1S96.] ^"i«^ [dishing. 



had possessed narrow terraces aud two or three considerable founda- 

 tions. The greater portion of this work, however, liad been removed 

 by Captain Whiteside — at a cost of more tlian ten tliousand dollars — for 

 use in the construction of a boulevard around the end of the island and 

 of crossroads through the marshy space it enclosed. Miles of shell- 

 road — the most beautiful in southwestern Florida — had thus been made, 

 yet still the shell material of this one old-time beginning merely, of a 

 key, had not thereby been wholly exhausted. Few relics, other than a 

 couple of skeletons and numerous shreds of pottery and fragments of 

 bx'oken shell tools, had been encountered during the demolition of the 

 structure ; yet it was plain that it had been built on low encircling 

 reefs up from the very level of the water as had all the others. 



Another work, quite similar to this, but still undisturbed, was found 

 by me straight across Carlos Bay, — as the body of water to the south and 

 west of Pine Island and at the mouth of the Caloosahatchee river was 

 called — on one of the inner marginal reefs of Sanybel Island, the lower 

 end of which formed here a great loop around the bay aud entrance re- 

 ferred to. At this point the ancient key -builders had succeeded in progress- 

 ing a stage or two further in the construction of one of their settlements 

 ere they had been, evidently in like manner as at the other places, over- 

 whelmed by some catastrophe. Such portions of the work as were 

 left — for some part of it had been destroyed and washed away by suc- 

 cessive storms — formed more of an enclosure of mangrove swamp than 

 did either of those last described. It had been considerably widened 

 and built up, at its middle, and again towards its western end. Well- 

 defined canals led in from among shell-bank enclosures within the man- 

 grove swamp to both of these built-up points, the westernmost termina- 

 ting in a diminutive inner court. At both pomts, too, the foundations of 

 mound-terraces had been begun. Digging in towards the middle of one 

 of these incipient terraces from the outer shore line, I encountered not 

 only numerous relics, but also large, flat fragments of breccia-like 

 cement. Further up, on the more level portion of this tei'race, I found 

 that the cement was continuous over a considerable space, but that the 

 bed thus formed abruptly ended along a line parallel with the western 

 edge or end of the elevation. At almost regular intervals along this 

 line occurred holes in the compact substratum of shell., formed by the 

 decaying of stout posts that had been set therein — as was shown by lin- 

 gering traces of rotten wood that occurred in each. Thus it appeared 

 that this flat bed of cement had once formed a thin vertical wall, or 

 rather the plastering of a timber-supported wall, probably the end of 

 some large building which had crowned the terrace, and that had fallen 

 in under the stress of some storm or as a result of other accident. 



To ascertain whether the w^orks here were, like the outworks of 

 Demorey's key, originally founded upon a shallow or submerged reef, 

 I caused a trench several feet long to be excavated down to between 

 eighteen inches and two feet below mean tide-level. I thus ascertained 



