106 A NATURALIST ON DESERT ISLANDS. 



difficult to land ^yithout running the risk of smashing 

 up the boat in which one goes ashore. One of the 

 launches used to tow us up to the soufch side of the island, 

 and then wait for the boat, while we scrambled ashore 

 as best we could. As soon as this was safely accomplished, 

 a little piece of cliff work had to be faced, during which 

 process the guns, collecting bags, and whatever gear we 

 were carrying, were hauled up from hand to hand ; and in 

 such wise we literally hauled ourselves up the jagged coral 

 rock and so gained the top of the island. 



Unfortunately, the moment you have congratulated 

 yourself on the accomplishment of this preliminary piece 

 of work, you find your troubles have only just begun. 

 The cliff where we used to land was bad enough, studded 

 as it was with thorny cactus, prickly bushes and creepers ; 

 but it was a mere luxury compared with, the horribly 

 jagged surface of coral limestone A^iiich you have to 

 scramble over, once you are fairly started on your 

 explorations. 



To make matters worse, this boot-destroying, leg-break- 

 ing surface is densely overgrown for some way in with low 

 white mangrove " bushes (? Laguncularia ? Conocarpus), 

 struggling to grow in what must have been for them very 

 arid and rocky conditions ; and through the short upright 

 stems of which you have to thread your ^vay with bent 

 back and perspiring bodj; . 



This sort of thing continues for long enough to make 

 you wonder if " the game is worth the candle ; " and then 

 you suddenly find yourself in the middle stretches of the 

 island, where the surface is seamed and channelled with 

 long and deep chasms running east and west. Some of 

 these we could stride across, others we could not nearly 

 jump across, some are three, some six, and others ten 

 feet deep. The walls of these troughs are nearly always 

 perpendicular, or at least that part of them which is visible ; 

 for the chasms have been filled to various levels with 

 vegetable mould, which has been formed hy the constant 

 dropping of leaves from the low forest trees which every- 



