KOLUMADULU AND HADDUMATI. 



407 



ends from the larger sandy inner areas, and in manj^ places isolated islets, or rock masses, 

 so formed are found in or just behind the boulder zone. Most islands are separated from 

 the atoll-lagoons by broad flats, and in but few are definite changes of any sort indicated 

 on this side. The sandy islands commonly show a slight washing away on their seaward faces 

 with often, where many islands are present, a growth of spits towards one another, or, if an 

 island be isolated, along the reef In some places, too, sand-banks have formed on the reefs, 

 and here and there an island seems to have been completely removed. The most marked 

 change was seen at Fahala in Kolumadulu, to which has become joined the small island to 

 the north. The south end has on the contrary been completely washed away, and in its place 



»-.i'«' 



u/"'' 



SCALE 



Fio. 106. Kolumadulu AtoU (a Utile altered from the Admiralty Chart). 



is a double velu in the centre of the reef the largest and deepest part to the E. Further 

 south the next seven islands mostly have to the east rocky points, which are being washed 

 away, while the sandy parts behind are growing out along the reef towards one another, so 

 that the south four have nearly fused. Isdu in Haddumati sends out a point to the boulder 

 zone north-east and another north-west, both of which are probably outgrowths, while the 

 north end of Dambidu has been cut off to form a fresh islet. The latter has also taken 

 place at Funadu, but the second island north of Ghang has been washed away. 



Within the lagoons of both atolls a mixture of .sand and mud is found, covering the 

 bottoms, on which a few Mollusca and Crustacea alone live. In some places there is a 

 covering of low green algae, and in others homy or siliceous sponges. Mr Forster Cooper 

 nowhere recorded a determinate change in depth, nor did I find any in a close-set line of 



