Hi,
I agree that the diagram could be clearer, but the idea is that the dashed line coming in from the right is hidden until it passes under the loading tipple and then emerges as the solid line extending to the left. Likewise the visible, solid line approaching from the right disappears under the tipple and continues as the hidden, dashed line to the left. The two are roughly parallel under the tipple and don't cross. There's a clearer diagram of the same concept on page 117 of John Armstrong's Creative Model Railroad Design (originally published in 1978 as John Armstrong on Creative Layout Design).
Both tracks passing under the tipple would have to be at the same elevation at that point, but the line that disappears there would obviously have to start downgrade to pass under itself, and the hidden line approaching from the right might be climbing up to the level of the tipple.
John would expect that some scenic trickery would be involved, which would probably be easiest to pull off if the tipple were close to eye level. One or both of the hidden tracks might reach the tipple inside some sort of building that isn't indicated in the diagram. John employed quite a few visual tricks on his own layout, and they're a feature of may of his layout designs.
Good luck with your layout,
Andy
Andy Sperandeo MODEL RAILROADER Magazine
Below is a unit train loading schematic as shown in John Armstrong's Track Planning for Realistic Operation p.69. Is the result of this plan to model a train entering the tipple from the southeast "empty" and then continuing through the tipple and out the northwest side "loaded"? If this is the desired effect, my question is regarding the hidden trackage of the "loaded" train approaching from the east while the visable "empty" train is also approaching the tipple from the southeast. How does one effectively hide the approaching "loaded" train (which appears to me to be in the line of viewing of the "empty" train) without also blocking the view of the "empty" train which is suppossed to be entering the tipple for a load? Is there a vertical difference in track height? This is all probably a lot easier than what I am envisioning, but thanks for the help.
Modeling the UP Sidney Sub in Nebraska & Wyoming in fall of 1999 in HO scale