MAYOR— ROSE ATOLL, AMERICAN SAMOA. 69 



appears that the cHmate was tropical when the sea stood at least 

 8 feet higher than at present and cut the bench around all the volcanic 

 islands of American Samoa. 



In the Funafuti boring the percentage of magnesium in the core 

 ranged from 4 per cent, at a depth of 4 feet to 16 per cent, at 15 

 and 26 feet, below which it declined to 3 per cent, at a depth of 

 60 feet. Judd attributes this high percentage of magnesivmi to the 

 supposed leaching out of calcium by the sea water, but we now 

 know that the surface waters of the tropical Pacific are super- 

 saturated in respect to calcium carbonate, and that calcium car- 

 bonate is therefore practically insoluble in this surface water. Judd 

 admits that there is much lithothamnium in this upper part of the 

 core of the boring, but unfortunately he made no analysis of the 

 magnesium contents of any lithothamnia at present growing upon 

 the Funafuti reef ; and. judging from the conditions at Rose Atoll, 

 I am inclined to believe that the magnesiiun in this upper part of the 

 Funafuti boring is due solely to its being largely composed of litho- 

 thamnium, and not to any leaching out of calcium carbonate. This 

 conclusion is supported also by the fact that in the Funafuti boring 

 between 100 feet and 637 feet in depth the magnesium carbonate 

 was nowhere greater than 5.4 per cent.; yet if calcium leached out 

 in water about 26 feet deep, why did it not leach out at these greater 

 depths where conditions of temperature and carbon dioxide are 

 more favorable for solution than on the surface? 



Wilkes, 1852, Narrative of the U. S. Exploring Expedition, Vol. 

 I, p. 155, states : 



Some boulders of vesicular lava were seen on the coral reef (of Rose 

 Atoll) they were from 20 to 200 pounds in weight and were found among 

 blocks of coral conglomerate. (See also Couthouy, 1844, Boston Journal of 

 Nat. Hist., vol. 4, p. 138.) 



I was unable to find any volcanic rock upon Rose Atoll, and it 

 seems probable that Wilkes or Couthouy mistook some dark-col- 

 ored scoreaceous-looking, weather-worn limestone boulders for lava. 



Summary. 

 The visible parts of the rim of Rose Atoll is composed of litho- 

 thamnium rather than of coral, and is apparently chiefly constituted 



