FEWKES] ARCHEOLOGICAL OBJECTS 181 



The ditferent forms of bowl handles and other fragments of pottery 

 that were excavated in the Cueva de las Golondrinas, near Manati, are 

 represented in plate lxxiii. The handles are in general similar and 

 evidenth^ belonged to coarse bowls, vases, and ollas. In similar forms 

 a raised ring of clay served all the purposes of a handle, but there 

 were often added grooves with adjacent elevations. The handle was 

 sometimes broad and flat, at other times narrow and round. One of 

 the specimens represented on this plate has two solid knobs on the 

 rim; another is perforated just below similar knobs. The edges of 

 the handles of many vessels are pinched into ridges that may be cor- 

 rugated, notched, or serrated. i 



Hardly any two handles are exactly alike; those on this plate rep- 

 resent only the more tj'pical forms. These show that there was an 

 abundance of red ware. The surface of this pottery in one or two 

 instances is smoothly polished. The majority of specimens are handles 

 of cooking pots, man_y of which are blackened by soot from fire. 



Every collection of Porto Rican ohjects contains l)urnt-clay heads 

 (figure 3-i), that have been broken 

 from some bowl, vase, or other 

 pottery object. These heads 

 vary in size, form, and other 

 particulars, and are often fash 

 ioned with considerable skill. 

 Not unusuaUy we find consider- 

 able fragments of flat dishes or 

 bowls attached to them, showing 

 that they are handles; but in 

 some instances they were evi- 

 dently low relief figures pressed 

 on the surfaces of the vessels. 



These objects are sometimes (-lUed zemis or idols, an identification 

 that is misleading, for there can hardly be a doubt that they are simply 

 portions of broken jars or vessels excavated from caves, shell heaps, 

 or dance inclosures. 



Plates Lxxiv and lxxv give a fair idea of the general forms of these 

 clay heads. The latter have little in common except their large round 

 eyes, fillets over the forehead, and mouths wide open. The general 

 cast of many of the specimens suggests monke}^ heads, but this resem- 

 blance is unintentional, being due rather to the method of working clay 

 into faces adopted by the ancient potters. It is impossible to identify 

 the great majoritj' of these figurines, and they may be regarded as 

 simply fantastic forms used for decorative purposes, having no further 

 import f)r meaning. 



Professor Mason gives several figures of fragments that illustrate 

 fairly well the general character of Porto Rican pottery. Five of 

 these are heads which once decorated the rims of bowls or vessels of 



