forms in which the sexes are either separate, or combined in the individuals, no other characters can be of greater importance in establishing primary divisions. As a means of further subdivision according to natural affinities, he suggests distinctive characters derivable from the auditory sacs and concretions, and from the oral, lingual and gastric dental organs.

In Mollusca, as in Fishes, the auditory concretions present themselves in one of two forms, viz. solitary lapilli, usually named otolithes, or groups of small granules of a rounded oval or irregular shape, which have been designated by the term otoconia. The lingual teeth are either set together on a short and broad lingual membrane, and form what the author calls a lingual pavement, or on a narrow longitudinal band termed lingual ribbon or strap. The latter usually consists of a median rachis flanked by two lateral portions or pleuræ ; but in some cases the rachis, and in others the pleuræ are absent, and the number of longitudinal rows of teeth in these divisions may also differ in different genera. The fore part of the lingual membrane is supported by cartilage, so curved and fashioned as to receive the lingual sac behind and form a basis to the tongue itself projecting in front. This lingual cartilage may consist of a single piece thinned in the middle line, or of two or four distinct pieces, similarly arranged and wrapped together by muscle and ligament. The oral dental organs or labial plates are disposed either horizontally or laterally. In the former case a single plate may occupy the upper lip, or there may be two guarding the aperture of the mouth, and corresponding with both upper and lower lip, but the lateral plates are always in pairs. Gastric teeth occur in the Aplysiada and Bullida.

After pointing out further differences in the form and arrangement of the dental apparatus in different genera, the author thus describes the mode of development of the lingual teeth. "The lingual sac at first appears as a little cæcal process appended to the inferior part of the œsophagus, where it joins the oral cavity. In the median line of the floor of this sacculus, a few minute plates disposed in a longitudinal row form the rudiment of the future rachis, and the progress of their development may be distinctly traced on examining them, seriatim, from before backwards, in which direction, as their growth advances, they acquire a more perfect form. The internal row of pleural plates now makes its appearance, their development proceeding in a similar way; and after this follow the others according to their posi

tion, the more internal arising first. Thus the whole ribbon of dental organs increases in length and breadth by additions made respectively to its anterior extremity and sides; and each transverse row gradually moving backwards by the continued development and growth of others anterior to it, causes elongation of the lingual sac, which only attains its perfect state when these processes are at an end. The idea, therefore, that the new teeth are developed from behind forwards and successively brought into use, as in sharks and rays among fishes, does not appear to me to be correct."

In the annexed Table a rough arrangement is given of a considerable number of genera grouped together by the characters above referred to. Although the author thinks it improbable that any genera opposed to each other in those fundamental particulars can be intimately related, yet the facts are not advanced as the basis of a new classification, but simply that they may yield their own weight, as so many available tests of affinity.

according to the characters of their dental organs.

The following Table exhibits a further subdivision of the first series,

Murex, Triton, and Ranella have always been associated together as members of one family by universal consent, and it must be confessed that the external resemblance between them is very remarkable; but on comparing the lingual and labial dental organs of Triton or Ranella with those of Murex, it will be at once perceived that the latter genus can have no immediate affinity with either of the former

genera.

The aperture of the proboscis in Murex is transverse, and armed with two horizontally arranged dental plates, connected laterally by the minute semi-calcified cells which line this part. The upper plate presents a rough palatal surface, with an anterior encrusted cutting border, much resembling that of the crescentic mandible of Limax or Helix; whereas the dental plates of Triton and Ranella consist of oblique rhombic cells identical in character with those of Cyclophorus or Natica, disposed laterally. Again, on comparing the lingual strap of Murex with that of Triton or Ranella, we remark, first, that each transverse row of the former consists of three members, viz. one in the rachis, and one in each pleura, while in the two latter cases the pleuræ present two additional elements; thus there are seven series of dental plates in the strap. The tongue-strap of Murex, moreover, is lengthy like that of Purpura and Ricinula, both of which genera are more closely allied to Murex than perhaps any others referred to its family.

The strap of Triton and Ranella, on the other hand, is comparatively much shorter, and singularly enough more nearly approaches that of Pileopsis or Vermetus, not only in general proportions, but also in the actual number and configuration of the dental plates and processes. Now with these facts before us, it will be scarcely worth while entering further into the characters of the lingual dentition of Murex, Triton, and Ranella, but the most superficial examination of the figures (which accompany the communication) will show that Murex must be separated from its assumed alliance with Triton and Ranella, while the close relationship of the two latter genera gains additional support.

On comparing the lingual dentition of the genus Cyrtulus with that of Tritonidea of Swainson (the Polia of Gray), both are found to be naturally allied by characters which very distinctly manifest a family relationship, and Swainson's genus Muricidea, with several

others, must also be referred to this group. The lengthy triserial ribbon of Cyrtulus, or Tritonidea, for example, exhibits no true or immediate affinity with the comparatively short and septiserial dental armature of Triton or Ranella. Thus the author is induced to dissent from Mr. Gray's view that Tritonidea is allied to Triton, but agrees with him that the Buccinidæ, forming an equally characteristic natural family, are very close at hand.

The lingual dentition, and in fact the whole anatomy of Terebra, most unequivocally refer it to the Conida, and not to the Buccinidæ, amongst which it is at present received.

The author has not been able to detect lingual cartilages of the usual character in Conus, Conorbis, or Terebra, but the walls of the tongue-sac are stout, tough, and distinctly cartilaginous in structure; indeed the whole organ, including its armature, very much resembles the dental cheek-pouches of some Pteropods.

The lingual ribbon of Pleurotoma is exceedingly minute, and the parietes of the sac are not of that dense and unyielding character which they exhibit in Conus, Conorbis*, and Terebra. Moreover, in Pleurotoma the little lingual membrane is supported by two rounded masses of cartilage composed of large spheroidal cells. The rachis appears to be absent altogether, and there is but a single row of elongated, slightly curved, and sharp-pointed teeth (differing considerably from those of Conus and Terebra) in the pleuræ.

The tongue-strap of Mitra, although remarkably short, is triserial like that of Murex, Purpura, &c. ; but the author has invariably found that in those Mitræ in which the sculpturing of the shell was transverse, the pleural teeth were simple, uncinate, and mobile, while in those species characterized by a smooth surface or longitudinal sculpturing, the dental processes were small, straight, and numerous, arising just within the posterior border of broad basal plates. This difference is exactly such as exists between the lingual dentition of the respective groups to which Murex and Tritonidea belong.

Harpa and Oliva are very closely allied, by the general configuration of the body and the characters of the lingual dentition, though it must be remembered that the tongue-strap in the former is so very

* Several specimens of a recent species of this genus (hitherto known only in a fossil state) were obtained from depths ranging between 10 and 20 fathoms, within the barrier reefs surrounding the Feejee Islands.