34 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 92 



and a nuich worn sherd with a rather elahorate incised design con- 

 sisting of a repeated double spiral ending in two dots. 



Our sample collection consists of half an open bowl and some i6 

 other sherds. These are all monochrome, ranging from brown to 

 red, grit-tempered and coarse in texture. Most of them are coated 

 with coral, which is almost a centimeter thick on some sherds, and 

 many other sherds at the site had been fastened into the rocks by 

 coral growth. The largest fragment is from an open bowl (i8 cm in 

 diameter) with a round bottom and a low concave neck which flares 

 out slightly at the top. There are rim fragments of several round pots 

 of medium size with low, slightly flaring rims, and several others 

 from small round bowls with no necks but thickened rims. No other 

 artifacts were noted at the site. 



Although the water in the accessible caves is fresh, there may be 

 some that is brackish. That this was a water hole used by the 

 Indians seems the most logical explanation. The whole site, however, 

 gives the impression of having once been a great cave that has falleti 

 in, and it is tempting to explain the presence of numerous broken 

 pots in inaccessible places in this way. Our own examination of the 

 site was too hurried to check this hypothesis. 



" EIGHTY ACRE " AND OTHER SITES 



This site was evidently not worked by the Boekelman Expedition, 

 but on Bird's composite photograph of the east end of the island he 

 locates a " mound with potsherds " in the vicinity. Mr. Waterhouse 

 took us to the place and is responsible for the statement that it covers 

 80 acres. It is certainly a large and important site and, being close 

 to Stuart Hill, is probably one of the main habitation places con- 

 nected with the stone causeways and other ceremonial features there 

 which are mentioned by Rose (1904). 



We examined an area of several acres on the edge of some low 

 hills (see map, fig. 2), all of which is covered with broken pottery 

 and kitchen refuse. There are a large number of low, irregular 

 mounds here which seem to be composed in considerable part of 

 refuse. They also contain human burials. We dug shallow pits into 

 two of these mounds, but although we found some scattered human 

 bones and teeth, along with sherds and other refuse, we did not 

 strike any definite burials. Mr. Waterhouse stated that both extended 

 and flexed burials, with the bones in rather firm condition, had been 

 dug up here. 



