On a another forum it has been suggested that for tunnelling the loading gauge width had to be less than the average width of a mule's rear-end and this is how the gauges were originally derived. This seems apocryphal.
Is there a reference to where this was designed and changed over time. I expect it would be in the Permanent way journal if I had online access.
Almost totally false.
The standard track gauge (not loading gauge) was popularised by Stephenson when he started building main line railways. He kept the track gauge used on the colliery tramways in the North East that he had worked on previously. Before steam locomotives these were horse drawn, so the track gauge would have been chosen so a horse could walk along between the rails while hauling wagons. It was probably also influenced by the size of the carts used on the roads, because wagons would naturally be built to the same size, and on some early railways the wagons could be used on roads as well. Other early railways used different track gauges, but most were between 4ft and 5ft.
The loading gauge is something different. It specifies the width and height of the trains themselves, and therefore also sets the structure gauge which is the width and height of bridges and tunnels next to and above the track. Those original wagons usually had the wheels on the outside, but quite soon trains started getting wider and overhanging the wheels. The loading gauge has to be wider than the track gauge but different countries have different loading gauge while using the same track gauge - some even have a loading gauge much larger than the UK while using a narrower track gauge. The UK has about the smallest loading gauge of anywhere for standard gauge track.
A version I've seen of this urban myth states that the space shuttle solid rocket boosters were built to a particular size so they could be transported by train from the factory to the launch site, and inaccurately tracks effect back to cause to claim this is based on width of a mule. From a quick search these have a diameter of 12.17ft, so all I can say is that they must have had some pretty big mules in America.
As to where the loading gauge came from, that's an interesting question that I can't directly answer but it probably also comes back to the dimensions adopted by Stephenson on his main lines, as some of his bridges and tunnels are still in use today. I recall reading that he set the Liverpool and Manchester track spacing so a train with a large load could run on the middle two rails, so their spacing must have been 4ft 8.5in between the nearer edges of the inner rails. This would have set the maximum train width to twice the track gauge less an amount for clearance, which pretty much agrees with the 9ft or so of the various UK loading gauges today. The spacing between the tracks was increased sometime later, and is normally now 6ft or more.
