GFF volume 121 (1999), pp. 243–252. Article The Lower Palaeozoic of the probable impact crater of Hummeln, Sweden MAURITS LINDSTRÖM, TOM FLODÉN, YNGVE GRAHN, STEFAN HAGENFELDT, JENS ORMÖ, ERIK F.F. STURKELL, and ROGER TÖRNBERG Lindström, M., Flodén, T., Grahn, Y., Hagenfeldt, S., Ormö, J., Stur- kell, E.F.F. & Törnberg, R., 1999: The Lower Palaeozoic of the probable impact crater of Hummeln, Sweden. GFF, Vol. 121 (Pt. 3, September), pp. 243–252. Stockholm. ISSN 1103-5897. Abstract: A 1.2 km wide, sub-circular hole in Lake Hummeln (57°22ʼN, 16°15ʼE) has been suggested to be an impact crater. The lake has fur- thermore been suspected to contain an outlier of Cambrian and Ordo- vician sedimentary rocks. A shallow seismic survey, followed by core drilling to 164.25 m below the lake surface, demonstrated the presence of a strongly slumped succession consisting, from below, of shattered crystalline basement rock, Lower and Middle Cambrian sandstone and claystone, Lower Ordovician limestone, and Middle Cambrian clay- stone. The lithogenesis and original stratigraphy are closely similar to coeval beds on Öland 40 km to the east. Dating was done through acritarchs in the Cambrian, and through conodonts and chitinozoans in the Ordovician. An impact is postulated although no quartz with PDF has been identified in the available lithologies. The drilling had to stop, apparently without reaching the crater floor. If dated by the oldest sedi- ments present, the impact would be Early Cambrian. A late Early or early Middle Ordovician age, however, is more likely because slump- ing after impact filled the crater with sediments aged Lower Cambrian through Lower Ordovician, and because the slumped sediments bear evidence of not having been originally deposited in a crater. Keywords: Impact, Hummeln, Sweden, slumping, Cambrian, Ordovi- cian, acritarchs, conodonts, chitinozoans. M. Lindström, T. Flodén, S. Hagenfeldt and R. Törnberg, Dept. of Ge- ology and Geochemistry, Stockholm University, SE-106 91 Stockholm, Sweden, e-mail maurits.lindstrom@geo.su.se; Y. Grahn, Faculdade de Geologia, Universidade de Estado do Rio de Janeiro, R. São Francis- co Xavier 524, Maracanã, 20559-900 Rio de Janeiro, R.J., Brazil; J. Ormö, International Research School of Planetary Sciences, Univer- sitaʼ dʼAnnunzio, Viale Pindaro 42, 65127 Pescara, Italy; E.F.F. Stur- kell, Nordic Volcanological Institute, Grensasvegur 50, 108 Reykjavik, Iceland. Manuscript received 21 August 1998. Revised manuscript ac- cepted 31 July 1999. In Swedish geological literature, Lake Hummeln in the parish of Kristdala, SE Sweden (57°22ʼN, 16°15ʼE; Fig. 1), is considered a possible outlier of Lower Palaeozoic sedimentary bedrock on pre-Cambrian crystalline basement and an astrobleme. A 164 m long drill core, described herein, proves the existence of Cam- brian and Ordovician rocks in a topographic basin for which an impact is, indeed, the most likely explanation. Palaeozoic rocks have been identified in large concentration in a glacial boulder train that extends southeastwards from the lake through the farm Humlenäs (Fig. 2). The boulders were first described by Hisinger (1826). Good descriptions were given by Linnarsson (1878) and Svedmark (1904), who concluded that the succession represent- ed by the boulders occurs in situ under the lake. The crystalline bedrock is porphyric granite belonging to the Proterozoic Trans- scandinavian Igneous Belt (TIB). A number of outcrops of this granite are described as severely brecciated, with the interstices between clasts either filled by gouge or with either calcite or flourspar cement. The Cambrian is reported to consist mainly of sandstone that is fine to coarse grained and well to poorly sorted. Like Cam- brian sandstones of the continuous Palaeozoic cover 25 km to the east, it contains Skolithos burrows. Whereas the sandstone has been assumed to be Lower Cambrian, the Middle Cambrian is represented by “green shale with Paradoxides oelandicus?” (Sved-mark 1904, p. 49) and “sandy limestone with Paradoxides Tes-sini” (=paradoxissimus). Dark grey, bituminous limestone nodules are reported to yield Agnostus pisiformis (Linnarsson 1878, p. 7) and Orusia lenticularis (Westergård 1947, p. 11) and, thus, belong to different levels of the Upper Cambrian. The Ordovician is represented by Lower Ordovician limestone of a medium bedded hemipelagic facies that is commonly re- ferred to as “orthoceratite limestone” (Lindström 1979). Two varieties are referred to, one being red, and placed in what is referred to as lower Arenig, and one being grey, and regarded as upper Arenig to Llanvirn. The grey variety contains glauconite and a highly diverse fauna that includes trilobites, as well as ses- sile benthos, such as bryozoans and big articulate brachiopods. The basal Ordovician Tremadoc Series is taken to be most likely missing (Westergård 1947). The Ordovician is closely similar to the coeval succession on northern Öland, some 40 km to the east. Moberg (1911) rejected the idea that the Hummeln boulders represent an autochthonous outlier of Lower Palaeozoic, situated in the lake. Although Mobergʼs claim, that the boulders came from northern Öland by Pleistocene ice transport, is contradicted by the circumstance that the flow of ice came from the northwest, it removed the Hummeln occurrence from further general con- sideration for several decades. Fredriksson & Wickman (1963) mentioned Hummeln as a possible impact crater. According to an earlier interpretation by Asklund (in Nordenskjöld 1944, p. 164), Hummeln owed its origin to explosive volcanism. Nordenskjöld (1944), who found much evidence of extensional tectonic movements in the area but no evidence of explosive volcanism, suggested that the basin was an abortive volcanic structure in a tensional tectonic set- ting. Svensson (1966) gave credence to the impact hypothesis of Fredriksson & Wickman by demonstrating that the lake cov- ers a 1.2 km wide subcircular topographic basin that is 61.5 m deep at its deepest point. His map reiterates previous mapping by Nor-denskjöld (1944) but is based on soundings with closer average spacing. New observations on the basement breccias, presented by Svensson (1966), did not add conclusively to the discussion. Wickman (1988) listed Hummeln among 15 possible impact structures in Sweden and published a gravimetric map of