Here are some close pictues of the right and left side of the socket in the base.
The right side shows a distinct seam between the socket piece and the base. Heavy grinding marks on the socket, but not much on the base. Not very clear if the grinding marks on the base were done at the same time as the socket marks.
The left side pictures tell a pretty clear story to me.
The red arrow points to the grinding marks on the socket. The green arrow points to the marks on the base. They are not quite parallel.
I've put a pointer up against the socket assembly in another picture. Again, red arrow, socket marks, green arrow base marks. What is significant here is that the pointer is stuck up against the ridge of the socket assembly. It is higher than the base.
If the wheel doing the grinding on the socket made the marks by the green arrow, it would have had to grind some of that ridge off. It did not.
If the wheel doing the grinding on the base and made the marks on the socket, it too would have ground off some of the ridge.
If you recall the seam you can see in the pictures from the top of the socket, there is a distinct start and stop to the insert for the socket.
My conclusion is that the socket part was cast separated, cleaned up with a grinder, then placed into the mold for the base. There likely was a recess in the patter to hold the piece as the sand was rammed.
I could be wrong, but it seems to be the simplest answer for what I see. It would be interesting to see more examples.
I'll bet you think this is the end of the story :D- no way, read continue on.
Tom