It is a common procedure to describe the fission-fragment mass distributions of fissioning systems in the actinide region by a sum of at least 5 Gaussian curves, one for the symmetric component and a few additional ones, together with their complementary parts, for the asymmetric components. These components have been attributed to the influence of fragment shells, e.g. in the statistical scission-point model of Wilkins, Steinberg and Chasman. They have also been associated with valleys in the potential-energy landscape between the outer saddle and the scission configuration in the multi-channel fission model of Brosa. When the relative yields, the widths and the mean mass-asymmetry values of these components are fitted to experimental data, the mass distributions can be very well reproduced. Moreover, these fission channels are characterised by specific values of charge polarisation, total kinetic energy and prompt-neutron yields. The present contribution investigates the systematic variation of the characteristic fission-channel properties as a function of the composition and the excitation energy of the fissioning system. The mean position of the asymmetric fission channels in the heavy fragment is almost constant in atomic number. The deformation of the nascent fragments at scission, which is the main source of excitation energy of the separated fission fragments ending up in prompt-neutron emission, is found to be a unique function of Z for the light and the heavy fragment of the asymmetric fission channels. A variation of the initial excitation energy of the fissioning system above the fission saddle is only seen in the neutron yield of the heavy fragment. The charge polarisation in the two most important asymmetric fission channels is found to be constant and to appreciably exceed the macroscopic value. The variation of the relative yields and of the positions of the fission channels as a function of the composition and excitation energy of the fissioning system obey a hidden systematics that can be explained by the number of states in the vicinity of the outer fission barrier as a function of mass asymmetry, if the potential is constructed as the sum of the macroscopic contribution of the compound nucleus and empirically determined fragment shells. This hidden systematics also explains the transition from asymmetric to symmetric fission around 226 Th and around 258 Fm.