XXXVII .
On the Formation of Islands .
By Alexander Dalrymple , Esquire .
Communicated by C. Morton , M. D. S. R. S. Received May 4 , 1767 .
THERE is not a part of natural history more curious , or perhaps to a navigator more useful , than an enquiry into the formation of islands .
The origin of islands , in general , is not the point to be discussed ; but of low , flat , islands in the wide ocean ; such as are most of those hitherto discovered in the vast Southsea .
Read July 2 , 1767 .
These islands are generally long , and narrow ; they are formed by a narrow bar of land , enclosing the sea within it ; generally , perhaps always , with some channel of ingress at least to the tide ; commonly , with an opening capable of receiving a canoe ; and frequently sufficient to admit even larger vessels .
The origin of these islands will explain their nature .
What led me first to this deduction was an observation of Abdul Roobin , a Sooloo pilot ; that all the islands , lying off the N.E. coast of Borneo , had shoals to the eastward of them .
These islands being covered to the westward by Borneo ; the winds from that quarter do not attack them with violence .
But the N.E. winds , tumbling in the billows from a wide ocean , heap up the coral with which those seas are filled .
This , obvious after storms , is perhaps , at all other times , imperceptibly effected .
The coral banks , raised in the same manner , become dry .
These banks are found of all depths , at all distances from shore , entirely unconnected with the land , and detached from each other : although it often happens they are divided by a narrow gut , without bottom .
Coral banks also grow , by a quick progression , towards the surface ; but the winds , heaping up the coral from deeper water , chiefly accelerate the formation of these into shoals and islands .
They become gradually shallower ; and , when once the sea meets with resistance , the coral is quickly thrown up by the force of the waves breaking against the bank ; and hence it is that , in the open sea , there is scarce an instance of a coral bank having so little water , that a large ship cannot pass over , but it is also so shallow that a boat would ground on it .
I have seen these coral banks in all the stages ; some in deep water , others with few rocks appearing above the surface , some just formed into islands , without the least appearance of vegetation , and others , from such as have a few weeds on the highest part , to those which are covered with large timber , with a bottomless sea , at a pistol shot distance .
The loose coral , rolled inward by the billows in lage pieces , will ground , and the reflux being unable to carry them away , they become a bar to coagulate the sand , always found intermixed with coral ; which sand , being easiest raised , will be lodged at top .
When the sand bank is raised by violent storms , beyond the reach of common waves , it becomes a resting place to vagrant birds , whom the search of prey draws thither .
The dung , feathers , &amp;c. increase the soil , and prepare it for the reception of accidental roots , branches , and seed , cast up by the waves , or brought thither by birds .
Thus islands are formed : the leaves and rotten branches , intermixing with the sand , form in time a light black mould , of which in general these islands consist , more sandy , as less woody ; and when full of large trees , with a greater proportion of mould .
Cocoa nuts , continuing long in the sea without losing their vegetative powers , are commonly to be found in such islands ; particularly as they are adapted to all soils , whether sandy , rich , or rocky .
The violence of the waves , within the Tropicks , must generally be directed to two points , according to the monsoons .
Hence the islands formed from coral banks must be long and narrow , and lie nearly in a meridional direction .
For even supposing the banks to be round , as they seldom are when large , the sea , meeting most resistance in the middle , must heave up the matter in greater quantites there than towards the extremities : and , by the same rule , the ends will generally be open , or at least lowest .
They will also , commonly , have soundings there , as the remains of the bank , not accumulated , will be under water .
Where the coral banks are not exposed to the common monsoon , they will alter their direction ; and be either round , extend in the parallel , or be of irregular forms , according to accidental circumstances .
The interior parts of these islands , being sea , sometimes form harbours capable of receiving vessels of some burden , and , I believe always , abound greatly with fish ; and such as I have seen , with turtle-grass and other sea-plants , particularly one species , called by the Sooloos Gammye , which grows in little globules , and is somewhat pungent , as well as acid , to the taste .
It need not be repeated , that the ends of those islands , only , are the places to expect soundings : and they commonly have a shallow spit running out from each point .
Abdul Roobin 's observation points out another circumstance , which may be useful to navigators ; by consideration of the winds to which any islands are most exposed , to form a probable conjecture which side has deepest water ; and from a view which side has the shoals , an idea may be formed which winds rage with most violence .

