94 J.S. Hmerson—Kilauea after the Eruption of 1886. 
equal height above the level of the top of Houlder’s monu- 
ment, while the cone which it had formed, like a little mount- 
ain, towered to a considerably greater height. The weak sides 
of the crater could not long resist the immense pressure of such 
a liquid column. As long ago as 1868, they had been badly 
rent, and it needed but a shight disturbance to cause them now 
to give way and allow the liquid free exit. The coincidence 
of the collapse of the lakes, following so closely upon the solar 
eclipse of April 5th, suggests, as Mr. C. J. Lyons, of the Gov- 
ernment Survey, has pointed out, the possible influence of the 
united pull of the sun, moon and Venus upon our planet as 
determining the time for the sides of the overstrained caldron 
to give way. Be thatasit may, the time had come, and the 
whole column sank, within a few hours, a vertical distance of 
from 587 to 597 feet, disappearing in the depths below. Mr-.: 
Maby observed that the fire first went out in New Lake, some 
little time before its final disappearance from Halema’uma’u. 
This would naturally be the case from its comparatively shal- 
low depth, as the molten mass was being drained away from 
the two lakes into the common reservoir or duct of exit 
beneath. 
While a portion of the escaping lava may have taken a 
southeasterly course, indicated by the large fissure in that 
direction, already alluded to, a number of indications and con- 
siderations lead to,the supposition that the greater part, follow- 
ing the line of least resistance, found its way into the great 
fissures of 1868, and, from the spongy nature of the district 
through which its course lay, readily found all the space 
needed to contain its entire volume without coming to the sur- 
face or entering the sea. It has been affirmed that the steam 
jets along the fissures of 1868 showed greater activity immedi- 
ately after the subsidence of the lakes than at any time for 
months before. If this be true, it becomes an indication 
strongly favoring the above supposition. 
The numerous cracks and fissures running parallel with the 
great steam cracks may then possibly be explained as follows: 
As the subterranean lava stream forced its way through these 
steam cracks, it would naturally compress and force upward | 
the spongy mass on either side of their walls. But when the 
swollen stream had passed along and this pressure was 
removed, a portion of the inelastic mass would fall back to 
nearly its former position, while another portion, adjusting 
itself to its new position, would remain separated, and a new 
crack would thus be formed. The escaping steam, however, 
would find its natural outlet through the fissures most directly 
over its course, instead of seeking exit by any. crack removed 
to one side of it. I have already remarked that this was 
