NavList:

I’ve been tossing this refraction business around since it was first mentioned.  I must admit I didn’t think seriously about in my initial calculations because I was just looking for ballpark figures, i.e., to the nearest couple of minutes of time and arc.  However, I did start to think about it when people started going into greater detail than I ever felt necessary.  This was the conclusion that I came to, but it seemed too simple to risk suggesting in such august company.  I first considered an observer at the centre of the Earth and the Earth having no atmosphere.  You could lie a long stiff wire from the touching point of the Sun through the touching point of the Moon to the centre of the Earth.  This mucked up my very first attempt at the initial problem posed.  The Moon is too close to the Earth, so you must allow for PinA, which gives us a different line for the wire.  However, once you’ve settled on that, the wire isn’t bent towards the Earth by the Earth’s atmosphere until it gets very close to the Earth, so the height observed of the touching point, and indeed the entire view of Sun and Moon, might need reducing by one minute of arc, but that’s all that’s affected.  DaveP