Vörös A. szerk.: Fragmenta Mineralogica Et Palaentologica 15. 1992. (Budapest, 1992)
was able to spread up to the whorl surface again. In the third case the shell secretion accelerated along the full length of the peristome. All feeding individuals were left too far from the mantle edge at the same time so the entire colony died. If the epibiont had been a commensalist (e.g. a filter feeder), it must have survived these crises. Animals producing elongated microborings of similar size occur in Sipunculida, Annelida, Porifera, Phoronidea and Bryozoa but only the last three groups build comparable ramifying systems. From the last three phyla, surface excavating forms occur only in Bryozoa. However, in certain modes of preservation the boring traces of the others can also appear superficially. The poriferan canals can be easily distinguished as they are not parallel-sided but rather similar to a pearl necklace or a less regular row of chambers. Their width decreases away from a usually well defined centre. Phoronideans bore cylindrical tunnels with circular openings to the surface. Though they are solitary animals, different generations of asexually born specimens can live in "false colonies" in ramifying boring systems. If the outer walls of such borings are dissolved, the excavation pattern may appear similar to that on the shells of Proconulus (e.g. Talpina VON HAGENOW, 1840). However, phoronidean borings are 0.2 - 0.38 mm wide (Voight 1972) so they are significantly larger than the excavations studied here. It is possible to conceive fossil phoronidean species of smaller size but many of the other observed characters of the canal system suggest a true colony builder. Therefore these traces seem to be most likely of bryozoan origin. Modern boring bryozoans - Terebriporidae, Immergentiidae (Ctenostomata) and Penetrantiidae (Cheilostomata) - produce characteristic combinations of delicate canals (stolons) and pits (zoecia) of larger diameter. Their fossil counterparts are also known from the Mezozoic but the trace fossil studied here does not show close relation to them. Comparable bryozoans occur among the Palaeozoic Ctenostomata. Ropalonaria (Ropalonariidae) is partly embedded in the shell of the host like the inferable Bakony species but morphologically it is similar to vinellids. Their tubular stolons bear pores interpreted as attachment points of zoecia. The excavations on the shell of Proconulus may be traces of such tubular, membranous or chitinous stolons which were partly embedded. Seule and Seule (1969) proved that some Cheilostomata (Penetrantia) also have boring ability. Because this bryozoan order also contains membranous and chitinous forms with creeping stolons, a cheilostomatan origin of the excavations on Proconulus is possible, too. Hary (1987) described Talpina-type ramifying borings of proven subsurface origin on Liogryphaea shells. Their dimensions are close to those of Talpina but Hary thinks the traces are comparable to Stomatopora (Tubuliporidae, Cyclostomata), However, this order contains genera with zoecia of calcareous outer wall only (Bassler 1953), therefore a cyclostomatan origin can be ruled out. CONCLUSIONS The bryozoan origin of the traces seems very likely, although the lower rank systematic position of the borer (Order Ctenostomata or Cheilostomata) has not been unambiguously determined. At generic and species level, the finds seem to
