182 CORAL AND ATOLLS 



thirteen feet at its highest point ; and in Pulu Atas, to the, 

 windward, rising to a " hill " of thirty feet. 



From this sea-bank the land falls to the centre of the 

 island, where in some instances there are many hollows below 

 the high spring-tide level. Towards the lagoon the land rises 

 again as an inconsiderable sand ridge, and thence falls gently 

 away to the lagoon. In Pulu Tikus, which is the island that 

 has undergone least change from the bare debris condition of 

 its first making, the greater part of the surface of the land is 

 bare coral fragments with a very sparse admixture of sand and 

 vegetable mould, and there has been very little building up of 

 the interior of the island. The result is that most of its 

 surface is raised but a foot or two above high-water level, 

 and a considerable part of it is below, and as the debris is 

 easily permeable to water, the burrows in which the Ka'peting 

 Balong lives are usually filled with salt water almost to their 

 mouths. Wherever a hole is dug, the water is found just 

 below the surface, and large puddles are formed in many 

 places at every high spring tide ; whilst the water in a fish- 

 pond, some thirty feet from the lagoon shore, follows all the 

 movements of the tide with but little delay. There are many 

 facts of great interest in connection with this sponge-like 

 condition of the islands, which permits the sea to fill them at 

 high tide, and leave them when the tide is low. 



It is this condition that gives rise to the curious pheno- 

 menon of rising and falling wells of fresh Avater, which can be 

 made on a great many of the islands, and to account for which 

 many suggestions have been put forward. It has been said 

 that the sea water at the height of the tide percolates into 

 the island, and, as it goes, it becomes filtered through the 

 coral so perfectly that in the middle of the island it becomes 

 quite fresh, and may there be tapped as good drinking water 

 in wells. That filtration by the accumulation of dead coral 

 could be so complete as to remove sodium chloride from its 

 solution is an assumption so bold that its acceptance could 

 only be justified by the careful exclusion of all other possible 

 causes of the freshening of the water : and as a matter of fact 



