350 CERTAIN ABORICxINAL REMAINS OF THE NW. FLORIDA COAST. 



Tampa bay. It is only fair to say, howevei', that owing to the hick of inhabitants 

 along that i^art of the coast, from whom inquiry could be made, certain mounds may 

 have escaped him. Besides, several years before, we had covered the territory from 

 Tampa almost to Clearwater Harbor (see outline map) with but negative results, 

 while two men in our employ, one of whom w^as very familiar with the coast, had 

 searched from Tampa to Anclote Key, finding only mounds previously located and 

 dug into by Mr. S. T. Walker, ^ whose researches were continued still farther north 

 without discovery of importance. 



Mr. Cashing, '' it is true, opened a mound at Tarpon Springs, in w^hich he found 

 fragments of interesting ware, but we believe this mound to have been an exception 

 to the general run of mounds from Cedar Keys southward. Just north of Cedar 

 Keys the great shell-heaps of the west coast begin, and neither on the east coast, 

 where large shell-heaps are throughout, nor in the territory of the great shell-heaps 

 on the west coast have we found the contents of mounds to be of much interest. 



As we have stated, our work of last season ended at the eastern extremitj' of 

 Choctawhatchee bay and began this year at the western end of St. Andrews bay, in 

 direct continuation (see map). 



During our season's work certain points were brought to our attention. 



Going eastward along the coast, we saw the waning influence of Alabama and 

 of the middle Mississippi district as to composition of ware, the admixture of shelP 

 entirely disappearing, none having been met with by us east of Choctawhatchee bay, 

 which marks also the easternmost limit of the polished, black ware of Mississippi, as 

 found by us. 



1 Smithsonian Report, 1879. 



- Proceedings American Philosophical Societ}-, Vol. XXV, No. 153. 



' In athoughtfnl article in the "American Antiquarian, " May and June, 1902, entitled "Primi- 

 tive Keramic Art in Wisconsin", the author of that paper is inclined to believe that a mineral sulistance, 

 and not pounded shell, is used to temper earthenware through parts of the mound region, where hitherto 

 we have been led to believe pounded shell was used. 



To determine this matter, we submitted to Prof. Harry F. Keller, Ph. D., a fragment of thick, 

 porous earthenware found by us in Alabama, which contained a considerable amount of what we have 

 always considered broken shell used for tempering. 

 Dr. Keller's report is as follows : 



Junfe 24, 1902. 

 "The analysis of the carefully picked material (shell) from the specimen of earthenware from 

 Alabama gave the following results: 



Insoluble in dilute acid, 3.09/^ 



(Silica and ferruginous clay) 

 CaO (Lime) " 54.07 



MgO (Magnesia) -19 



MnO (Manganous oxide) -09 



Fe.Oj (Ferric oxide) -11 



CO, (Carbonic anhydride) 41.58 



PjOj (Phosphoric anhydride) -62 



Moisture -H 



99.86 

 "The small amounts of the oxides of iron and maganese may be derived from adhering matrix. 



The insoluble residue contains traces of carbonaceous matter." 



Here we have almost pure carbonate of lime, showing the scaly fragments in the earthenware to 



be shell. 



