240 NEW YORK STATE MUSEUM 



a long, very narrow " hog back," a tongue of land which f urms a 

 continuation of the point itself aild slopes gradually down from the 

 flat above to the creek bank just above the mouth of the brook. A 

 narrow path may still be seen along the crest of the " hog back," 

 with steep declivities on either side ; handy for getting down to the 

 water and at the same time easy of defense. 



Before taking up a description of the earthwork let us glance at 

 the conditions which made the site a suitable one for a stronghold or 

 a place of habitation. Its commanding position with views of both 

 Cattaraugus and Clear creeks and the broad expanse of valley, the 

 steepness of the banks and its consequent inaccessibility, the ease 

 with which the narrow neck could be fortified and the point thus 

 shut off from the rest of the plateau, the good spring on the side hill 

 inside the wall and within 10 feet of the top, all recommend the 

 place from a military standpoint. It was, moreover, a place where 

 all the necessaries of life were within easy reach: the above- 

 mentioned spring, the creek with its teeming fish, the good corn 

 lands close by, everything was favorable. Perhaps the field, now 

 washed away, at the base of the bluff was used as a corn field, or 

 perhaps the level tract of the plateau, or of the the flood plain across 

 the creek. It does not seem strange that the ancient fort-builders 

 chose this location. 



The archeological environment should also be looked into before 

 taking up the details. The Silverheels stronghold, the scene of the 

 1903 exploration, lies about 3^ miles down the creek and on the 

 opposite side on a point of the terrace where the flood plain is con- 

 siderably narrower. On the same side as the Double Wall Fort but 

 only lyi miles below, at the mouth of Big Indian creek, is another 

 similar work, the Burning Spring Fort, which is much older than 

 the Silverheels site. Many arrow points, flakes and the like have 

 been found by the Indians on the now destroyed field at the foot 

 of the bluff at Double Wall Fort, which may point to the previous 

 existence of a camp or even a village there instead of a corn field. 

 Up stream from Double Wall, the first site of which I have definite 

 knowledge is the now obliterated semicircular work on Point Peter 

 above Gowanda and fully 8 miles distant. Across the creek, 

 however, there are several sites within a radius of a mile or so. 

 Immediately across is a low alluvial terrace, part of the flood plain, 

 which shows slight evidences of occupation in the shape of occa- 

 sional chips, points, net sinkers and potsherds with a few fire-cracked 

 stones scattered over several acres. There are other sites of this 



