Popular Science Monthly/Volume 44/April 1894/New Lights on the Problem of Flying


SOME of the readers of The Popular Science Monthly may remember that in November, 1888, I published an article in which I tried to show the physical impossibility of a true flying machine—i. e., one which could both lift and propel itself without the help of a balloon. The article was widely commented upon, the only objection urged being the threadbare one that more wonderful things than this have come to pass and will come to pass again. Since that time, however, a very elaborate series of experiments by Prof. Langley has thrown so much and so new light on the whole problem of flying, that I am forced to reconsider and modify somewhat the conclusions then reached. Having been asked to contribute a paper on the subject of flying to the World's Congress of Aëronautics, I, a few months ago, reviewed the whole subject in the light of these new experiments. The pressure of other duties at that time prevented me from putting my thoughts in final form, and I laid aside my notes. But I feel that I owe it to myself, as well as to The Popular Science Monthly, that I should again express my views under the changed conditions. This article, therefore, may be regarded as substantially what I would have given at the World's Congress if I had had time then to prepare it.

But this time I find it necessary to take up the subject from a more general point of view than before. My theme now is the problem of flying, both natural and artificial. I begin, therefore, with some discussion of the flight of birds.

The bird's wing has two distinct functions, viz., that of a propeller and that of an aëroplane. Both of these functions are performed by the wing in ordinary flight, but in different relative proportions according to the size of the bird and the extent of its wings. In insects and in small birds the wings act almost wholly as propellers. In large birds with great expanse of wings, except in rising, they act mainly as an aëroplane. This difference between small and large birds is fully recognized in my previous paper, but I did not then appreciate its supreme importance. Now it is on the properties of an aëroplane that the new light has been mainly shed, and it is therefore of its function in flight that I shall have most to say. First, however, a very few words on the bird's wing as a propeller.

The structure of a bird's wing is a marvel of exquisite contrivance—a wonderful combination of lightness, elasticity, and strength. The hollow quill, the tapering shaft, the vane composed of barbs clinging together by elastic hooks, making thus an impermeable yet flexible plane—all this has been often insisted on by writers on design in Nature. But there are two points not so often noticed which especially concern us here.

1. Of the two vanes of each feather, the hinder one is much the broader. This, together with the manner of overlap, causes

PSM V44 D763 Longitudinal section of the wing plane and cross section of three feathers.jpg
Fig. 1.—Longitudinal Section of the Wing Plane and Cross-section of Three Feathers. a, shaft; v, v', vane.

the feathers to rotate and close up into an impervious plane in the downstroke, and to open and allow the air to pass freely through in the upstroke, as shown in the figure (Fig. 1). This structure and arrangement produce the greatest possible effectiveness of the downstroke and the least possible loss in recovery for another stroke.

2. The plane of the wing is supported not along the middle, but along the extreme anterior border, as shown in Fig. 2, which

PSM V44 D763 Diagrammatic cross section of a bird wing.jpg
Fig. 2.—Diagrammatic Cross-section of Bird's Wing. a, wing bones; b, plane.

is a diagrammatic cross-section of the wing. The effect of the down stroke is to tip up the wing behind, as shown in Fig. 3. The whole force of the stroke, a b, is resolved into two components—one, a c, sustaining, and the other, b c, propelling onward. In easy flight, therefore, only downward flapping is necessary, although

PSM V44 D763 Diagrammatic cross section of a bird wing during downstroke.jpg
Fig. 3.—Diagrammatic Cross-section of Bird's Wing during Downstroke. a, b, whole force; a, c, part sustaining; b, c, part propelling.

in rapid flight doubtless the stroke is also a little backward.

The same admirable adaptation is carried out in every part of the bird. The whole bird is an exquisitely constructed flying machine. The smallness of the head, the feet, and the viscera, the lightness and the strength of the bones, all show that everything is subordinated to this one supreme function. In comparing a bird with an artificial flying machine it is necessary to bear this in mind.

But it is in the use of the wing as an aëroplane that the most wonderful feats of bird locomotion consist. If we are ever to achieve artificial flight it must be by the application of the principles underlying these. There are four of these feats of bird flight which require special notice as bearing on the subject of artificial flight. These are hovering, poising, soaring, and sailing.

Hovering. There is some confusion in the use of this term. It always refers to a maintenance of a body in one position in the air; but this may be done in two ways either by vigorous flapping of the wings, or else, under certain conditions, with no motion of the wings at all. This latter, however, I shall call poising, and confine the term hovering to the former. In this sense hovering is seen in many insects and in the humming bird, and, among larger birds, in the sparrow hawk (Falco sparverius) and in the osprey (Pandion haliætus). In these it is seen that in hovering the body is inclined upward, and the stroke of the wing is decidedly forward as well as downward. The reason of this is, as already explained, that downward strokes give onward motion. But the main thing to be observed in large birds is the violent struggles necessary to maintain position compared with the ease of onward flight. This difference furnishes the key to the properties of an aëroplane, and was, I believe, first explained by Marey. In maintaining the body in the same position, as in hovering, the air gives way under each stroke of the wing, creating a downward current, thus greatly diminishing the effectiveness of the downstroke and increasing the loss in recovery or upstroke. In progressive flight, on the contrary, and more and more as the progress is more rapid, every phase of the downstroke is on new air. The wing strikes on three feet or six feet or ten feet or twenty feet of air, according to the velocity of progress, with every stroke. The air has not time to give way before the wing passes on to new, unmoved air. But if it is difficult to maintain one position, as in hovering, it is evidently still more difficult, on this principle, to raise the body directly upward. This explains the difficulty experienced by a large bird like the condor in rising, and yet the ease and grace of progressive flight when well up. We will see hereafter the great importance of this principle, as shown by the experiments of Langley.

Poising. By this term I mean the maintenance of a fixed position with outstretched, motionless wings. During my boyhood I was fond of field sports of all kinds, and therefore a constant and accurate observer of the flight of birds; and yet, during all that time, I never saw this feat. The reason was, that I lived in a perfectly flat country. I saw it for the first time when, at the age of fifteen, in going to college, I moved to a rolling country. It is best seen in a bare rolling country, like much of the western portion of the United States. The most perfect poising I have ever seen done was by the red-tailed hawk (Buteo montanus), on the bare, rolling lava plains of eastern Oregon. The conditions absolutely necessary are a rolling country and a steady breeze. The bird places himself above the brow of a hill with face to the wind. As long as the wind remains steady the bird retains his position, with outstretched wing, motionless.

The explanation is as follows: As already said, the bird places himself facing the wind just above the brow of a hill. The wind is deflected upward by the slope of the hill. The bird places his aëroplane (wings and tail) in a plane inclining slightly downward, but not so much inclined as the slope of the hill, so that the wind still strikes the under side of the aëroplane. In this position the force of gravity would carry him downward and forward, while the wind would carry him upward and backward. The bird skillfully adjusts the position of the aëroplane so that these two opposite forces shall exactly balance one another. As long as the wind remains steady his position is unchanged. If the wind changes in direction or in velocity, he wiggles himself a little, perhaps flaps once or twice, until he finds a new position of equilibrium, and again remains steady. This explanation is, I believe, complete.

Soaring. It is well known that many large and long winged birds, such as vultures, hawks, pelicans, etc., will sweep about in wide circles with motionless, outstretched wings, not only maintaining their level, but rising in ascending spiral until they disappear from view. I have often watched their easy, graceful motion for hours, and am quite sure that it is accomplished without any expenditure of energy at all commensurate with the work of elevation. How is it done? There is no problem of bird-flight upon which so much has been written, and so little of any value. Let us see first what are the necessary conditions.

1. Every careful observer must have noted that the bird slopes downward along one half of the circle, so as to acquire high velocity, and then rises along the other half to a higher level than that from which he descended. How can he rise higher?

2. Every clear thinker must see that this feat is impossible, and every careful observer must have noted that it is never done—in still air. For if air is still, even if there were no friction and no tendency to fall toward the ground, the most that the velocity acquired by the down slope could do would be to carry the bird back to the same level again. Therefore, in still air the bird must descend instead of ascend. A necessary condition, therefore, is wind; and, indeed, a soaring bird always drifts with the wind. The spiral is never upright, but always inclined to leeward. In soaring, the bird slopes downward with the wind, then turns and rises, facing the wind. How does wind help him?

3. The feat is physically impossible in an even current of air. It is strange that this is not seen at once; and yet excellent writers have become confused on this point. I confess that until recently I have been confused myself. It really follows as a necessary consequence of the last conditions for an even current is still air to the bird immersed in it, precisely as the earth is practically still to us dwelling on it. If a cloud should intervene between the bird and the earth, it would be impossible for him to know whether he was in a current or not.

4. Therefore, in order to rise in a spiral without doing work by flapping, there must be differential currents, which the bird takes advantage of to do the work.

Explanation. Now, there are such differential air currents. "Wind, like all other currents, increases in velocity from bottom upward. Experiments show that on a grass meadow the velocity eight feet above ground is double of that at one foot, and the velocity goes on increasing upward. A gentle breeze on the plain becomes a furious wind on the mountain-top. Like all other currents, too, there are differential currents side by side, the velocity along some stream lines being greater than along others. Also in air currents, especially, the velocity varies in time i. e., the wind blows in puffs. These differential currents, both side by side and in altitude in time, would be evident at once if we could see the air. Now, the bird feels these invisible differential currents, and skillfully uses them to lift himself. In soaring, the bird slopes downward with the wind, acquiring thus great velocity, passes into a lower current of less velocity, then turns facing the wind, and shoots up a slope which carries him higher than the level of the start, then turns again in a current of still higher velocity, then descends again along a slope and repeats the same cycle.

To explain more definitely: Observe (1) that the lines of a bird are so fine that the front resistance is almost zero. There is practically only skin friction, which is also small. Observe (2) that with large aëroplane and rapid motion the fall by gravity is also very small. This is proved by experiments of Langley, to be described presently. Therefore, if the differential force of the air currents through which he circles is precisely equal to the skin friction plus the downward tendency, the bird will just rise to the level of the starting point; if greater, he will rise above that level. In order to rise, therefore, the differential force of the successive currents must be greater than this small, amount. As the bird is immersed in the current, and if he uses the whole available differential force in rising, none is left over for progress against the wind. He therefore drifts with the wind.

I have spoken thus far only of differential currents in altitude, for these can always be depended upon, but there may be also differential currents side by side. These might be utilized in the same way. The same may be said also of differential currents in time—i. e., successive puffs or gusts. The bird may take advantage of these. If so, he would slope down with the gust and come back and rise in the interval.

5. Sailing.—Many large birds fly with alternate intervals of flapping and sailing. But in such cases the bird always loses either velocity or height during the sail, which it recovers only by flapping. There is nothing remarkable in this. But some sea birds which live almost continuously on the wing and usually in a high wind, acquire an almost incredible expertness in the use of the wings as an aëroplane, and sometimes go for hours and over many miles of space without flapping once. The most wonderful bird in this regard is probably the albatross. On several voyages from Oregon to San Francisco I have watched these birds with their long, narrow wings rigidly extended, skimming the surface of the sea, then rising and wheeling and swooping, and again skimming, but without moving a feather for hours. I will briefly describe the phenomenon as I have seen it. The explanation will be brought out as I proceed.

I will suppose a wind aft, as was the case in most of my observations. The bird follows the boat, skimming the very surface of the sea perhaps for several hours; then, finding that he is losing ground, wheels about, facing the wind, shoots up to forty or fifty feet above the sea, then turns again with the wind, swoops down a steep incline, acquiring great speed both by the high velocity of the upper stratum and by the descent, then skims the surface again, and quickly overtakes the boat, to repeat the same evolution. With head wind these evolutions are more frequent. As before, he skims the surface behind the boat, but more quickly begins to lose ground; then rises and then wheels and swoops downward, leaving the boat; then, having acquired the necessary velocity, he again turns and skims the surface in air of small velocity, and in spite of head wind overtakes the boat, to repeat the same evolution. For hours these evolutions are repeated, the wings remaining motionless, with only varying position toward the wind.

Here, again, the bird takes advantage of the great difference of velocity between the lower and the upper strata of the air. Doubtless, also, with head wind, advantage is taken of eddies in the wake of the boat. In all cases, if the wind slackens, the bird flaps now and then. If it stops, he flaps all the time. It can not be done in still air.

With the wind on the side the evolution must of course be different, but, as I have not watched this case, I will describe it as given by a writer in Nature (Nature, vol. xliii, p. 223, 1891).

Fig. 4 represents the stern of the boat and the circle of the bird's path. The arrows show the directions of the motion of the PSM V44 D768 Direction of wind the boat and the circling bird.jpgFig. 4.—Arrows show Direction of Wind, of the Boat, and of the Circling Bird. boat, of the wind, and of the bird. In the circle the strong line represents the higher portion of the circular sweeps and the light line the lower portion, almost in contact with the water's surface. It is seen, then, that the bird swoops down with the wind, skims the surface, and then rises against the wind. The sweep of the bird is here represented as a circle, because the boat is standing still; but really it is an advancing spiral, following the moving boat. The explanation here is exactly the same as in soaring, except that the differential force of the air currents is utilized in progress instead of in rising.

One more example: On the ferry boat going across the bay to San Francisco, with strong wind ahead and a little to the right (i. e., coming through the Golden Gate), I have several times seen a gull place himself behind on the left, just opposite the hind deck, and maintain his position with motionless wings for a half mile or more—I say motionless with confidence, because he was so near that I could see his eyes wink.

In this case I feel sure that the motion of the boat created an eddy in which the air was still or perhaps moving in contrary direction—i. e., with the boat, and perhaps also a little upward. If the air currents had been visible, I have no doubt the explanation would have been obvious. What I could not see, the bird felt and skillfully utilized.

Now, if birds, even the largest and heaviest of them, can thus play and gambol in the fiercest wind with the greatest ease and grace, and without serious expenditure of energy, but only by the skillful use of wings as an aëroplane, why can not we by artificial means—i. e., by a machine—do the same? The article before referred to was written to show why we can not.

The outline of the argument, the reader will remember, was as follows:

1. There is a limit to the size and weight of any locomotive machine, whether natural or artificial. This limit is the result of the law that while the strength of material and force of all kinds, whether by muscular contraction or by steam pressure, increase as the square of the diameter of muscle or of piston, the weight of the machine varies as the cube of the diameter of all parts. Thus with increasing size, weight must quickly overtake and pass beyond strength. This limit varies with the kind of machine. The limit of an efficient walking machine was probably reached in the largest land animals of previous geological times. For a rolling machine like a locomotive engine or a bicycle, where the weight is supported on wheels, or in a swimming machine where the weight is supported by water, and where, therefore, in both cases the whole energy is expended in progression, the limit is much higher; and, therefore, a locomotive engine and a whale may be heavier than any walking animal.

2. The limit of weight of a flying machine is very much lower than that of either a swimming, rolling, or a walking machine. The limit of an efficient, manageable flying machine is in fact reached in the largest birds, such as in the condor among long-winged and the bustard among short-winged birds, and to our surprise we find it only about fifty pounds. The condor can barely lift himself from the ground, although when well up he sails with ease and grace. There are, indeed, still larger birds, like the ostrich, but they can not fly. True, their wings are rudimentary, but they have become so only because these birds have passed beyond the flying limit.

3. Now, a bird is admirably constructed for economy of force. Not only is everything sacrificed to the one supreme object of flying, but the animal machine, using fats and starch for fuel and getting energy through the mechanism of nerve and muscle, is admittedly more economical and efficient—i. e., will develop more force and do more work for the same weight of fuel and machine—than any artificial machine yet devised. It seems hopeless to surpass it. Therefore, the weight of a machine that will be able to lift itself in the air can not exceed fifty or a hundred pounds.

4. But it is idle to talk of a flying machine with fuel and engineer and freight being less than many times this limit. Therefore, a flying machine which is anything more than a toy is impossible.

Such is a bare outline of the argument which seemed then—and to a large extent seems now—irrefutable. But Langley's recent experiments certainly put the question in a new and somewhat more hopeful light; and renewed reflection on the whole subject, in the light of these experiments, has led me to some important modifications of my previous views. What, then, if any, are the fallacies in the above argument?[1]

1. We have spoken of two functions in the wing, viz., as propeller and aëroplane. These two are of necessity united in the bird. The bird's wing was not made at once—a special organ for a special purpose—but by modification of limbs. But there was only one pair of limbs to spare for the purpose of flying, and therefore one organ, the wing, was used for both functions. Now, there is probably a disadvantage in this; for short wings with great rapidity and sweep of the stroke are probably the best propellers, while long wings are undoubtedly the best aëroplanes. Thus in long-winged birds, like the condor, the ability to rise quickly and fly rapidly is sacrificed to easy, graceful, long-continued circling and sailing; while short-winged birds, like the turkey and bustard, rise more easily and fly more rapidly, but never remain long on the wing. There seems little doubt that the limit of weight in rising is higher in short-winged birds; and the great flightless birds were of this kind before they passed beyond the limit of flight and their wings became rudimentary. Now, it is needless to say that in artificial flying these two functions may and will be separated. The propeller will be used wholly for lifting and onward progress. I suppose, therefore, that the limit of weight may be raised higher than I have placed it. But if this were all, I can not think that it could be pushed much, if at all, beyond one hundred pounds. But this is not all.

2. We have said that the animal body using fats and starch as fuel, and getting force through the mechanism of nerve and muscle, is more economical—i. e., will generate more force and do more work with the same weight of fuel and machine than any artificial contrivance yet devised or likely to be devised. This is certainly true; but there is another important element here concerned, viz., the intensity of the force—i. e., the amount of force developed and work done in a given time. This depends on the rate of combustion of the fuel. Now, there is a strict limit to the rate of combustion of fuel, and therefore of development of force, in the animal body. This limit may doubtless be greatly overpassed in an artificial machine. But be it remembered that this entails greater weight of fuel and of all parts of the machine engaged in the generation and transmission of force. What we want more specially is (1) a material stronger in proportion to weight than bone. This, no doubt, we have in steel tubes.[2] Here, then, is a positive gain. But (2) we also want a force more intense than muscular contraction, which is, I believe, about a hundred to a hundred and twenty pounds per square inch of cross-section. This can doubtless be surpassed, but not without increased weight of containing and transmitting parts. If anything can be gained in this direction (which is doubtful), against it must be set over the greater economy of the natural machine. The problem is an exceedingly complex one, and can be solved only by careful experiments. But let us admit that, by greater strength of material and greater intensity of force, the limit of weight of machine and fuel which can be lifted in the air may be pushed to several hundred pounds. This, I am sure, is as far as we can go on this score.

3. But the most important new light is found in the effect of motion on the sustaining power of an aëroplane, and the greatest flaw in my previous reasoning is the imperfect recognition of this principle.

As already stated, this principle was first brought out by Marey, and was alluded to in my previous paper, but its supreme importance was not fully appreciated until the experiments of Langley. In his hands it becomes almost a new principle, and one which must modify not only our theory of flying, but even our theory of projectiles. Langley's experiments bring out the unexpected result that in air a body does not fall the same distance in a given time whether it falls straight downward from rest or is affected with horizontal motion—that its motion in the latter case is not a resultant of distance of downward fall from rest and horizontal motion. The same is true of all bodies, but the difference is greatly exaggerated in the case of aëroplanes. According to his experiments, a thin aëroplane of material two thousand times the specific gravity of air, say aluminum, in perfectly horizontal position and free to fall, would take four times as much time to fall a certain distance if moving horizontally twenty feet per second as it would if falling directly downward from rest. With still greater velocities the time of falling a given distance is greater and greater, until it may become almost inappreciable. The reason is plain. The aëroplane falling straight downward must press the air out of its way. It takes time to do this. Now, if it is moving horizontally edge on, before the air can move appreciably, the plane is already on to new still air. Or, to put it more definitely, supposing the aëroplane to be one foot square, then, if falling straight downward, one square foot section of air must be moved in a given time; but if moving onward twenty feet per second, then twenty square feet must be started in motion in the given time. Thus with increasing velocity the air becomes more and more rigid because more and more must be started in motion in given time, until, if velocity is infinite, the air becomes immovably rigid.[3]

We have spoken thus far of a perfectly horizontal plane moving edge on, and therefore with no front resistance. But if the plane be slightly inclined upward in the direction of motion, then the onward motion would tend to sustain the plane. The whole air pressure may be resolved into two parts, one resisting onward progress and one sustaining the plane; and when this latter is equal to the weight the plane will not fall at all. Now, as velocity increases, less and less inclination is necessary to get the requisite sustaining force. But with less inclination comes also less front resistance. Thus at very high velocity the aëroplane may be placed nearly horizontal with proportionally small front resistance and yet sufficient sustaining power. Thus it follows from this important principle that instead of force increasing as the square of the velocity attained (or even higher rate), as in a steamboat, the increase of force with increasing velocity is unexpectedly moderate. This, of course, applies only to the aëroplane. Resistance to the attached machine follows the usual law. But this is small in comparison with the sustaining power of the aëroplane. Therefore, once get a flying machine, even one of great weight, with its aëroplane well up in the air and moving onward, and there seems to be no physical impossibility of sustaining it indefinitely and giving it by means of suitable propellers a great velocity, say of forty to sixty miles per hour.

In the light of this new principle (for such it may be called) Langley and Maxim have constructed models of flying machines, and expect eventually to solve the problem of flying. A small model of a machine which he calls an aerodrone (air-runner) has been constructed by Langley, and was to have been exhibited at Chicago at the World's Congress of Aeronautics. This model was fifteen feet long, with two transverse aeroplanes forty feet from tip to tip. Whether it was so exhibited or not I do not know. Maxim, it is said, is now constructing a flying machine on a large scale in London, but has not attempted yet to launch it. In both of these the aëroplane slightly inclined is the main reliance for sustaining when once in motion; so that the whole power of the engine and propellers is concentrated on rising and progress through the air.

Now, in the light of these experiments, what may we reasonably expect in the near future?

There are many difficulties in the way of success, which, of course, these men clearly see and will try to provide for. These are mainly three, viz.: (1) Difficulty of rising; (2) stability in progress; and (3) safety in alighting. We take these in succession.

1. Rising.—Every word I have said in my previous paper, only modified as to limit of weight, applies here still and without abatement. It seems to be impossible for any machine, natural or artificial, of greater weight than at most a few hundred pounds, to lift itself straight up in the air, or even to maintain itself in the same place like a hovering bird, by the force of propellers alone and without the aid of a balloon. Therefore, there must be some device other than, or in addition to, propellers to raise the machine in the act of starting. But observe, I said straight up. Many birds can not rise so. They must rise at very gentle incline. They must get onward motion before their wings can get full effect on the air. It is said that the mode of taking the condor is to build a pen, say, forty to fifty feet in diameter and six feet high, and put a carcass in the middle of it. The condor alights, but can not again rise at an angle which will take him over the fence. Many heavy-bodied, short-winged ducks rise from the water at so small an angle that they must use both feet and wings for thirty to forty feet in order to get onward motion enough to give effectiveness to their wings by coming in contact with larger masses of still air, as already explained. It follows, therefore, that the flying machine must have some station device to start it. It may be an elevator, but more probably it will be machine rollers on a railway. With aëroplane spread and slightly inclined and propellers directed a little backward, velocity might be got sufficient to sustain and finally with the help of the propellers to raise the machine. As far as I can learn, this is the plan of Maxim.

Viewed in the light of the new principle, there is certainly nothing impossible in this. But every machine is liable to accidents. It is absolutely necessary that we should be able to stop and go on again. Suppose in mid-flight anything should go wrong and it is necessary to alight, how is the machine to rise again without the station device? A locomotive machine that can not stop anywhere and again resume its journey is an impracticable one. This, I think, will prove the greatest of all the difficulties.

2. Stability in Progress.—Once fairly up, as already seen, there is no reason why a moving aëroplane should not sustain a heavy flying machine indefinitely if nothing disturbs its equilibrium. Therefore, once up, we might hope for success in still air, or even possibly in a perfectly uniform current. But air currents are extremely variable in time (puffs and gusts) and in space—i. e., air streams of varying velocities and varying directions. When we see the frantic evolutions of a badly made kite, or of any kite if the steadying string breaks, we are warned of the danger of our aeroplane at high speed and with variable wind, unless skillfully managed, perhaps by means of several independent propellers and adjustable aëroplanes. In the bird we have the last perfection of skill acquired by constant practice and inherited through successive generations. Even if the science of aviation were perfect, the exquisite art necessary to manage such a machine seems almost hopelessly unattainable.

3. Safety in Alighting.—If the last--i. e., stability in progress—be attained, I suppose this also may be. In still air, by checking the velocity by the use of the propellers, the aeroplane would let down the machine with all the gentleness desirable. With head wind, also, there is no reason why alighting should not be successful. With the wind aft, it would be necessary to turn about and face the wind, as a bird does under similar circumstances. A 'parachute, with tubular opening atop, descends with perfect steadiness.[4]

4. To all these difficulties we must add the enormous hazard of a first attempt, the apparent impossibility of approaching success gradually, and thus practicing the difficult art of managing with safety.

Conclusion. Under present lights, therefore, it is no longer justifiable to say, as I have previously done, that a flying machine is physically impossible. I therefore retract that expression. But the engineering difficulties are enormous and possibly insurmountable. At the present time the nearest approach to success in aerial locomotion is still to be found in the French dirigible balloon—i. e., a balloon propelled and steered by machinery—and for some time to come the best success may still be looked for in that direction. In fact, the art of managing a true flying machine is so refined and the skill required so great, and in the absence of such skill the danger of a first attempt is so extreme, that probably the only way to achieve true flight would be by the use first of a dirigible balloon, and then gradually to decrease the sustaining gas and substitute the aëroplane and propeller.

In the distant future, and by means of such gradual approaches, the engineering difficulties in the way of a true flying machine may be finally overcome. If so, then we may look for the greatest success in the direction of the work of Langley and Maxim.

Addendum, January 23, 1894.

The above article was finished and sent to the publisher some time in October, 1893. In the January number of the American Journal of Science Prof. Langley published an account of another epoch-making series of experiments bearing on this subject. In his previous series he showed the enormous importance of onward movement in the sustaining power of an aëroplane. In this he shows the enormous variation of velocity in air currents from moment to moment. The whole air is in a violent turmoil from varying currents. In the above article I have shown that soaring and sailing are impossible without differential air currents; but the amount of difference of velocity of these currents shown by Prof. Langley was wholly unexpected. These experiments, therefore, show that the supply of force from this source available to the bird or to the flying machine is far greater than previously supposed. While they do not seriously vitiate any of my conclusions, they certainly place the subject of artificial flight in a still more hopeful light.

Mr. Bruce, of the Dundee antarctic whaling fleet, describes the whole of the district south of 60° south latitude as strewn with icebergs, which become very numerous south of 62°. On one day the author counted at one time from deck sixty-five of great size, besides many smaller ones. The highest berg seen from his vessel, the Balaena, was about two hundred and fifty feet high; but many were not more than seventy or eighty feet, the average possibly being about one hundred and fifty feet high. All these bergs are tabular, or weather-worn varieties of the tabular forms. They become pierced with caves, and these are sometimes connected with funnel holes, through which, as the swell beats up the caves, immense columns of spray are projected. They may be finely castellated, pillared, or arched. One was beautifully conical. The base of the bergs was colored pale brown by marine organisms, and other brown streaks were seen beyond the water-level. No luminous glow was observed. "Clothed in mist, they raise their mighty snow-clad shoulders to a stately height, or shine forth brilliantly in the sun. Although they are of the purest white, yet they glow with color. The crevices exhibit rich cobaltic blue, and everywhere are splashes of emerald green."
  1. Among these possible fallacies or oversight of my previous article I have not thought it worth while to mention the difference between the reciprocating motion of a wing and the steady pressure of a screw propeller such as would probably be used in any artificial machine; because I believe that, in the comparison, what is lost in the bird's wing in recovery for another stroke, is gained in the application of the force in the direction of greates efficiency.
  2. It is probably a mistake to suppose that aluminium or any alloy of that metal is stronger, weight for weight, than steel.
  3. A striking illustration of this principle is seen in the extreme rigidity of the jet issuing from the nozzle of a hydraulic pipe. The water is under a pressure of three hundred or four hundred feet head, and is projected with a velocity which would cut in two a man's body. If the jet is struck with a crowbar, the bar rebounds as it would from steel. In penetrating, say, half an inch, the bar encounters an immense quantity of water at once. It is evident also that the same principle must apply to all bodies moving in the air, and therefore also to projectiles. There is, then, a kind of truth in the popular notion that velocity holds up or prevents the fall of a rifle ball—not, indeed, that the velocity itself holds up the ball, as popularly supposed (for it would not do so in a vacuum), but that the air is more effective in sustaining a moving body than one falling directly downward from rest.
  4. The Chinese have most ingeniously utilized this principle in the construction of little kites shaped like a bird, with wings and tail. These require no long, steadying tail, because the wings are made tubular at the tips, and the outrush of the air keeps the kite steady.