After LM and SEM examination of representative specimens, and after revisiting all previously published descriptions and illustrations, we found different morphological characters that are possibly informative on the phylogenetic position of Dinogeophilus. These are summarised in Table 1 and described in the following lines. A complete revised diagnosis of Dinogeophilus is provided in Appendix 2.
Table 1. Morphological characters that are considered informative on the phyletic position of Dinogeophilus within either Geophilidae or Schendylidae. Characters are listed in anatomical order, anterior to posterior.
Cephalic plate. When comparing the polygonal reticulation of the head of different species of Geophilomorpha (Moretto et al., 2015), the average area of the scutes in Dinogeophilus (about 100 μm2; Fig. 1A) is well within the range of variation estimated for the Schendylidae to the exclusion of the ballophilines (about 80-120 μm2, calculated on five species), whereas it is remarkably higher than the range of variation estimated for the Geophilidae (40-80 μm2, calculated on 9 species) with the only exception of a subgroup distinguishable as Ribautiinae (80-100 μm2, measured in five species).
Labrum. Along the posterior margin of the labrum, one or two projections are present on each lateral part (Fig. 2A). In ventral view, these projections are subtriangular, stout and pointing mesally. These features distinguish the so-called ‘denticles’ (present exclusively on the labrum of the Himantarioidea, which includes Schendylidae) from other kinds of projections. The condition found in Dinogeophilus (denticles very few, inconspicuous, and limited to the lateral part of the labral margin) is similar to the condition documented in some Schendylidae, especially in ballophiline species classified either in Ballophilus Cook, 1896 or in Ityphilus Cook, 1899 (e.g., Ribaut, 1914; Pereira et al., 1994). Instead, it is different from the most common and putatively basal condition found in Geophilidae (intermediate ‘tubercles’ and lateral ‘bristles’) and other derived conditions found in this family (Bonato et al., 2014). It is quite different with respect to cases of remarkable reduction of labral projections, e.g. in the aphilodontines (e.g., Silvestri, 1909b; Attems, 1929).
Fig. 2. Dinogeophilus oligopodus, SEM. A. Mouth, with surrounding labrum, mandibles and maxillae; arrow-heads indicate labral denticles. B. Left pretarsus of the second maxillae; arrow-heads indicate filaments. C. Leg-bearing segment 3. D. Leg-bearing segment 28. E. Glandular pores on metasternite of leg-bearing segment 7. F. Pretarsus of the left leg of the pair 10; arrow-heads indicate the accessory spines. G. Distal part of the right leg of the ultimate pair; arrow-heads indicate two of the large setae associated with internal vesicular structures (Fig. 1D). H. Gonopods. All pictures are in ventral view. Scale bars: 5 µm (B, E, F); 20 µm (A, G, H); 50 µm (C, D). Micrographs from a 5.5 mm long male (C, E) and a 5.0 mm long female (A-B, D, F-H), both from La Plata, 19.xii.1985.
Mandible. The distal margin of the mandible of Dinogeophilus bears a single row of projections (Silvestri, 1909b; Pereira, 1984; Fig. 2A). Shape, size and sclerotization of the projections vary only slightly and gradually along the row, without any abrupt transition suggesting a composite origin of the apparently single lamella. Moreover, we could not recognise any rudiment of other lamellae. The presence of a single pectinate lamella on the mandible is a well-established synapomorphy of the Geophiloidea, while two lamellae have been invariantly recognized in all Schendylidae and more than two in all other Geophilomorpha (Bonato et al., 2014). However, the mandibles are quite variable within the Schendylidae: the two lamellae may be aligned and contiguous, with one often more sclerotized and modified in shape, or instead imbricated and similar to each other; in some species (e.g., Plesioschendyla confossa Ribaut, 1923), the two lamellae are aligned but weakly distinguishable from each other (Ribaut, 1923).
Second maxillary pretarsus. The pretarsus of the second maxillae of Dinogeophilus has a rounded tip, which is distinctly flattened (Fig. 2B). Three projections emerge from the basal part of the pretarsus, one on the ventral side, the other two on the dorsal side. These projections are elongate and point distally. A shallow-domed sensillum is present at about mid-length on the ventral side of the pretarsus. A more or less evident flattening of the pretarsus towards the tip is a synapomorphy of the Himantarioidea, although weakly manifested in some derived clades (Chalande and Ribaut, 1909). The presence of elongate projections (‘filaments’) like those in Dinogeophilus is common to different geophilomorph families, among which the Schendylidae, and could be a basal condition in the Adesmata, which comprises Geophiloidea and Himantarioidea (Bonato et al., 2014). On the contrary, no filaments are present in some other families, including Himantariidae and Geophilidae, possibly because of convergent evolution (Bonato et al., 2014).
Ventral pore-fields . Clusters of glandular pores are present on the ventral side of the trunk of Dinogeophilus (Fig. 2C), but only on the approximately anterior fourth of the trunk (Fig. 2D) and to the exclusion of the first leg-bearing segment. The microstructural features of these pores (Fig. 2E) and their arrangement suggest that they are homologous to the so-called pore-fields known in most Adesmata. In particular, in Dinogeophilus a single pore-field is present on each metasternite, in a subcentral position, and can be described as approximately subelliptical and remarkably elongate longitudinally (Fig. 2C). For their sub-central position on the metasternites, the pore-fields of Dinogeophilus resemble most closely those found in some Schendylidae (e.g., in some species of Ityphilus, where pore-fields are however rarely elongate longitudinally; Pereira, 2013b). Conversely, they are very different from the putative basal condition within the Geophilidae (pore-fields wider than long, centred on the posterior part of the metasternite, and often accompanied by additional anterior pore-fields; Bonato et al., 2014), as well as from all the many derived conditions found in some Geophilidae (Turcato et al., 1995).
Legs. The locomotory legs of Dinogeophilus are provided with a total of three accessory spines: besides a single anterior spine, two posterior spines are usually detectable, even though one is often shorter and narrower than the other (Fig. 2F). While in most centipedes the pretarsi of the legs are provided with only two accessory spines, one anterior and one posterior, in all Mecistocephalidae and all Schendylidae they bear a third spine, close to the posterior one, or even more additional spines (Bonato et al., 2014). Such conditions could be interpreted as independently evolved in the two families.
Ultimate leg-bearing segment. In Dinogeophilus the legs of the ultimate pair are distinctly different from those of all other pairs: all articles are conspicuously inflated and especially the prefemur is distinctly bulging on its mesal side at its distal end; additionally, the leg ends with a single short spine instead of a claw-like pretarsus (Fig. 2G). The same condition is common to male and female adults, without obvious sexual dimorphism. The legs of the ultimate pair are variously modified in most Epimorpha, but the particular combination of modified features found in Dinogeophilus (legs mesally inflated and ending with a spine, in both sexes) is found only in some schendylids, especially in species belonging to some of the largest genera, like Schendyla Bergsøe and Meinert, 1866, Pectiniunguis Bollman, 1889 and Schendylops Cook, 1899 (see, e.g., Brolemann, 1930; Pereira and Minelli, 1996). Conversely, somehow similar conditions are very rare among the diverse Geophilidae, like e.g. in Dignathodon Meinert, 1870 (however without mesal bulges, and with some trace of claws; Brolemann, 1930).
Gonopods. In Dinogeophilus females, the gonopods are two paired appendages, touching each other at their bases but distinctly separated. The contour of each gonopod is stoutly rounded and no intermediate articulation is detectable (Fig. 2H). In shape, structure and relative position, these gonopods resemble those commonly found in the females of most Schendylidae (Brölemann and Ribaut, 1912). Indeed, paired, uni-articulate, rounded-tipped gonopods could be a synapomorphy of a subgroup of Schendylidae. Conversely, in all female Geophilidae the pair of gonopods is invariantly represented by a single, significantly shorter lamina, at most shallowly bilobate. In most other geophilomorphs, instead, female gonopods are separate, but usually bi-articulated at full development and pointed at the tip (Bonato et al., 2014).
Summing up, in Dinogeophilus we detected many characters (referring to different body parts) that may correspond to synapomorphies for the Schendylidae, or for a subgroup of Schendylidae, or for the parental superfamily Himantarioidea. Conversely, in Dinogeophilus we found a single character (referring to the shape of the mandible) that could be interpreted as a synapomorphy for the superfamily Geophiloidea. Additionally, Dinogeophilus shares with Schendylidae other characters that are represented in Geophilidae by a different state, which is recognised as a synapomorphy of the latter.