JTJKES-BEOWITE : ON THE MTTILIDJE. 219 



anterior side, and often compressed on the posterior side. Smooth 

 or concenti'ically striated, seldom ribbed,^ but in some species feebly 

 striated on the posterior slope. Umbones obtuse, anterior, but seldom 

 terminal. Anterior margin expanded in front of the umbones, but 

 quite smooth. Hinge-line without teeth or crenulations. Anterior 

 adductor scar larger than in Mt/tilus ; posterior scars united, the 

 byssal portion long and narrow, the adductor part rounded and not 

 bulging upwards as in Mytilus. 



According to Bernard the development of the hinge-line in Modiola 

 differs in several respects from that of Mytilus (see references on 

 p. 211). In the nepionic stage, the crenulations of the provinculum 

 ai'e finer and closer than in Mytilus ; in the nealogic stage, dysodont 

 teeth appear on the posterior extension of the hinge-line, but are 

 long, slender, and oblique, not short and vertical as in Mytilus. 

 Moreover, the extension of the ligament is more rapid than the 

 growth of these teeth, so that it quickly covers those that are first 

 formed, and then prevents the formation of others. 



As regards the anterior side, the few dysodont teeth that are here 

 developed are obliterated by the backward inflexion of that side, and 

 no fresh teeth are developed on the margin below them, as in the case 

 of Mytilus, so that both sides of the hinge-line remain toothless. 



The development of the hinge of Lithodomus is in all respects 

 similar to that of Modiola. 



There are several shells which have been described as belonging to 

 the genus Mytilus which seem to me more properly referable to 

 Modiola, because they all have a smooth, toothless, expanded margin, 

 below or in front of the umbones. One of these is the large Mytilus 

 tortus, for which Gray created the subgenus Stavelia. It certainly 

 will not come under any of the subgenera of Mytilus which I have 

 defined above, but I cannot see that it differs in any particular from 

 the diagnosis of Modiola, and it has much resemblance in form to the 

 well-known M. harhata of the British coast. The peculiar twist of 

 the anterior side does not seem to me a sufficient character for 

 establishing a subgenus, especially as no other species with a similar 

 twist has yet been found, so that there is no group of twisted shells. 

 Moreover, it is generally admitted that the nearest ally of M. tortus 

 is M. horridus, and the alliance is indeed so close that Mr. E. A. 

 Smith informs me that the former may be regarded as a twisted form 

 of the latter. I am glad to say also that he agrees with me in 

 referring both species to the genus Modiola. 



H. von Ihering, in the paper before mentioned, not only failed to 

 recognise the true aflSnities of these two species, but actually grouped 

 them with M. hirsutus, and made the latter the type of a new 

 subgenus of Mytilus for which he proposed the name Trichomya. In 

 my opinion M. hirsutus differs so greatly in all essential characters 

 that it does not even belong to the same genus as the shell which 

 should be called Modiola torta. It has a crenulated hinge-line, and is 



if. demissus (= M. plicatula, Lamk.) is a true Modiola, though ribbed. 



