family Craspedosomidæ includes both the subfamilies Lysiopetalidæ and Craspedosomidæ of Newport, although these are placed by the latter in different families. Whether these two groups belong together or not I cannot say, since I have never seen a specimen of the Craspedosomidæ of Newport, The Lysiopetalidæ, as defined in this memoir, includes that of Mr. Newport, but whether it reaches still further I cannot certainly say, on account of the differences in the characters used. The distinguishing mark of the Polydesmidæ is commonly supposed to be the possession of lateral lamina or side plates. Now, on glancing over a selected series of specimens, it will be seen that there is a regular gradation in size, from the largest lateral lamina to those which are merely rudimentary. Further, in some species of Lysiopetalidæ, the whole surface is strongly keeled, and the keel corresponding in position to the lateral lamina is so enlarged as to equal in size the smaller lateral lamina. The groups thus coming, as regards this character, in contact, or at least very close proximity.
If a segment of a species of each of the three groups is examined, it will be found that fixed characters can be drawn from the varying development of the sterna of the subsegments. In the Polydesmidæ both of the sterna are thoroughly developed, so that the segment constitutes a perfect ring. In the Julidæ the sterna of the posterior subsegments are atrophied, so that each segment is emarginate posteriorly. In the Craspedosomidæ both sterna are reduced to their minimum, and not consolidated with the scuta, so that the ring is left altogether incomplete. I have placed the Polydesmidæ below instead of above the other families, because they have both the organs of special sense and the locomotory apparatus least developed.
Ord. LYSIOPETALIDÆ.
Sterna minima, cum scutis haud conjuncta.
Sterna very small, not conjoined with the scuta.
There is much obscurity hanging around the genera of this family; so much that, in the absence of known representatives and types, it is impossible to clear it up.
Genera, with precisely opposite characters, families widely separated, have been tortured out of what claims to be one species.
In "Europe Méridionale," M. Risso described a genus under the name of Calipus, the type of which is a European species, which, M. Gervais says, belongs to the genus Lysiopetalum. Risso's generic characters apply nearly equally well to any Chilognath. Yet, if Gervais is right, his name should have the preference. In Recueil (p. 42), M. Brandt indicated the genus Lysiopetalum, and afterwards (p. 90), the genus Spirostrephon. The description of the former is mostly made up of what I consider family cha-
