Cystoblastus and the origin of the Hemicosmitoida (Echinodermata: Blastozoa)

All previous descriptions and interpretations of thecal plating in Cystoblastus Volborth depict a closed infralateral circlet, whereas actual specimens have that plate circlet open in three places. Cystoblastus is redescribed using Estonian specimens. It has a blastoid-like theca composed of a pelvis of four basals and five infralaterals, plus a vault composed of nine alternating laterals and radials. The radials are deeply cleft to accommodate the ambulacra, as in many blastoids, but lack underlying lancets, or any trace of blastoid deltoid plates. Cystoblastus vault plating resembles the ‘lateral’ circlet of hemicosmitoid Rhombifera. So, recognition of plate homologies between glyptocystitoid and hemicosmitoid (= dichoporite) rhombiferans was attempted. One unique interpretation accounts for all the hemicosmitoid plates without contravening any positional constraints imposed by glyptocystitoid plating. Thus, the new interpretation is not invalidated by conventional tests of homology. We used the new plate homologies to investigate the phylogeny of glyptocystitoids and hemicosmitoids, utilizing twelve genera and 27 parsimony-informative characters. Cladistic analysis using PAUP produced eight equally parsimonious trees. A strict consensus tree revealed the hemicosmitoids as highly derived dichoporites, with Hemicosmites as sister to the Caryocrinitidae and Thomacystidae. Cystoblastus is a sister group to hemicosmitoids in a trichotomy including Glyptocystella and Pirocystella. This lineage was derived from the Cheirocrinidae. Thus, Cystoblastus is not related to blastoids, as suggested by Otto Jaekel, but is a critical link between glyptocystoid and hemicosmitoid dichoporites, and the Dichoporita is a valid taxon.