4.2. The Development of the Pelvic Skeleton in Extant ChondrichthyansHere, we have documented the development of the pelvic skeleton of C. milii (Figure 10). At the earliest point of skeletogenesis examined here in a stage 30 C. milli embryo, the pelvic skeleton, the girdle and basipterygium and fin radials, initially consist of a single condensation of mesenchymal cells (Figure 4). This condensation is diffuse and not easily detected with histological staining until stage 31 (Figure 5). However, this condensation can be visualised in detail at stage 30 through nanoCT imaging (Figure 4A,B). Our findings make it clear that by this stage much of the patterning of the pelvic skeleton, particularly the pelvic girdle, has already taken place prior to chondrogenesis. To our knowledge there is only one contemporary study of the morphology of fin skeleton development in cartilaginous fish, examining C. milii using clearing and alcian blue staining [41], which examined a stage 29 embryo. At this stage the only visible structures were a series of mesenchymal rods and no significant change occurred by stage 30 [41]. Given the level of skeletal development observed in stage 30 in the present study it is likely that more of the skeleton has formed by stage 29 than just these rods. The morphology of the earlier development of the pelvic skeleton could be determined in future investigations using nanoCT imaging.Some historical studies of pelvic development of scyliorhinid sharks, whose developmental stages are approximately equivalent with C. milli [46,63], agree with our findings on the early development of the pelvic fin and girdle consisting of a single mesenchymal condensation [14,21,22]. For instance, in his study of the development of the skeleton of paired fins of the nursehound (Scyliorhinus stellaris) [14], Balfour states that both the pectoral and pelvic fin skeletons are first visible within the “indifferent mesoblast” and are only distinguished histologically from the surrounding cells by being “more concentrated,” and their borders are not “strongly marked,” ([14] p. 665). Both sets of fins are also observed to arise “simultaneously and continuously with the pectoral and pelvic girdles,” ([14] p. 665). Further, in agreement with our findings at stage P (stage 31 [63]) the girdle and basipterygium of S. stellaris are still continuous and the fin radials are continuous with the basipterygium. Likewise, in Wiedersheim’s study of three other sharks, the early pelvic skeleton of the small-spotted cat shark (Scyliorhinus canicula) and blackmouth catshark (Pristiurus [Galeus] melastomus) (15 mm and 19 mm respectively) are also described as consisting of an condensation of cells within the fin mesoderm [21]. In slightly older S. canicula embryos the pelvic girdle and “free extremities” are described as a “single coherent mass” ([21] p. 28) and the pelvic fin and girdle are also initially continuous in the Spiny dogfish (Squalus acanthias) ([21] p. 30). No observations are made on the condensations of the clasper skeleton.A point of disagreement between these historical descriptions is the development of the fin skeleton, i.e., the basipterygium and fin radials. Balfour [14] describes the fin skeleton during early development (before stage 31 sensu [63]) in S. stellaris as consisting of a “bar [basipterygium]” whose outer surface consists of a “thin plate” extending into the fin that by stage P (stage 31 sensu [63]) has begun to differentiate into fin radials and notes that this occurs before chondrification [14]. In contrast, separate studies of P. melastomus by Wiedersheim [22] and S. canicula by Dohrn [21] suggest that the basipterygium is formed by union of individual radii or “cartilage rays”. Wiedersheim also states that the pelvic girdle itself is also a product of this fusion [21]. One of these authors, Wiedersheim [21], contends that Balfour’s observations are mistaken as they have been conducted in specimens too old to see this fusion. There are some anatomical differences between the pelvic fins of chimaeras and selachians. Chimaeras lack any radials that articulate with the pelvic girdle such as the propterygium found in selachian pelvic fins. However, despite this slight anatomical difference the development of the fins in C. milli does appear to corroborate that the basipterygium is derived from the fusion of radii in early development. The only other study of the development of the fin skeleton of C. milli has revealed, through the use of clearing and alcian blue staining, that in stage 29 embryos, mesenchymal rods are situated within the fin, and that the basipterygium is present by stage 31 and articulates laterally with these rods [41]. Using nanoCT imaging, we have visualised the development of the mesenchymal fin skeleton at stage 30 (Figure 3), which cannot be visualised with standard histological stains. Our findings demonstrate that the basipterygium forms from the proximal fusion of these “rods” supporting the hypothesis proposed by Dohrn [22] and Wiedersheim [21], and indicate that the pelvic fins of extant chimaeras and selachians occur through the same developmental mechanisms. The developmental origin of the pelvic girdle and its relationship with the fin radials during early development (prior to stage 30) remain unknown. Our data provide no evidence for Wiedersheim’s assertion that the pelvic girdle is also a result of the fusion of radii [21]. Further investigation using nanoCT imaging and specific condensing mesenchyme markers [64] on younger embryos of C. milli and other chondrichthyans should resolve these questions.Despite similarities in the formation of the basipterygium, the timing of the segmentation of the pelvic skeleton differs between C. milii and S. canicula. In both taxa the pelvic girdle and basipterygium have chondrified prior to segmentation (Figure 6) [21], but segmentation begins in stage 32 in C. milii and stage 30 in S. canicula. The segmentation of the girdle and basipterygium in S. canicula is described as resulting from the formation of a “resorption zone” between the structures [21]. Resorption is not known to occur in chondrichthyans, which are believed to be incapable of repairing or remodelling their endoskeleton [65,66,67], but see [68,69,70]. Therefore, it is unlikely that resorption occurs in the subdivision forming these joints. The difference in alcian blue staining and pre-chondrogenic morphology of the tissue in the developing joints in C. milii at stage 32 (Figure 6) may indicate that joint formation/segmentation is the result of the repression of chondrogenesis as has been suggested in some tetrapods [71]. Our findings suggest that the fin joints of chimaeras are synarthotic, being separated by cartilaginous tissue as with the fin joints of some elasmobranchs [72,73].To our knowledge the morphology of pelvic clasper skeletons has only been examined in C. milii [41] and not in any other chondrichthyan. Through clearing and alcian blue staining the earliest observation of the anterior and posterior clasper cartilages in C. milii is stage 33 [41]. In contrast, in the present study the formation of the pelvic claspers, both the primary anterior and posterior cartilages, was detected at stage 31 using standard histological staining (Figure 5F–H, AC, PC) as diffuse extensions from the condensation of the basipterygium. Despite this slight difference, which is likely a result of the different methods employed, the rest of the development of the pelvic claspers observed by clearing and staining agrees with our own with respect to the development of the clasper skeleton, including not observing the secondary components of the anterior clasper cartilage [41]. As the secondary components of the anterior clasper cartilage have not been observed embryonically it is likely these form and develop post hatching (i.e., after stage 36). Our findings concur with observations of clasper development based on external morphology in another chimaera, H. colliei [74], in which claspers can be identified by stage 31. Conversely, in a separate study of the external development of C. milli, claspers were first observed at stage 35 [46]. The morphology of clasper development in other chimaeroids and other chondrichthyans remains unknown and requires further study. Future investigations should employ multiple methods to gain a more comprehensive view of the development of these structures.Prior to the present study the development of pre-pelvic tenacula has only been examined externally in both C. milli [46] and H. colliei [74]. Rudiments of the tenacula were observed externally as “bulges anterior to the pelvic fin,” at stage 30 in C. milli [46] and stage 31 in H. colliei [74]. We also found that the rudiments of this structure are present by stage 31 (Figure 5), with the grappler being identifiable by stage 32, and all elements being identifiable by stage 36 (Figure 6E,F, TG). The slight differences between our results and those based on external morphology of C. milii [46] may be due to individual variation or the bulges observed in a stage 30 C. milii [46] may have been concentrations of mesenchymal tissue prior to the formation of the condensations of the tenacula components.