214 PftOCEEDINGS OF THE MALACOLOGICAL SOCIETY. 



case. The true Mytilus is generally smooth, but there are a few 

 striated and ribbed species which can hardly be excluded from the 

 genus. In the case of Modiola, even when all the species which have 

 a denticulate hinge are removed, there still remains at least one species, 

 If. demisstis, better known as M. pUcatula, which is a true edentulous 

 Modiola, and yet is strongly ribbed. Lastly, in the Mytilaster group, 

 there is every gradation from a smooth shell to divaricate striation. 



In studying the shells of Mytilidae, therefore, the position of the 

 umbones and the surface sculpturing of the shell may be regarded 

 as features of secondary importance. If, on the other hand, we look 

 to the characters of the hinge, and to the number and position of 

 the muscular impressions, we shall find that these afford a much more 

 satisfactory means of distinguishing genera and subgenera from one 

 another. 



I have studied the hinge of a considerable number of species oi Mytilus, 

 and have found that, though in some species the number of teeth varies, 

 yet it does so within narrow limits, while in other species there is 

 little or no variation. Further, I find that there are few, if any, species 

 of Mytilus which are really edentulous, and that the dental characters 

 are more or less correlated with the other special characters which 

 have been mentioned, so that the teeth are really as useful a basis 

 for diagnosis as they are in any other family. 



In the common Mytilas edulis, and in all the closely allied species, 

 such as Galloprovincialis, borealis, DunJceri, latissimus, and planulatus, 

 the teeth are connected with a series of small riblets on the anterior 

 side of the shell. These riblets descend from the extreme border 

 of the shell, which is incurved against the hinge-line beneath the 

 umbo ; this border is thickened so as to form a kind of hinge-plate, 

 and the teeth arise from the plate opposite the grooves which separate 

 the riblets. All these details can be clearly seen on any half-grown 

 British M. edidis from 37 to 50 millimetres long (1^ to 2 inches). 

 The position of the teeth thus accords with the origin attributed 

 to them by Bernard ; they represent the ends of the primitive 

 internal ribs, corresponding with the grooves or depressions between 

 the external ribs, which were developed in the nealogic stage of the 

 shell. There is no trace of such riblets on the posterior side even 

 of very young shells, and, if teeth were ever formed on this side of 

 the hinge, they have all been obliterated by the elongation and 

 upward growth of the ligament. 



The number of these anterior teeth in this group of shells varies 

 even in the same species. In M. edulis it varies from 3 to 6 in each 

 valve, but so far as my experience goes 4 is the usual number, and, 

 when 6 occur, three of them are generally very small, and are placed 

 close together, so that they only occupy the space of two in a more 

 normal specimen. In M. Galloprovincialis the normal number appears 

 to be 3, though occasionally there are as many as 5. In what I take 

 to be M. uHjiulatus, of which I have seen specimens from the Falkland 

 Islands and from Vancouver, the number of teeth varies from 2 to 

 4 or 5, and seems largely to depend on the extent to which the 

 anterior side is inflexed beneath the umbo, the degree of inflexion 



