V °j889. n '] PROCEEDINGS OF THE NATIONAL MUSEUM. 233 



groups. As we gradually become acquainted with the mutability of 

 the adductor muscles, the gills? the arrangements for retracting the 

 siphons and other factors in the mechanics of these organisms, the 

 classification based upon their mutations has gradually ceased to satisfy 

 students, though one phase or another of it may still retain a place in 

 ordinary text-books. 



To cite a few examples: It wdl be remembered that the most persistent 

 of the early systems for classifying these animals was based on the 

 number of adductor muscles or the scars upon the shell by which they 

 might be traced. At first the groups of Mouomyarians, or forms with 

 one adductor like the oyster, and Dimyariaus with two adductors, like 

 the ordinary edible clam, seemed sufficiently well distinguished. Later 

 when transitional forms like the mussel and its allies were carefully 

 studied, a new group, Heteromyaria, was erected for those which would 

 not fit into either of the others. 



But when it is considered that there are forms like Dimya, in which 

 with a mouomyarian organization two distinct adductors are found, one 

 at each end of the shell ; that in Chlamydoconcha we have a specially 

 modified anijial with no adductors at all; that in ^[>lUc>•ia we have the 

 young (not larval) animal typically dimyarian yet becoming in its adult 

 stage as typically monomyarian in its muscular apparatus as an oyster; 

 then it is sufficiently evident that better and more fundamental diag- 

 nostic characters should be found or the so-called orders given up. 



Again, an attempt has been made to use the characters of one of the 

 most mutable parts of the whole organism, namely the gill, as a basis 

 for primary divisions of the group. I have shown elsewhere, I venture 

 to think conclusively, that this selection is ill-advised and can not suc- 

 cessfully solve the problem. 



The simplicity or sinuation of the pallial line has been regarded as a 

 character of high importance and has heen used as diagnostic of divis- 

 ions of primary importance. I have recently shown that, in certain 

 groups, long siphons may exist with a simple pallial line, as in Cu.spi- 

 daria; that in species without long siphons, members of the same fam- 

 ily Poromyidce, and perhaps of the same genus, may show a simple or a 

 strongly sinuated pallial line according to the modifications of certain 

 muscular elements which certainly can not be claimed to have any high 

 systematic importance. 



The question is further complicated by the fact that certain characters 

 which iu general are indicative of very early evolutionary divergencies, 

 may be simulated or assumed as very modern special modifications 

 brought about iu animals of diverse groups by natural selection under 

 the influence of special circumstances. Species thus modified will very 

 naturally be classed with those who bear the same or similar characters 

 as the early result of very ancient ancestral divergencies, and, as a 

 consequence, other characters not harmonizing, the systems are thrown 

 into confusion. These are the difficulties among which the sum total 



