GENERAL DISCUSSION. 59 



began to appear."" At 640 feet the percentage of magnesium carbonate was 26. .33, and it 

 increased, vvitli some (iuctuations, to the linal deptii of 1,114 feet. At 950 feet it reached 43 

 per cent and at the lowest depth it was 41.05 per cent. From 640 feet downward the rock was 

 essentially dolomite, although it contained an excess of calcite. Pure dolomite contains 45.65 

 per cent of magnesium carbonsite, a figure that was very nearly approached. 



The column of rock represented by the Funafuti boring thus appears to be divisible into 

 three fairly definite zones. The surface layer is about 25 fi^et thick, and its composition is 

 directly determined by the org(inisms living above it. In tliis zone the evidence of concen- 

 tration by leaching is quite clear. From 25 feet down to 640 feet the rock is essentially a 

 limestone, with very little magnesia. The lowest zone, from 640 feet downward, is dolomite, 

 and the dividing line between this and the limestone above is very distinct. Within 3 feet 

 the proportion of magnesium carbonate in the rock rises from 2.44 to 26.33 per cent. 



To account for all the differences in the Funafuti column does not fall within the scope 

 of this investigation, even if it were possible to explain them. Possibly the limestone of the 

 middle zone was laid down during a period when nonmagnesian organisms were relatively 

 much more abundant than they are now. This supposition, however, does not account for 

 the sudden change from limestone to dolomite in passing from the second to the lowest zone. 



In order to partly explain these changes we venture to offer some purely speculative sug- 

 gestions, believing that speculation is legitimate if it tends to stimulate investigation or to 

 provoke a closer scrutiny of existing evidence. 



The lowest portion of the Funafuti rock is, of course, the oldest, and it contains fragments 

 of Lithothamnion and other organisms which flom-ish abundantly only at moderate depths. 

 Magnesia was then concentrated in the rock, in part directly from livmg forms and in part 

 by leaching, as at present. The thickness of the rock shows that it must have been deposited 

 during a long period of depression, which may have submerged the island to a depth at which 

 few of the magnesian organisms, especially the algie, could thrive. A prolonged rest, a period 

 of equilibrium, then followed, dm'ing which very little rock was formed, and in this period 

 much of the dolomitization took place. 



The period of rest was succeeded by one of elevation, which brought the dolomitic rock 

 again to the surface, when reef building began anew, but with relatively fewer magnesian 

 organisms than formerly. Between 552 and 660 feet the nomnagnesian Halimeda is the main 

 constituent of the cores. The new rock, then, was less magnesian than the older, and the 

 sharp break between the two zones becomes intelligible. Magnesian organisms were not ex- 

 tinct, for theu' remains appear throughout the Funafuti cores, but they were much less abun- 

 dant than at first. Whether this supposition is true or not might be determined by a quanti- 

 tative study of the thin sections of the rock, which ought to be still in existence. The published 

 records of them seem to be purely qualitative, except in so far as they indicate the frequency 

 of occurrence of the different organisms. They do not show their relative quantities.'" At 

 present magnesian organisms predominate, and their composition is reflected in the composition 

 of the surface limestone. 



Our assumptions regarding changes of sea level at Funafuti are not altogether imaginary. 

 In their report upon the geology of the island T. W. E. David and G. Sweet" assert that 



the surface geological evidence collected by as proves, in our opinion, that several oscillating vertical movements 

 of the above have taken place in the immediate past at Funafuti, and wo should not. therefore, be surprised if the 

 evidence gained from the core shows that mo^■ements of the shore line in both directions have occurred at Funafuti 

 at earlier epochs. 



The chemical and algal evidence reinforce the physiographic evidence. On the formation 

 of " coral reefs " during subsidence or elevation there is an abundant literature, which we can not 

 attempt to summarize. We are dealing with a specific instance from a single point of view. 

 The subject is one over which there has been much controversy. Our principal assumption, 



" See the petrographic report by C. G. Cullis in The atoll of Funafuti, pp. 392-i20. 

 '• The atoU of Funafuti, tables on pp. 336-361. 

 n David, T. W. E., anil Sweet, G., idem, p. 88. 



