BOTTOM SAMPLES, MURRAY ISLAND, THE BAHAMAS, AND FLORIDA. 283 



to England from the Marquesas; Kellerman found from 3,000 upward per 

 cubic centimeter, but the greatest number per cubic centimeter appears to 

 be off the west side of Andros Island, where, as already stated, it was 

 160,000,000 per cubic centimeter. As Marquesas Atoll is relatively small 

 and as there are tidal currents across the lagoon, tidal currents aided by wind 

 might remove more fine material than is done on the west side of Andros 

 Island; and the removal of the fine material might reduce the number of 

 bacteria. Other possible factors might operate. 



Specimen No. 1 10 is from 9 feet below the surface on which No. 115 was 

 taken. Specimen No. 1 10 in sizing is nearly the same as specimen No. 88 from 

 the west side of Andros Island, having 54.6 per cent of particles of silt and 

 clay size, while No. 88 has 49.8 per cent. Specimen 87 has 61.2 per cent of 

 particles of the same size. The deeper material in the core is finer in grain 

 than that on the surface, indicating a change in conditions not now under- 

 stood. 



The chemical analysis of specimen No. 98 (see page 269) shows 2.88 per 

 cent of MgC0 3 , nearly the same as that for specimen No. 87, from Andros 

 Island. Dr. Cushman furnishes the following notes: 



No. 95. Marquesas Lagoon, Florida, small mangrove key east of northwest entrance, 

 inside the lagoon. Worn material with a few pelecypods and fewer gastropods, 

 a very few ostracod valves and few foraminifera. 



Foraminifera: 



Polystomella striatopunctata, Orbiculina adunca, few. 



frequent. Quinqueloculina. 



No. 98. Marquesas Lagoon, Florida, east side. Material includes fragmental calcium 

 carbonate, gastropod and pelecypod shells, numerous ostracod valves, and 

 numerous foraminifera. 



Foraminifera: 



Orbiculina adunca, few. Polystomella crispa, few. 



Obitolites marginalis, few. Polystomella striatopunctata, fre- 



Articulina sagra, few. quent. 



Quinqueloculina reticulata, few. Numerous species of Quinqueloculina and 



Quinqueloculina agglutinans, few. Triloculina. 



Triloculina linneiana, few. 



Besides the organisms noted by Dr. Cushman, Halimeda is very abun- 

 dant in places in the lagoon, and the atoll rim is largely composed of more or 

 less broken Halimeda joints. There are almost no corals in the lagoon; 

 there is an occasional specimen of Maandra areolata, and, although I seem 

 to have no notes on them, there are almost certainly some colonies of branch- 

 ing species of Porites. 



Samples from Tortugas Lagoon. 



Specimen No. 97 is from Tortugas Lagoon, obstruction buoy, near 

 northwest entrance to Fort Jefferson Channel; depth about 7 fathoms. 

 Many samples have been collected within the Tortugas Lagoon and outside 

 the atoll perimeter, and when properly worked up will add much to the 

 knowledge of the bottom deposits in this area. Only sample No. 97 will be 



