S. convexa n. sp. has rounded whorls throughout, but there are numerous fine spirals of secondary and higher orders arranged in groups as obsolete cinguli, corresponding to those of related species. It is therefore not so primitive a type as at first might appear.

The characteristic feature of this group is the presence of spiral cinguli which make the spire-whorls bicarinate or even tricarinate; but the cinguli are sometimes obsolete, especially on the body. The spirals are occasionally nodular or moniliform, but are generally smooth, and there are never axial ribs. The aperture differs from most of the S. papulosa group in having a definitely limited inner lip of regular width, while the callus of the outer lip is thick and rounded in cross-section. The sinuation of the outer lip is shallow, sometimes obsolete, but the posterior edge of the callus generally shows its bisinuate character, which, with the curved columella, indicates a relationship closer to Struthiolaria s. str. than to any of the other groups. If a sectional or subgeneric name is required it will be Pelicaria Gray, 1857, with Buccinum vermis Martyn as type (see remarks below under Tylospira).

As already pointed out, the bicarinate spire is foreshadowed in the third conch-volution of S. subspinosa and S. cincta.

The canaliculate suture of such species as S. canaliculata, S. fossa, and S. zelandiae must be considered as a gerontic feature parallelling a similar development in Tylospira coronata (Tate).

It is possible that exception may be taken to the specific recognition of some of the forms described below. No subspecific, mutational, or varietal divisions are used in this paper; but it must be understood that the relations between some of the species in a group are much closer than those between others. After all, a species is a purely artificial division, and in palaeontology especially a grading is found between different forms, so that in a good series one can trace the gradual change which produces what is commonly termed a “new species.” The placing of the specific boundary must always be a difficulty, and the better the collection the harder it is to decide; but that two different shells can be connected by a series is no reason why the extremes should not be separated specifically, especially if the change goes on throughout a considerable lapse of time.

Although the arrangement proposed in this paper is by no means final, it will be of much more use to the stratigrapher than the previous one.