According the the abstract of a paper by Johannes Mehl (1984): Radula und Fangarme bei Michelinoceras sp. aus dem Silur von Bolivien, Palaeontologische Zeitschrift, Band 58, Heft 3/4 . p. 211-229, which some of you may have seen already, Sulurian Michelelinoceras found in Bolivia had seven radular teeth per row, as common in Ammonoids and Coleoids, not the nine found in Nautilus, and ten arms of which two were long tentacles. This seems to corroborate the conclusion by Flower in a 1955 paper in the Journal of Paleontology that certain imprints in the Upper Ordovician Corryville beds of Cincinnati were produced by the tentacles of Orthonybyoceras (=Treptoceras), later determined to be an orthocerid. In his paper, Flower speculated that Orthonybyoceras may have had ten tentacles (not differentiating between them and arms). This certaintly indicates that the orthocerid, e.g. Michelinoceras, animal was rather a neocephalopod, as is shown in some more recent taxonomies, in spite of having a basically nautiloid shell. Flower also wondered (I'm recalling this from memory) which was the more primitive or ancestral, many arms or tentacles as in Nautilus or few (8 or 10) as in modern coleoids.