NATURAL FISH-TRAPS 139 



think nothing of taking a flying leap across the dry 

 rocks into another chain of pools. The shallower 

 pools are often filled with writhing masses of brittle- 

 stars, but it is difficult to get a perfect specimen, 

 owing to their habit of threading their long snaky 

 arms through and through the crannies of the rocks. 

 In a single morning one might see hundreds of 

 specimens of one species (Ophiothrix longipeda), and 

 this gives one an impressive lesson as to the manner 

 in which coral-reefs are augmented and solidified, by 

 the death and burial of the shoals of calcareous 

 animals that are generated upon them. 



There are in Great Coco, connected with some of 

 the mangrove swamps, some little creeks, which get 

 filled now and again at high spring-tide, and become 

 dry, or nearly dry, in the intervals between the spring- 

 tides. In one of these, which still contained a little 

 very foul seawater, I found a multitude of small species 

 of sea-perches {Pristiponia^ Therapon, and Dules), most 

 of which were dead and decomposing, the few survivors 

 being almost at the last gasp. This observation throws 

 some light on two important kinds of natural phenomena, 

 namely, modes of origin of fossils, and modes in which 

 freshwater faunas are initiated and recruited from the 

 sea. Here, in fact, was seen, inland, the early founda- 

 tions of a bed of marine fossils ; and all we have to 

 do, in order to complete the story, is to imagine the 

 heap of victims being augmented at each successive 



