The characters of existing genera are discussed and it is shown that confusion has arisen because the anterior segments have been telescoped and there are two different systems of numbering them. It is proposed that the separation of genera be based primarily on the number of gills and the number of uncigerous thoracic segments. The main diagnostic characters of 49 type-species have been tabulated and it is suggested that 16 genera which are mostly monotypic be sunk. Keys and definitions of the remaining 33 genera are provided.