Agriochoerus antiquus (Leidy 1850; Late Eocene to Oligocene; 38–16mya, skull) was similar to a juvenile Cornwallius.The upper canine is a vestige but the lower canine remains large.

Lophiaspis maurettei(Depéret 1910; Vautrin et al. 2014, Early Eocene 43mya; skull length 20+ cm) Originally Lophiaspis was considered a basal tapir and perissodactyl close to Lophiodon. Here Lophiaspis nests with Grangeria. The skull is flattened and wide. Premolars are absent. Incisors are reduced.

Cornwallius sookensis (originally Desmostylus sookensis, Hay 1923, Cornwall 1922; Beatty 2006a, b; Early Oligocene, 25 mya) was orginally and traditionally considerd a desmostylian. Here it nests between Neoparadoxia and Desmostylus. The nares are anterior in location. Note the resemblance (lack of a downturned snout) on the juvenile to Janjucetus. Either that or the juvenile was misidentified.

Neoparadoxia cecilialina (Barnes 2013, 2.2m long; Miocene, 20-10 mya) was a desmostylian derived from a sister toCambaytherium and was basal to Behemotops and Desmostylus. No prior studies have recovered the hippo and desmostylian relationship, but several prior papers have noted the convergent size, body type and lifestyle with hippos. Rather most studies ally desmostylian with sirenians and elephants.

The nares continue to move to the dorsal skull on Paleoparadoxia, ideal for breathing while otherwise underwater. Note the vertical orientation of the pelvis and procumbent incisors. Only four fingers were present here, as in whale paddles and mesonychid hooves.

Paleoparadoxia tabatai (Reinhart 1959, 2.2m long; Miocene, 20-10 mya) The origin of baleen might be with such genera.