In the history of natural science, many hypotheses made on the basis of deductive reasoning and common sense have later proven to be inconsistent with observational data. One example involves the conodont genus Polygnathus: many species thereof are cosmopolitan in distribution; ergo common sense dictates that all were pelagic creatures. However, in lower Middle Devonian limestones (Columbus and Delaware) of north-central Ohio, Polygnathus species are stratigraphically segregated, apparently owing to lack of post-mortem transport, and are easily correlated with facies indicative of different energy levels and substrate texture. The middle Columbus' Marblehead Member largely reflects shoal conditions favoring P. costatus. Large elements of P. linguiformis show preference for low-energy conditions, especially offshore below normal wave base (Venice Member and basal Delaware) but also in a near-shore-lagoon setting. The P. angustipennatus Group (including P. angusticostatus and P. intermedius) occupies a section over 4 m thick in the lower Delaware consisting of extremely fine-grained argillaceous limestone (all other species of Polygnathus are absent). These correlations reflect preferences based on energy levels and substrate sensitivity that would be of little concern to pelagic taxa but important to nektobenthic ones. The most significant evidence supporting this view is a 3-meter section in the middle of the Delaware containing abundant carbonaceous matter but not a single specimen of the aforementioned species; the organic matter means anoxic conditions and an understandable absence of nektobenthic creatures. On the basis of similar data, Barnes and Fåhræus (1975) concluded that many cosmopolitan Ordovician conodonts other than simple-cone genera were benthic or nektobenthic. It is possible that the species discussed above were filter feeders with bodies anchored in the substrate, as is the case with lancelets.