82 



THREE CRUISES OF THE " BLAKE. 



more active tidal circulation through them formerly than is now 

 taking place. An examination of the cross-sections in the direc- 

 tion of the prevailing winds, and in the direction of the tides, 

 shows at a glance the mound-shaped mass which forms the base, 

 risino- from the general level of twenty-five to thirty fathoms, 

 with its abrupt side facing the east. There are also seen the 

 deep furrows, more or less broad, which have been scooped out 

 of this mass by the action of the currents, such as those passing 

 through the Southwest Channel. 



The corals which give to the reefs their peculiar physiognomy 

 are the extensive patches of Madrepora (principally M. cervi- 

 cornis), the clusters of the two common species of Porites (P. 



furcata and P. clavm^ia) (Fig. 

 48) covering more or less the 

 shallow tracts of coarse sand, 

 and Mmandinna areolata grow- 

 ing between the patches of ma- 

 rine lawns formed by a species 

 of Thalassia, with occasional 

 patches of Anadyomene. In other 

 parts of the reefs large holothu- 

 rians (MuUeria) lie scattered on 

 the bottom, while in somewhat 

 deeper regions are pockets filled 

 Fig. 48. — Pontes ciavaria. (Agassiz.) -with large Diadematidse. Im- 

 mense masses of nuUipores (Udotea and Hahmeda, Figs. 49, 50) 

 and corallines grow on the shallowest flats, on the tops of the 

 branches of madrepores which have died from exposure to the 

 air, either because they have grown up to the surface and so 

 have become exposed by extreme low tides, or because strong 

 winds have blown the Avater from the flats. The destructive 

 effect which an extremely low tide has on a growing reef is well 

 shown on the flats to the southward of Fort Jefferson, where 

 the upper part of the branches of a certain size reaching up to 

 a given level are frequently killed off by low tides. Exposure 

 to the action of the sun even for a very short time is sufficient 

 to kill them. The extreme sensitiveness of all corals to atmos- 

 pheric action is well known, so that it becomes plain, as has been 



