. 1276-1279.
. 1967, The genesis of basaltic magmas: Contr. Mineralogy and Petrology,
v. 15, p. 103-190.
Hamilton, E. I., 1965, Distribution of some trace elements and the isotopic
composition of strontium in Hawaiian lavas: Nature, v. 206, no. 4981, p.
251-253.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
520 STUDIES IN VOLCANOLOGY

Hess, H. H., 1960, Stillwater igneous complex, Montana: Geol. Soc. America
Mem. 80, 230 p.
Hill, D. P., 1966, Crustal structure of Hawaii from seismic-refraction studies
(abs) : Hawaiian Acad. Sci. Proc., 41st Ann. Meeting, p. 22-23.
Jackson, E. D., 1961, Primary textures and mineral associations in the ultramafic
zone of the Stillwater complex, Montana: U.S. Geol. Survey Prof. Paper
358, 106 p.
. — — 1966, "Eclogite" in Hawaiian basalts: U.S. Geol. Survey Prof. Paper 550-
D, p. 151-157.
Kinoshita, W. T., 1965, A gravity survey of the island of Hawaii: Pacific Sci.,
v. 19, p. 339-340.
Kinoshita, W. T., and Okamura, R. T., 1965, A gravity survey of the island of
Maui, Hawaii: Pacific Sci., v. 19, p. 341-342.
Kuno, H., Yamasaki, K., Iida, C., and Nagashima, K., 1957, Differentiation of
Hawaiian magmas: Japanese Jour. Geology and Geography, v. 28, p. 179-218.
Kushiro, I., and Kuno, H., 1963, Origin of primary basalt magmas and classifica-
tion of basaltic rocks: Jour. Petrology, v. 4, p. 75-89.
Kushiro, I., and Yoder, H. S., Jr., 1965, The reactions between forsterite and
anorthite at high pressures: Geophys. Lab. Ann. Rept. 1964-1965, Carnegie
Inst. Washington Year Book 64, p. 89-94.
Larsen, E. S., 1940, Pétrographie province of central Montana: Geol. Soc.
America Bull., v. 51, p. 887-948.
Larsen, E. S., 3rd, and Gottfried, D., 1960, Uranium and thorium in selected
suites of igneous rocks: Am. Jour. Sci., v. 258-A, p. 151-169.
Lessing, P., and Catanzaro, E. J., 1964, Sr S7 /Sr 8G ratios in Hawaiian lavas: Jour.
Geophys. Research, v. 69, p. 1599-1601.
Macdonald, G. A., 1940, Petrography of the Waianae Range: Hawaii Div.
Hydrography Bull. 5, p. 63-91.
1942, Petrography of Maui: Hawaii Div. of Hydrography, Bull. 7, p. 275-
334.
1949a, Hawaiian pétrographie province: Geol. Soc. America Bull., v. 60,
p. 1541-1596.
. 1949b, Petrography of the island of Hawaii: U.S. Geol. Survey Prof. Paper
214D, p. 51-96.
— — . 1960, Dissimilarity of continental and oceanic rock types: Jour. Petrology,
v. 1, p. 172-177.
1963a, Relative abundance of intermediate members of the oceanic basalt-
trachyte association — a discussion: Jour. Geophys. Research, v. 68, p.
5100-5102.
1963b, Physical properties of erupting Hawaiian magmas: Geol. Soc.
America Bull., v. 74, p. 1071-1078.
Macdonald, G. A., Davis, D. A., and Cox, D. C., 1960, Geology and ground-
water resources of the island of Kauai, Hawaii: Hawaii Div. Hydrography
Bull. 13,212 p.
Macdonald, G. A., and Katsura, T., 1961, Variations in the lava of the 1959
eruption in Kilauea Iki: Pacific Sci., v. 15, p. 358-369.
. 1962, Relationship of pétrographie suites in Hawaii: The crust of the
Pacific Basin: Am. Geophys. Union, Geophys. Mon. 6, p. 187-195.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
M A C D O N A L D — C O M P O S I T I O N AND ORIGIN OF HAWAIIAN LAVAS 521

1964, Chemical composition of Hawaiian lavas: Jour. Petrology, v. 5,


p. 82-133.
Macdonald, G. A., and Powers, H. A., 1946, Contribution to the petrography of
Haleakala volcano, Hawaii: Geol. Soc. America Bull., v. 57, p. 115-124.
1968, A further contribution to the petrology of Haleakala Volcano, Ha-
waii: Geol. Soc. America Bull., v. 79, p. 877-888.
Manghnani, M. H., and Woollard, G. P., 1965, Ultrasonic velocities and related
elastic properties of Hawaiian basaltic rocks: Pacific Sci., v. 19, p. 291-295.
Moore, J. G., 1965, Petrology of deep-sea basalt near Hawaii: Am. Jour. Sci.,
v. 263, p. 40-52.
Muir, I. D., and Tilley, C. E., 1957, Contributions to the petrology of Hawaiian
basalts, I. The picrite-basalts of Kilauea: Am. Jour. Sci., v. 255, p. 241-253.
Muir, I. D., and Tilley, C. E., 1957, Contributions to the petrology of Hawaiian
basalts, I. The picrite-basalts of Kilauea: Ah. Jour. Sci., v. 255, p. 241-253.
Murata, K. J., 1960, A new method of plotting chemical analyses of basaltic rocks:
Am. Jour. Sci., v. 258-A, p. 247-252.
Murata, K. J., and Richter, D. H., 1966, Chemistry of the lavas of the 1959-60
eruption of Kilauea Volcano, Hawaii: U.S. Geol. Survey Prof. Paper 537-A,
26 p.
Nockolds, S. R., 1954, Average chemical compositions of some igneous rocks:
Geol. Soc. America Bull., v. 65, p. 1007-1032.
O'Hara, M. J., 1965, Primary magmas and the origin of basalts: Scottish Jour.
Geol., v. 1, p. 19-40.
Oxburgh, E. R., 1964, Physical evidence for the presence of amphibole in the
upper mantle and its petrogenetic and geophysical implications: Geol. Mag.,
v. 101, p. 1-19.
Powell, J. L., and DeLong, S. E., 1966, Isotopic composition of strontium in
volcanic rocks from Oahu: Science, v. 153, no. 3741, p. 1239-1242.
Powers, H. A., 1935, Differentiation of Hawaiian lavas: Am. Jour. Sci., 5th ser.,
v. 30, p. 57-71.
1955, Composition and origin of basaltic magma of the Hawaiian Islands:
Geochim. et Cosmochim. Acta, v. 7, p. 77-107.
Richter, D. H., and Moore, J. G., 1966, Petrology of the Kilauea Iki lava lake,
Hawaii: U.S. Geol. Survey Prof. Paper 537-B, 26 p.
Richter, D. H., and Murata, K. J., 1961, Xenolithic nodules in the 1800-1801
Kaupulehu flow of Hualalai volcano: U.S. Geol. Survey Prof. Paper 424-B,
p. 215-217.
1966, Petrography of the lavas of the 1959-60 eruption of Kilauea Volcano,
Hawaii: U.S. Geol. Survey Prof. Paper 537-D, 12 p.
Rittmann, A., 1963, Vulkanismus und Tektonik des Atna: Geol. Rundschau, v. 53,
p. 788-800.
Roedder, E., 1963, Liquid carbon dioxide inclusions in ultramafic xenoliths in
Hawaiian basalts (abst): Internat. Union Geodesy and Geophysics Proc.
(Berkeley), v. 7, p. 77.
Ross, C. S., Foster, M. D., and Myers, A. T., 1954, Origin of dunites and of
olivine-rich inclusions in basaltic rocks: Am. Mineralogist, v. 39, p. 693-737.
Schairer, J. F., and Yoder, H. S., Jr., 1960, The nature of residual liquids from
crystallization, with data on the system nepheline-diopside-silica: Am. Jour.
Sci., v. 258-A, p. 273-283.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018
522 STUDIES IN VOLCANOLOGY

Schilling, J., and Winchester, J. W., 1966, Rare earths in Hawaiian basalts:
Science, v. 153, p. 867-869.
Shand, S. J., 1939, On the staining of feldspathoids, and on zonal structure in
nepheline: Am. Mineralogist, v. 24, p. 508-510.
Stearns, H. T., 1940a, Four-phase volcanism in Hawaii (abs.): Geol. Soc. America
Bull., v. 51, p. 1947-1948.
1940b, Supplement to the geology and ground-water resources of the island
of Oahu, Hawaii: Hawaii Div. Hydrography Bull. 5, p. 3-55.
1942, Origin of Haleakala Crater, Island of Maui, Hawaii: Geol. Soc.
America Bull., v. 53, p. 1-14.
1946, Geology of the Hawaiian Islands: Hawaii Div. Hydrography Bull. 8,
106 p.
Stearns, H. T., and Macdonald, G. A., 1946, Geology and ground-water resources
of the island of Hawaii: Hawaii Div. Hydrography Bull. 9, 363 p.
1947, Geology and ground-water resources of the island of Molokai, Hawaii:
Hawaii Div. Hydrography Bull. 11, 113 p.
Stearns, H. T., and Vaksvik, K. N., 1935, Geology and ground-water resources of
the island of Oahu, Hawaii: Hawaii Div. Hydrography Bull. 1, 479 p.
Strange, W. E., Machesky, L. F., and Woollard, G. P., 1965, A gravity survey of
the island of Oahu, Hawaii: Pacific Sci., v. 19, p. 350-353.
Tatsumoto, M., 1966a, Isotopic composition of lead in volcanic rocks from Hawaii,
Iwo Jima, and Japan: Jour. Geophys. Research, v. 71, p. 1721-1733.
— — . 1966b, Genetic relations of oceanic basalts as indicated by lead isotopes:
Science, v. 153, no. 3740, p. 1094-1101.
Tilley, C. E., 1950, Some aspects of magmatic evolution: Geol. Soc. London Quart.
Jour., v. 106, p. 37-61.
Tilley, C. E., Yoder, H. S. Jr., and Schairer, J. F., 1965, Melting relations of
volcanic tholeiite and alkali rock series: Geophys. Lab. Ann. Rept. 1964-1965,
Carnegie Inst. Year Book 64, p. 69-82.
Turekian, K. K., and Kulp, J. L., 1956, The geochemistry of strontium: Geochim.
et Cosmochim. Acta, v. 10, p. 245-296.
Wager, L. R., and Deer, W. A., 1939, Geological investigations in east Greenland,
Part III, The petrology of the Skaergaard intrusion, Kangerdlugssuaq, East
Greenland: Medd. om Gronland, Bd. 105, no. 4, Copenhagen, 352 p.
Wagner, L. R., and Mitchell, R. L., 1953, Trace elements in a suite of Hawaiian
lavas: Geochim. et Cosmochim. Acta, v. 3, p. 217-223.
Washington, H. S., 1923, Petrology of the Hawaiian Islands; II. Hualalai and
Mauna Loa: Am. Jour. Sci., 5th ser., v. 6, p. 100-126.
Washington, H. S., and Merwin, H. E., 1922, Augite of Haleakala, Maui, Hawaiian
Islands: Am. Jour. Sci., 5th ser., v. 3, p. 117-122.
White, R. W., 1966, Ultramafic inclusions in basaltic rocks from Hawaii: Min-
eralog. und Petrog. Mitt., v. 12, p. 245-314.
Williams, H., 1954, Problems and progress in volcanology: Geol. Soc. London
Quart. Jour., v. 109, p. 311-332.
Winchell, H., 1947, Honolulu series, Oahu, Hawaii: Geol. Soc. America Bull., v.
58, p. 1-48.
Yoder, H. S., Jr., and Tilley, C. E., 1962, Origin of basalt magmas: an experi-
mental study of natural and synthetic rock systems: Jour. Petrology, v. 3,
p. 342-532.
CONTRIBUTION N O . 1 6 2 , H A W A N INSTITUTE OF GEOPHYSICS.

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/962568/mem116-0477.pdf


by UNAM user
on 18 September 2018

Anda mungkin juga menyukai