322 
A tlantis Once More, 
[June, 
was found covered with volcanic deposits. He contends that 
“ when these connecting ridges extended from America to 
Europe and Africa they shut off the flow of the tropical 
waters to the north ; there was then no Gulf-stream ; the 
land-locked ocean that laved the shores of Northern Europe 
was then intensely cold, and the result was the Glacial 
Period. When the barriers of Atlantis sunk sufficiently to 
permit the natural expansion of the heated water of the 
tropics to the north, the ice and snow which covered Europe 
gradually disappeared ; the Gulf-stream flowed around At- 
lantis, and it still retains the circular motion first imparted 
to it by the presence of that island.” He adds : — “ It does 
not follow that at the time Atlantis was finally ingulfed the 
ridges connecting it with America and Africa rose above the 
water-level. These may have gradually subsided into the 
sea, or have gone down in cataclysms such as are described 
in the Central American books. The Atlantis of Plato may 
have been confined to the Dolphin ridge.” 
Here, then, we have a novel theory of the Glacial Epoch. 
Is it sufficient to account for the phenomena ? It is admitted 
that the removal of the Gulf-stream would lower the tem- 
perature of Western Europe, which is now decidedly higher 
than might naturally be expected from its latitude. But for 
glaciation two faCtors are necessary— a low temperature and 
abundant atmospheric precipitation. Were the Gulf-stream 
suspended it is extremely probable that the quantity of rain, 
mist, and snow experienced in Britain, Holland, Denmark, 
Norway, and Sweden would decrease. Further, we have to 
account not merely for glaciation in Europe, but in North 
America, where it appears to have been, if anything, more 
severe than in the eastern continent, and in the southern 
hemisphere, where it has been traced as far as 3 0 south lati- 
tude. It is exceedingly improbable that the arrest of the 
Gulf-stream, or of any analogous current, should induce 
glaciation in Nicaragua and at Pernambuco. 
It is scarcely conceivable that the Gulf-stream could retain 
for many thousands of years a circular motion imparted to 
it by an island no longer in being. 
We come to the testimony of the Flora and Fauna. Mr. 
Donnelly points to the identity of the Palseardfic and Neardlic 
forms as an important piece of evidence in his favour. Ad- 
mitting their close approximation, — which, indeed, leads 
some authorities to include these two great regions under 
one, — we must remark that it extends chiefly to the more 
boreal forms. A land-connedtion between the two conti- 
nents in the genial latitudes of the Azores, Madeira, and the 
