122 THREE CRUISES OF THE ^ BLAKE." 
Pacific, and of their fossil remains, if any are to be found, taken 
in connection with the bathymetric knowledge we have lately 
acquired, would go far towards solving these problems. The 
flora alone can hardly give us the necessary facts. 
The fossil marine fauna, while it can do something to help us 
to reconstruct the probable route of the oceanic currents, gives 
us no clue to land connections. Neither does the distribution 
of the birds, nor that of the reptiles or of the land shells, if we 
are to judge by the case of the West Indies, Galapagos, and 
Bermudas. The simplest assumption we can make for these 
islands is that the drift carried by the existing currents for a 
long period of time is amply sufficient to account for the pre- 
sence of their few reptiles, mollusks, and even their mammals, 
and for the distribution of their flora, without the existence at 
any time of a direct land connection.’ 
In the case of New Zealand — an island for a long period of 
time disconnected from continental masses, except perhaps by 
its northern extension toward Australia — we account for the 
South American elements of its fauna by the action of the cur- 
rents. As yet we know positively of no South American mam- 
mal element, and the birds have undoubtedly, as has been 
concluded by Captain Hutton, been derived from the north. 
The same would hold good of the antarctic elements. This 
seems a more natural explsuation than the attempts made to 
cite plateaux of not less than two thousand fathoms in depth as 
proof of a former continental extension, when, for aught we 
know, these plateaux may be forming and increasing at the 
present day. No better example of the fallacy of же reason- 
ing can be given than the phenomena of the growth of the 
great н plateaux of Yucatan and Fonda, which we 
1 The changes constantly going on over 
the continental shelf illustrate admira- 
bly what is meant by a former land con- 
nection of continental islands. In a 
letter to Professor Benjamin Peirce, 
Superintendent of the United States Coast 
Survey, Professor Agassiz thus referred, 
in 1867, to the shoals and islands lying 
to the eastward and southward of Cape 
Cod: “The drift of the coast of New 
England formerly extended many miles 
beyond the present shores, and is still 
slowly washed away by the action of 
tides, winds, and currents. This sheet of 
drift once extended in unbroken conti- 
nuity from Cape Ann to Cape Cod, and 
farther south ; . it is constantly di- 
minishing, and, in centuries to соте, the 
whole peninsula of Cape Cod may disap- 
pear." 
(See Fig. 60.) 
