made by their wings. They flew about fifteen yards from the ground, and as their pinions were urged not much faster than those of the heron,1 I had abundant leisure for studying their movements. The sight was very imposing, and as novel as it was grand. I had seen nothing before, and certainly have seen nothing since that could convey a more elevated conception of the prowess and guiding power which birds may exert. What particularly struck me was the perfect command they seemed to have over themselves and the medium they navigated. They had their wings and bodies visibly under control, and the air was attacked in a manner and with an energy which left little doubt in my mind that it played quite a subordinate part in the great problem before me. The necks of the birds were stretched out, and their bodies to a great extent rigid. They advanced with a steady, stately motion, and swept past with a vigour and force which greatly impressed, and to a certain extent overawed, me. Their flight was what one could imagine that of a flying machine constructed in accordance with natural laws would be.2

The Natural Wing, when elevated and depressed, must move forwards. It is a condition of natural wings, and of artificial wings constructed on the principle of living wings, that when

I have frequently timed the beats of the wings of the Common Heron (Ardea cinerea) in a heronry at Warren Point. In March 1869 I was placed under unusually favourable circumstances for obtaining trustworthy results. I timed one bird high up over a lake in the vicinity of the heronry for fifty seconds, and found that in that period it made fifty down and fifty up strokes ; i.e. one down and one up stroke per second. I timed another one in the heronry itself. It was snowing at the time (March 1869), but the birds, notwithstanding the inclemency of the weather and the early time of the year, were actively engaged in hatching, and required to be driven from their nests on the top of the larch trees by knocking against the trunks thereof with large sticks. One unusually anxious mother refused to leave the immediate neighbourhood of the tree containing her tender charge, and circled round and round it right overhead. I timed this bird for ten seconds, and found that she made ten down and ten up strokes; i.e. one down and one up stroke per second precisely as before. I have therefore no hesitation in affirming that the heron, in ordinary flight, makes exactly sixty down and sixty up strokes per minute. The heron, however, like all other birds when pursued or agitated, has the power of greatly augmenting the number of beats made by its wings.

2 The above observation was made at Carlow on the Barrow in October 1867, and the account of it is taken from my note-book.

forcibly elevated or depressed, even in a strictly vertical direction, they inevitably dart forward. This is well shown in fig. 81.

If, for example, the wing is suddenly depressed in a vertical direction, as represented at a b, it at once darts downwards and forwards in a curve to c, thus converting the vertical down stroke into a down oblique forward stroke. If, again, the wing be suddenly elevated in a strictly vertical direction, as at cd, the wing as certainly darts upwards and forwards in a curve to e, thus converting the vertical up stroke into an upward oblique forward stroke. The same thing happens when the wing is depressed from e to f, and elevated from g to h. In both cases the wing describes a waved track, as shown at eg, gi, which clearly proves that the wing strikes downwards and forwards during the down stroke, and upwards and forwards during the up stroke. The wing, in fact, is always advancing; its under surface attacking the air like a boy's kite. If, on the other hand, the wing be forcibly depressed, as indicated by the heavy waved line a c, and left to itself, it will as surely rise again and describe a waved track, as shown at c e. This it does by rotating on its long axis, and in virtue of its flexibility and elasticity, aided by the recoil obtained from the air. In other words, it is not necessary to elevate the wing forcibly in the direction c d to obtain the upward and forward movement c e. One single impulse communicated at a causes the wing to travel to e, and a second impulse communicated at e causes it to travel to i. It follows from this that a series of vigorous down impulses would, if a certain interval were allowed to elapse between them, beget a corresponding series of up impulses, in accordance with the law of action and reaction; the wing and the air under these circumstances being alternately active and passive. I say if a certain interval were allowed to elapse between every two down strokes, but

this is practically impossible, as the wing is driven with such velocity that there is positively no time to waste in waiting for the purely mechanical ascent of the wing.

That the

f

g'

m

k

n

FIG. 82.

FIG. 83.

Figs. 82 and 83 show that when the wings are elevated (e, f, g of fig. 82) the body falls (s of fig. 82); and that when the wings are depressed (h, i, j of fig. 83) the body is elevated (r of fig. 83). Fig. 82 shows that the wings are elevated as short levers (e) until towards the termination of the up stroke, when they are gradually expanded (f, g) to prepare them for making the down stroke. Fig. 83 shows that the wings descend as long levers (h) until towards the termination of the down stroke, when they are gradually folded or flexed (i, j), to rob them of their momentum and prepare them for making the up stroke. Compare with figs. 74 and 75, p. 145. By this means the air beneath the wings is vigorously seized during the down stroke, while that above it is avoided during the up stroke. The concavo-convex form of the wings and the forward travel of the body contribute to this result. The wings, it will be observed, act as a parachute both during the up and down strokes. Compare with fig. 55, p. 112. Fig. 83 shows, in addition, the compound rotation of the wing, how it rotates upon a as a centre, with a radius m bn, and upon a c b as a centre, with a radius kl. Compare with fig. 80, p. 149.-Original.

ascent of the pinion is not, and ought not to be entirely due to the reaction of the air, is proved by the fact that in flying creatures (certainly in the bat and bird) there are distinct

elevator muscles and elastic ligaments delegated to the performance of this function. The reaction of the air is therefore only one of the forces employed in elevating the wing; the others, as I shall show presently, are vital and vitomechanical in their nature. The falling downwards and forwards of the body when the wings are ascending also contribute to this result.

The Wing ascends when the Body descends, and vice versâ.As the body of the insect, bat, and bird falls forwards in a curve when the wing ascends, and is elevated in a curve when the wing descends, it follows that the trunk of the animal is urged along a waved line, as represented at 1, 2, 3, 4, 5 of fig. 81, p. 157; the waved line a ce gi of the same figure giving the track made by the wing. I have distinctly seen the alternate rise and fall of the body and wing when watching the flight of the gull from the stern of a steam-boat.

The direction of the stroke in the insect, as has been already explained, is much more horizontal than in the bat or bird (compare figs. 82 and 83 with figs. 64, 65, and 66, p. 139). In either case, however, the down stroke must be delivered in a more or less forward direction. This is necessary for support and propulsion. A horizontal to-and-fro movement will elevate, and an up-and-down vertical movement propel, but an oblique forward motion is requisite for progressive flight.

In all wings, whatever their position during the intervals of rest, and whether in one piece or in many, this feature is to be observed in flight. The wings are slewed downwards and forwards, i.e. they are carried more or less in the direction of the head during their descent, and reversed or carried in an opposite direction during their ascent. In stating that the wings are carried away from the head during the back stroke, I wish it to be understood that they do not therefore necessarily travel backwards in space when the insect is flying forwards. On the contrary, the wings, as a rule, move forward in curves, both during the down and up strokes. The fact is, that the wings at their roots are hinged and geared to the trunk so loosely, that the body is free to oscillate in a forward or backward direction, or in an up, down, or oblique direction. As a consequence of this freedom of movement,

and as a consequence likewise of the speed at which the insect is travelling, the wings during the back stroke are for the most part actually travelling forwards. This is accounted for by the fact, that the body falls downwards and forwards in a curve during the up or return stroke of the wings, and because the horizontal speed attained by the body is as a rule so much greater than that attained by the wings, that the latter are never allowed time to travel backward, the lesser movement being as it were swallowed up by the greater. For a similar reason, the passenger of a steam-ship may travel rapidly in the direction of the stern of the vessel, and yet be carried forward in space,-the ship sailing much quicker than he can walk. While the wing is descending, it is rotating upon its root as a centre (short axis). It is also, and this is a most important point, rotating upon its anterior margin (long axis), in such a manner as to cause the several parts of the wing to assume various angles of inclination with the horizon.

Figs. 84 and 85 supply the necessary illustration.

In flexion, as a rule, the under surface of the wing (fig. 84 a) is arranged in the same plane with the body, both being in a line with or making a slight angle with the horizon (x x).1

1 It happens occasionally in insects that the posterior margin of the wing is on a higher level than the anterior one towards the termination of the up stroke. In such cases the posterior margin is suddenly rotated in a downward