360 
PACIFIC SCIENCE, Vol. XV, July 1961 
A still later group of lavas, which may be 
called the nephelinic suite, consists characteris- 
tically of nepheline basalt and melilite-nepheline 
basalt, but includes also basanites, and alkali oli- 
vine basalts ("linosaites”) in which nepheline 
is present in the norm though not in the mode. 
The nephelinic suite is in general separated 
from the rocks of the other groups by a pro- 
found erosional unconformity (Stearns, 1946: 
22 ). 
THEORIES OF ORIGIN OF HAWAIIAN 
ROCK SUITES 
There appears to be little or no question that 
the tholeiitic magma originates in the upper 
part of the earth’s mantle, probably at a depth 
of about 30 or 40 mi., and that variations within 
the suite are largely or entirely the result of crys- 
tal differentiation. The same degree of certainty 
does not extend to the alkalic suite. Macdonald 
has previously attributed the formation of the 
more alkalic members of the alkalic suite to dif- 
ferentiation of a parent magma corresponding 
approximately in composition to the average 
basalt of Kilauea (Macdonald, 1949^: 92; 
Stearns and Macdonald, 1946: 205), or to an 
average of basalts from all Hawaiian volcanoes 
(Macdonald, 19497?: 1569) . The former average 
corresponds with tholeiite only slightly under- 
saturated with silica. The latter average included 
alkali olivine basalts, and therefore is a little 
higher in alkalies and less saturated with silica 
than the Kilauean average. Calculations indi- 
cated that the alkalic rocks could be derived 
from either parent by crystal differentiation. To 
do so, however, it was necessary to hypothesize 
the separation of a large amount of pyroxene 
(both diopside and hypersthene) from the 
magma. Derivation of alkali olivine basalt per 
se was not considered, because it was not recog- 
nized as an independent rock type. The possibil- 
ity of other differentiation processes, such as al- 
kali transfer by volatiles, was also recognized. 
Tilley (1950: 44-45) also attributed the var- 
ious members of the alkalic rock suite to crystal 
differentiation of tholeiite, separation of hypers- 
thene in place of olivine producing the alkali 
olivine basalt. Powers (1955) agreed to the im- 
portance of the movement of olivine crystals in 
producing the variations among the tholeiitic 
basalts, but pointed out, as indeed he had earlier 
(Powers, 1935), that crystal differentiation alone 
is inadequate to produce alkali olivine basalt 
from a saturated tholeiite. 
It certainly is true that desilication of a 
magma by crystal differentiation can only result 
from the removal of crystals containing more 
silica than the magma. The removal of pyroxene 
from tholeiite can perpetuate a state of under- 
saturation in silica, but cannot bring it about. 
No mineral containing more silica than a satu- 
rated tholeiite magma is likely to form and sep- 
arate except during the very latest stages of 
crystallization. Provided, however, that a dis- 
tinctly undersaturated tholeiitic magma exists as 
a liquid, crystallization of pyroxene can not only 
perpetuate the undersaturation, but increase it. 
Murata (I960) has suggested that alkali olivine 
basalt is derived in this way from undersatu- 
rated tholeiitic magma. 
Kuno et al. (1957: 212) agree with Powers 
that crystal differentiation cannot produce alkali 
olivine basalt from tholeiite. Instead, they as- 
sume the existence of two independent primary 
basalt magmas, tholeiite and alkali olivine basalt, 
formed by partial melting of peridotite at differ- 
ent levels in the mantle, incongruent melting of 
pyroxene at the higher level supplying extra 
silica for the tholeiite. More recently, Kuno 
(I960) has hypothesized still a third primary 
magma, high-alumina basalt, produced by melt- 
ing in the mantle at a depth intermediate to the 
other two levels. 
For several reasons, the existence of two pri- 
mary basalt magmas in Hawaii appears improb- 
able. Chemical analyses demonstrate a complete 
intergradation of the two types in composition 
( Figs. 1 and 3 ) . Furthermore, it seems unlikely 
that melting at a deeper level in the mantle 
would produce a magma richer in alkalies than 
would melting at a shallower level. If anything, 
the reverse would be expected. These arguments 
are far from conclusive but, particularly when 
it is remembered that the rocks of the alkalic 
suite comprise only a very small proportion of 
the total, they do suggest that in some way alkali 
basalt magma is produced from tholeiitic 
magma, rather than having a wholly independent 
origin. 
