JUKES-BROWNE : ON THE MYTILID^. 215 



varying greatly in this species. In M. BunJceri and M. crassitesta 

 (both from Japan) this inflexion is extreme, the umbones are terminal, 

 and the teeth of the adult shell are reduced to 2 in each valve. 



In the group typified by M. afer^ M. perna, If. smaraffdimis, and 

 M. latus, the anterior side is always more inflexed than in M. edulis 

 and M. Galloprovincialis ; the riblets are fewer, and the teeth are never 

 more than 2 in number, frequently appearing as 2 in one valve 

 and 1 in the other. The two teeth do not always occur in the same 

 valve, some species having them in the left {M. smaragdinus), others 

 in the right valve {M. latus and M. perna). The single tooth of one 

 valve always fits into the space between the two of the other valve. 

 The final stage of the suppression of these anterior teeth seems to 

 be reached in M. Magellmiicus, M. crenatus, and their allies. In 

 these species the anterior margin is so inflexed and twisted under 

 the hinge-line that the teeth of one valve seem to have been com- 

 pressed into a single tooth or ridge which fits into a hollow or groove 

 in the other valve, the teeth of the latter having become entirely 

 obsolete. In these species the umbones are more than terminal, for 

 in most cases they project beyond the anterior side. 



To sum up, therefore, out of those species of the genus Mijtilus 

 which do not possess denticulations behind the ligament, three 

 groups can be based on the dilferences of the anterior teeth. In the 

 first group the umbones are seldom terminal, there being generally 

 a small anterior inflation which bears from 2 to 6 small teeth. In 

 the second group the umbones are terminal, the dental area is com- 

 pressed, and there are never more than 2 teeth, generally 1 opposed 

 to 2. In the third the umbones project beyond the anterior side, and 

 the dental formula is 1 opposed to 0. 



Passing now to the differences in the number and position of the 

 adductor and retractor scars, we find that they are important, because 

 they are correlated with other differences both in the shell and in 

 the animal. So far as I can ascertain, the first detailed account of 

 the anatomy of a mussel of the M. latus type was that by Mr. A. Purdie, 

 published in 1887 by the New Zealand Government as one of the 

 Studies in Biology at the Colonial Museum. 



Mr. Purdie compared the anatomy of M. latus, M. edulis, and 

 M. Magellanicus from New Zealand material, and the following are 

 his conclusions: — "While M. edulis and M. Magellanicus agree so 

 closely as to leave no question of their being members of one genus, 

 there is a very considerable interval between them and M. latus. 

 To pronounce definitely upon the above question [are M. latus and 

 edulis congeneric ?] would require an intimate knowledge of a great 

 number of species of Mytilineae, when it is possible that all intermediate 

 stages between the above distinct forms might be found. My study 

 of the internal structure has extended only to the three species 

 mentioned in this paper, but so far as this serves to decide, there 

 seems to be occasion for separating M. latus from the others, if not 

 as a genus, at any rate as a subgenus." He gives a complete tabular 

 view of the differences between M. latus and M. edulis, of which the 

 following are the most important : — 



VOL. YI. — MARCH, 1905. 16 



