EASTER ISLAND. 61 



The tools used in carving the images are large stone hammers made out 

 of hard lava rocks, and smaller tools of obsidian which must have been used 

 for carving the wooden idols and cutting the wooden tablets. 



The stone images seem to be of very different age. Those found scat- 

 tered over the island are of coarser workmanship than those found at Rana 

 Roroka, all of which liave a more animated expression. The difference may 

 perhaps be due to tlie longer weathering and exposure of the images dropped 

 on the way to their respective platforms. 



The older images distributed on the platforms have large square heads, 

 fitted to carry the red tufa crowns, wliile the conical lieads of the more re- 

 cent images found at Rana Roroka seem to preclude their ornamentation 

 with crowns. 



At many points of the island are found ruins of extensive settlements, 

 suggesting a former extensive population. On the west coast, west of Kota- 

 taki Mountain, tlie "Mohican" found the remains of stone houses, ellip- 

 tical in shape, built of rough stones, witli doorways facing the sea, with a 

 niche at the rear end covered with loose stones forming an arcli supported 

 by a fairly shaped keystone. We found the renniants of such an elliptical 

 house near the shore of La Perouse Bay (PI. 48, fig. i), and I have given on 

 PI. 49, fig. 2, a figure of the arch of one of the niches with its keystone. 

 Geiseler has given a restored figure (1. c. PI. 5) of such an elliptical 

 stone house, as well as sections of some of the stone houses at Oronaco on 

 the west face of Rana Kao (1. c. Pis. 9-11), showing the narrow passages 

 leading into them. We also found the remnants of a stone house in the 

 crater of Rana Kao, on the edge of the lake (PI. 49). The ruins of similar 

 extensive .settlements of stone houses exist also at TonQ;ariki and in the 

 vicinity of tlie larger platforms all along the coast of Easter Island. But 

 by far the most interesting settlement is that of Orongo on the western rim 

 of Rana Kao. Sketch plans of the location of these stone houses on the 

 sharp rim of the crater have been given by Paymaster Thomson (1. c. p. 479), 

 and by Geiseler (1. c. PI. 20). 



On Pis. 45-48, fig. a, I have given views of the most characteristic and 

 best preserved of the stone houses. Plate 45 shows a stone house witli a 

 double entrance. Plate 46 is a characteristic stone house with a single en- 

 trance. Plate 47 shows a stone house with two entrances widely separated, 

 with a second house placed at right angles to the larger one. Plate 48, 

 fig. 2, is another large stone house with two entrances. The houses are all 



