was a liquid drain of the
molten lava underneath toward the lake, by means of which a great
subterranean cavity was formed as far back as the mountain; that the
crust on top, being of insufficient strength to bear its own great
weight, must have fallen in as the whole mass cooled, and thus created
this vast crack in the earth.

[Illustration: OUTLINE VIEW OF THINGVALLA.]

I incline to the first of these theories myself, as the most
conformable to the contractile laws of heat. There is also something
like practical evidence to sustain it. A careful examination of the
elevations and depressions on each wall of the gap satisfied me that
they bear at least a very striking analogy. Points on one side are
frequently represented by hollows on the other, and even complicated
figures occasionally find a counterpart, the configuration being
always relatively convex or concave. This would seem to indicate very
clearly that the mass had been forcibly rent asunder, either by the
contractile process of heat, or a convulsion of the earth. The most
difficult point to determine is why the bottom should be so flat and
regular, and what kept the great mass on each side so far intact as to
form one clearly-defined fissure a hundred feet wide and nearly five
miles in length? This, however, is not for an unlearned tourist like
myself to go into very deeply.

How many centuries have passed away since all this happened the first
man who "gazed through the rent of ruin" has failed to leave on
record--if he ever knew it. The great walls of the fissure stood grim
and black before the old Icelandic Sagas, just as they now stand
before the astonished eyes of the tourist. History records no material
change in its aspect. It may be older than the Pyramids of Egypt; yet
it looks as if the eruption by which it was caused might have happened
within a lifetime, so little is there to indicate the progress of
ages. I could not but experience the strangest sensations in being
carried so far back toward the beginning of the world.

At the distance of about a mile up the "Jau" a river tumbles over the
upper wall of lava, and rushes down the main fissure for a few hundred
yards, when it suddenly diverges and breaks through a gap in the
inferior wall, and comes down the valley on the outside toward the
lake.

During my stay at Thingvalla I walked up to this part of the
Almannajau, and made a rough sketch of the waterfall.

From the point of rocks upo