A Tax~Omic And Lf.Pidological Study. Of ''Lhe .South African Heterosoma Ta With Special R.Eference To The Value Of Scale Characters • In Identification Amd Classification

During the course of some years of research into growth-rate and age of South African fishes of commercial importance, amongst wbich are included various species of Heterosomata, 1 the scales of sucb species received a great deal of attention.· In a very short time the characteristics of the sea les thus . examined became so familiar, that it was found poseible to identify any particular species solely from its scales. In a great many instances scale-samples, labelled with the local common name of the fish f'rom which they were taken, were received from interested persons engaged in the commercial fisheries in various ways. Very often in such cases, the actual species could not with certainty be. ascertained from the local common name appli 3d 9 as, a long the 2000 miles of coastline of the Union, the same species may have a great variety of names, differing from place to plaoe,, or totally different species may have thP. same name at widely separated places. In euch obscure cases, due to the falriilierity acquired by experience, poeitive identifications were readily' rapidly and conveniently made by examination of the sea le-samples. It was considered, that if the diagnostics of scales could be of so much assistance in identifying the individual species of a heterogeneous collection of epectea, such as that composing the South African f'ood-fisbee, it would be of considerable interest and value to determine to what extent ecalecharac ters could be utilised in identification and taxonomy, both-independently and accessory to normal morphological diagnos tics. The desiderata of such a study was considered best ~--- fulfilled by a homogeneous group of fishee in which varying degrees of rela tionshipe between sub-groups and individual species were to be found. For the purpoees of a South African etudy of this nature, the order Heterosomata, which is represented by all f'ive of the recognised families and a goodly number of subfamilies, genera and species in South African eeae, was selected. A secondary but nevertheless important reason for tipping the balance in favour of' t.his selection is to be attributed .. to. the large amount ·of material available for study in the almost complete collections of l