of the limited nature of the material, I have not been able to follow up Barnard's suggestions (1932) about the shape of the epistome. I have sorted out the New Zealand material available and find myself with what I consider four distinct species, one of these occurring in two forms. There are two difficulties which occur with this material. The first is in deciding which characteristics are of specific value within the genus; and the second is in establishing the precise relationships of the New Zealand material with previously described species. Generally speaking, I have accepted the same criteria as Schellenberg (1931). In particular, I have accepted the stability of the epimeral plates despite suggestions to the contrary by Barnard (1930: 365) and others because it has not yet been proved that their shape is variable in this genus although there are a number of records of differences. I do not discount these records, but the general similarity of species suggests the possibility of different species being confused in these particular instances. The remarkable stability of the 3rd epimeral plate throughout the order as a specific criterion makes me hesitate to accept variability without very good reason. The dorsal spination is more open to criticism, but it does seem to have some considerable specific value, although it is a characteristic to be used with caution. The number of setae on the inner plate of the first maxilla used by Pirlot (1939) seems to me, after reviewing my material, to have also a limited application. However, some other characteristics appear to have value, and to have been overlooked previously.

The serration or otherwise of the second peraeopod sideplate posterior margin appears constant as does the spination or setation of the propod. The number of teeth on the dactylos of the gnathopods varies slightly but does seem to be of some specific value. I do not think the number of teeth on the outer plate of the first maxilla is sufficiently constant to be used as a specific criterion.

Certain information which I would like for the older established species is not available from the literature; consequently I cannot be completely happy about the following identifications but, if my assumptions as to the stability of the epimeral plates are correct, most of the difficulties disappear.