e seen, I wanted to see too, and disregarding
his admonition, I pretty soon found myself standing beside my companion.

The great cleft into which we were peering was about six feet wide at
the bottom, coming together some twenty feet above our heads, having
been apparently widened at the base by the action of the water, which,
being here ankle-deep, rushed foaming over and around the many blocks of
lava with which the channel was encumbered. As far as we could see, the
fissure led straight away without a bend; and Joe was for trying to
walk down it at once. I suggested, however, that we leave that for the
present and try another plan.

"Look here, Joe," said I. "If we try to do that we shall probably get
pretty wet, and stand a good chance besides of hurting our feet among
the rocks. Now, I propose that we go down to the ranch again, get our
rubber boots, and at the same time bring back with us my father's
compass and the tape-measure and try to survey this water-course. By
doing that, and then by following the same line on the surface, we may
be able to decide whether it is really this stream which keeps 'the
forty rods' so wet."

"I don't think there can be any doubt about that," Joe replied; "but I
think your plan is a good one, all the same, so let us do it."

We did not waste much time in getting down to the ranch and back again,
when, pulling on our rubber boots, we proceeded to make our survey. It
was not an easy task.

With the ring at the end of the tape-measure hooked over my little
finger, I took a candle in that hand and the compass in the other, and
having ascertained that the course of the stream was due southeast, I
told Joe to go ahead. My partner, therefore, with his arm slipped
through the handle of the lantern and with a pole in his hand with which
to test the depth of the stream, thereupon started down the passage,
stepping from rock to rock when possible, and taking to the water when
the rocks were too far apart, until, having reached the limit of the
tape-measure, he made a mark upon the wall with a piece of white chalk.

This being done, I noted on a bit of paper the direction and the
distance, when Joe advanced once more, I following as far as to the
chalk-mark, when the operation was repeated.

In this manner we worked our way, slowly and carefully, down the
passage, the direction of which varied only two or three degrees to one
side or the other of southeast, until, having advanced a li