that he has abandoned it. Should this criterion be replaced and, if so, how? Or should the two genera be united?

There are undoubtedly some species which are more or less intermediate between Orchestia and Talorchestia. These are almost without exception those showing terrestrial tendencies. The typical littoral species are easily separated. For this reason, I believe the two genera should be kept separate. It seems to me that we are here dealing with two closely allied genera showing considerable degree of convergence, and that relationships are better indicated by retaining both genera.

Where the first gnathopod is intermediate, the other characteristic which Stebbing gives for Talorchestia, the constriction of the second peraeopod dactylos, usually holds. Apart from these two characteristics, there seem to be no others which will apply to more than a few species. Chilton's fairly extensive diagnosis of the genus is applicable to the three species which he described from New Zealand, but is not universally applicable. Most of the characteristics which have been used to separate the genera in accounts of small faunas are unsatisfactory when applied to the genera as a whole.

The situation seems to be this. There is a definite Orchestia complex and a definite Talorchestia complex. These can easily be distinguished by one or both, usually both, of Stebbing's characteristics. Between these there is an intergrading series which can be assigned to one or other of the typical complexes on whichever characteristic is present, and on general facies. In general, the truly terrestrial species belong to Orchestia (in which I include Paiorchestia), and Talitrus.

It may be preferable to reword the generic diagnosis after Chilton (1917): “Like Orchestia, except that gnathopod 1 in the female is simple instead of being subchelate. Peraeopod 2 has the finger notched or otherwise modified and differing from that of peraeopod 1”.

Stephensen (1948) lists the species of Talor chestia as then known. To these should be added T. maicuzzi Ruffo. (1950).

I have omitted Talorchestia patersoni Stephensen (1938) from this paper because I consider it a species of Orchestia. It resembles Orchestia species more closely in general facies, the second peraeopod dactylos is not markedly constricted, and the gnathopods are neither definitely subchelate nor definitely simple. In habitat, it is truly terrestrial, being a leafmould inhabitant.