Based on the black paint on the ends of one face and the black paint on the upper prism, looks to me as if the loose prism slide under the upper prism and you have it oriented. I am not familiar with this prism setup but it does remind me of the Kodak Medalist system. In that case, the two prisms create a true split, meaning and upper and lower image. Not a convergent image like your image in the beginning shot. So consider if this is the actual system- two images stacked on top where one side moves left to right and 'focus' is when they converge- a vertical line in the top image continues into the bottom image. It is by adjusting the mirrors feeding this system that you adjust the vertical position of the images. So the top and bottom image are continuous. Looking at a triangular shape will make this clear.
Not certain if there needs to be a tight cemented joint between the upper and lower prisms. On the Medalist thisis a cemented pair and the line between the two images gets ugly if this joint fails. But I wasn't seeing the top of the loose prism showing signs of an optical joint.
Well, I could be very wrong about all of this, just extrapolating from other cameras (Kodak Chevron has something similar but there is a small area of convergence where the top and bottom image meet, so you have both- a true split with a thin band of convergence).