304. 
demonstration, all that I have to offer in support of my 
opinion are arguments founded on probability and analogy. - 
“If we may judge from the form and structure of the 
coralline productions, the number of species endowed with 
the power of fabricating them must be very considerable. 
All of them agree, however, in one respect; they commence 
their work at a central point, and extend it in all directions 
indiscriminately. The law which governs the operations of 
an individual or a family in this respect, is no doubt equally 
operative on the whole society; and the aggregate ought to 
exhibit, on reaching the surface, a reef of a circular form, 
and coextensive at least with the base whereon it rests 
But Coral-reefs rarely exhibit this form, or any other that 
can be easily defined. Their forms are usually the most 
irregular that can be imagined; such, indeed, as can only be - 
attributed to a corresponding irregularity in the base on which — 
they are reared. 
* The coral-worms build perpendicularly, and only on 
fixed solid foundation. This is well exemplified in the Island 
of Bourbon; where a deep, precipitous, rocky coast on one 
side, and on the other, a shelving shore of water-worn 
pebbles, have hitherto prevented the establishment of these 
fabrics. The fact, too, that harbours and the mouths of 
rivers in tropical regions are invariably clear of living coral, 
though liable to be choked up with its débris, is illustrative 
of the same principle, j sie. 
* If the Coral-reefs were built up from the bottom of the 
sea, how are we to account for the partial manner in wh 
they are distributed? The shores of intertropical continents 
and islands are generally fringed with them; and we 
them running in interrupted chains from one groupe of islands 
to another, in the same manner.as submarine rocks 87? 
known to do in those parts of the sea that lie beyond the 
range of the coral-worms; but the ocean at large, is, as ft 
as we know,'entirely clear of them. — di Mn 
_ * Detached .reefs are sometimes met with, of ve or 
 "üperlicia! extent, one hundred yards, or perhaps less. . 47 
