450 Prof. Carl Vogt on 
We now know, by the discovery at Solenhofen of some per- 
fectly preserved wings of Pterosaurs, that their wing was 
formed by a stiff membranaceous border, finely folded and 
rather narrow, attached to the exceedingly elongated fifth 
digit, and extending to the anterior part of the body without 
reaching the hind-feet. I have before me a photograph, of 
the natural size, of the wing of Rhamphorhynchus gemmingki, 
which shows this structure most conclusively. After these 
discoveries, we may say that all the restorations of Pterosaurs 
found in books are wrong. The other digits, four in number, 
are slender, free, and armed with sharp-edged claws; while 
the fifth digit, so strong and so elongated, to which the 
membrane is attached, bears no nail at all. 
In the Bats, four digits are fine, elongated, pointed at the 
end, and arranged like the ribs of an umbrella, while the fifth, 
the thumb, is free, short, and alone armed with a claw. 
Step by step I have followed the development of the wing 
in the embryos of Bats. At first the manus of these animals 
is formed exactly like that of all other Mammals with five 
digits, and is so like the pes that it is difficult to distinguish 
the two limbs when they are separated from the body. The 
embryonic tissue wholly envelops the digits, which do not 
protrude from the circumference of the primitive disk. Later 
the various parts are differentiated ; but while at the anterior 
extremity the membrane which unites the digits follows them 
as they by degrees are lengthened, behind it stops on the 
feet, where the digits become free by passing through it. The 
flying-membrane of Bats is thus no new structure, but only 
the primitive membrane uniting the digits developed equally 
with them. 
We have, then, three stages of adaptation to flight by means 
of a membrane. 
The first is represented in the volant Squirrels and Mar- 
supials : the skin of the body is alone concerned, the skeleton 
taking no part in it: the organ of flight is still passive. As 
soon as it becomes active, osseous parts are needed as sup- 
ports on the one hand, and on the other muscles to put the 
levers in motion. Furthermore, arrangements for extending 
