262 
PACIFIC SCIENCE, Vol. XII, July, 1958 
eral years ago (Steams and Macdonald, 1946: 
258) and the construction of wells to provide 
an alternative or supplementary water supply 
was suggested. Such wells should be kept 
within the line of the proposed lava diversion 
barrier, protected as far as possible from 
lava flows. 
INADEQUACY OF AERIAL BOMBING 
The use of explosives to alter the course of 
lava flows was first suggested by the late 
Lorrin A. Thurston in the early 1920’s, and 
was elaborated and made specific by Jaggar 
(1931, 1936). The idea of emplacing the ex- 
plosive by means of aerial bombs was sug- 
gested by the late Guido Giacometti at the 
time of the 1935 eruption. 
There are three general ways in which bomb- 
ing can divert lava flows: (1) by breaching a 
lava tube in a pahoehoe flow, (2) by breaching 
an open channel in an aa flow, or (3) by 
breaking down the walls of the cone at the vent 
(Finch and Macdonald, 1949; 1951: 128-132). 
(For a discussion of the characteristics of aa 
and pahoehoe flows, see Macdonald, 1953.) 
(1) At first the main feeding streams of all 
flows are in open channels, but after the first 
few hours or days of activity the main stream 
of a pahoehoe flow crusts over and develops a 
roof. Thereafter it flows through a tube, from 
a few feet to as much as 50 feet in diameter, 
resembling a great pipe or subway. Bombs 
dropped on this tube may break it open, 
clogging the tube partly with debris from the 
shattered roof and partly with viscous aa lava 
resulting from the violent agitation of the 
fluid lava in the tube. The clogging may 
cause an overflow from the tube at that point 
and a consequent diversion of the main feed- 
ing stream of the flow. If the diversion is sev- 
eral miles upstream from the former advancing 
flow front, several days may pass before the 
front of the new flow reaches as great a dis- 
tance from the vent as had the earlier flow front. 
(2) The main feeding river of an aa flow re- 
mains largely open, but repeated overflows 
gradually build up natural levees on each side 
of the stream, and after the first few days the 
stream commonly is flowing at a level several 
feet higher than the adjacent land surface. 
Breaking down the levee by bombing permits 
the liquid to escape from the old channel and 
start a new flow. The removal of part or all of 
the supply of liquid lava from the old channel 
causes the advance of the old front to slow 
greatly or stop altogether, and it may be sev- 
eral days before the new front reaches a point 
as far from the vent as that reached by the old 
one. At that time bombing can be repeated 
if necessary. 
(3) Commonly the pool of liquid lava in the 
cone, which feeds the flow, is at a level several 
feet above the ground surface adjacent to the 
cone. As with the aa levees, breaking down 
of the walls of the cone allows the lava to 
spill out laterally, starting a new flow and 
depleting the supply of lava feeding the 
previous flow. 
The last method, suggested independently 
by Finch (1942) and the writer (Macdonald, 
1943), has not yet been tried, although the 
natural breakdown of the cone walls during 
the 1942 eruption produced essentially the 
same effect that would be brought about by 
bombing. The first method was employed un- 
der the direction of Jaggar in 1935, and the 
second under the direction of Finch in 1942. 
In neither case did the bombing wholly divert 
the flow, but in both it was demonstrated that 
the methods can be successful under favorable 
circumstances. However, bombing methods 
can be used only where topography is favor- 
able and at times when the lava flow has formed 
well- developed tubes or channels between 
elevated levees, or when a large cone of ap- 
propriate shape has been built at the vent. 
Furthermore, the bombs must be very ac- 
curately placed to produce the desired effects, 
and this in turn requires good visibility of the 
targets from the air. During times of eruption 
visibility is often very poor over the flows in 
any area because of the clouds of volcanic 
fume and smoke from burning forests. In the 
area southwest of Hilo visibility is apt to be 
