576 PROCEEDINGS OF THE AMERICAN ACADEMY. 



from six to ten fathoms deep. The bottom (at Station 404, one fathom), 

 just off the beach at Trunk Island, is grayish white, mottled with red, 

 brown, or yellow ; it is composed of coarse sand, the largest fragments 

 15 mm. in diameter, and about 1 per cent fine ooze. This bottom dif- 

 fers markedly from any hitherto mentioned. There are very few Foram- 

 inifera, not more than 2 per cent ; echinoid plates and spines, coralline 

 algae, alcyonarian spicules, moUusks, corals, and polyzoans together form 

 not more than 20 per cent ; the remaining 75 per cent consists almost 

 wholly of limestone fragments, ranging in diameter from 1 to 15 mm., of 

 rounded and irregular shapes, and variously stained red and brown. 

 They are the coarsest detritus from the neighboring cliffs. 



In four fathoms of water, between Rabbit Island and the shore (Sta- 

 tion 470), the sand is of the same color and composed of the same ele- 

 ments, but much finer, and with 5 per cent fine ooze. There are no 

 large fragments, and the particles are of fairly uniform size, 1-5 mm. in 

 diameter. The sand decreases in coarseness from shoal water near the 

 shores to greater depths, with an increase of fine ooze ; a sample from 

 eight fathoms consists of very fine sand with 30 per cent ooze, and a 

 few large fragments of dead coral. Living organisms seem to have 

 added very little to the sands of Harrington Sound. The undercut 

 and water-worn condition of the cliffs, and the huge limestone masses 

 recently detached from them, bear witness to the effects of surf action on 

 soft friable limestone, even in so limited an area as the Sound. Enor- 

 mous quantities of detritus, both coarse and fine, must be constantly 

 added to the beaches, the former remaining in shallow water, the latter 

 being swept out and gradually deposited in greater depths beyond the 

 reach of wave action. The bottom in this sound is more extensively 

 derived from broken-down limestone than that of any other extended 

 area which I have examined, and I believe that it is being added to with 

 greater rapidity. 



The most important and widespread of all the formations of sand that 

 are now taking place occur on the northern side of the main islands, 

 between them and the nearly continuous boundary reef. Both Heilprin 

 and Agassiz have shown that this area consists of a series of sunken 

 lagoons, bounded by more or less continuous ridges, the " Flats." At 

 only one spot, the " North Rocks," do any of the flats rise above sea- 

 level, but the depth in many places is so slight that they are nearly 

 awash at low tide. The numerous ledges and shoal patches, although 

 covered by a luxuriant growth of corals and gorgonians. are not of coral 

 origiu, but are ledges of the ordinary aeoliau rock, over which the corals 



