THE FLYING-SQUIRRELS OF ASIA AND AFRICA 237 
the whole skeleton of the fore-limbs be so modified as to 
form a wing, as in bats, it is difficult to see how ordinary 
mammals could be endowed with the power of taking flying 
leaps save by the development of an expanse of skin along 
the sides of the body in the manner which obtains in the 
true flying-squirrels, the scale-tailed flying-squirrels, the 
flying-phalangers, and, it may be added, the flying-lemurs. 
The development of flying membranes in all these four 
groups of mammals has, in fact, taken place quite inde- 
pendently, and affords an interesting example of what is 
known as parallelism in development. Such parallelisms 
are due, so to speak, to the poverty of possibilities in 
the way of modification of animal structures. As already 
said, the simplest and most obvious way of endowing an 
ordinary four-limbed mammal with the power of taking 
flying leaps is by the development of lateral expansions 
of skin. Similarly, the only easily conceivable method by 
which a primitive short-limbed and many-toed hoofed 
mammal could be converted into one cut out for speed, 
like a horse or a gazelle, is by reducing the number of 
the digits and increasing the length of the lower segments 
of the limbs. Accordingly, we find parallelism in this 
respect between the horses and the zebras on the one hand, 
and the gazelles, antelopes, and deer on the other. 
But the parallelism is by no means exact in this latter 
case, as indeed would be naturally expected if the lines 
of evolution were distinct; and the structure of the lower 
portion of the limb of a horse differs essentially from the 
same part in a gazelle. 
Neither is the parallelism exact in the case of the two 
groups of flying-squirrels. In the flying-squirrels of Europe 
and Asia, such as the one depicted in the plate, the 
flying membrane, or parachute, is merely a lateral expansion 
