184 THE ANIMALS OF LAKES AND RIVERS 
the sea as well as in fresh water. Similarly, the bivalve 
molluscs of the lakes and streams have certainly had 
marine ancestors. Most of the higher crustacea live in 
the sea, and the few crayfish and prawns which live 
in fresh water are certainly descended from ancestors 
which lived in the open ocean. 
It is otherwise with most of those numerous forms 
which, though they live in water, are yet adapted for 
breathing air. The otter which haunts the salmon 
rivers has lungs no less than its distant relative the 
polecat ; the water-tortoises have lungs like the land 
forms ; the water-beetles and the many insect larvae 
found in pools, the water-spider, the water-mites, the 
pond-snails, are all air-breathers, and their peculiar 
respiratory organs would be inexplicable if we could 
not assume that their immediate ancestors lived on 
land. In these cases there is no reason to believe that 
the animals have ever sought the sea since the time 
when their far-off ancestors acquired terrestrial charac- 
ters. Therefore in possessing some of these forms 
the fresh-water fauna is not poorer but richer than 
the sea, which has hardly any insects (cf. p. 176), no 
air-breathing molluscs, no true spiders, and so on. 
Again, while most amphibians have a fresh-water 
larval stage, and some pass much of their life in water, 
no living amphibian is ever found at any period of its 
life in the sea. 
But when we come to consider the groups which are 
richly represented in the sea, the tale is different. The 
echinoderms are exclusively marine—no starfish, sea- 
urchin, brittle-star, nor sea-lily is known from fresh 
water. No cuttle hunts its prey in the depths of the 
great lakes, nor creeps along river bottoms. The poly- 
chaete worms are almost exclusively marine, and the 
