

Each of them is based on the following assumptions: (1) Neocoleoidea uniting extant Decabrachia and Octobrachia is monophyletic (= proostracum-bearing coleoids) (2) extinct Belemnitida and Diplobelida are stem decabrachians (3) proostracum-less coleoids (Hematitida, Donovaniconida, Aulacoceratida) represent stem-neocoleoids (4) Ammonoidea and Bactritoidea are stem coleoids. Since a priori character state reconstructions are still ambiguous, we suggest and discuss three different evolutionary scenarios. New palaeontological discoveries in the recent past afford to review and reassess origin and homology of suckers, sucker rings, hooks, and cirri. The cephalopod arm armature is certainly one of the most important morphological innovations responsible for the evolutionary success of the Cephalopoda. This strongly supports earlier ideas on a close relation-ship between Cretaceous Decabrachia and belemnites and simultaneously challenges opinions that Decabrachia origi-nated in the Carboniferous. is derived from diplobelid-like belemnoids and gave rise for the Decabrachia or at least groenlandibelid spiru-lids. A phylogenetic approach suggests that Longibelus gen. These specimens belong to Longibelus as do specimens from the Aptian of Caucasus (previously described as 'Naefia' kaba-novi). Finally, we introduce the first records of coleoids from Zululand (South Africa). New material from the Albian of India likewise assignable to Longibelus is described for the first time. Also, we can reinterpret material from the Cenomanian of India as Longibelus gen. Similarly, two specimens from the Maastrichtian of Mexico previously determined as N. neogaiea from the Maastrichtian of Chile, we found one specimen that unambiguously belongs to Longibelus gen. Besides a redescription of Longibelus ('Naefia') matsumotoi, we describe the first Ma-astrichtian occurrences of this species from Hokkaido (northern Japan) and Alaska. neogaiea, the type species of Naefia, and 'N.' matsumotoi, the new genus Longibelus has been erected. Diplobelida and 'Naefia' matsumotoi, however, exhibit a mosaic of decabrachian and belemnoid characters. These diagnostic characters are as follows: (1) presence/ absence of a mural flap (2) position of the siphuncle (3) shape of the dorsal soft tissue attachment scar (4) pres-ence/absence of tabular nacre in the conotheca (5) pres-ence/absence of a rostrum proper (6) presence/absence of a narrow rod-like proostracum and (7) presence/absence of a caecum. Seven characters (apical angle, chamber length, dorsal and ventral sutures, orientation of septa, direction of the dorsal part of the sep-tal neck, primordial rostrum) are not or less diagnostic, whereas the seven remaining characters can be reliably used to distinguish between the Decabrachia on the one hand and Belemnitida and Aulacocerida on the other hand. A comparative analysis of shell structures of well-preserved specimens including types and new material of Cretaceous spirulids Groenlandibelus, Naefia and Cyrtobelus, as well as selected taxa of aulacocerid, belemnitid and diplobelid be-lemnoids, revealed a set of 14 characters. One central problem concerns a clear differentiation between belemnoids and early spirulids. This is due to a poor understanding of character complexes relating to the shell, which causes difficulties in establishing homolo-gies among different taxa. The phylogenetic origin and the timing of orig-ination of the Decabrachia are controversial.
