ISre.] OttJ [dishing. 



wide, straight-edged sandy island on the other, bordered tlie inlet that 

 led straight eastward a mile or more to the majestic cocoannt grove that 

 fronted Collier's Bay and Key Marco. I will not describe the key 

 greatly in detail, for an admirable contour map of it, made with great 

 care by Mr. Wells M. Sawyer, artist of the expedition I later conducted 

 to the place, is furnished herewith. The key, like Battey's Landing, 

 like Johnson's key, and many other places of the kind, was now more 

 or less connected with contiguous land ; yet obviously, when built and 

 occupied, it had stood out in the open waters. It was not even yet 

 joined to Caxunbas Island, at the northwestern angle of which it stood, 

 save by a wide and long mangrove swamp that was still washed daily 

 by high tide. As may be seen by the plan, — on Plate XXX, — 

 a number of long, straight and narrow canals, terminating in little 

 court-like landings and short graded ways, stretched in from the 

 Avestern side, the lower end of which was enclosed and extended 

 by a massive, level-topped sea-wall, now used as a wagon road, 

 reaching nearly a quarter of a mile into the mangrove swamps, and 

 indicating that w'hen it was Imilt, this had been the stormward side, 

 which it had therefore been necessary to protect. There were 

 other indications that the extensive sand bank or island which now 

 fronted the key across Collier's Bay on this gulf-w^ard side, as Avell 

 as the long reaches of mangrove swamp to the southAvard, had all 

 been formed, in the main, since the date of its occupancy. This Avas 

 notably the case Avith many other keys in the neighborhood of Key Marco, 

 which keys formed, with the intermediate mangrove islets, — mere seg- 

 regated sections of swamp they appeared, scarcely rising above the tide 

 level, — the northernmost of the great archipelago of the Ten Thousand 

 Islands. Explorations among these border islands, within a radius of 

 from fifteen to twenty miles around Key Marco, demonstrated the fact 

 that on an average about one in every five of them was an ancient 

 shell settlement or key proper like Marco and the others already 

 described, and that the loAv-lying intermediate islets had mostly been 

 formed on shoals caused by drift, around and betAveen these obstructions 

 built by man, smce the time of their occupation. Again, around each 

 one of these more southerly shell keys or settlements, the fringe of the 

 mangrove SAvamps Avas far deeper, or wider, than around the more 

 northerly keys, indicating that a much greater time had elapsed since 

 their abandonment ; time enough for the formation of many miles of 

 sand bank, and the growth thereon of the mangrove sAvamps around 

 and batween them. Marco inlet, or the eastAvard and southward exten- 

 sion of Big Marco Pass, formed to the northeast and east of Key Marco 

 a comparatiA^ely wide, deep bay. The edge of the key along this bay 

 had evidently been worn aAvay to some extent, so that its eastern face 

 afforded in places sectional views of its structure that told the same 

 story with regard to this key that my excavations had told with regard 

 to Demorey's and the little keys in the neighborhood of St. James City ; 



