286 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



northern material. Zache (13) remarked that among the Tertiary 

 sands in Brandenburg are bowlders and sands of northern material 

 in a few places, possibly representing a still older glaciation, but so 

 far as I am aware this is unconfirmed. 



ARCIIE0L0GICAL PERIODS. 



The correlation of these three glacial periods with the archeo- 

 logical stages is still somewhat uncertain. In 1910 R. R. Schmidt 

 (14) placed the Aurignacian stage in the postglacial period, and in 

 1913 the same author (15), with the assistance of E. Koken and 

 A. Schliz, made an elaborate attempt to correlate the diluvial stages 

 with the archeological sequence mainly on the basis of cave explora- 

 tions in Swabia (southwest Germany). He finds that the Mous- 

 terian, Aurignacian, Solutrean, Magdelenian, and xVzilian-Tardeno- 

 isian stages follow immediately one upon another, with no break 

 or hiatus; the accompanying diluvial fauna shows that below the 

 Mousterian and at the beginning of the Magdalenian are two beds 

 with high-arctic rodents, indicating two deteriorations of climate. 

 He considers that there is no room for an interglacial between these 

 two arctic beds, so that the Mousterian belongs to the maximum of 

 the Wiirm glacial period, and the Magdalenian to the Buhlstadium. 

 Gagel, however (review in Geologisches Centralblatt, vol. k 2(), pp. 449- 

 451), points out that this correlation is invalidated by the erroneous 

 age determinations of some beds in the north German diluvium, e. g., 

 an interglacial bed at Markleeberg, near Leipzig, ascribed to the last 

 interglacial period, undoubtedly belongs to the first. This error was 

 also pointed out by F. Wiegers in 1913 (16), who found Mousterian 

 implements in a calcareous tufa at Ehringsdorf, associated with 

 Elephas antiquum and a warm fauna, and therefore interglacial. 



Gagel describes (17) from a bowlder sand in West Holstein flint 

 implements apparently of early neolithic type (ax, scraper, thin, 

 long knife, etc.). The implements are in situ 40 to GO centimeters 

 deep in bowlder sand lying on upper bowlder clay at Michaelisdonn ; 

 although ' they lie among well-roiled pebbles the}' are quite sharp 

 angled. If the age determination is correct, it carries the neolithic 

 period back into the last glacial period. 



In 1913 (18) Gagel also summarized the facts which throw light 

 on the position of the paleolithic stages in the glacial sequence. Prac- 

 tically the only definite horizon is that given by the very character- 

 istic knives of the Levallois stage in the younger Acheulian, which 

 have been found in association with interglacial deposits at Hundis- 

 burg near Neuhaldensleben, northwest of Magdeburg, in Saxony (by 

 Wiegers) and near Leipzig. At Hundisburg the implement bed con- 

 tains also an interglacial fauna of snails, mussels, and great diluvial 

 mammals. Above this lies a bowlder clay covered by an important 



