oC 
April 19, 1883] 
systematic movement within itself : the rearward skirts of the 
cloud are climbing up its sloping back, often with little rolling 
curls : the top of the cloud is rolling over like a breaker in the 
act of tumbling on the beach. Each cloud is in fact the scene 
of a travelling vortex in the air revolving round a more or less 
horizontal axis. And the persistency with which such a cloud 
will preserve its individual identity indicates the persistency of 
the vortex. In the comparatively lower regions of the air it 
may be difficult to demonstrate the presence of such vortices, but 
similar causes are present, and it cannot be doubted that similar 
effects are produced and that such vortices, possessing a propor- 
tionate degree of persistency, are generated in those regions of 
the air which are within the range of the habitual flight of 
cireling birds. 
Let us see what effect these conditions (1) and (2) separately 
would have upon the circling flight of a bird. 
(1.) First, let us take horizontal currents increasing in velocity 
the higher they are above the earth ; and suppose a bird at the 
highest point of one of its gyrations, when it has mounted against 
the wind and is wheeling to one side or the other, preparatory to 
the descent with the wind which is to give it sufficient velocity 
for another rise (but which could not enable it to rise to the same 
height as before if the air had no internal movemen’, for there 
would be no self renewing force to neutralise the ever-new force 
of gravity and the perpetual friction of the air). Let us regard 
the air at the level of the bird, at this turning-point, as s¢2//, 
Then, relative to this point, the lower strata of air have a hori- 
zontal velocity in the opposite direction to the wind (as perceived 
on earth) ; and the bird in falling apparently down the wind will 
really be meeting stronger and stronger adverse currents, and 
when it has reached the lowest point of the ‘‘circle,’’ it will 
have a greater horizontal velocity relative to the air at that level 
than if the whole air through which it has fallen had been still. 
Therefore, in virtue of its greater horizontal velocity relative to 
the air (which is accompanied by increased air-resistance), the 
bird will be subject to a greater force upon its wing-surface, and 
will therefore be able to mount higher (ce¢erds paribus) than if it 
had fallen through still air. But (instead of ‘‘ ceteris paribus’’) 
suppose the bird, as it rises, wheels gradually round and faces 
the wind. Then, in rising, it will enter successive strata of air 
having successively greater and greater velocity relative to itself 
(the bird) than if the air had no internal movement, and therefore 
the air-resistance, which is the lifting force, will ever be greater 
than that due to the height gained by the bird if in still air; and 
therefore the bird will be able to rise yet higher. But this 
manceuvre of wheelins to face the wind in rising will cost some 
time, during which gravity ceases not to act ; it will also cost some 
friction and a slight loss of horizontal velocity, and the question is 
whether these los-es are sufficient to destroy the advantage above 
described. This is a problem for the mathewaticians to solve. 
It seems difficult to imagine that within the narrow limits of 
the actual rise and fall of the bird at the different phases of its 
circle, there should be sufficient difference of velocity of upper 
and lower air-currents, to account for such a gain of elevation as 
Mr. Peal mentions (from 10 to 20 feet at each lap), We require, 
however, to know the vertical height of the bird’s fall and sub- 
sequent rise. I have not seen any estimate of this, but, judging 
from Mr. Peal’s diagram, the bird’s fall appears no greater than 
its gain of elevation (10 or 20 feet). 
Still it appears from the foregoing considerations that the bird 
will gain support by falling with the wind and rising against it, 
when the upper wind is stronger than the lower. 
This result suggests that a bird might with like effect make 
use of two collateral currents of different velocity. Suppose two 
currents, fast and slow, side by side, flowing in the same direc- 
tion. The bird may fall with the slow current, and so acquire 
a certain horizontal velocity. Then let it wheel round against 
the swift current, and it will be able to rise against it to a height 
due to the greater horizontal velocity between bird and air. 
Having reached full height, let it again wheel round into the 
slow current and recover by a sloping descent therein the hori- 
zontal velocity it has lost, which, when recovered, will enable it 
to mount again against the fast current. 
Thus it would appear that a bird can take advantage of alternate 
fast and slow currents, whether collateral or superposed, rising 
against the fast and falling with the slow, to maintain itself in 
the air, while partaking in the general drift of the wind, without 
flapping its wings. 
(2.) In the next case to be considered, we have to deal not with 
horizontal currents, but with the rotatory currents of rolling 
NATURE 
591 
masses of air. A mass of air rolling about a horizontal axis will 
have descending currents in its front, and ascending currents in 
its rear. The former can be of no use to the bird for the 
purpose of support. The bird must keep in the rear of the roll, 
where it will find an upward slantig current. In a high wind 
this current would probably be strong enough to support the 
bird in motionless poise (relative to the earth), but this could 
only be for a few seconds, because the whole vortex is travelling 
rapidly with the wind (of which it forms a part) and would 
speedily pass and leave the poised bird behind at the mercy 
of the downward currents in the van of the next advancing 
vortex. How then is the bird to remain in the upward current, 
and at the same time to maintain a high velocity relative to the 
air in which it moves? It can only be done by circling. The 
bird must face the current in rising, and as it approaches at once 
the outskirts of the current and the limits of its own momentum 
(relative to the air) it must wheel round (—indeed it must have 
begun to wheel while rising—) and fall down the wind, for the 
double purpose of recovering its spent velocity and of overtaking 
the receding vortex. 
In falling down the wind, the bird will pass out of a stronger 
into a weaker current, and will be able to take advantage of the 
difference (regarded horizontally), just as in the case (already 
considered) of horizontal currents of different velocity. But 
regarded vertically the descent into the weaker current will be a 
disadvantage. However, it is clear that under these conditions 
there will be no difficulty about the bird’s support in air by a 
circling flight without stroke of wing. 
But there is still a difficulty with regard to the progressive 
ascent of the bird. Mr. S. E. Peal (NATURE, vol. xxiii. p. 10) 
testifies that the pelican, adjutant, vulture, and cyrus rise circling 
from 100 or 200 to as much as 8coo feet. Can it be supposed 
that a rolling vortex of air would have equal range or climb to 
such a height? Swirls formed at the edge of a deep stream of 
water are seen to be drawn obliquely away from the side towards 
mid-stream, and I suppose that an aérial vortex with horizontal 
axis will in like manner be drawn obliquely upwards into the 
more rapid air. Moreover I remark that Mr. Peal’s observa- 
tions were made on the coast, and that his diagram represents 
the birds as rising on a wind blowing up the country towards 
the hills. Such a wind would have a general upward slant, 
and any rolling of the air would have the same slant to begin 
with and to rise from, so that a bird keeping to the (supposed) 
vortex would rise with it to the same height. 
The same principles which we have found useful in dealing 
with the regular and rhythmical phenomenon of circling flight 
will, I think, help us to understand the general case of irregular 
sailing flight, like that of the albatross following a ship, as 
described by so many writers (¢.g. the Duke of Argyll, “‘ Reign 
of Law,” fifth edition, pp. 153-4). This general case may be 
accounted for by (1) irregular alternations (either in strength 
or direction) of horizontal air-currents ; or (2) irregular upward 
currents. 
Currents alternating in strength are equivalent, in relation to 
any intermediate point, to currents alternating in direction. 
To take an extreme, almost imaginary, case : let us sup; ose a 
bird on outspread wings exposed alternately to the force of 
exactly opposite winds. To each in turn the bird will offer the 
sloping under-surface of its wings, and by each in turn it will be 
at once uplifted and pushed back, but each will counteract the 
backward push of the other, while each will reinforce the other's 
uplifting effort. The result will be that the bird will rise in a 
wavy line without any effort of its own beyond what is required 
to keep its wings rigid, and to present them favourably to the 
alternate winds. } 
Now suppose the whole air to be travelling horizontally in a 
direction at right angles to the two opposite currents. This 
supposition will not affect the lifting power of those opposite 
currents, but it will make it necessarv for the bird (if it is not to 
be swept away by the travelling air) to sacrifice some of the 
height it might gain for the sake of making head against the 
general drift of the wind. This is no longer an extreme or 
imaginary case, but one of very frequent occurrence. It is 
simply that of oscillating gusts in a high wind, The air is full 
of sidelong rushes of wind (probably parts of neighbouring 
vortices). See how the vane of a weathercock oscillates. A 
sidelong rush means fresh velocity relative to the bird in a new 
direction. The bird by a tilt of the wing can instantly convert 
that fresh air-pressure into a lifting force and rise upon it. And 
if these rushes of wind come alternately (as in an irregular 
