According to Newton's laws of motion, a moving object not acted on by any force continues to travel in a straight line. For objects near the earth's surface there is, however, always the force of gravity pulling directly downwards, and it is this that brings things back to earth. But there is no similar sideways force to deflect anything, and so as a rule the path in which an object moves freely will be expected to lie in a vertical plane.

But any cricketer knows that on occasion this no longer holds, and that the ball may move out sideways from this vertical plane, and do what is called swerve in the air. This is what is meant by the term "swing bowling". We are not here concerned with what a ball may do on striking the ground, but only with effects during its flight through the air.

It seems probable that swerve bowling has come into more frequent use during the present century, simply because the rule allowing access to the all-important new ball, not only at the beginning of an innings, but at the latest after 200 runs have been scored, was first introduced only in 1907. Hitherto a new ball made its appearance only at the start of each innings. The practical effect is that on average a new ball is now available every 130 runs or so, as a few sample matches will show, instead of after the average length of an innings, which is more like 230 runs (Note :- this was the case when Raymond Lyttleton gave this talk. The rule has since been altered.)

Swerve occurs not only in cricket, but also in golf, tennis and baseball, not to mention other games. But in these, swerve is produced by spinning the ball, whereas in cricket the ball can be made to swerve without imparting any significant degree of spin to it, solely because of the presence of the seam on an otherwise smooth surface, as we shall explain.

If by some magic cricket or any of these games could be played in a vacuum, then Newton's laws would operate directly, and no swerve at all could occur. So we must look for some effect arising from the presence of the air for the cause. It is not always realised just how large the forces are that are exerted by the air on an object. The atmosphere presses on everything with a force of around one kilogram per square centimetre of surface. So on a cricket ball, the opposing pressures on any two hemispheres are each about 40 kg. If by any means some kind of pressure difference, as between the left-hand side of the ball and the right-hand, could be brought about, so that these opposing forces failed to balance by even as little as one part in a thousand, a sideways force of 40g would come into play. This is about a quarter of the weight of a cricket ball, and so could easily make the path curve sidewards appreciably. The deviation would be about 30cm during the motion lasting half a second, which is roughly the time the ball is in the air for fast bowling.

But in order to see how any difference of pressure on opposite sides of the ball could arise, it is first necessary to understand in some detail what happens to the near air as the ball travels through it. We can see that the air must be pushed aside and parted as the ball moves along, but in fact, the air does not simply slide by the ball and join up again at the back in an otherwise undisturbed way. Whenever a fluid such as air streams past a solid body, the particles of fluid actually in contact with the body adhere to it and move at exactly the same speed as the body itself. In other words, the air immediately in contact with the ball is at rest on its surface. On the other hand, not very far away, the air is quite obviously more or less undisturbed - just as for a boat, the water only a few centimetres from the side appears to flow by almost undisturbed. On the forward part of the ball at any rate then, where it is breasting its way through the air, this means that in an extremely thin layer just outside the surface the motion of the air changes from that of the ball itself to almost no movement at all. For the speeds a cricket ball is bowled at, which range roughly from 60 to 130 kilometres an hour, this boundary layer, as it is termed, is less than a millimetre deep, but its thickness is least at the front of ball and gradually increases round the sides. We can picture it as a very thin skin clinging to the ball at its inner side but slipping by with the air at its outside.

If the speed is extremely small, this boundary layer will cling to the surface almost round to the rearmost point before it finally leaves the ball in a narrow stream behind it. But when the speed is increased to anything like the rates that occur in cricket, things are very different, and the layer breaks away about half-way round the ball at its sides. Immediately behind the ball, however, the air is violently disturbed, and forms an irregular eddying wake that gradually diffuses back into the ordinary air as it drifts away behind the ball. This wake is exactly contained within the boundary layer that streams off the sides of the ball. The existence of this turbulent wake can be demonstrated visually in the experiments in which the ball is held at rest and the air made to stream past it, by introducing smoke into the air immediately behind the ball.

Now imagine the speed of the air past the ball to go on steadily increasing. Then the place on the surface where this smooth boundary layer flow ceases gradually moves forward on the ball. But there is a limit beyond which it never comes. This is not quite half-way round the ball from the front - actually, it is about 80 degrees from the nose of the ball. The eddying wake streams off at a tangent from this part of the ball to form a gently widening region trailing away behind and gradually melting back into the undisturbed air. The energy to produce the turbulence in this wake is supplied by the ball's motion, which accordingly experiences a resistance slowing it down. One might think that resistance would come mainly from the drag of the air slipping by at the sides. But this is not so. The drag is negligible compared with the effect of the low pressure region immediately behind the ball, which can be thought of as a partial vacuum sucking the ball back.

If the speed is still further increased, an entirely unexpected thing happens next. The boundary layer now begins to creep further round towards the back again as the speed rises, before finally breaking away to form the edge of the wake. The wake behind the ball grows much narrower as a consequence, and measurements show that once this stage is passed the resistance to the motion begins to decrease quite abruptly. At sufficiently high speeds, the boundary layer reaches three-quarters the way round the ball, that is to within 45 degrees of the rearmost point before peeling off.

For any particular size of ball there is a "critical speed" at which these things begin to happen, and above which the resistance force suddenly starts to drop. For a perfectly smooth metal sphere the size of a cricket ball, the boundary layer begins to extend rearwards again when the speed reaches 145 kilometres an hour. The upper limit, when the boundary layer has got round to the back as far as it can, occurs at about twice this speed. This is for a smooth ball. But for a rough ball, as we shall see, this critical speed may be considerably less.

Now increase of speed above this critical velocity is not the only way in which the boundary layer can be made to cling further round the surface. This can be achieved at speeds much less than the critical one by attaching small raised ridges to the otherwise even surface of the ball; in other words by roughness at suitable parts of the surface. At first sight one might think any such irregularity would so disturb the flow near the surface as to incite separation of the boundary layer, but, in fact, just the reverse happens. A thin wire ridge, for instance, placed transverse to the air-stream round the forward part of the ball, results in the boundary layer adhering further round than it otherwise would, for a given air speed, before lifting off into the wake. The breadth of the turbulent wake is again consequently narrowed, and more important still the air resistance opposing the motion of the ball is lessened owing to the surface irregularity. This is the reason why grooves and dimples are patterned on the surface of a golf ball. Were they not there, the ball could not be driven anything like as far, because air resistance to a smooth ball, at these critical speeds, is more than four times what it is on a ridged ball.

Ball moving below the critical speed. Symmetrical Flow.