Vv CHANGE FROM A REPTILE TO A BIRD 57 
it cannot be too strongly insisted upon that they are 
not occasional but unceasing, so that it is highly 
probable that two, that would be useless unless they 
appeared simultaneously, might occur together in the 
same individual. It is only reasonable to suppose 
that some of the reptiles from which our birds have 
sprung were born not only with a fore limb that, 
fringed with scale-like feathers, might act as a make- 
shift for a wing, but also, by a happy coincidence, with 
hind legs stronger than those of their contemporaries. 
Quite apart from such coincidences, a very slight 
power of flight, due to a modification of the fore limb 
not sufficient to incapacitate it for walking, would be 
highly advantageous to the birdlike reptile, or reptile- 
like bird. When menaced by a snake as he sat upon 
a tree, he would flutter to another tree, perhaps feebly, 
descending much as he went, his wings acting as a 
clumsy parachute. Still, he would save his life. Here, 
then, is a stage in advance accomplished. Why 
not after this a development of the hind limbs, making 
an upright posture possible? And when this was 
attained, why not a further development of wings? 
And why should not an improvement in the internal 
organs follow closely upon the visible changes? 
The fact is that in this case it does not seem necessary 
to assume an absolute simultaneity of variations. 
But supposing it is held by any one that modifica- 
tions arising singly could not have advanced the 
reptile to a bird, and that for simultaneous variations 
we must not trust to mere coincidence, we must appeal 
to what is called correlated variation. Many instances 
of this are given by Darwin. Cats, which are entirely 
