Wible et al. (2007)
described Maelestes gobiensis (PSS-MAE 607), as a Late Cretaceous (75-71mya, Fig. 1) eutherian nesting outside of the Placentalia (Fig. 2). Wikipedia reports epipubes were present, but they are not listed by Wible et al. who list: “an incomplete skull, left mandible, atlas, axis, twelve thoracic vertebrae, eight partial ribs, incomplete scapula, clavicle, humerus, proximal radius and ulna, and incomplete astragalus.”
the large reptile tree (LRT) nests Maelestes between Asioryctes and Onychodectes at the base of the tenrecs (Fig. 3). The Wible team tested many more mammals (Fig. 2), some far from complete. The Wible team nested Maelestes within a few nodes of Asioryctes + Ukhaatherium in a cladogram lacking many modern taxa. This nesting somewhat matches the LRT, which includes Leptictis + Ukhaatherium and Andrewsarchus at the base of the tenrecs. In the Wible et al. tree Leptictis nested several nodes away, but still not in the Placentalia. The Wible et al. tree (Fig. 2) also nested the wombat, Zalambdalestes deep with the Eutheria, but also not within the placentalia. So, as before, Wible et al. are finding clades between marsupials and placentals while the LRT (Fig. 3) does not.
According to Wible et al.
DNA studies postulate the origin of placentals between 129-78 mya. “This controversy has yet to be addressed by a comprehensive phylogenetic analysis that includes all well-known Cretaceous fossils and a wide sample of morphology among Tertiary and recent placentals. Here we report the discovery of a new well-preserved mammal from the Late Cretaceous of Mongolia and a broad-scale phylogenetic analysis. Our results exclude Cretaceous fossils from Placentalia, place the origin of Placentalia near the K/T boundary in Laurasia.”
the earliest placentals in the LRT include Megaconus (170 mya) and the multituberculates. Based on comparisons to Monodelphis and Ptilocercus the skull shape and size of Maelestes is more primitive than any of the above named Early Cretaceous through Jurassic placentals.
Wible et al. count only three molars for the uppers and lower. I see four uppers and three lowers. Molar number is not a scored trait in the LRT. It just rides along with the other scored traits. Asioryctes has four upper and lower molars. Onychodectes also has four uppers. The lower molar data is not available at present.
To their credit
and distinct from several prior stuides, Wible et al. nest Carnivora as basal placentals and Xenarthra as derived and close to Paenunugulata (elephants and kin) matching the LRT. Unfortunately, they also nest whales with artiodactyls as sisters to Carnivora, which does not match the LRT.
Wible JR, Rougier GW, Novacek MJ and Asher RJ 2007. Cretaceous eutherians and Laurasian origin for placental mammals near the K/T boundary.” Nature, 447: 1003-1006.