> and if they turn off the track the length of the wheel changes But is doesn't actually change. A different part of the wheel comes into contact with the rail

> and if they turn off the track the length of the wheel changes But is doesn't actually change. A different part of the wheel comes into contact with the rail, but that different part always had a higher linear speed. It just wasn't in contact with the rail until the turn. But with these bearings, it appears that the balls themselves actually do speed up or slow down to maintain separation from one another. In that case, there must be a force coming from somewhere to effect those changes.