266 
PACIFIC SCIENCE, Vol. XII, July, 1958 
city toward the sea (Sartorius, 1880: 252-253). 
Eventually the lava broke through a weak part 
of the wall and flowed into the city. It should 
be noted, however, that the breach occurred 
in a part of the wall that lay essentially at 
right angles to the course of advance of the 
flow, and hence was acting as a dam rather 
than as a diversion barrier. 
The foregoing illustrations are ample to 
demonstrate that thin masonry walls, and even 
ordinary loose stone walls such as are built as 
fences along land boundaries, commonly are 
able to withstand the pressure of lava flows 
without being pushed over. As Mason and 
.Foster (1953) have pointed out, such pressure 
usually is no more than the hydrostatic pres- 
sure that the lava is capable of exerting against 
the wall (and it will be shown that this is only 
a portion of the theoretical hydrostatic pres- 
sure) . In some instances the forward momen- 
tum of a flow may result in sufficient pressure 
to push over ordinary stone walls or even 
masonry walls. Examples of this are known at 
Etna. However, even the relatively high ve- 
locity of the Kaohe flow during the 1950 
eruption of Mauna Loa was not sufficient to 
disturb materially the loose stone wall along 
the highway. Fortunately, also, on the gentle 
slopes in the vicinity of Hilo lavgggows are 
likely to be slow moving, thus reducing es- 
sentially to zero the risk of the momentum- 
pressure of a flow pushing over even a very 
frail wall. 
EXPERIENCE WITH BARRIERS DURING 
1955 ERUPTION 
The most recent attempts to control lava 
flows in Hawaii by means of walls were made 
during the 1955 eruption of Kilauea. Ac- 
counts of the eruption have been, or will be, 
published elsewhere (Macdonald and Eaton, 
1955, and in preparation). 
The first possible need for a diversion bar- 
rier arose on the evening of March 3, when a 
big aa flow from the vents -near Puu Kii 
reached a low divide at the head of a shallow 
valley that led toward the village of Kapoho. 
Fig. 4. Wooden plank set in the path of a pahoehoe 
flow on the flank of Kilauea volcano on March 13, 
1955, diverting the flow. 
Had the flow spilled over the divide and en- 
tered the valley, it probably would have fol- 
lowed the valley to Kapoho. A former railroad 
embankment 8 to 10 feet high, currently used 
as a truck roadbed, lay across the top of the 
divide. The lava reached that embankment 
and' piled up as much as 15 feet above it, but 
was deflected southward by it, away from 
Kapoho. Although the top of the flow stood 
high above the level of the top of the embank- 
ment, the movement of the flow was gov- 
erned by the lower liquid portion, on which 
the top was merely carried along. The be- 
havior of the flow in this instance clearly dem- 
onstrated that under favorable circumstances 
Fig. 5. Bulldozers constructing an earthen barrier in 
an attempt to keep lava from reaching the Iwasaki 
camp during the eruption of Kilauea on March 21, 1955. 
