Voyages of 'Discovery and Survey. 329 



they could be easily floated over extensive flats covered by the sea. 

 It would appear that these pumice pebbles are mingled with the coral 

 conglomei'ates, and were noticed in the coral rock of Raine''s Island, 

 The accumulation of pumice at one time moi-e than at another 

 would depend upon the nature of the supply, and would last no 

 longer than the eruptions continued which produced it in some situa- 

 tion whence it could be drifted to the Australian coast and be ground 

 into pebbles. That there has been elevation and depression of the 

 land and adjacent sea-bottoms in many localities in different parts of 

 the world, producing the effects connected with coral reefs which 

 have been pointed out by Mr Darwin, is exceedingly probable : in- 

 deed, respecting the elevation of coral reefs we have abundant proofs, 

 and therefore there is little reason to doubt depression, which the 

 accumulation of rocks of all geological ages shews to have been very 

 common and often very extensive. Upon such points the geologist 

 can have little doubt, but he may sometimes doubt the application of 

 depression to every case of a barrier reef, seeing the power of exten- 

 sion, through long-continued time, of the reefs outward into deep 

 water. The study of reef-making corals shews us how they seek 

 clean water and the surf, and the manner in which they avoid waters 

 charged with mineral matter in mechanical suspension. Hence we 

 commonly find them either attached to, or at a distance from, the 

 shore, according to the clearness of the water, though certainly here 

 and there they seem to struggle hard with adverse conditions in this 

 respect. Where by their inci-ease the surf fails them, they seem to 

 be soon covered by other corals and Nulliporse ; the latter Mr Dar- 

 win points out as especially creeping over them towards the surf. 

 When we consider that many a volcanic island rises through the sea 

 by accumulating erupted masses of ashes, cinders, and molten rock, 

 and that the permanency of its continuance above the sea-level de- 

 pends upon its power to resist the cutting and levelling action of 

 the breakers, we perceive that such an accumulation may be either 

 cut down entirely by this action, as that of Graham Island, which 

 came through the waters of the Mediterranean in 1831, and now re- 

 mains a shoal, or be notched by it, if the mass be sufficiently hard 

 to resist entire removal. When partly cut back, the matter removed 

 is distributed over the talus of the general volcanic protuberance, not 

 very materially increasing its angle of slope, since it would be piled 

 over the talus much in the same manner as the detrital coral is accu- 

 mulated in front of the coral reefs. The island, reduced to a shoal, 

 would have the matter distributed over the flanks of the latter in 

 the same manner, considering the accumulation to have been conical, 

 though moie flat than the cones formed in the air. Supposing no 

 subsidence, and this is iiy no means necessary, such a volcanic mass 

 as Graham Island might be reduced to a shoal, the depth of which, 

 beneath the surface, would cori'espond with the power of the waves 

 during heavy gales of wind to remove ash and cinders from the top 



