148 THE FLIGHT OF BIRDS 
slight one perceptible ifygne approached close to the 
edge of the hill.” 
What is going on in the air high aloft when soaring 
takes place over a dead level I imagine to be this. 
There is a wind sweeping over the plain, and at the 
outset it is horizontal. Coming into contact with 
an up-current from the heated surface below, it is 
deflected upward. The soaring bird, then, gets 
support not only from the ascending column of air, 
but from the wind to which the ascending column 
gives an upward trend. Were the heated air 
mounting from below the sole support, birds might 
soar in an almost dead calm, and that is a thing 
which all observers agree does not take place. They 
first ascend to some height—two or three hundred 
feet—by beating with their wings, and then the 
performance begins. 
The reason of the spiral movement is not, I believe, 
far to seek. There are over the plain regions of 
ascending and regions of descending air. It is 
essential that the bird should not pass beyond the 
boundaries of the upward stream that maintains 
and lifts him. Had the wind over the whole extent 
of the plain an upward incline, then the Kites and 
the rest might soar, if the term will stand this strain- 
ing of its use, in a straight line like the Eagles, whose 
majestic advance, without deviation to right or left, 
I have already described ; or, to take an example 
more commonly seen, like the Gull that follows a 
steamer, poised on an up-draught over the stern. 
The up-and-down currents on which birds depend 
for soaring are sometimes very formidable to an 
aviator, who in a few seconds may pass through an 
