No.1700 (MC)

(C- ; partial test by Jacobi )

The position is almost orthodox, except that white Pawn h5 cannot be explained orthodox-wise.
From initial position, a Pawn has to be transfered from a-file to h-file, by way of a capture with change of color for each change of file.
Thus, the 4 missing white pieces (wQd1, wBf1, wSb1 and wPa2) were captured by a×b, ç×d, é×f and g×h and the 3 missing black pieces (bQd8, bRa8 and bSb8) were captured by b×ç, d×é and f×g. wPa2 had to promote without capture in order to be captured (Ceriani-Frolkin).
"Horizontal zigzag Excelsior" by black Pawn a7.

I am rather confident that the problem is correct (with "human error" margin)