Mantodea Burmeister, 1838 are a small order of insects with a very poorly known fossil record. While fossils of stem Dictyoptera Leach, 1818 (the super-order that comprises Blattodea and Mantodea) and Blattodea Latreille, 1810 are very common and found in all geological stages from the mid-Carboniferous to the present (Zhang et al. Reference Zhang, Schneider and Hong2013), those of Mantodea are more rarely found and the oldest reliable fossils date from the mid-Cretaceous (Grimaldi Reference Grimaldi2003). The Jurassic family Juramantidae Vršanský, Reference Vršanský2002 is based on a fragment of the cubito-anal area of a Dictyoptera, too incomplete and insufficient to be accurately attributed to the Mantodea (Vršanský Reference Vršanský2002, figs 6 and 7; Grimaldi Reference Grimaldi2003). Moreover, there is no crown Mantodea fossil known from the Mesozoic period, and they are very rare in the Cenozoic (see Cui et al. Reference Cui, Evangelista and Béthoux2018).
In this context, the discovery of a crown Mantodea is very important because it greatly helps in dating the clade. Ross (Reference Ross2019) described a new genus and species of Mantodea, Protohierodula crabbi, from the latest Eocene of the Isle of Wight (UK), and assigned it to the Manteidae (=Mantidae; see Carpenter Reference Carpenter, Moore and Kaesler1992) family, the largest family of Mantodea with more than half of extant mantises species. Svenson & Whiting (Reference Svenson and Whiting2009) considered the family polyphyletic. This new fossil was assigned to the Manteidae on the basis of the general shape of the forewing, which is very close to that of some extant genera in this family (Ross Reference Ross2019). As mentioned by Ross, there is no forewing character allowing an attribution of a mantis to the Manteidae, not only because the current classification is mainly based on foreleg spines but also because the forewings of the crown mantises are very similar in their venations.
The general shape of the wings of P. crabbi, with a ‘coarse archaedictyon' and ‘gently curved' subcostal (Sc) vein and radial (R) vein that diverge in the basal half of the wing (Ross Reference Ross2019), can be found in a large number of Mantodea, including non-Manteidae mantises, such as Hymenopodidae (Fig. 1). However, this fossil wing is not similar to those of the Chaeteessidae Fig. 2a), where the area between Sc and R (sensu Ross Reference Ross2019) is broader and with numerous crossveins. It is also not similar to those of the Mantoidae (Fig. 2c), which are thinner, smaller and with fewer branches in all areas; and to those of Metallyticidae (Fig. 2b), which have no archaedictyon and possess a clearly higher number of A2 branches. Due to these facts, this wing could reliably be assigned to Artimantodea (see Svenson & Whiting Reference Svenson and Whiting2009), as we find that its attribution to Manteidae is not supported by any synapomorphy. Thus, we consider P. crabbi as an Artimantodea incertae sedis sit. nov.

Figure 1 Right forewing of extant Chloroharpax modesta (Gerstaecker, 1883) (Artimantodea: Hymenopodidae) in dorsal view (Muséum national d'Histoire naturelle (MNHN), Paris). Abbreviations: Sc=subcostal vein; R=radial vein; CuA=cubitus anterior vein; CuP=cubitus posterior vein; M=median vein; A1=first anal vein; A2=second anal vein.

Figure 2 Forewings of non-Artimantodea extant Mantodea: (a) Chaeteessa valida Perty, 1883; (b) Metallyticus violaceus Burmeister, 1838; (c) Mantoida brunneriana Saussure, 1871 (MNHN collection). Abbreviations: Sc=subcostal vein; R=radial vein; CuA=cubitus anterior vein; CuP=cubitus posterior vein; M=median vein; A1=first anal vein; A2=second anal vein.
Despite the fact that this identification is less precise, P. crabbi remains one of the oldest, if not the oldest, known Artimantodea fossil. A few Mantodea are recorded from the Middle Eocene Baltic amber, but none are correctly described (von Spahr Reference von Spahr1992). As a consequence, this fossil is very important and interesting to clarify the evolution of all the order.