Historic Littoral Cones — M oore and Ault 
near the surface of the aa flows; every effort 
was made to break out fresh rock and to avoid 
oxidized zones. A modern analysis is available 
(Tilley and Scoon, 1961) for one of the three 
flows (1868), and this is shown in Table 1. 
The purpose of these analyses was to investigate 
any chemical changes which occurred as the 
result of contact of incandescent molten lava 
with sea water. 
In Table 1 the differences between the weight 
percent of the oxides of littoral cone ash and 
feeding lava are shown. These differences show 
virtually no chemical interchange between the 
lava and the sea water. Except for iron, MgO 
is the only oxide which shows an average dif- 
ference of more than 1% between feeding lava 
and littoral cone ash in all three littoral cones. 
However, Si0 2 differs in a comparable, though 
smaller, degree in the opposite sense, suggesting 
that this difference in MgO is due to olivine 
control. Apparently the lava-flow material sam- 
pled had slightly less olivine than the littoral 
cone ash, and this difference in the amount of 
olivine is believed to be fortuitous. 
Macdonald (1955:35) has had lava analyzed 
from above and below the tidal zone for two of 
the 1955 flows of Kilauea. These analyses also 
show virtually no interchange and no significant 
change in MgO or Si0 2 content. 
All the analyses, however, do show a very 
interesting change in the oxidation state of the 
iron. The littoral cone ash is invariably higher 
in FeO and lower in Fe 2 0 2 than is the corre- 
sponding feeding lava flow, the littoral cone ash 
averaging about 1.2% more FeO and 1.4% less 
Fe 2 0 3 than the feeding flow. Likewise, the tidal 
zone 1955 lava is higher in FeO and lower in 
Fe 2 0 ;! than is the subaerial part of the same 
flow (Macdonald, 1955:35), but the differences 
are smaller, averaging only about 0.5%. The 
reason for this difference, apparently, is that the 
rapid quenching of the water-chilled lava in- 
hibits oxidation, whereas oxidation of iron pro- 
ceeds in the subaerial part of the flow long after 
it has solidified, but while it is still hot. The 
smaller difference between tidal zone lava and 
its feeding flows, as compared with littoral cone 
ash and its feeding flows, suggests that the less 
drastic chilling of the tidal zone lava has allowed 
some oxidation to proceed in it, whereas that 
in the littoral cone ash was largely prevented. 
9 
Washington (1923:415-416) has pointed out 
that the more glassy forms of a lava flow contain 
a higher proportion of ferrous to ferric iron 
than do the more crystalline phases of the same 
flow. He has shown further that FeO is uni- 
formly higher relative to Fe 2 0 3 in the pahoehoe 
form of lava flow than in the aa form; this 
difference is apparently due to the more glassy 
character of pahoehoe as compared with aa. 
Figure 3 is a compilation of modern analyses in 
which the Fe 2 0 3 /(Fe 2 0 3 + FeO) ratio is 
plotted for different kinds of flows, for pumice, 
and for littoral cone ash. These new data clearly 
support Washington’s concept that iron in aa 
flows is more highly oxidized than that in pa- 
hoehoe flows. They also show that littoral cone 
ash and pumice from the Kilauea Iki eruption 
( 1959) on the average are slightly less oxidized 
than the average pahoehoe flows. 
All the feeding flows of the littoral cones 
are aa flows, yet the chilled littoral cone ash is 
in general less oxidized than pahoehoe lava and, 
with pumice, is some of the least oxidized of 
historic tholeiitic Hawaiian lava. Presumably, 
the characteristic high-oxidation state of the aa 
flows had not developed when the flow was in 
motion or when the littoral cones were built. 
Apparently the aa lava becomes highly oxidized 
rather late in the cooling history of the flow, 
probably after it solidifies, but before it is 
entirely cool. The greater thickness of the aa 
flows, as well as the insulating layer of rubble 
on the surface, would cause them to cool more 
slowly than pahoehoe flows, and hence they 
would be subject to oxidation for a longer 
period. 
The state of oxidation of the iron in truly 
juvenile, unaltered Hawaiian tholeiitic lava is 
not known. However, the littoral cone glassy 
ash and basaltic pumice, both of which are dras- 
tically quenched, include the least oxidized of 
historic lava (Fig. 3) and may represent most 
closely the unoxidized lava. 
MECHANISM OF FORMATION 
The most important single factor in the for- 
mation of a littoral cone is that a flow of suf- 
ficient volume enters the sea. At most only a 
small amount, probably never more than 5%, 
