from High Level Terraces of Thames Valley. 53 



which the river has undercut its banks since the Pliocene period. 

 The deposits of the highest terrace are Pliocene, and no flint 

 implements have been found in them. The second terrace deposits 

 are early Pleistocene and contain Strepyan forms; the third, or 

 30 metres, terrace gravels have Strepyan forms at their base, and in 

 the sands near the top implements assignable to early Chelles. 

 Upper Chelles types are plentiful in the gravels of the lowest terrace. 

 Sweeping downward from the plateau over these terraces are two 

 mantles of loam, an earlier and a later, called respectively the older 

 and the younger Loess. Each of these Loess deposits is separable 

 into upper, middle, and lower beds — those of the younger Loess being 

 called locally the Ergeron ; each seam is underlain by a bed of 

 pebbles, while the surface of each is an ancient soil. 



Calcareous concretions, known as Poupeesor Loess-puppchen, occur 

 at the base of both Upper and Lower Loess, those of the latter being 

 larger than the piippchen of the Upper Loess. 



The ancient Loess consists in upward succession of (a) the Union 

 rouge sableux, (b) the Union d points noirs, and (c) the Union rouge. 

 In gravels at the base of the Union rouge sableux the whole series of 

 forms assigned to the St. Acheul stages are found, but no implements 

 of any kind have so far been found in beds b and c. 



In the younger Loess, and especially in the pebble-beds at the 

 base, the Le Moustier types have been recorded ; some occurrences 

 of implements of the later cave periods also occur in this younger 

 Loess. 



Most of the Chelles implements are coups-de-poing. This fact is 

 so striking that it led M. de Mortillet 1 to conclude that no other 

 type of implement was made during that period, but more recent 

 finds show that in the Somme Valley as in that of the Thames flake- 

 implements also occur, though rarely, with the other forms. One of 

 these 2 is practically identical in every detail with that figured on 

 PI. II, Fig. 2, and further corroborates the inference drawn from the 

 English evidence. These curved scrapers possess the characteristics 

 of the racloirs of Aurignacian age found at Chatelperron and to 

 some extent the broad points from the Abri Audi. They are of 

 clumsier technique, and are larger, but evidently designed to meet 

 a similar need. 



In the following notes a brief description is given of the implements 

 figured upon Plate II. 3 



Fig. 1. — Large flake of yellow-mottled flint, unrolled, with broad rounded 

 end flaked to form a scraper, the under (flat) face trimmed at end ; bulb of 

 percussion and flat striking platform or butt. A median ridge runs half-way 

 from the rounded end, and the side edges are irregularly worked. Length 4 • 1 in. , 

 breadth 3-1 in., thickness 1-1 in. From the lower part of the Middle Gravel 

 at Barnfield Pit, Swanscombe, Kent, for which see Archceologia, lxiv, 185. 

 A specimen pairing with this is in the British Museum. It came from the 

 Somme Valley drift at St. Acheul, and has the following dimensions : 



1 Bull. Soc. d'Anthr., Paris, 1887, ser. in, pt. x, p. 173. 



2 V. Commont, L' Anthropologic, x\x, p. 551, fig. 40, 1908. 



;! [The Editor regrets that the name of the author has, by an accident, been 

 inserted at the foot of Plate II instead of that of his friend, Mr. B. A. Smith, 

 who kindly made the drawings for him. — Ed.] 



