Morphometric statistics (“log sem”) based on standard errors of the m-coefficient (sem) associated with general equations of the form y = mx + c have been calculated from least squares regression analyses of cranial measurements of two chimpanzee species: Pan troglodytes (currently distributed north of the Congo river) and P.paniscus (the bonobo) distributed south of that river in the Democratic Republic of Congo (DRC). A phenetic tree has been obtained from UPGMA analysis of the log sem matrix for the two species of Pan. This study focusses primarily on P. paniscus. The phenetic tree is related to locality data for individuals collected in various regions of the DRC. A few specimens attributed to P. troglodytes are “misfits” in the sense that they occur within the P. paniscus group. An explanation is provided by invoking a scenario in which there was not always a geographical barrier in the form of the Congo river, and hybridisation could occur. Genetic evidence points to at least two periods within the Pleistocene during which hybridisation took place. It is proposed that log sem values based on anatomy are indirectly sensitive to variability in gene pools. Alpha taxonomy (assuming clear boundaries between taxa) is not necessarily always possible, hence the need for a probabilistic species definition (“sigma taxonomy”), without assuming boundaries (Thackeray, 2018). These concepts are relevant to challenges associated with the taxonomy of early Pleistocene hominins, especially if there is no clear boundary between species representing Australopithecus and early Homo.