Bibliographical Notices. 151 



shell of Pectuncidns (like P. violaceus and P. glycymeris), smoothed 

 down, and bored at the umbo. In short, says H. Karstcn, we found 

 nearly, if not quite, the same conditions as described by Do Taillefer 

 and Saussure ('Archives Sc. Phys, Nat.' 1870) at Veyricr and Ville- 

 neuve on the Lake of Geneva, aiid by Von Fraas at Hchusscnreid, 

 and quite the same objects, only more sparingly, as were found close 

 by on the south-cast side of the Reyath, near Thayingcn* (' Neues 

 Jahrbuch fiir Min. GcoL u. Paliiont.' 1874, pp. 2Go-2G8). As at 

 the places mentioned, and at many others worked out in the Depart- 

 ment of Dordogne and in Belgium, the remains of human households 

 are found in this so-called civilization-bed (Culturschieht), without 

 any trace of pottery, under turf-, tuff-, and breccia-deposits, so at 

 the Rosenhalde this bed yields no evidence at all of the existence 

 and use of cooking-vessels. From the entrance nearly to the middle 

 of the cave this bed was streaked grey and black, and contained a 

 larger proportion of flint knives ; and some charcoal, burnt bones, 

 and flat pieces of limestone and sandstone, burnt red, here clearly 

 indicated a fireplace or hearth. At the left side, towards which the 

 bods gently sloped, the implements and chips were particularly 

 abundant- The lx)undary between this implement-bed (1 foot thick 

 on an average) and the loam beneath is not definite ; and probably 

 the early cave-dwellers here trod many of their refuse things into 

 the loam softened in rainy weather by drip-water. 



4. This lower loam, brownish yellow in colour, was very thin in the 

 back part, and about a foot thick in the fore part of the cave. It 

 had none of the small angular limestone fragments, but contained 

 numerous irregularly shaped nodules, rough to the touch, and mostly 

 penetrated by crystalline veins. Together with flints and small 

 nodules of Bohnerz (concretionary oxide of iron), these nodules 

 occur of all sizes, and belong apparently to the same category as 

 some very large blocks (one measuring half a cubic metre) which 

 were noticeable in the upper beds. The flint nodules have a white 

 chalk-like crust, as much as 4 lines thick. Some fragmentary bones 

 and molar teeth of Mammoth found in the cave appear to have come 

 from this bed, if, indeed, they do not belong to the lowest part of the 

 bed with flint knives and reindeer-bones. 



5. In the back part of the cave, under the loam was a local deposit 

 of tough white clay, without bones or stones, similar to the mamma- 

 liferous fire-clay and pottery-clay on the top of the Keyath. 



Among the several subjects of interest discussed in this memoir, 

 the author gives his reasons for believing that the cave-folk were 

 eannibaLs, on account of the split marrow-bones and the peculiarly 

 fractured condition of a piece of human skull found at the Rosen- 

 halde — thus accepting the conclusions arrived at by Spring studying 

 the Chavaux cave, by Jarrigou on the cave near Montesquieu- 

 Avantes, and by Yirchow (Address, ' Naturf. Ver. Wiesbaden,' 1873). 

 Remarks also on the probable history of the several deposits, com- 

 parisons of the contents of the Rosenhalde cave with those of the 



* See also above, p. 148. 



