ANNUAL REPORT-EARLY INVESTIGATIONS. 
soundings south of the mainland of Florida was an uncovered 
coral ground, upon which the deep-water species were just 
beginning to spread. But we may trace the change farther. 
There was a time when neither the southern bluffs of the con¬ 
tinent, nor Long Key within the Everglades, nor even the Ev¬ 
erglades themselves, existed; when, therefore, the Gulf Stream 
had a broad communication with the Atlantic and the south¬ 
ern shores of the United States extended in almost unbroken 
contiguity from west to east, from the shores of Texas and 
Louisiana to St. Augustine. At that time the Gulf channel 
was, in reality, a broad bay, as broad as the Gulf itself, desti¬ 
tute of all those obstructions which now cause the tropical 
current to follow such a circuitous course between the West 
India islands, through the Caribbean Seas, and around the 
peninsula of Florida. The influence which the Gulf Stream 
has upon the climate of the Atlantic is so well known, that its 
connexion with the changes which the current itself has un¬ 
dergone within a comparatively recent period cannot be over¬ 
looked.” 
“We have seen how successfully several rests have been 
formed, more or less parallel, within the limits of the penin¬ 
sula of Florida, as well as beyond the mainland. We have 
seen, also, how these parallel or concentric reefs have been 
gradually transformed into mainland by the accumulation of 
coral sand and mud with other loose materials, and also that 
the keys are now slowly annexed to the mainland by the same 
process.” 
A series of rock samples obtained by Agassiz in the 
eonrse of his investigations of the keys were examined 
by Horsford and reported upon in two papers, the first 
of which was published in the Proceedings of the Ameri¬ 
can Association for the Advancement of Science, and 
with some changes in the American Journal of Science 
(100). The second paper (101), was called out by criti¬ 
cism of the first by Professor Dana, and is occupied with 
a defense of his earlier paper. The chief conclusions of 
these papers are: 
“1. That the submerged or oolitic rock has been solidified 
by the infiltration or finely powdered (not dissolved) carbon¬ 
ate of lime, increasing the points of contact; and the intro- 
