The actual force of air resistance exerted on a cricket ball is surprisingly large. With a critical speed of about 30 metres per second, just above this the air resistance directly opposing the motion is about half the weight of the ball. But just below it, the opposing force becomes nearly twice the weight of the ball. The precise value of the critical speed depends very much on the condition of the surface of the ball, and for a rough ball might well be still lower. For instance, if it were 20 metres per second, that is about 70 kilometres an hour, the air resistance would be about a quarter the weight of the ball above> the critical speed, and equal to the weight below it. So for a bowler operating at and near the critical speed of the ball, considerable difference of trajectory would result from very minor changes in the starting speed. This shows that the claims one hears about the ball being flighted in the air, of its being made to dip suddenly, and such like, although not usually stated in precise terms, may nevertheless have a definite basis in mechanics. All these effects, and that of swing itself, depend on the fact that the density of the air is just right for them to happen. If it were very much denser or rarer, none of these peculiarities would enter. So if cricket is played on Mars, swing bowling and flighting the ball will be unknown in that thin atmosphere.
To come now to the effect of the seam. As you know, a cricket ball has a prominent band of six lines of stitches running parallel to each other right round it and holding the two halves together. This band is about 2cm wide, and when the ball is new the four outer lines of stitches stand out about half a millimetre - that is, something approaching the depth of the boundary layer. Also, when the ball is new, the rest of the surface, nearly 90 per cent of the area, is smooth and shiny. As the ball is used, the seam gradually becomes flattened down, and the rest of the surface loses its shininess.
If we imagine the air streaming horizontally past the ball, with the general plane of the seam upright and in the direction of motion through the air, then everything will be perfectly symmetrical on the two sides of the ball, and there can be no sideways force. But things become very different if the plane of the seam is turned to one side round the vertical axis. Suppose, for example, that the forward part of the seam is turned towards the bowler's left, so that the plane of the seam runs from mid-on towards the slips; that is, turned at say 30 degrees to the direction of motion through the air. It is at this point that we come to the crucial effect that the seam can have. The roughness provided by the stitches at the front, which are now slightly to the left, will operate to maintain the boundary layer flow more than half-way round on this left-hand side of the ball. But on the right, the stitches will be too far round at the back to have any influence on the boundary layer, which will already have broken away somewhere about half-way round the ball. The flow on the two sides is no longer symmetrical, and we arrive at a situation in which a sideways force can occur. On the left, where the flow is undisturbed over a greater range of boundary layer, the total air pressure is less than it is on the right, and the ball accordingly experiences a force to the left that will deviate it sideways and make it move towards the slips. This force can rise to nearly half the weight of the ball on occasion when everything is just right for it. If the seam is turned the other way, with its plane running from mid-off down to fine-leg, the situation is reversed, and the sideways force is to the leg side.
Ball moving below the critical speed, but flow rendered unsymmetrical by
the effect of the seam. Ball viewed from directly above.
We can see also at this point that a rough ball will not swerve. This is because its surface roughness will operate equally everywhere to keep the boundary layer fully extended and quite symmetrical on the two sides, and there is no further effect that the seam can have. We can also predict that a ball sufficiently roughened on one side, but smooth on the other, could be made to swerve even though it had no seam at all.
We should perhaps explain how it is that the seam can remain fixed in the right position as the ball moves through the air and does not stray away to spoil the effect. The reason is simply the slight amount of rotation, almost in the backward direction, that is automatically given to the ball as it leaves the bowler's hand. The seam merely turns in its own general plane just as the rim of a wheel does. This, of course, may not always happenm and then the ball will not swerve.
One of the most intriguing features of swing bowling is the so-called "late swerve" in which the sideways deviation shows little or no signs of occurring until late in the ball's flight, usually as it begins to dip downward towards the ground. There are two possible causes that could bring this about; both could independently produce a late effect, and if operating together would combine to produce it in greater amount.
The first arises from the mere position of the seam relative to the air-flow past the ball. We have so far thought of the ball as travelling along a horizontal straight line with the plane of the seam vertical but turned slightly to one side. But, in fact, the later part of the flight, just before the ball reaches the ground, is inclined distinctly downwards at an angle of some 20 degrees or so, the exact value depending on the speed. If, therefore, the seam is in the most suitable position for swerve during this downward part of the path, it would necessarily be in a less suitable position at the early horizontal part, and vice versa. So by altering the grip, the bowler could take advantage of this changing direction of motion and obtain the best effect late in the flight.
The second independent cause could arise from the existence of this critical speed above which, as we have seen, the ball will not swerve at all because the place of separation of the boundary layer is already so far towards the rear that surface roughness in the shape of the seam can have no effect, and the flow is perfectly symmetrical on the two sides of the ball. The seam cannot affect this situation. But the resistance of the air will diminish the speed and bring it to values at which swerve effects can come into play. So, if the bowler starts the ball off at the merest fraction above the critical speed, swerve may occur in the later part of the flight only, when the speed has dropped below the critical amount, and the seam thereby given its chance to operate.
Clearly, by combining these possible ways of producing late swerve the greatest effect will be obtained. This would be a question of experimenting not only with the grip, but also with the speed, and obviously it would be a very elusive thing to get all the factors right, as is well known to be only too true.
The foregoing explanation of swerve is based on the results of practical experiments with various-sized spheres in the airstream of a wind-tunnel. Even under the more or less ideal conditions so provided, it is found that the critical speed is highly sensitive to the surface roughness, and also to seemingly minor irregularities in the airstream itself, such as the slightest degree of inherent turbulence. These factors are found to be capable of reducing the critical speed by fully 50 per cent.
There seems little doubt that it is features corresponding to these that explain why it is that a bowler may at one time succeed hugely in obtaining swerve and yet at another fail altogether to get the effect. There are strong indications that humid, sultry weather somehow affords the most suitable conditions, and one may conjecture that this may involve very small scale irregularities of motion in the air, or it may affect suitably the surface smoothness of the ball, or, of course, even do both. It has even been suggested that humidity might swell the seam stitches to make them even more effective in relation to the depth of the boundary layer.
Another strange thing is that an individual new ball, apparently indistinguishable from any other new ball, may not swerve at all, in so far as this can be established on the cricket field. So evidently the various factors, if indeed they have all yet been appreciated, combine together in an obscurely complicated way, and it is impossible at present to say in detail precisely what are the best conditions, of atmosphere and surface of the ball, conducive to the effect. In problems of this kind, when several factors may be operating to produce or hinder a single main result, the analysis of the underlying causes can be one of the most difficult things to carry out. Unfortunately, the theoretical treatment of such problems is beyond the resources of mathematical analysis, at any rate the present time, and, of course, no amount of mere verbal discussion of the various factors can ever finally settle their individual contributions. The only remaining way to study the matter is by direct experiment, in which the various factors, supposing them to have been appreciated, are under control and capable of being varied at will. Such experiments are, in fact, at present being carried out by means of a wind-tunnel with the object of achieving a fuller understanding of this intricate question of swerve. With a wind-tunnel the air-speed can be adjusted to any desired value, while the ball can be dropped vertically through the stream to give a situation completely equivalent to a ball projected through the air. Beside the speed, the things that can be varied are the angle of the seam to the air-flow, the size of the stitches, the smoothness of the ball and, of course, of each side of it separately, and even the humidity of the air flowing through the tunnel. These are the factors that appear most likely to be involved, but possibly there may be others not yet thought of, whose existence will gradually emerge if it does not prove possible to interpret the results on the basis of the assumed causes only.
If all the factors can eventually be isolated, it would then become possible, instead of leaving it to chance whether a ball is suited to swerve bowling or not, to have a standard specification for a cricket ball, just as at present there are standards for size and weight. But perhaps this is a dangerous matter to bring up, suggesting as it does the possibility of secret researches to discover how to make cricket balls that will swerve, and also ones that will not.