1896.] o6v [Gushing. 



which was almost as much overgrown with luxuriant and forbidding 

 vegetation as had been the wilder key first explored, I climbed high up 

 among the skinny and crooked limbs of a gigantic gumbo limbo that 

 grew directly from the inner edge of this elevation. Luckily, great fes- 

 toons of tough vines clung to the lower limbs of this tree, for in shifting 

 my position I slipped and fell, and was caught by these vines, to the sal- 

 vation of my bones probably, since by the force of the fall some of the 

 vines were torn away, revealing the inner side of this platform and the 

 fact that it was almost vertically faced up with conch-shells ; their 

 larger, truncated and spiral ends, laid outward and in courses so regular, 

 that the effect was as of a mural mosaic of volutes. I hastily tore away 

 more of the vines, and found that this faced-up edge of the platform 

 extended many feet in either direction from the old gumbo limbo. I 

 may say here, that on occasion of two later visits I cleared the facade 

 of this primitive example of shell architecture still more ; was enabled, 

 indeed, when I last visited the place — since I was then accompanied by 

 a considerable force of workmen — to entirely expose its inner side 

 and its southern end. Thus was revealed — even more completely than 

 is shown in Plate XXIX, — a parallelogrammic and level platform, 

 some three and a half feet high and twelve yards in width, by nearly 

 thrice as many in length. It was approached from the inner side by a 

 graded way that led obliquely along the curved ascent up from the man- 

 grove swamp, to a little step-like, subsidiary platform half as high and 

 some twelve feet square, which joined it at right angles, just beyond 

 the point shown at the extreme right of the picture here given. The top 

 of this lesser step, and the approaches to either side of it, were paved 

 with very large, uniform-sized clam-shells, laid convex sides upward, and 

 as closely and regularly as tiles. The lower or southern end of the main 

 platform was rounded at the corners, and rounded also on either side of 

 the sunken ascent midway, in which the longer of the graded ways I have 

 described terminated. Contemplating the regularity of this work, its 

 central position, and its evident importance as indicated by the several 

 graded ways leading to it from distant points, I could not doubt that it 

 had formed the foundation of an imposing temple-structure, and this 

 idea was further carried out by the presence at its northern end of two 

 small, but quite prominent altar-like mounds. 



Descending from the end of the platform down along the main 

 graded way — the one which divided the terraces from the central group 

 of high mounds — I found that at more than one point, the sides of this 

 deep, regular path, had also been faced up with conch-shells, though 

 none of the courses were now, to any extent, in place. 



At the foot of the inner and parallel sided, sunken or graded way — 

 the one descending from between two of the great central mounds — I 

 caused an excavation to be made between the two straight banks or ridges 

 of shell that extended thence far out into the mangrove swamp, in order 

 to ascertain whether this supposed canal had really been such ; that is, 



PROC. AMER. PHILOS. SOC. XXXV. 153. 2 Q. PRINTED JUNE 2, 1897. 



