The endolithic life habit has evolved several times in the Bivalvia, e.g., in the superfamilies Modiolopsoidea (family Modiolopsidae), Mytiloidea (family Mytilidae, subfamily Lithophaginae), Arcoidea (family Arcidae), Cardioidea (family Tridacnidae), Veneroidea (family Petricolidae), Gastrochaenoidea, Myoidea (family Myidae), Hiatelloidea, and Pholadoidea (families Pholadidae and Teredinidae) (Otter, 1937; Yonge, 1963; Cox, 1969; Carter, 1978; Kleemann, 1980). The oldest known definite bivalve borings are Ordovician slotlike depressions in stromatoporoids, made by the facultative nestler/borer modiolopsid Corallidomus Whitfield, 1893 [1895] (Pojeta and Palmer, 1976; Wilson and Palmer, 1988). Pojeta and Palmer (1976) and Morton (1990) cited Corallidomus as the ancestor of the Lithophaginae. However, lithophaginids more likely evolved from modiolinid mytilids, which in turn evolved from modiolopsids (Fang and Morris, 1997; Fang, 1998; Carter et al., 2000). Lithophaga-like shells are known from the Carboniferous and Permian, but these Upper Paleozoic examples are not known to have been endolithic (Kleemann, 1983, 1990, table 2). Wilson and Palmer (1998) attributed certain borings without associated shells in Upper Carboniferous limestone cobbles to lithophaginids because the borings are widest near their anterior end and they lack a constricted neck. However, if, as Wilson and Palmer suggested, these borings are not posteriorly truncated, then their openings are relatively much wider than typical lithophaginid borings. One boring illustrated by Wilson and Palmer (1998, fig. 4) has an irregular anterior cross-sectional shape that is unlike lithophaginid and gastrochaenid borings. This irregularity recalls the Early Ordovician ichnospecies Gastrochaenolites oelandicus Ekdale and Bromley, 2001, which is similarly vase-shaped. Ekdale and Bromley (2001) noted that the latter boring predates known endolithic bivalves, so they attributed it to an unknown invertebrate.

Modern pinnoideans, or “pen shells,” are shallowly to deeply endobyssate, equivalved, edentulous, ham-shaped bivalves with a single, slightly submerged, opisthodetic ligament. Malacologists have generally regarded pinnids as closely related to the pterioid superfamily Pterioidea, especially its Paleozoic family Pterineidae (e.g., Zittel, 1927, p. 446; Thiele, 1935, p. 803; Pojeta, 1978; Waller, 1978). As early as 1890, Jackson suggested that pinnids evolved from the Silurian-Permian pterineid Leptodesma Hall, 1883. More recently, Pojeta (1978, p. 239) speculated that they evolved from the pterineid Pteronitella Billings, 1874 (a Silurian genus similar to Leptodesma) through the morphologically intermediate Devonian Palaeopinna Hall, 1883. The hinge and ligament of Palaeopinna remain unknown, so its family-level placement is uncertain (Newell and LaRocque, 1969, p. N301). However, Leptodesma and Pteronitella have hinge teeth, inequivalved shells, and a duplivincular ligament similar to other pterineids (Newell and LaRocque, 1969, p. N299–N301; Pojeta, 1978, p. 239; Pojeta and Runnegar, 1985, fig. 16D; Carter, 1990a, p. 205; 1990b, p. 334).

The rowleyellinae is a small subfamily of meristelloid athyridid brachiopods, previously with only one genus, Rowleyella Weller, 1911 (Alvarez and Brunton, 1995; Alvarez et al., 1998; Alvarez and Rong, 2002). In this paper we report a new meristelloid athyridid brachiopod genus Muhuathyris from a newly discovered, very diverse Tournaisian brachiopod fauna in South China. It is noteworthy that this newly discovered fauna is slightly older than that reported earlier from the same outcrop (Balinski, 1999). A full discussion of the newly discovered fauna is currently in preparation. Morphological features of Muhuathyris n. gen. suggest that it is probably an evolutionary intermediate between Camarophorella Hall and Clarke, 1893 and Rowleyella Weller, 1911. The new genus differs from Camarophorella and Rowleyella only in the nature of the structures beneath the septalium in the dorsal valve. In Camarophorella a shoe lifter is present in front of or beneath the septalium. In Rowleyella a cruralium rests on the dorsal median septum under the septalium and extends to the front of septalium. Lateral to the dorsal median septum, two thin and short plates, asymmetrically disposed, are present posteriorly under the cruralium, which were considered by Carter (1991) as a very posteriorly placed shoe lifter. In the new genus a structure consisting of two plates lateral to the dorsal median septum is attached to the dorsal surface of septalium near the crural bases. This structure extends from the dorsal beak to the posterior of the adductor muscle fields. It may be developed as two vertical plates lateral to the median septum, or, when the plates converge medially, as a sessile or median septum-supported cruraliumlike platform. The structure primarily acts as support to the crural bases and septalium rather than a dorsal adductor muscle platform, but may also serve as part of dorsal adductor muscle attachments in some individuals. It seems probable that this structure might be homologous to the cruralium in Rowleyella. For this reason we put the new genus within the subfamily Rowleyellinae.