The traditional subclass Protobranchia holds together with characters (the protobranch gills, the taxodont hinge in Nuculoidea and Nuculanoidea) which are plesiomorphic in Bivalves. There is debatable evidence that Nuculanoidea form a clade with lamellibranch bivalves rather than with the other protobranchs (Giribet & Wheeler, 2002). Pending further studies the three groups could be maintained as unassigned superfamilies of the Bivalvia but it is just as useful to keep Protobranchia with rank of subclass as long as it is not demonstrated to be polyphyletic (note that Tree of Life still uses Protobranchia with equal rank to Pteriomorpha, Paleoheterodonta, Heterodonta...). Internal relationships are not straightforward, there are not so many taxa so just go directly to superfamily rank. Xenoturbella to be ousted from here, with a note about Israelsson’s hypothesis (1999: Proceedings of the Royal Society 266B, 835-841.