154, PERCY SLADEN TRUST EXPEDITION. 
covered with Pemphis, many dead stumps standing in the water and indicating loss. 
The loop at the south end is now more definitely a barachois, and behind in the land we 
found a slime-covered pool, round which snipe were feeding. Tracing from this we 
could see how the present island of St. Joseph was formed by the junction of two islets, 
and it was obvious that a third, Benjamin, would, under present conditions, soon join 
up*. St. Joseph is covered with tall coconuts, some of which were weighted down and 
killed by the large nests of Pelecanus crispus, of which there was a numerous colony. 
How far this bird is a wanderer we do not know. It was its breeding-season, and we 
saw plenty on the reefs and in the lagoon of the atoll. None, however, were to be seen 
at Darros or any of the other reefs of the Amirantes that we visited. Another common 
Fig. 43. 
St. Joseph, Amirante Group. View of east shore of Pelican Island, showing five former beaches. 
bird, likewise breeding, was the booby (Sula piscator), found so abundantly before at 
St. Pierre, and wandering daily from St. Joseph over every reef in the group. 
Approaching the north-east of Pelican Island we came across a sand-flat with five 
lines of sandstone, parallel to the beach behind, and showing former beaches which had 
been washed away one after another (fig. 43). The whole island has quite changed since 
the chart was made in 1882. and is gradually disappearing, as evidenced in its steep 
beaches with small cliffs of 2 to 4 feet. Its south end was a mass of puffin-holes, into 
which one plunged at every step. Chien and Cascassaye are wasting on their lagoon 
* It is possible that Benjamin was once part of St. Joseph. The latter island was then washed through, while 
now it is apparently joining up again. 
