April s. 1888] 



NATURE 



547 



change which is continually in progress. Large masses of sand 

 are in the space of a month deposited in one spot to be swept 

 away during the next month and deposited in another. Every- 

 where there is evidence that the sea has encroached upon the 

 land, or that the land has in its turn gained upon the sea. In one 

 place numerous dead and fallen cocoa-nut palms show where 

 old-established land has been carried away ; in an adjoining spot 

 tracts of sand, either bare or covered with a scanty growth of 

 young shrubs, show where the combined action of wind and 

 waves has added a new piece to the island. Within the lagoon 

 the currents are constantly changing in force and direction, and 

 their every change affects the growth of coral in their track. In 

 estimating the structure of the atoll these changes should be kept 

 in mind, although their complexity makes it far more difficult to 

 arrive at a correct conclusion. 



In the course of my investigations I learnt to distinguish the 

 following kinds of coral rock formed by the action of the waves 

 or wind, or both combined. 



Firstly, reef rock, a tolerably homogeneous mass of compacted 

 coral debris, the component parts of which are so thoroughly 

 infiltrated with carbonate of lime held in solution in the sea-water 

 that the masses of fragments of coral composing the rock are 

 rarely distinguished from one another. This form of rock 

 exhibits a fine horizontal stratification ; it is invariably formed 

 under the sea or between tide marks. 



Secondly, boulder rock, formed just above high-tide mark by 

 means of the masses of coral which are transported across the 

 reefs by the waves and are piled up to form the low rampart 

 already alluded to. The interstices of the boulders are soon 

 filled up with coral dfbris and sand, and are cemented together 

 by the spray. Such rock is only formed on the seaward shores, 

 and invariably shows a stratification dipping downwards towards 

 the sea. 



Third!)', shingle rock, which may be of two kinds. The first 

 kind is horizontally stratified, and is scarcely distinguishable from 

 reef rock, except in its finer texture ; it is formed below water or 

 between tide marks by the agglomeration of small pieces of 

 broken coral, among which are included numerous shells of 

 mollusks, remains of Crustacea, Echinoderms, &c., and in the 

 more sheltered parts of the lagoon it may include considerable 

 masses of dead Madrepores embedded in their natural position 

 in the rock. This rock is of a looser texture than the reef rock. 

 The second kind of shingle rock is formed above high-water 

 mark by the action of the waves. It is entirely composed of small 

 fragments, and exhibits a fine stratification dipping seawards at 

 an angle. 



Lastly, there is the sand rock, formed above water by the action 

 of the wind. Wherever masses of fine sand are piled up within 

 reach of the spray they are gradually compacted, and form a 

 friable rock, the stratification of which dips seaward. 



In many parts of the island I observed that the land was 

 composed of stratified reef or shingle rock, the strata of which 

 were perfectly horizontal, and did not dip down towards either 

 shore. Having observed the manner in which the different 

 kinds of coral rock were formed, I was at a loss to understand 

 how such horizontally stratified masses could have been formed 

 in their present position above high-water mark, and could only 

 believe that they were originally formed as reef or shingle rock 

 below high-water mark, and had been subsequently raised to 

 their present position. I was thus led to believe that a slight 

 elevation hnd taken place, and this belief was strengthened by 

 a study of the formation of East Islet. This islet is about 800 

 yards long, and nearly 100 yards broad ; its westermost extremity 

 is composed of masses of sand piled up on the underlying reef 

 rock, and in this place there is a clump of high trees {Hentandia 

 peltata). The eastern and by far the larger part of the islet is of 

 different formation. The even surface of the soil is covered with a 

 low scrub, but bears no trees nor cocoa-nut palms. It forms a low 

 plateau, the surface of which does not slope down towards the 

 lagoon, but is perfectly horizontal, and stands 4 feet above the 

 very highest spring tides. The shore on the lagoonward side 

 shows an abrupt fall of 6 feet to the reef, which in this place 

 extends for a distance of 50 yards towards the lagoon, and is only 

 left uncovered at the lowest spring tides. At the eastern extremity 

 of the island there is no reef, but from i^ to 2 fathoms of water 

 are found within a few yards of the shore. This point is exposed 

 to the ocean, and a strong and constant current sets against it, 

 so that it is undergoing a considerable amount of erosion. On 

 the north or seaward side the reef again extends outwards from the 

 shore, the latter differing from the inner shore in the presence of a 



talus of large boulders which have been thrown up against it by the 

 waves. Wells have been sunk in numerous parts of the island, 

 though, for some reason which I cannot explain, water is only 

 found in one of them. Numerous pits, some of which are 9 

 feet deep, have also been dug for the purpose of planting cocoa- 

 nuts. These pits and wells expose the interesting structure of 

 the superficial part of the island. Beneath a thin surface layer 

 of sand and mould lies a horizontal layer of stratified shingle 

 rock, in which large embedded coral masses may occasionally be 

 distinguished ; this layer is about z\ feel thick. Beneath is a 

 layer of loose coral sand about 18 inches thick, and beneath that is 

 another layer of coral rock of the same character as the first, 

 and about 3 feet thick. Beneath this is another layer of friable 

 sand lying on the solid reef rock into which the excavations did 

 not penetrate. These layers lie perfectly horizontally, and do 

 not dip in any direction. They crop out above the reef on the 

 steep eastern and southern shores, and as the loose sand is 

 wa'-hed out by the waves the overhanging layer of rock breaks 

 off and falls down in large masses. The cen'.ral parts of this 

 area are absolutely beyond the reach of any waves at the present 

 time, and as the strata of rock and sand run evenly through it 

 there is no evidence of its having been formed by successive 

 additions of material through the action of the waves. Nor can 

 it possibly have been .formed under the surface of the water 

 unless it has since been raised to its present position, for, as I 

 have said, its upper surface is 4 feet above the level of high 

 spring tides. On one occasion when the tide rose to an ab- 

 normal height and invaded several parts of the main island, I 

 saw that the water reached to within 3 feet of the top of the shore, 

 but even then the whole of the upper stratum of coral rock was 

 well above the waves. It is scarcely credible that an even 

 layer of shingle rock could have been formed above the highest 

 high-water mark. 



My belief in elevation is farther strengthened by the 

 following facts, communicated to me by M. Spurs, a resident 

 for twenty-five years at Diego, an ardent naturalist, and much 

 interested in coral formations. 



A small shore crab of the genus Ocyptts is always to be 

 found on the sandy flats between high and low water marks. 

 These crabs, as is well known, form numerous gallei^ies in the fine 

 muddy sand, which they line with seaweed, &c. , to prevent 

 their falling in. These galleries open to the surface by short 

 passages placed perpendicularly, the mouths of which open 

 only a few inches above the level of low tide. This crab is 

 only found on the shore between tide marks ; on the dry land 

 its place is taken by Gearcimis, another genus of crab, which 

 forms different burrows. In the west part of East Islet there is 

 an aggregate of friable, scarcely compacted sand, which has 

 somewhat the appearance of half- dried clay. It lies 5 feet 

 above high-water mark, and was found by M. Spurs, during 

 some excavations which he had to make for the purpose of 

 constructing a slip for boats, to be riddled with the sea- 

 weed-lined galleries of Ocypiis, evidently long since disused 

 and empty. 



Having made this observation on East Island, M. Spurs made 

 a search in similar formations on the main island, and found, 

 he tells me, precisely the same facts in several instances, 

 aggregates of sand lying at some distance above high-water 

 mark, riddled with the abandoned burrows of Ocypiis. Now, 

 since the burrows of Ocypiis are quite characteristic, and could 

 not have been mistaken by so £;ood an observer as M. Spurs for 

 those of another species, and since they are in the present day 

 only found between tide marks, these observations afford a 

 further presumption in favour of a slight elevation having' 

 recently taken place. In any case they preclude the idea of any 

 subsidence being in progress, as Mr. Darwin fancied to be the 

 case in the Keeling atoll. M, Spurs further informs me that, 

 during the time that he was superintendent of the oil company's 

 estate, he caused more than 30,000 pits to be dug on the main 

 island for the purpose of planting cocoa-nut palms, and that he 

 frequently observed in different localities the same alternate 

 layers of sand and rock that I have described as existing on East 

 Island. These alternations of sand and rock would suggest 

 alternations of very slight subsidence with very slight elevation, 

 rather than a single movement of upheaval, yet on the sup- 

 position that all the layers were formed beneath the water, as 

 their horizontal stratification leads me to believe, I can venture 

 on the following explanation. The mass of rock which forms 

 the base upon which the islets and other dry land rests is solid 

 reef rock, and the whole floor of the lagoon is similarly formed. 



