The historical development of the dosimetry systems for Hiroshima and Nagasaki is outlined from the time immediately after the A-bomb explosions to the publication of the dosimetry system DS86 in 1987, and the present status of the so-called Hiroshima neutron discrepancy is summarized. Several long-lived radionuclides are discussed with regard to their production by neutrons from the A-bomb explosions. With the exception of 63Ni, these radionuclides have not, up to now, been measured in samples from Hiroshima and Nagasaki. Two of them, 63Ni in copper samples and 39Ar in granite samples, were predominantly produced by fast neutrons. 63Ni can be determined by accelerator mass spectrometry with a gas-filled analyzing magnet. It should be measurable, in the near future, in copper samples up to 1500 m from the hypocenter in Hiroshima. 39Ar can be measured in terms of low-level beta-counting. This should be feasible up to a distance of about 1000 m from the hypocenter. Three radionuclides, 10Be, 14C, and 59Ni, were produced predominantly by thermal neutrons with smaller fractions due to the epithermal and fast neutrons, which contribute increasingly more at larger distances from the hypocenter. State-of-the-art accelerator mass spectrometry is likely to permit the determination of 10Be close to the hypocenter and of 14C up to a distance of about 1000 m. 59Ni should be detectable up to a distance of about 1000 m in terms of accelerator mass spectrometry with a gas-filled magnet. The measurements of 10Be, 14C, 39Ar, 59Ni -- and potentially of 131Xe -- can be performed in the same granitic sample that was already analyzed for 36Cl, 41Ca, 6Co, 152Eu, and 154Eu. This will provide extensive information on the neutron spectrum at the specified location, and similarly complete analyses can conceivably be performed on granite samples at other locations.