Although the holotype of Chasmosaurus russelli Sternberg, 1940, from the upper Dinosaur Park Formation of southern Alberta, shares features with other Chasmosaurus Lambe, 1914 specimens from Dinosaur Provincial Park as well as various chasmosaurine taxa from the southwest United States, it also possesses unique features of the premaxilla and frill epiossifications. An iterative phylogenetic analysis, scoring only the holotype (CMNFV 8800), produces three consensus topologies, one of which places CMNFV 8800 closer to Chasmosaurus than to Pentaceratops, another in which CMNFV 8800 is closer to Pentaceratops than to Chasmosaurus, and a final topology in which these relationships are unresolved. Under current definitions, CMNFV 8800 cannot be accommodated within Chasmosaurus or any other known chasmosaurine genus. Consequently, we erect Cryptarcus gen. nov. to receive it. Cryptarcus russelli comb. nov. may represent a migrant originating from the southern “ Pentaceratops clade”, but it may also be part of a lineage more closely related to Chasmosaurus in which Pentaceratops-like features (e.g., frill wider anteriorly than posteriorly, deep parietal embayment, and first epiparietals nearly touch at the midline) evolved convergently. None of our phylogenetic topologies support an anagenetic trend from Cryptarcus russelli through Pentaceratops to produce Terminocavus Fowler and Freedman Fowler, 2020, as has been recently hypothesized.
Holmes et al. (Thu,) studied this question.