396 ON THE GROWTH OF 



to masses. Whenever the summit of this new island can 

 remain constantly uncovered by the sea, and the waves 

 can no longer destroy what they themselves have contri- 

 buted to form, then its extent is enlarged, and its edges 

 are gradually raised by the successive addition of sand. 

 According to the direction of the winds and currents they 

 may long remain sterile ; but if the seeds of vegetables be 

 transported to them from the neighbouring shores, by the 

 action of these two causes, then, in latitudes favourable to 

 their development, we presently see these islands becoming 

 covered with verdure, the successively accumulated remains 

 of which form layers of soil, which contribute to the ele- 

 vation of the surface. 



But, in order that this phenomenon of growth be ac- 

 complished, the distance from land must not be too great, 

 because then the vegetables cannot get so easily to the 

 islets in question, which then remain almost always bare 

 and sterile. And for this reason what navigators report 

 of those madrepore Islands of the Great Ocean, which are 

 covered with verdure, and are yet at a great distance from 

 any known land, has always appeared to us extraordi- 

 nary ; and that so much the more, that, in those vast 

 spaces, the violence of the waves, which nothing can 

 break there, must disturb the operations of the zoophytes. 

 We do not, however, deny the existence of these islands, 

 which it would be interesting carefully to examine anew ; 

 for, whenever navigators meet with low islands between the 

 Tropics, they do not hesitate, in compliance with the ge- 

 nerally received opinion, to say that they consist of ma- 

 drepores. Yet how many islands, which scarcely rise a- 

 bove the surface of the water, recognise no such origin ? 

 We may mention, as an example, the Island of Boni, si- 



