STONE IMPLEMENTS FROM SHORES OF LAKE MICHIGAN. 599 



about rightly proportioned, and as the collections represent a number 

 of sites this relation can be relied upon as evidence that waste and 

 products are fairly accounted for. As already stated, the greatest care 

 was taken to collect all visible trap waste on the sites studied. Later 

 visits to the sites have yielded little additional material. The preva- 

 lence of outer flakes seems to indicate designed flake production rather 

 than the blocking out of blades, especially as it is associated with 

 general use and specialization of outer flakes. There are excellent 

 reasons for assigning nearly all flaked cobblestone nuclei to the unde- 

 signed waste, though there may be occasional instances of used and 

 shaped nuclei. 



Methods of shaping. — Thus far in the description of flaking opera- 

 tions direct percussion has been the method of working implied. In 

 the chert flaking of the region trap hammer stones have been used, 

 other stones rarely serving this function. Hundreds of sites yield trap 

 pebbles battered in portions of the circumference and associated with 

 chert flakes and rejects. That similar hammers may have served at 

 Benton in the flaking of trap is possible; as elsewhere, they are present 

 in the mixed refuse of the sites, but chert flakes and rejectage are like- 

 wise present. There is, however, a distinction to be drawn between 

 two classes of hammer stones on the Benton sites — a distinction which 

 might easily result from the different properties of the two materials 

 flaked, the trap being only of medium hardness and remarkably tough; 

 the chert hard, but brittle. From the great number of trap hammers in 

 certain localities associated with chert refuse only, the incidental effect 

 upon the hammer in shaping that material is shown to be nearly con- 

 stant, many light or moderate blows with the softer rock upon hard, 

 flinty angles having produced finely battered portions of its margin. 

 Distinct from these are a small number of stones on the Benton sites 

 which have been subjected to few and exceedingly hard blows, produc- 

 ing coarse and irregular battering (Plate II, b). In size and shape 

 such battered stones were like those selected for flake making by the 

 aborigines. 



Experiments inflaJcing. — When flaking is attempted experimentally 

 with fresh stones from the beach, blows of sufficient force to flake a 

 trap cobblestone uDar the trap hammer in precisely the same manner as 

 the rougher specimens of the refuse have been marred. The blow nec- 

 essary to flake trap smashes a chert stone into innumerable pieces 

 which have no resemblance to flakes. Flaking of the hammer rarely 

 occurs in working chert experimentally, and from investigations on 

 many sites it is clear that it rarely occurred in aboriginal manufacture. 



On account of the force required in flaking trap bowlders, it proved 

 to be a ditficult matter to hold the stone which was to be flaked in the 

 hand. Both hands were numbed by repeated shocks and often con- 

 tused by glancing strokes, the perpendicular blow upon the edge per- 

 mitting little chance of escape for the fingers of either hand in case of 

 a deviation in direction on the part of the hammei-. The rebound like- 



