548 MOUND EXPLORATIONS. 



port jtalisades, it yet was liigli enough for a breastwork, and probably 

 served as such. A single gateway sufficed for some small inclosures, 

 but there were usually more. These works vary greatly in size, some 

 having an area of large dimensions. Quite rarely they present no 

 marks of occupation. While often on commanding spots, they are fre- 

 quently overlooked by some near eminence, and are occasionally found 

 on low lands or in swam^js. If the situation has natural defenses, as 

 steep banks, either of ravines or streams, these parts may be left 

 apparently open. 



Stockades and embankments have been found near together, and 

 palisades may have had supports of earth, piles of wood, or cross tim- 

 bers, such as the Hurons and Iroquois commonly used when first known. 

 The first would remain, the last two would leave no trace. There is 

 little direct evidence that palisades surmounted the earthworks, and 

 reasons can be given for this. Besides cross timbers, other supports 

 were sometimes used, suggestive of the bank of earth. The Seneca 

 village visited by La Salle in 1669 had "palisades 12 or 13 feet high, 

 bound together at the top and supported at the base behind the pali- 

 sades by large masses of wood of the height of a man." 



A well-preserved trace of an oblong stockade near Cazenovia, New 

 York was recently examined by the writer. The line of the palisade was 

 a trench a foot wide and deep, in which the posts were set at intervals 

 of about 30 inches fi-om center to center. No holes were dug, but the 

 sjiace between was refilled. By the settling of the ground and the 

 decay of the posts the trench became distinct again. With so slight a 

 hold the need of binders at the top and cross timbers at the sides 

 becomes evident. With these appliances there may sometimes have 

 been no digging at all. 



. As a rule gateways show no signs of defense, though some there 

 must have been. In a few instances these are found. The double 

 walled inclosure in Shelby, Orleans county, New York, has not its 

 inner and outer gateways opposite, but the inner gateway is protected 

 by the outer wall. A large earthwork, nearly a parallelogram, de- 

 scribed by Mr. T. A. Cheney,' had a wide gateway, " with elevated 

 mounds upon each side, to giuird the entrance." These formed really 

 an outer and inner wall. An overlapping wall, forming a gateway, 

 occurred in Macomb county, Michigan, but a simple inside barrier of 

 wood may have been the usual defense, protecting the approaches 

 within. 



The outside ditch is rarely lacking in earthworks, and rarely found 

 with stflckades, if at all. In one instance, in Michigan, the trench 

 appears ]iartly within and partly without, as though a matter of indif- 

 ference, and convenience may commonly have governed its position 

 more than considerations of defense, being simply the places where 

 earth for the wall was most easily procured. 



'Auc. Mon. in western New York, p. 44, PI. vii. 



