14 TRANSACTIONS OF THE [oct. 13, 
very great. If we could have a good view of the proper bottom 
of a banana-hole, we might be able to account for their forma- 
tion; but, unfortunately, the bottom is always filled with a deposit 
of earth or blocks of coral rock, and generally covered with vege- 
tation. 
It is not improbable that the deep cylindrical ones were formed 
in the same manner as pot-holes. And others might have ori- 
ginated in the same way as the spouting-holes, where the waves 
undermine the shore and afterward break an opening in the 
rock above. Should some of the boiling-holes, described above, 
become elevated and their bottoms filled up with fallen blocks 
of coral rock and deposits of earth, they would form banana- 
holes. The holes in the cracks at Fresh Creek can be easily ex- 
plained, but if all banana-holes were formed in this manner we 
should find them in a line with others, which I was tald was the 
case, but I was never able to satisfy myself that it was so. The 
caves are often the result of the former action of the sea, and 
some of them have probably been washed out by rain-water; 
but in either case, should a portion of the roof fall in, it would 
make a banana-hole if small, or, if large, an ocean-hole like the 
one near Nicols Town. The horizontal passages are evidently 
washed out by water, but whether by the sea or rain-water I do 
not know, but I believe either might have accomplished it. It 
is difficult to understand how underground channels could be 
formed under water, yet the boiling-holes prove that such exist ; 
but there is no means of determining whether they were formed 
under the present circumstances or at some previous period 
when the land might have been elevated. 
I was told that holes were as common under the water as they 
were on the land, but did not myself observe this to be the case ; 
but then my opportunities for observation in this direction were 
limited. I infer from the facts I have given that banana-holes 
and caves pass gradually into each other, and that they have been 
formed by the action of the sea-water and afterwards modified 
by the action of rain-water, aided by the products of the decom- 
posing vegetable material and in some cases by the falling-in of 
the roofs of the caves. 
Liffects of Vegetation on the Surface. 
One of the facts that I noticed shortly after my arrival in the 
Bahamas was the occurrence of great numbers of blocks of coral 
rock scattered irregularly over the ground, and I first thought 
that they were the result of the excessive erosion that I saw tak- 
ing place around me. But on some of the cays—as on Goat 
Cay, described above—where the erosion was most rapid, there 
