'124 SMITHSONIAN MISCELLANEOUS COLLECTIONS [vOL. 45 



these objects an intermediate morphological link between mammi- 

 form stones and masks, although more closely allied to the former. 



Two specimens of this hitherto unknown type of idols occur in 

 the Archbishop's collection. The first (plate xliv, 3), made of. light 

 brown stone, has a shallow eye, an elongated mouth, and fore-legs 

 cut on the sides in low relief. The second example (plate xliv, 4) 

 is even more elaborately made, the details of the jaw being more 

 completely worked out. In this specimen the fore-legs are not rep- 

 resented, but the raised forehead and throat ridges peculiar to other 

 mammiform images are well shown. The eye sockets are deep, 

 the nostrils appear in relief, and there are sviperficial markings sug- 

 gesting teeth. 



In studying the form and position of parts of this type it is evi- 

 dent that it is practically the same as that to which belong the pre- 

 ceding two with conical projections on top of the head, so that any 

 valid objection to a theory of the use of the objects belonging to this 

 type applies also to the others. 



The specimen next to be considered (plate xliii, 1,2) also has the 

 tripointed form of the mammiform zemi, but it lacks the conoidal 

 elevation, and in that respect is more like a mask. It resembles the 

 third type, or the two specimens last mentioned, except that the 

 mouth, instead of replacing the conical elevation, is situated on one 

 side, the nose being extraordinarily flattened. This specimen, like 

 the last two, came from Santo Domingo ; it was purchased from Sr. 

 Zeno Gandia and formerly belonged to the Gabinete de Lectura 

 at Ponce, Porto Rico. 



The author also purchased in Porto Rico a rude stone head, re- 

 sembling in certain respects the one last mentioned, but differing 

 from it in having a projection at the top. A corresponding pro- 

 tuberance forms the neck, suggesting that the stone may have been 

 lashed to some other object, such as a stone ring. A beautiful stone 

 of the third type, in which the nose takes the place of the conoid pro- 

 jection, was purchased from Seiior Gandia. Its lower, slightly con- 

 cave surface has been fitted to one of the Porto Rican collars, as 

 shown in plate xli, 2. 



In the absence of information regarding the use of these tripointed 

 or mammiform stones here identified as idols, it has been suggested 

 that they were merely highly ornamented mortars, the object when 

 in use being reversed — the conoidal projection being inserted in the 

 ground for stability, and the slightly concave surface thus brought 

 uppermost. This theory is advocated by Im Thurn, a generally 

 excellent authority on account of his intimate knowledge of- related 

 tribes. 



