Genus URDYELLA gen. nov.

Type species. Urdyella unicorna n. sp.

Type stratum and locality. Luolou Formation, Laren road cut, Guangxi Province, China .

Etymology: Named after Severine Urdy.

Diagnosis. Short coniform-to-segminiscaphate P 1 element with a long cusp, usually at least twice the length of the adjacent denticle, a very short anterior process, and a broadly excavated basal cavity.

Remarks. Te relationship of this genus to Cornudina is uncertain (Orchard, 2005, 2007). P 1 elements similar to the P 1 element of Urdyella n. gen. are often considered as belonging to Cornudina . Yet, as explained by Orchard (2005), the holotype of Cornudina ( O. breviramulus) appears to be a P 2 element. Both Kozur and Mostler (1971, p. 11) and Sweet (in Clark et al., 1981, p. W155) placed Cornudina with Chirodella as a multi-element Chirodella . Te holotype of Chirodella (Metalonchodina triquetra) is an S 2 element. Te holotypes of both Cornudina and Chirodella are from Muschelkalk (Middle Triassic) collections made by Tatge (1956), who described no elements like those described by Orchard (2005) as P 1. ‘ Chirodella’ sensu formo does not occur in Orchard’s nor in Koike’s Spathian collections of Cornudina . For this reason, Koike (1996) and Orchard (2005) regarded the two genera as unrelated: Koike reconstructed Cornudina as an apparatus consisting of P 1 and P 2 elements only ( Cornudina breviramulis) or of P 1 elements only ( Cornudina igoi), whereas Orchard reconstructed Cornudina ? as having an octomembrate apparatus whose P 2 element vaguely resembles the holotype of Cornudina, but he questioned the very validity of that name for Spathian forms that might be unrelated to the ‘true’ Middle Triassic Cornudina .

‘ Chirodella’ sensu formo does not occur in our Smithian collections of Urdyella n. gen. and neither does ‘ Cornudina’ sensu formo. Hence we suggest that Cornudina -likeP 1 elements found in the Smithian, and possibly those found in the Spathian, belong to Urdyella n. gen. not to Cornudina .

Based on multi-element considerations, Orchard grouped both ‘ Cornudina’ and Spathicuspus within the same subfamily Cornudininae . Because Spathicuspus seemed to occur first, he hypothesized that ‘ Cornudina’ evolved from Spathicuspus near the Lower/ Middle Triassic boundary through overall shortening of the P 1 element. Since Urdyella occurs already in the Smithian, we suggest that Spathicuspus may have evolved from Urdyella and the Spathian forms of ‘ Cornudina’ either belong to Urdyella, derived directly from Urdyella, or as assumed by Orchard, evolved from Spathicuspus, although the latter hypothesis seems less likely. Urdyella itself may have evolved from Discretella via forms like Discretella ? n. sp. B. Discretella has a much shorter cusp and a larger anterior process. Furthermore, the basal cavity in Urdyella is flatter and not as broadly excavated as in Cornudina .