The light becomes stronger as the sun rises higher, the daylight streams through the fictitious crater, and I shall soon be able to estimate the cavern’s dimensions.  This is how I calculate: 

Exteriorly the island of Back Cup, which is as nearly as possible circular, measures two hundred and fifty yards in circumference, and presents an interior superficies of about six acres.  The sides of the mountain at its base vary in thickness from thirty to a hundred yards.

It therefore follows that this excavation practically occupies the whole of that part of Back Cup island which appears above water.  As to the length of the submarine tunnel by which communication is obtained with the outside, and through which the tug passed, I estimate that it is fifty yards in length.

The size of the cavern can be judged from these approximate figures.  But vast as it is, I remember that there are caverns of larger dimensions both in the old and new worlds.  For instance in Carniole, Northumberland, Derbyshire, Piedmont, the Balearics, Hungary and California are larger grottoes than Back Cup, and those at Han-sur-Lesse in Belgium, and the Mammoth Caves in Kentucky, are also more extensive.  The latter contain no fewer than two hundred and twenty-six domes, seven rivers, eight cataracts, thirty two wells of unknown depth, and an immense lake which extends over six or seven leagues, the limit of which has never been reached by explorers.

I know these Kentucky grottoes, having visited them, as many thousands of tourists have done.  The principal one will serve as a comparison to Back Cup.  The roof of the former, like that of the latter, is supported by pillars of various lengths, which give it the appearance of a Gothic cathedral, with naves and aisles, though it lacks the architectural regularity of a religious edifice.  The only difference is that whereas the roof of the Kentucky grotto is over four hundred feet high, that of Back Cup is not above two hundred and twenty at that part of it where the round hole through which issue the smoke and flames is situated.

Another peculiarity, and a very important one, that requires to be pointed out, is that whereas the majority of the grottoes referred to are easily accessible, and were therefore bound to be discovered some time or other, the same remark does not apply to Back Cup.  Although it is marked on the map as an island forming part of the Bermuda group, how could any one imagine that it is hollow, that its rocky sides are only the walls of an enormous cavern?  In order to make such a discovery it would be necessary to get inside, and to get inside a submarine apparatus similar to that of the Count d’Artigas would be necessary.

In my opinion this strange yachtsman’s discovery of the tunnel by which he has been able to found this disquieting colony of Back Cup must have been due to pure chance.

Now I turn my attention to the lake and observe that it is a very small one, measuring not more than four hundred yards in circumference.  It is, properly speaking, a lagoon, the rocky sides of which are perpendicular.  It is large enough for the tug to work about in it, and holds enough water too, for it must be one hundred and twenty-five feet deep.