48 BULLETIN 14 7, UNITED STATES NATIONAL MUSEUM 



series of similar shell vessels accompanying- child burials in another 

 section of the midden at San Juan. (Cat. No. 341006, U.S.N.M.) 

 With the child burial and the accompanying shell dish was a 2-com- 

 partment pottery vessel in red ware (2, pi. 14). Both shell dish and 

 pottery vessel are unique in Tainoan collections. The shell vessel 

 fiaiired as 4 of Plate 9 is polished, regularly shaped by cutting, but 

 is sharpened along the edge at one end to a knifelike thinness. The 

 vessel might be called a spoon. Other shell vessels accompanying 

 child burials at San Juan are flattened and are more distinctly in- 

 tended as plates. No. 4 is 4.2 centimeters (1.6 inches) wide, 5.2 

 centimeters (2 inches) long, and 2.6 centimeters (1 inch) deep. 



Several slightly worked large conch shells dug up at San Juan 

 are possibly unfinished vessels. 



Shell gouges and celts. — Reference has been made to fossile concli 

 shell gouges and celts. These figured in Plate 8 were uncovered on 

 the peninsula both at Anadel and at San Juan, one also from the 

 cave at Boca del Infierno. With the exception of 3 (Cat. No. 

 341041, U.S.N.M.), the shell implements conform to th£ Barbados 

 type referred to by Doctor Fewkes as the shoehorn type. This type 

 is unusual elsewhere in the West Indies, although it appears again 

 in the Floridan shell middens. Nos. 1, 2, 3, and 4 are entire and are 

 semifossilized ; 5 is broken off at the smaller end and is much more 

 mineralized. It is therefore heavier than either of the others. All 

 of the shell gouges from Samana are, however, of the same general 

 type. No. 4 was unearthed at San Juan; 1, 2, and 3 at Anadel; 

 while 5 comes from cave 4 of the San Lorenzo area. These shell 

 gouges were fashioned from the lip of large conch shells {Stromhu^ 

 gigas), smoothed to a celt form by crumbling and grinding, and 

 provided with a semisharp cutting edge at the base of the wider 

 end. Most of the convolutions of the lip of the shell remain un- 

 worked, as may be seen in 3, 4, and 5. In 1 and 2 these have been 

 entirely ground away in obtaining the necessary bevel for shaping 

 the cutting edge at the base. The largest of the shell gouges or 

 celts collected, 4, is 19.2 centimeters (7.6 inches) long and 6.6 centi- 

 meters wide. It is possible that 3 is not a celt but had been used 

 for another unknown purpose. Its edges have been worked to an 

 oval contour, but the end surfaces have not been cut away to a 

 beveled cutting edge. It may be an unfinished specimen. It is 8.7 

 centimeters (3.4 inches) long and 3 centimeters (1.2 inches) wide. 



Several specimens of shell gouges were collected at Anadel. The 

 large central whorl of the conch had been carefully removed by 

 fracturing and the edges shaped to a roughly triangular outline. 

 The cutting edge was obtained by grinding the basal end at a bevel 

 extending from the base to the thick and narrow upper end. The 



