FEWKE.s] AECHEOLOGICAL SITES 87 



The largest shell heaps visited b\- the author were those near Coamo 

 springs, on the farm of Sefior Vincente Usera. who fii'st called the atten- 

 tion of the author to these remains. The heaps are very extensive, 

 covering considerable portions of the bluff near his house on the 

 left bank of the Coamo river. At this place there is an embankment 

 20 or more feet above the river in which are layers of shells with 

 fragments of i^otterj'. On the surface shells occur for a hundred feet 

 or more from the edge of the embankment. These shell heaps would 

 well repay systematic excavation. The numerous fragments collected 

 were practically identical in character with those from Cayito and 

 resemble those from shell heaps in -lamaica." 



Caves 



Porto Kico has many caves, some of which are of great size and 

 beauty. Many of these caves have not been explored to their full 

 extent and but a few have been entered bj' archeologists. The only 

 sj'stematic attempts to discover aboriginal objects concealed in their 

 floors are those of M. Pinart in the neighborhood of Arecibo and of 

 the author in caves near Manati. 



It is impossible to enumerate all the caves on this island, iov they 

 occur along the whole north coast and in the mountains wherever the 

 soft limestone comes to the surface. Some are inaccessible: many are 

 entered from a level, others from the top. The majority bear evi- 

 dence of great age; others are more modern. Many caves are simple 

 shelters with overhanging cliffs, the entrance being open to the sky; 

 others have ver3- narrow passageways opening into large chambers 

 which communicate by means of passages with smaller rooms extend- 

 ing deep into the bowels of the earth. 



There are caves showing evidences of human occupation, near Aguas 

 Buenas, Bayamon, Ciales, Manati, Arecibo, Corosal, Caguas, Adjuntas, 

 Utuado, Coyuco, Juana Diaz, and in many other localities. These par- 

 ticular ones are mentioned, not as exhausting the possible list, but as 

 those visited by the author or those from which Indian objects are 

 known to have been obtained. As a rule, these caves have many stal- 

 actites and stalagmites, which have obscured such evidences of occu- 

 pation as mural pictographs. Niches that are evidently artiiicial, plat- 

 forms, and other artificial stiuctures occur quite frequently. All the 

 caves visited by the author had been previously entered by others, and 

 whatever movable objects, such as idols or pottery, formerly stood in 

 these niches or on these platforms had long before been removed. 



The floors of the Porto Kican caves ofler the most promising places 

 for archeological examination, and for that purpose the author chose 

 the cave called the "Cueva de las Golondrinas," situated not far from 



n Edith Blake, the Norbrook Kitchen Midden. Victoria Quarterly, ii, Kingston, Jamaica. 1890. 



