How wilt thou reason with them, how refute
Their idolisms, traditions, paradoxes?
John Milton, Paradise Regained, 1671
On the Use of Ichthyology
Utility will always be found to depend more on the degree of attention to any subject connected with science, than on the nature of the subject itself; yet it is a common remark that this, or that, is important or frivilous, according as we happen to be acquainted with it. When we find any branch of science regarded as useless, we may be assured that, contrary to ordinary expectation, it will prove the most productive field we can enter. Science, indeed, can only be useful where it has been cultivated, and its principles worked out; practical results will then follow in proportion to the pains taken to develop them.
John McClelland (1839: 457)
On the basis of an ontogenetic series of Indostomus paradoxus, we test characters that have been proposed for the phylogenetic placement of this enigmatic taxon. Contrary to previous authors, we found that the body armor of Indostomus differs from that of syngnathoids greatly and it closely resembles that of gasterosteoids in many unique details. The body plates originate from two different sources, that is, the endoskeleton (proximal-middle radials of dorsal and anal fin, neural and hemal spines, pelvic cartilages) and the exoskeleton (postcleithra, lateral body plates, sternal plate). The median bone in the ethmoid region develops from two centers and most likely represents the nasal bones that fuse during ontogeny with each other and with the vomer. Identity of the opercular bones is clarified, and it is demonstrated that Indostomus has an interopercle. The single pterygoid bone is the ectopterygoid. A parietal is lacking. There is only one cartilaginously preformed hypural element in the caudal fin. There is no parhypural, but a similar structure, termed the pseudoparhypural by us, develops as membranous outgrowths of the single hypural and the ural centrum. The pectoral radial plate fuses to the scapulocoracoid cartilage, and the pectoral radials ossify within that fused plate without prior fragmentation of the plate into individual radials, being specializations of the pectoral girdle that we think to be shared with all gasterosteoids. Indostomus shares with other gasterosteiforms the modification of the tripartite occipital condyle into an articulation of the basioccipital and the first centrum through loss of the articulation between exoccipitals and the first centrum in all developmental stages. Indostomus lacks distal radials in all pterygiophores supporting fin spines at all developmental stages, a character shared with other gasterosteiforms, mastacembelids, and probably other smegmamorphs. We conclude that Indostomus is a gasterosteoid gasterosteiform.