Vol. 6, 1920 PALAEONTOLOGY: W. D. MATTHEW 
17 
was able to walk with the body well raised off the ground are the follow- 
ing: (1) Inability to flex strongly the elbows, evidenced by the characters 
of the distal facet of the humerus. (2) Ability to turn the femur well 
forward, like that of carnivorous dinosaurs. (3) Detailed resemblances 
of the femur to that of Stegosaurus, assuredly a straight-limbed reptile. 
(4) Total dissimilarity between the humerus and femur of sauropods and 
those of primitive Permian reptiles. 
PLATO'S ATLANTIS IN PALAEOGEOGRAPHY 
By WiIvIvIam Dii^IvHr Matthew 
American Museum of Natural History, New York City 
Communicated by H. F. Osborn. Read before the Academy, November 11, 1919 
The classic story of Atlantis has been supposed by some writers to be 
a genuine tradition, and to be supported by scientific evidence, which they 
believe indicates a land bridge across the Atlantic in former geologic times. 
The island of Atlantis as described by Plato they suppose to be a remnant 
of this bridge. Examination of the story in detail shows that it is a fable, 
and that the scientific evidence does not lend any support whatsoever to 
it nor vice versa. 
The existence of a transatlantic bridge in Tertiary or pre-Tertiary times 
is a legitimate .scientific question, which should not be confused with the 
fabulous story of Plato. The present writer does not, however, believe 
that such bridges are necessary. Their existence during the Tertiary 
period cannot apparently be reconciled with the known history of verte- 
brate evolution on the two sides of the Atlantic. The supposed arguments 
in their favor from the distribution of certain lower animals and plants 
can all be otherwise explained; and the arguments that favor a general 
permanency of the ocean basins afford grave objections to any such bridges, 
especially of such recent date, geologically. 
Transatlantic bridges in pre-Tertiary times are theoretically more 
plausible simply because there is less positive evidence against them. 
The evidence adduced in their favor appears to be similar in kind to that 
for Tertiary land bridges, and not any more convincing. If in the one case 
it must be otherwise interpreted, then it should be so in the other. 
The conformation of the bottom of the Atlantic is very little suggestive 
of any former bridges, except in the extreme north (from Newfoundland 
to Ireland, or across Greenland, Iceland and Scandinavia). A long, 
depressed, irregular trough or series of "deeps" on each side fronts the con- 
tinental borders which are areas of elevation and erosion. These deeps 
may have been formed by the withdrawal of material that has gone to 
build up the lighter fragmented rocks of the continental platform borders, 
