The histogram of the J distribution h2JIntrins is currently plotted for some spins g_dPlotSpMax only, which may not be the same number as the spins created in he continuum bins, g_nConSpbMax:
https://github.com/fzeiser/RAINIER_Fork/blob/31cfda33869c32fa2d787a0b8559a733dbc97ef2/RAINIER.C#L1464-L1468
This may also be an issue when you populate spins with the bExFullRxn population method: If the number of spins g_nExPopI in the populated spins table is not equal to g_nConSpbMax, unexpected things may happen.